From: Simon Glass <simon.glass@canonical.com> Copy extents.c and ext4_extents.h from Linux v6.18 fs/ext4 directory. These files implement the extent-based block mapping used by ext4. Co-developed-by: Claude Opus 4.5 <noreply@anthropic.com> --- fs/ext4l/ext4_extents.h | 268 ++ fs/ext4l/extents.c | 6189 +++++++++++++++++++++++++++++++++++++++ 2 files changed, 6457 insertions(+) create mode 100644 fs/ext4l/ext4_extents.h create mode 100644 fs/ext4l/extents.c diff --git a/fs/ext4l/ext4_extents.h b/fs/ext4l/ext4_extents.h new file mode 100644 index 00000000000..c484125d963 --- /dev/null +++ b/fs/ext4l/ext4_extents.h @@ -0,0 +1,268 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright (c) 2003-2006, Cluster File Systems, Inc, info@clusterfs.com + * Written by Alex Tomas <alex@clusterfs.com> + */ + +#ifndef _EXT4_EXTENTS +#define _EXT4_EXTENTS + +#include "ext4.h" + +/* + * With AGGRESSIVE_TEST defined, the capacity of index/leaf blocks + * becomes very small, so index split, in-depth growing and + * other hard changes happen much more often. + * This is for debug purposes only. + */ +#define AGGRESSIVE_TEST_ + +/* + * With EXTENTS_STATS defined, the number of blocks and extents + * are collected in the truncate path. They'll be shown at + * umount time. + */ +#define EXTENTS_STATS__ + +/* + * If CHECK_BINSEARCH is defined, then the results of the binary search + * will also be checked by linear search. + */ +#define CHECK_BINSEARCH__ + +/* + * ext4_inode has i_block array (60 bytes total). + * The first 12 bytes store ext4_extent_header; + * the remainder stores an array of ext4_extent. + * For non-inode extent blocks, ext4_extent_tail + * follows the array. + */ + +/* + * This is the extent tail on-disk structure. + * All other extent structures are 12 bytes long. It turns out that + * block_size % 12 >= 4 for at least all powers of 2 greater than 512, which + * covers all valid ext4 block sizes. Therefore, this tail structure can be + * crammed into the end of the block without having to rebalance the tree. + */ +struct ext4_extent_tail { + __le32 et_checksum; /* crc32c(uuid+inum+extent_block) */ +}; + +/* + * This is the extent on-disk structure. + * It's used at the bottom of the tree. + */ +struct ext4_extent { + __le32 ee_block; /* first logical block extent covers */ + __le16 ee_len; /* number of blocks covered by extent */ + __le16 ee_start_hi; /* high 16 bits of physical block */ + __le32 ee_start_lo; /* low 32 bits of physical block */ +}; + +/* + * This is index on-disk structure. + * It's used at all the levels except the bottom. + */ +struct ext4_extent_idx { + __le32 ei_block; /* index covers logical blocks from 'block' */ + __le32 ei_leaf_lo; /* pointer to the physical block of the next * + * level. leaf or next index could be there */ + __le16 ei_leaf_hi; /* high 16 bits of physical block */ + __u16 ei_unused; +}; + +/* + * Each block (leaves and indexes), even inode-stored has header. + */ +struct ext4_extent_header { + __le16 eh_magic; /* probably will support different formats */ + __le16 eh_entries; /* number of valid entries */ + __le16 eh_max; /* capacity of store in entries */ + __le16 eh_depth; /* has tree real underlying blocks? */ + __le32 eh_generation; /* generation of the tree */ +}; + +#define EXT4_EXT_MAGIC cpu_to_le16(0xf30a) +#define EXT4_MAX_EXTENT_DEPTH 5 + +#define EXT4_EXTENT_TAIL_OFFSET(hdr) \ + (sizeof(struct ext4_extent_header) + \ + (sizeof(struct ext4_extent) * le16_to_cpu((hdr)->eh_max))) + +static inline struct ext4_extent_tail * +find_ext4_extent_tail(struct ext4_extent_header *eh) +{ + return (struct ext4_extent_tail *)(((void *)eh) + + EXT4_EXTENT_TAIL_OFFSET(eh)); +} + +/* + * Array of ext4_ext_path contains path to some extent. + * Creation/lookup routines use it for traversal/splitting/etc. + * Truncate uses it to simulate recursive walking. + */ +struct ext4_ext_path { + ext4_fsblk_t p_block; + __u16 p_depth; + __u16 p_maxdepth; + struct ext4_extent *p_ext; + struct ext4_extent_idx *p_idx; + struct ext4_extent_header *p_hdr; + struct buffer_head *p_bh; +}; + +/* + * Used to record a portion of a cluster found at the beginning or end + * of an extent while traversing the extent tree during space removal. + * A partial cluster may be removed if it does not contain blocks shared + * with extents that aren't being deleted (tofree state). Otherwise, + * it cannot be removed (nofree state). + */ +struct partial_cluster { + ext4_fsblk_t pclu; /* physical cluster number */ + ext4_lblk_t lblk; /* logical block number within logical cluster */ + enum {initial, tofree, nofree} state; +}; + +/* + * structure for external API + */ + +/* + * EXT_INIT_MAX_LEN is the maximum number of blocks we can have in an + * initialized extent. This is 2^15 and not (2^16 - 1), since we use the + * MSB of ee_len field in the extent datastructure to signify if this + * particular extent is an initialized extent or an unwritten (i.e. + * preallocated). + * EXT_UNWRITTEN_MAX_LEN is the maximum number of blocks we can have in an + * unwritten extent. + * If ee_len is <= 0x8000, it is an initialized extent. Otherwise, it is an + * unwritten one. In other words, if MSB of ee_len is set, it is an + * unwritten extent with only one special scenario when ee_len = 0x8000. + * In this case we can not have an unwritten extent of zero length and + * thus we make it as a special case of initialized extent with 0x8000 length. + * This way we get better extent-to-group alignment for initialized extents. + * Hence, the maximum number of blocks we can have in an *initialized* + * extent is 2^15 (32768) and in an *unwritten* extent is 2^15-1 (32767). + */ +#define EXT_INIT_MAX_LEN (1UL << 15) +#define EXT_UNWRITTEN_MAX_LEN (EXT_INIT_MAX_LEN - 1) + + +#define EXT_FIRST_EXTENT(__hdr__) \ + ((struct ext4_extent *) (((char *) (__hdr__)) + \ + sizeof(struct ext4_extent_header))) +#define EXT_FIRST_INDEX(__hdr__) \ + ((struct ext4_extent_idx *) (((char *) (__hdr__)) + \ + sizeof(struct ext4_extent_header))) +#define EXT_HAS_FREE_INDEX(__path__) \ + (le16_to_cpu((__path__)->p_hdr->eh_entries) \ + < le16_to_cpu((__path__)->p_hdr->eh_max)) +#define EXT_LAST_EXTENT(__hdr__) \ + (EXT_FIRST_EXTENT((__hdr__)) + le16_to_cpu((__hdr__)->eh_entries) - 1) +#define EXT_LAST_INDEX(__hdr__) \ + (EXT_FIRST_INDEX((__hdr__)) + le16_to_cpu((__hdr__)->eh_entries) - 1) +#define EXT_MAX_EXTENT(__hdr__) \ + ((le16_to_cpu((__hdr__)->eh_max)) ? \ + ((EXT_FIRST_EXTENT((__hdr__)) + le16_to_cpu((__hdr__)->eh_max) - 1)) \ + : NULL) +#define EXT_MAX_INDEX(__hdr__) \ + ((le16_to_cpu((__hdr__)->eh_max)) ? \ + ((EXT_FIRST_INDEX((__hdr__)) + le16_to_cpu((__hdr__)->eh_max) - 1)) \ + : NULL) + +static inline struct ext4_extent_header *ext_inode_hdr(struct inode *inode) +{ + return (struct ext4_extent_header *) EXT4_I(inode)->i_data; +} + +static inline struct ext4_extent_header *ext_block_hdr(struct buffer_head *bh) +{ + return (struct ext4_extent_header *) bh->b_data; +} + +static inline unsigned short ext_depth(struct inode *inode) +{ + return le16_to_cpu(ext_inode_hdr(inode)->eh_depth); +} + +static inline void ext4_ext_mark_unwritten(struct ext4_extent *ext) +{ + /* We can not have an unwritten extent of zero length! */ + BUG_ON((le16_to_cpu(ext->ee_len) & ~EXT_INIT_MAX_LEN) == 0); + ext->ee_len |= cpu_to_le16(EXT_INIT_MAX_LEN); +} + +static inline int ext4_ext_is_unwritten(struct ext4_extent *ext) +{ + /* Extent with ee_len of 0x8000 is treated as an initialized extent */ + return (le16_to_cpu(ext->ee_len) > EXT_INIT_MAX_LEN); +} + +static inline int ext4_ext_get_actual_len(struct ext4_extent *ext) +{ + return (le16_to_cpu(ext->ee_len) <= EXT_INIT_MAX_LEN ? + le16_to_cpu(ext->ee_len) : + (le16_to_cpu(ext->ee_len) - EXT_INIT_MAX_LEN)); +} + +static inline void ext4_ext_mark_initialized(struct ext4_extent *ext) +{ + ext->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ext)); +} + +/* + * ext4_ext_pblock: + * combine low and high parts of physical block number into ext4_fsblk_t + */ +static inline ext4_fsblk_t ext4_ext_pblock(struct ext4_extent *ex) +{ + ext4_fsblk_t block; + + block = le32_to_cpu(ex->ee_start_lo); + block |= ((ext4_fsblk_t) le16_to_cpu(ex->ee_start_hi) << 31) << 1; + return block; +} + +/* + * ext4_idx_pblock: + * combine low and high parts of a leaf physical block number into ext4_fsblk_t + */ +static inline ext4_fsblk_t ext4_idx_pblock(struct ext4_extent_idx *ix) +{ + ext4_fsblk_t block; + + block = le32_to_cpu(ix->ei_leaf_lo); + block |= ((ext4_fsblk_t) le16_to_cpu(ix->ei_leaf_hi) << 31) << 1; + return block; +} + +/* + * ext4_ext_store_pblock: + * stores a large physical block number into an extent struct, + * breaking it into parts + */ +static inline void ext4_ext_store_pblock(struct ext4_extent *ex, + ext4_fsblk_t pb) +{ + ex->ee_start_lo = cpu_to_le32((unsigned long) (pb & 0xffffffff)); + ex->ee_start_hi = cpu_to_le16((unsigned long) ((pb >> 31) >> 1) & + 0xffff); +} + +/* + * ext4_idx_store_pblock: + * stores a large physical block number into an index struct, + * breaking it into parts + */ +static inline void ext4_idx_store_pblock(struct ext4_extent_idx *ix, + ext4_fsblk_t pb) +{ + ix->ei_leaf_lo = cpu_to_le32((unsigned long) (pb & 0xffffffff)); + ix->ei_leaf_hi = cpu_to_le16((unsigned long) ((pb >> 31) >> 1) & + 0xffff); +} + +#endif /* _EXT4_EXTENTS */ + diff --git a/fs/ext4l/extents.c b/fs/ext4l/extents.c new file mode 100644 index 00000000000..ca5499e9412 --- /dev/null +++ b/fs/ext4l/extents.c @@ -0,0 +1,6189 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright (c) 2003-2006, Cluster File Systems, Inc, info@clusterfs.com + * Written by Alex Tomas <alex@clusterfs.com> + * + * Architecture independence: + * Copyright (c) 2005, Bull S.A. + * Written by Pierre Peiffer <pierre.peiffer@bull.net> + */ + +/* + * Extents support for EXT4 + * + * TODO: + * - ext4*_error() should be used in some situations + * - analyze all BUG()/BUG_ON(), use -EIO where appropriate + * - smart tree reduction + */ + +#include <linux/fs.h> +#include <linux/time.h> +#include <linux/jbd2.h> +#include <linux/highuid.h> +#include <linux/pagemap.h> +#include <linux/quotaops.h> +#include <linux/string.h> +#include <linux/slab.h> +#include <linux/uaccess.h> +#include <linux/fiemap.h> +#include <linux/iomap.h> +#include <linux/sched/mm.h> +#include "ext4_jbd2.h" +#include "ext4_extents.h" +#include "xattr.h" + +#include <trace/events/ext4.h> + +/* + * used by extent splitting. + */ +#define EXT4_EXT_MAY_ZEROOUT 0x1 /* safe to zeroout if split fails \ + due to ENOSPC */ +#define EXT4_EXT_MARK_UNWRIT1 0x2 /* mark first half unwritten */ +#define EXT4_EXT_MARK_UNWRIT2 0x4 /* mark second half unwritten */ + +#define EXT4_EXT_DATA_VALID1 0x8 /* first half contains valid data */ +#define EXT4_EXT_DATA_VALID2 0x10 /* second half contains valid data */ + +static __le32 ext4_extent_block_csum(struct inode *inode, + struct ext4_extent_header *eh) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + __u32 csum; + + csum = ext4_chksum(ei->i_csum_seed, (__u8 *)eh, + EXT4_EXTENT_TAIL_OFFSET(eh)); + return cpu_to_le32(csum); +} + +static int ext4_extent_block_csum_verify(struct inode *inode, + struct ext4_extent_header *eh) +{ + struct ext4_extent_tail *et; + + if (!ext4_has_feature_metadata_csum(inode->i_sb)) + return 1; + + et = find_ext4_extent_tail(eh); + if (et->et_checksum != ext4_extent_block_csum(inode, eh)) + return 0; + return 1; +} + +static void ext4_extent_block_csum_set(struct inode *inode, + struct ext4_extent_header *eh) +{ + struct ext4_extent_tail *et; + + if (!ext4_has_feature_metadata_csum(inode->i_sb)) + return; + + et = find_ext4_extent_tail(eh); + et->et_checksum = ext4_extent_block_csum(inode, eh); +} + +static struct ext4_ext_path *ext4_split_extent_at(handle_t *handle, + struct inode *inode, + struct ext4_ext_path *path, + ext4_lblk_t split, + int split_flag, int flags); + +static int ext4_ext_trunc_restart_fn(struct inode *inode, int *dropped) +{ + /* + * Drop i_data_sem to avoid deadlock with ext4_map_blocks. At this + * moment, get_block can be called only for blocks inside i_size since + * page cache has been already dropped and writes are blocked by + * i_rwsem. So we can safely drop the i_data_sem here. + */ + BUG_ON(EXT4_JOURNAL(inode) == NULL); + ext4_discard_preallocations(inode); + up_write(&EXT4_I(inode)->i_data_sem); + *dropped = 1; + return 0; +} + +static inline void ext4_ext_path_brelse(struct ext4_ext_path *path) +{ + brelse(path->p_bh); + path->p_bh = NULL; +} + +static void ext4_ext_drop_refs(struct ext4_ext_path *path) +{ + int depth, i; + + if (IS_ERR_OR_NULL(path)) + return; + depth = path->p_depth; + for (i = 0; i <= depth; i++, path++) + ext4_ext_path_brelse(path); +} + +void ext4_free_ext_path(struct ext4_ext_path *path) +{ + if (IS_ERR_OR_NULL(path)) + return; + ext4_ext_drop_refs(path); + kfree(path); +} + +/* + * Make sure 'handle' has at least 'check_cred' credits. If not, restart + * transaction with 'restart_cred' credits. The function drops i_data_sem + * when restarting transaction and gets it after transaction is restarted. + * + * The function returns 0 on success, 1 if transaction had to be restarted, + * and < 0 in case of fatal error. + */ +int ext4_datasem_ensure_credits(handle_t *handle, struct inode *inode, + int check_cred, int restart_cred, + int revoke_cred) +{ + int ret; + int dropped = 0; + + ret = ext4_journal_ensure_credits_fn(handle, check_cred, restart_cred, + revoke_cred, ext4_ext_trunc_restart_fn(inode, &dropped)); + if (dropped) + down_write(&EXT4_I(inode)->i_data_sem); + return ret; +} + +/* + * could return: + * - EROFS + * - ENOMEM + */ +static int ext4_ext_get_access(handle_t *handle, struct inode *inode, + struct ext4_ext_path *path) +{ + int err = 0; + + if (path->p_bh) { + /* path points to block */ + BUFFER_TRACE(path->p_bh, "get_write_access"); + err = ext4_journal_get_write_access(handle, inode->i_sb, + path->p_bh, EXT4_JTR_NONE); + /* + * The extent buffer's verified bit will be set again in + * __ext4_ext_dirty(). We could leave an inconsistent + * buffer if the extents updating procudure break off du + * to some error happens, force to check it again. + */ + if (!err) + clear_buffer_verified(path->p_bh); + } + /* path points to leaf/index in inode body */ + /* we use in-core data, no need to protect them */ + return err; +} + +/* + * could return: + * - EROFS + * - ENOMEM + * - EIO + */ +static int __ext4_ext_dirty(const char *where, unsigned int line, + handle_t *handle, struct inode *inode, + struct ext4_ext_path *path) +{ + int err; + + WARN_ON(!rwsem_is_locked(&EXT4_I(inode)->i_data_sem)); + if (path->p_bh) { + ext4_extent_block_csum_set(inode, ext_block_hdr(path->p_bh)); + /* path points to block */ + err = __ext4_handle_dirty_metadata(where, line, handle, + inode, path->p_bh); + /* Extents updating done, re-set verified flag */ + if (!err) + set_buffer_verified(path->p_bh); + } else { + /* path points to leaf/index in inode body */ + err = ext4_mark_inode_dirty(handle, inode); + } + return err; +} + +#define ext4_ext_dirty(handle, inode, path) \ + __ext4_ext_dirty(__func__, __LINE__, (handle), (inode), (path)) + +static ext4_fsblk_t ext4_ext_find_goal(struct inode *inode, + struct ext4_ext_path *path, + ext4_lblk_t block) +{ + if (path) { + int depth = path->p_depth; + struct ext4_extent *ex; + + /* + * Try to predict block placement assuming that we are + * filling in a file which will eventually be + * non-sparse --- i.e., in the case of libbfd writing + * an ELF object sections out-of-order but in a way + * the eventually results in a contiguous object or + * executable file, or some database extending a table + * space file. However, this is actually somewhat + * non-ideal if we are writing a sparse file such as + * qemu or KVM writing a raw image file that is going + * to stay fairly sparse, since it will end up + * fragmenting the file system's free space. Maybe we + * should have some hueristics or some way to allow + * userspace to pass a hint to file system, + * especially if the latter case turns out to be + * common. + */ + ex = path[depth].p_ext; + if (ex) { + ext4_fsblk_t ext_pblk = ext4_ext_pblock(ex); + ext4_lblk_t ext_block = le32_to_cpu(ex->ee_block); + + if (block > ext_block) + return ext_pblk + (block - ext_block); + else + return ext_pblk - (ext_block - block); + } + + /* it looks like index is empty; + * try to find starting block from index itself */ + if (path[depth].p_bh) + return path[depth].p_bh->b_blocknr; + } + + /* OK. use inode's group */ + return ext4_inode_to_goal_block(inode); +} + +/* + * Allocation for a meta data block + */ +static ext4_fsblk_t +ext4_ext_new_meta_block(handle_t *handle, struct inode *inode, + struct ext4_ext_path *path, + struct ext4_extent *ex, int *err, unsigned int flags) +{ + ext4_fsblk_t goal, newblock; + + goal = ext4_ext_find_goal(inode, path, le32_to_cpu(ex->ee_block)); + newblock = ext4_new_meta_blocks(handle, inode, goal, flags, + NULL, err); + return newblock; +} + +static inline int ext4_ext_space_block(struct inode *inode, int check) +{ + int size; + + size = (inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header)) + / sizeof(struct ext4_extent); +#ifdef AGGRESSIVE_TEST + if (!check && size > 6) + size = 6; +#endif + return size; +} + +static inline int ext4_ext_space_block_idx(struct inode *inode, int check) +{ + int size; + + size = (inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header)) + / sizeof(struct ext4_extent_idx); +#ifdef AGGRESSIVE_TEST + if (!check && size > 5) + size = 5; +#endif + return size; +} + +static inline int ext4_ext_space_root(struct inode *inode, int check) +{ + int size; + + size = sizeof(EXT4_I(inode)->i_data); + size -= sizeof(struct ext4_extent_header); + size /= sizeof(struct ext4_extent); +#ifdef AGGRESSIVE_TEST + if (!check && size > 3) + size = 3; +#endif + return size; +} + +static inline int ext4_ext_space_root_idx(struct inode *inode, int check) +{ + int size; + + size = sizeof(EXT4_I(inode)->i_data); + size -= sizeof(struct ext4_extent_header); + size /= sizeof(struct ext4_extent_idx); +#ifdef AGGRESSIVE_TEST + if (!check && size > 4) + size = 4; +#endif + return size; +} + +static inline struct ext4_ext_path * +ext4_force_split_extent_at(handle_t *handle, struct inode *inode, + struct ext4_ext_path *path, ext4_lblk_t lblk, + int nofail) +{ + int unwritten = ext4_ext_is_unwritten(path[path->p_depth].p_ext); + int flags = EXT4_EX_NOCACHE | EXT4_GET_BLOCKS_PRE_IO; + + if (nofail) + flags |= EXT4_GET_BLOCKS_METADATA_NOFAIL | EXT4_EX_NOFAIL; + + return ext4_split_extent_at(handle, inode, path, lblk, unwritten ? + EXT4_EXT_MARK_UNWRIT1|EXT4_EXT_MARK_UNWRIT2 : 0, + flags); +} + +static int +ext4_ext_max_entries(struct inode *inode, int depth) +{ + int max; + + if (depth == ext_depth(inode)) { + if (depth == 0) + max = ext4_ext_space_root(inode, 1); + else + max = ext4_ext_space_root_idx(inode, 1); + } else { + if (depth == 0) + max = ext4_ext_space_block(inode, 1); + else + max = ext4_ext_space_block_idx(inode, 1); + } + + return max; +} + +static int ext4_valid_extent(struct inode *inode, struct ext4_extent *ext) +{ + ext4_fsblk_t block = ext4_ext_pblock(ext); + int len = ext4_ext_get_actual_len(ext); + ext4_lblk_t lblock = le32_to_cpu(ext->ee_block); + + /* + * We allow neither: + * - zero length + * - overflow/wrap-around + */ + if (lblock + len <= lblock) + return 0; + return ext4_inode_block_valid(inode, block, len); +} + +static int ext4_valid_extent_idx(struct inode *inode, + struct ext4_extent_idx *ext_idx) +{ + ext4_fsblk_t block = ext4_idx_pblock(ext_idx); + + return ext4_inode_block_valid(inode, block, 1); +} + +static int ext4_valid_extent_entries(struct inode *inode, + struct ext4_extent_header *eh, + ext4_lblk_t lblk, ext4_fsblk_t *pblk, + int depth) +{ + unsigned short entries; + ext4_lblk_t lblock = 0; + ext4_lblk_t cur = 0; + + if (eh->eh_entries == 0) + return 1; + + entries = le16_to_cpu(eh->eh_entries); + + if (depth == 0) { + /* leaf entries */ + struct ext4_extent *ext = EXT_FIRST_EXTENT(eh); + + /* + * The logical block in the first entry should equal to + * the number in the index block. + */ + if (depth != ext_depth(inode) && + lblk != le32_to_cpu(ext->ee_block)) + return 0; + while (entries) { + if (!ext4_valid_extent(inode, ext)) + return 0; + + /* Check for overlapping extents */ + lblock = le32_to_cpu(ext->ee_block); + if (lblock < cur) { + *pblk = ext4_ext_pblock(ext); + return 0; + } + cur = lblock + ext4_ext_get_actual_len(ext); + ext++; + entries--; + } + } else { + struct ext4_extent_idx *ext_idx = EXT_FIRST_INDEX(eh); + + /* + * The logical block in the first entry should equal to + * the number in the parent index block. + */ + if (depth != ext_depth(inode) && + lblk != le32_to_cpu(ext_idx->ei_block)) + return 0; + while (entries) { + if (!ext4_valid_extent_idx(inode, ext_idx)) + return 0; + + /* Check for overlapping index extents */ + lblock = le32_to_cpu(ext_idx->ei_block); + if (lblock < cur) { + *pblk = ext4_idx_pblock(ext_idx); + return 0; + } + ext_idx++; + entries--; + cur = lblock + 1; + } + } + return 1; +} + +static int __ext4_ext_check(const char *function, unsigned int line, + struct inode *inode, struct ext4_extent_header *eh, + int depth, ext4_fsblk_t pblk, ext4_lblk_t lblk) +{ + const char *error_msg; + int max = 0, err = -EFSCORRUPTED; + + if (unlikely(eh->eh_magic != EXT4_EXT_MAGIC)) { + error_msg = "invalid magic"; + goto corrupted; + } + if (unlikely(le16_to_cpu(eh->eh_depth) != depth)) { + error_msg = "unexpected eh_depth"; + goto corrupted; + } + if (unlikely(eh->eh_max == 0)) { + error_msg = "invalid eh_max"; + goto corrupted; + } + max = ext4_ext_max_entries(inode, depth); + if (unlikely(le16_to_cpu(eh->eh_max) > max)) { + error_msg = "too large eh_max"; + goto corrupted; + } + if (unlikely(le16_to_cpu(eh->eh_entries) > le16_to_cpu(eh->eh_max))) { + error_msg = "invalid eh_entries"; + goto corrupted; + } + if (unlikely((eh->eh_entries == 0) && (depth > 0))) { + error_msg = "eh_entries is 0 but eh_depth is > 0"; + goto corrupted; + } + if (!ext4_valid_extent_entries(inode, eh, lblk, &pblk, depth)) { + error_msg = "invalid extent entries"; + goto corrupted; + } + if (unlikely(depth > 32)) { + error_msg = "too large eh_depth"; + goto corrupted; + } + /* Verify checksum on non-root extent tree nodes */ + if (ext_depth(inode) != depth && + !ext4_extent_block_csum_verify(inode, eh)) { + error_msg = "extent tree corrupted"; + err = -EFSBADCRC; + goto corrupted; + } + return 0; + +corrupted: + ext4_error_inode_err(inode, function, line, 0, -err, + "pblk %llu bad header/extent: %s - magic %x, " + "entries %u, max %u(%u), depth %u(%u)", + (unsigned long long) pblk, error_msg, + le16_to_cpu(eh->eh_magic), + le16_to_cpu(eh->eh_entries), + le16_to_cpu(eh->eh_max), + max, le16_to_cpu(eh->eh_depth), depth); + return err; +} + +#define ext4_ext_check(inode, eh, depth, pblk) \ + __ext4_ext_check(__func__, __LINE__, (inode), (eh), (depth), (pblk), 0) + +int ext4_ext_check_inode(struct inode *inode) +{ + return ext4_ext_check(inode, ext_inode_hdr(inode), ext_depth(inode), 0); +} + +static void ext4_cache_extents(struct inode *inode, + struct ext4_extent_header *eh) +{ + struct ext4_extent *ex = EXT_FIRST_EXTENT(eh); + ext4_lblk_t prev = 0; + int i; + + for (i = le16_to_cpu(eh->eh_entries); i > 0; i--, ex++) { + unsigned int status = EXTENT_STATUS_WRITTEN; + ext4_lblk_t lblk = le32_to_cpu(ex->ee_block); + int len = ext4_ext_get_actual_len(ex); + + if (prev && (prev != lblk)) + ext4_es_cache_extent(inode, prev, lblk - prev, ~0, + EXTENT_STATUS_HOLE); + + if (ext4_ext_is_unwritten(ex)) + status = EXTENT_STATUS_UNWRITTEN; + ext4_es_cache_extent(inode, lblk, len, + ext4_ext_pblock(ex), status); + prev = lblk + len; + } +} + +static struct buffer_head * +__read_extent_tree_block(const char *function, unsigned int line, + struct inode *inode, struct ext4_extent_idx *idx, + int depth, int flags) +{ + struct buffer_head *bh; + int err; + gfp_t gfp_flags = __GFP_MOVABLE | GFP_NOFS; + ext4_fsblk_t pblk; + + if (flags & EXT4_EX_NOFAIL) + gfp_flags |= __GFP_NOFAIL; + + pblk = ext4_idx_pblock(idx); + bh = sb_getblk_gfp(inode->i_sb, pblk, gfp_flags); + if (unlikely(!bh)) + return ERR_PTR(-ENOMEM); + + if (!bh_uptodate_or_lock(bh)) { + trace_ext4_ext_load_extent(inode, pblk, _RET_IP_); + err = ext4_read_bh(bh, 0, NULL, false); + if (err < 0) + goto errout; + } + if (buffer_verified(bh) && !(flags & EXT4_EX_FORCE_CACHE)) + return bh; + err = __ext4_ext_check(function, line, inode, ext_block_hdr(bh), + depth, pblk, le32_to_cpu(idx->ei_block)); + if (err) + goto errout; + set_buffer_verified(bh); + /* + * If this is a leaf block, cache all of its entries + */ + if (!(flags & EXT4_EX_NOCACHE) && depth == 0) { + struct ext4_extent_header *eh = ext_block_hdr(bh); + ext4_cache_extents(inode, eh); + } + return bh; +errout: + put_bh(bh); + return ERR_PTR(err); + +} + +#define read_extent_tree_block(inode, idx, depth, flags) \ + __read_extent_tree_block(__func__, __LINE__, (inode), (idx), \ + (depth), (flags)) + +/* + * This function is called to cache a file's extent information in the + * extent status tree + */ +int ext4_ext_precache(struct inode *inode) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + struct ext4_ext_path *path = NULL; + struct buffer_head *bh; + int i = 0, depth, ret = 0; + + if (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) + return 0; /* not an extent-mapped inode */ + + ext4_check_map_extents_env(inode); + + down_read(&ei->i_data_sem); + depth = ext_depth(inode); + + /* Don't cache anything if there are no external extent blocks */ + if (!depth) { + up_read(&ei->i_data_sem); + return ret; + } + + path = kcalloc(depth + 1, sizeof(struct ext4_ext_path), + GFP_NOFS); + if (path == NULL) { + up_read(&ei->i_data_sem); + return -ENOMEM; + } + + path[0].p_hdr = ext_inode_hdr(inode); + ret = ext4_ext_check(inode, path[0].p_hdr, depth, 0); + if (ret) + goto out; + path[0].p_idx = EXT_FIRST_INDEX(path[0].p_hdr); + while (i >= 0) { + /* + * If this is a leaf block or we've reached the end of + * the index block, go up + */ + if ((i == depth) || + path[i].p_idx > EXT_LAST_INDEX(path[i].p_hdr)) { + ext4_ext_path_brelse(path + i); + i--; + continue; + } + bh = read_extent_tree_block(inode, path[i].p_idx++, + depth - i - 1, + EXT4_EX_FORCE_CACHE); + if (IS_ERR(bh)) { + ret = PTR_ERR(bh); + break; + } + i++; + path[i].p_bh = bh; + path[i].p_hdr = ext_block_hdr(bh); + path[i].p_idx = EXT_FIRST_INDEX(path[i].p_hdr); + } + ext4_set_inode_state(inode, EXT4_STATE_EXT_PRECACHED); +out: + up_read(&ei->i_data_sem); + ext4_free_ext_path(path); + return ret; +} + +#ifdef EXT_DEBUG +static void ext4_ext_show_path(struct inode *inode, struct ext4_ext_path *path) +{ + int k, l = path->p_depth; + + ext_debug(inode, "path:"); + for (k = 0; k <= l; k++, path++) { + if (path->p_idx) { + ext_debug(inode, " %d->%llu", + le32_to_cpu(path->p_idx->ei_block), + ext4_idx_pblock(path->p_idx)); + } else if (path->p_ext) { + ext_debug(inode, " %d:[%d]%d:%llu ", + le32_to_cpu(path->p_ext->ee_block), + ext4_ext_is_unwritten(path->p_ext), + ext4_ext_get_actual_len(path->p_ext), + ext4_ext_pblock(path->p_ext)); + } else + ext_debug(inode, " []"); + } + ext_debug(inode, "\n"); +} + +static void ext4_ext_show_leaf(struct inode *inode, struct ext4_ext_path *path) +{ + int depth = ext_depth(inode); + struct ext4_extent_header *eh; + struct ext4_extent *ex; + int i; + + if (IS_ERR_OR_NULL(path)) + return; + + eh = path[depth].p_hdr; + ex = EXT_FIRST_EXTENT(eh); + + ext_debug(inode, "Displaying leaf extents\n"); + + for (i = 0; i < le16_to_cpu(eh->eh_entries); i++, ex++) { + ext_debug(inode, "%d:[%d]%d:%llu ", le32_to_cpu(ex->ee_block), + ext4_ext_is_unwritten(ex), + ext4_ext_get_actual_len(ex), ext4_ext_pblock(ex)); + } + ext_debug(inode, "\n"); +} + +static void ext4_ext_show_move(struct inode *inode, struct ext4_ext_path *path, + ext4_fsblk_t newblock, int level) +{ + int depth = ext_depth(inode); + struct ext4_extent *ex; + + if (depth != level) { + struct ext4_extent_idx *idx; + idx = path[level].p_idx; + while (idx <= EXT_MAX_INDEX(path[level].p_hdr)) { + ext_debug(inode, "%d: move %d:%llu in new index %llu\n", + level, le32_to_cpu(idx->ei_block), + ext4_idx_pblock(idx), newblock); + idx++; + } + + return; + } + + ex = path[depth].p_ext; + while (ex <= EXT_MAX_EXTENT(path[depth].p_hdr)) { + ext_debug(inode, "move %d:%llu:[%d]%d in new leaf %llu\n", + le32_to_cpu(ex->ee_block), + ext4_ext_pblock(ex), + ext4_ext_is_unwritten(ex), + ext4_ext_get_actual_len(ex), + newblock); + ex++; + } +} + +#else +#define ext4_ext_show_path(inode, path) +#define ext4_ext_show_leaf(inode, path) +#define ext4_ext_show_move(inode, path, newblock, level) +#endif + +/* + * ext4_ext_binsearch_idx: + * binary search for the closest index of the given block + * the header must be checked before calling this + */ +static void +ext4_ext_binsearch_idx(struct inode *inode, + struct ext4_ext_path *path, ext4_lblk_t block) +{ + struct ext4_extent_header *eh = path->p_hdr; + struct ext4_extent_idx *r, *l, *m; + + + ext_debug(inode, "binsearch for %u(idx): ", block); + + l = EXT_FIRST_INDEX(eh) + 1; + r = EXT_LAST_INDEX(eh); + while (l <= r) { + m = l + (r - l) / 2; + ext_debug(inode, "%p(%u):%p(%u):%p(%u) ", l, + le32_to_cpu(l->ei_block), m, le32_to_cpu(m->ei_block), + r, le32_to_cpu(r->ei_block)); + + if (block < le32_to_cpu(m->ei_block)) + r = m - 1; + else + l = m + 1; + } + + path->p_idx = l - 1; + ext_debug(inode, " -> %u->%lld ", le32_to_cpu(path->p_idx->ei_block), + ext4_idx_pblock(path->p_idx)); + +#ifdef CHECK_BINSEARCH + { + struct ext4_extent_idx *chix, *ix; + int k; + + chix = ix = EXT_FIRST_INDEX(eh); + for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ix++) { + if (k != 0 && le32_to_cpu(ix->ei_block) <= + le32_to_cpu(ix[-1].ei_block)) { + printk(KERN_DEBUG "k=%d, ix=0x%p, " + "first=0x%p\n", k, + ix, EXT_FIRST_INDEX(eh)); + printk(KERN_DEBUG "%u <= %u\n", + le32_to_cpu(ix->ei_block), + le32_to_cpu(ix[-1].ei_block)); + } + BUG_ON(k && le32_to_cpu(ix->ei_block) + <= le32_to_cpu(ix[-1].ei_block)); + if (block < le32_to_cpu(ix->ei_block)) + break; + chix = ix; + } + BUG_ON(chix != path->p_idx); + } +#endif + +} + +/* + * ext4_ext_binsearch: + * binary search for closest extent of the given block + * the header must be checked before calling this + */ +static void +ext4_ext_binsearch(struct inode *inode, + struct ext4_ext_path *path, ext4_lblk_t block) +{ + struct ext4_extent_header *eh = path->p_hdr; + struct ext4_extent *r, *l, *m; + + if (eh->eh_entries == 0) { + /* + * this leaf is empty: + * we get such a leaf in split/add case + */ + return; + } + + ext_debug(inode, "binsearch for %u: ", block); + + l = EXT_FIRST_EXTENT(eh) + 1; + r = EXT_LAST_EXTENT(eh); + + while (l <= r) { + m = l + (r - l) / 2; + ext_debug(inode, "%p(%u):%p(%u):%p(%u) ", l, + le32_to_cpu(l->ee_block), m, le32_to_cpu(m->ee_block), + r, le32_to_cpu(r->ee_block)); + + if (block < le32_to_cpu(m->ee_block)) + r = m - 1; + else + l = m + 1; + } + + path->p_ext = l - 1; + ext_debug(inode, " -> %d:%llu:[%d]%d ", + le32_to_cpu(path->p_ext->ee_block), + ext4_ext_pblock(path->p_ext), + ext4_ext_is_unwritten(path->p_ext), + ext4_ext_get_actual_len(path->p_ext)); + +#ifdef CHECK_BINSEARCH + { + struct ext4_extent *chex, *ex; + int k; + + chex = ex = EXT_FIRST_EXTENT(eh); + for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ex++) { + BUG_ON(k && le32_to_cpu(ex->ee_block) + <= le32_to_cpu(ex[-1].ee_block)); + if (block < le32_to_cpu(ex->ee_block)) + break; + chex = ex; + } + BUG_ON(chex != path->p_ext); + } +#endif + +} + +void ext4_ext_tree_init(handle_t *handle, struct inode *inode) +{ + struct ext4_extent_header *eh; + + eh = ext_inode_hdr(inode); + eh->eh_depth = 0; + eh->eh_entries = 0; + eh->eh_magic = EXT4_EXT_MAGIC; + eh->eh_max = cpu_to_le16(ext4_ext_space_root(inode, 0)); + eh->eh_generation = 0; + ext4_mark_inode_dirty(handle, inode); +} + +struct ext4_ext_path * +ext4_find_extent(struct inode *inode, ext4_lblk_t block, + struct ext4_ext_path *path, int flags) +{ + struct ext4_extent_header *eh; + struct buffer_head *bh; + short int depth, i, ppos = 0; + int ret; + gfp_t gfp_flags = GFP_NOFS; + + if (flags & EXT4_EX_NOFAIL) + gfp_flags |= __GFP_NOFAIL; + + eh = ext_inode_hdr(inode); + depth = ext_depth(inode); + if (depth < 0 || depth > EXT4_MAX_EXTENT_DEPTH) { + EXT4_ERROR_INODE(inode, "inode has invalid extent depth: %d", + depth); + ret = -EFSCORRUPTED; + goto err; + } + + if (path) { + ext4_ext_drop_refs(path); + if (depth > path[0].p_maxdepth) { + kfree(path); + path = NULL; + } + } + if (!path) { + /* account possible depth increase */ + path = kcalloc(depth + 2, sizeof(struct ext4_ext_path), + gfp_flags); + if (unlikely(!path)) + return ERR_PTR(-ENOMEM); + path[0].p_maxdepth = depth + 1; + } + path[0].p_hdr = eh; + path[0].p_bh = NULL; + + i = depth; + if (!(flags & EXT4_EX_NOCACHE) && depth == 0) + ext4_cache_extents(inode, eh); + /* walk through the tree */ + while (i) { + ext_debug(inode, "depth %d: num %d, max %d\n", + ppos, le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max)); + + ext4_ext_binsearch_idx(inode, path + ppos, block); + path[ppos].p_block = ext4_idx_pblock(path[ppos].p_idx); + path[ppos].p_depth = i; + path[ppos].p_ext = NULL; + + bh = read_extent_tree_block(inode, path[ppos].p_idx, --i, flags); + if (IS_ERR(bh)) { + ret = PTR_ERR(bh); + goto err; + } + + eh = ext_block_hdr(bh); + ppos++; + path[ppos].p_bh = bh; + path[ppos].p_hdr = eh; + } + + path[ppos].p_depth = i; + path[ppos].p_ext = NULL; + path[ppos].p_idx = NULL; + + /* find extent */ + ext4_ext_binsearch(inode, path + ppos, block); + /* if not an empty leaf */ + if (path[ppos].p_ext) + path[ppos].p_block = ext4_ext_pblock(path[ppos].p_ext); + + ext4_ext_show_path(inode, path); + + return path; + +err: + ext4_free_ext_path(path); + return ERR_PTR(ret); +} + +/* + * ext4_ext_insert_index: + * insert new index [@logical;@ptr] into the block at @curp; + * check where to insert: before @curp or after @curp + */ +static int ext4_ext_insert_index(handle_t *handle, struct inode *inode, + struct ext4_ext_path *curp, + int logical, ext4_fsblk_t ptr) +{ + struct ext4_extent_idx *ix; + int len, err; + + err = ext4_ext_get_access(handle, inode, curp); + if (err) + return err; + + if (unlikely(logical == le32_to_cpu(curp->p_idx->ei_block))) { + EXT4_ERROR_INODE(inode, + "logical %d == ei_block %d!", + logical, le32_to_cpu(curp->p_idx->ei_block)); + return -EFSCORRUPTED; + } + + if (unlikely(le16_to_cpu(curp->p_hdr->eh_entries) + >= le16_to_cpu(curp->p_hdr->eh_max))) { + EXT4_ERROR_INODE(inode, + "eh_entries %d >= eh_max %d!", + le16_to_cpu(curp->p_hdr->eh_entries), + le16_to_cpu(curp->p_hdr->eh_max)); + return -EFSCORRUPTED; + } + + if (logical > le32_to_cpu(curp->p_idx->ei_block)) { + /* insert after */ + ext_debug(inode, "insert new index %d after: %llu\n", + logical, ptr); + ix = curp->p_idx + 1; + } else { + /* insert before */ + ext_debug(inode, "insert new index %d before: %llu\n", + logical, ptr); + ix = curp->p_idx; + } + + if (unlikely(ix > EXT_MAX_INDEX(curp->p_hdr))) { + EXT4_ERROR_INODE(inode, "ix > EXT_MAX_INDEX!"); + return -EFSCORRUPTED; + } + + len = EXT_LAST_INDEX(curp->p_hdr) - ix + 1; + BUG_ON(len < 0); + if (len > 0) { + ext_debug(inode, "insert new index %d: " + "move %d indices from 0x%p to 0x%p\n", + logical, len, ix, ix + 1); + memmove(ix + 1, ix, len * sizeof(struct ext4_extent_idx)); + } + + ix->ei_block = cpu_to_le32(logical); + ext4_idx_store_pblock(ix, ptr); + le16_add_cpu(&curp->p_hdr->eh_entries, 1); + + if (unlikely(ix > EXT_LAST_INDEX(curp->p_hdr))) { + EXT4_ERROR_INODE(inode, "ix > EXT_LAST_INDEX!"); + return -EFSCORRUPTED; + } + + err = ext4_ext_dirty(handle, inode, curp); + ext4_std_error(inode->i_sb, err); + + return err; +} + +/* + * ext4_ext_split: + * inserts new subtree into the path, using free index entry + * at depth @at: + * - allocates all needed blocks (new leaf and all intermediate index blocks) + * - makes decision where to split + * - moves remaining extents and index entries (right to the split point) + * into the newly allocated blocks + * - initializes subtree + */ +static int ext4_ext_split(handle_t *handle, struct inode *inode, + unsigned int flags, + struct ext4_ext_path *path, + struct ext4_extent *newext, int at) +{ + struct buffer_head *bh = NULL; + int depth = ext_depth(inode); + struct ext4_extent_header *neh; + struct ext4_extent_idx *fidx; + int i = at, k, m, a; + ext4_fsblk_t newblock, oldblock; + __le32 border; + ext4_fsblk_t *ablocks = NULL; /* array of allocated blocks */ + gfp_t gfp_flags = GFP_NOFS; + int err = 0; + size_t ext_size = 0; + + if (flags & EXT4_EX_NOFAIL) + gfp_flags |= __GFP_NOFAIL; + + /* make decision: where to split? */ + /* FIXME: now decision is simplest: at current extent */ + + /* if current leaf will be split, then we should use + * border from split point */ + if (unlikely(path[depth].p_ext > EXT_MAX_EXTENT(path[depth].p_hdr))) { + EXT4_ERROR_INODE(inode, "p_ext > EXT_MAX_EXTENT!"); + return -EFSCORRUPTED; + } + if (path[depth].p_ext != EXT_MAX_EXTENT(path[depth].p_hdr)) { + border = path[depth].p_ext[1].ee_block; + ext_debug(inode, "leaf will be split." + " next leaf starts at %d\n", + le32_to_cpu(border)); + } else { + border = newext->ee_block; + ext_debug(inode, "leaf will be added." + " next leaf starts at %d\n", + le32_to_cpu(border)); + } + + /* + * If error occurs, then we break processing + * and mark filesystem read-only. index won't + * be inserted and tree will be in consistent + * state. Next mount will repair buffers too. + */ + + /* + * Get array to track all allocated blocks. + * We need this to handle errors and free blocks + * upon them. + */ + ablocks = kcalloc(depth, sizeof(ext4_fsblk_t), gfp_flags); + if (!ablocks) + return -ENOMEM; + + /* allocate all needed blocks */ + ext_debug(inode, "allocate %d blocks for indexes/leaf\n", depth - at); + for (a = 0; a < depth - at; a++) { + newblock = ext4_ext_new_meta_block(handle, inode, path, + newext, &err, flags); + if (newblock == 0) + goto cleanup; + ablocks[a] = newblock; + } + + /* initialize new leaf */ + newblock = ablocks[--a]; + if (unlikely(newblock == 0)) { + EXT4_ERROR_INODE(inode, "newblock == 0!"); + err = -EFSCORRUPTED; + goto cleanup; + } + bh = sb_getblk_gfp(inode->i_sb, newblock, __GFP_MOVABLE | GFP_NOFS); + if (unlikely(!bh)) { + err = -ENOMEM; + goto cleanup; + } + lock_buffer(bh); + + err = ext4_journal_get_create_access(handle, inode->i_sb, bh, + EXT4_JTR_NONE); + if (err) + goto cleanup; + + neh = ext_block_hdr(bh); + neh->eh_entries = 0; + neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode, 0)); + neh->eh_magic = EXT4_EXT_MAGIC; + neh->eh_depth = 0; + neh->eh_generation = 0; + + /* move remainder of path[depth] to the new leaf */ + if (unlikely(path[depth].p_hdr->eh_entries != + path[depth].p_hdr->eh_max)) { + EXT4_ERROR_INODE(inode, "eh_entries %d != eh_max %d!", + path[depth].p_hdr->eh_entries, + path[depth].p_hdr->eh_max); + err = -EFSCORRUPTED; + goto cleanup; + } + /* start copy from next extent */ + m = EXT_MAX_EXTENT(path[depth].p_hdr) - path[depth].p_ext++; + ext4_ext_show_move(inode, path, newblock, depth); + if (m) { + struct ext4_extent *ex; + ex = EXT_FIRST_EXTENT(neh); + memmove(ex, path[depth].p_ext, sizeof(struct ext4_extent) * m); + le16_add_cpu(&neh->eh_entries, m); + } + + /* zero out unused area in the extent block */ + ext_size = sizeof(struct ext4_extent_header) + + sizeof(struct ext4_extent) * le16_to_cpu(neh->eh_entries); + memset(bh->b_data + ext_size, 0, inode->i_sb->s_blocksize - ext_size); + ext4_extent_block_csum_set(inode, neh); + set_buffer_uptodate(bh); + unlock_buffer(bh); + + err = ext4_handle_dirty_metadata(handle, inode, bh); + if (err) + goto cleanup; + brelse(bh); + bh = NULL; + + /* correct old leaf */ + if (m) { + err = ext4_ext_get_access(handle, inode, path + depth); + if (err) + goto cleanup; + le16_add_cpu(&path[depth].p_hdr->eh_entries, -m); + err = ext4_ext_dirty(handle, inode, path + depth); + if (err) + goto cleanup; + + } + + /* create intermediate indexes */ + k = depth - at - 1; + if (unlikely(k < 0)) { + EXT4_ERROR_INODE(inode, "k %d < 0!", k); + err = -EFSCORRUPTED; + goto cleanup; + } + if (k) + ext_debug(inode, "create %d intermediate indices\n", k); + /* insert new index into current index block */ + /* current depth stored in i var */ + i = depth - 1; + while (k--) { + oldblock = newblock; + newblock = ablocks[--a]; + bh = sb_getblk(inode->i_sb, newblock); + if (unlikely(!bh)) { + err = -ENOMEM; + goto cleanup; + } + lock_buffer(bh); + + err = ext4_journal_get_create_access(handle, inode->i_sb, bh, + EXT4_JTR_NONE); + if (err) + goto cleanup; + + neh = ext_block_hdr(bh); + neh->eh_entries = cpu_to_le16(1); + neh->eh_magic = EXT4_EXT_MAGIC; + neh->eh_max = cpu_to_le16(ext4_ext_space_block_idx(inode, 0)); + neh->eh_depth = cpu_to_le16(depth - i); + neh->eh_generation = 0; + fidx = EXT_FIRST_INDEX(neh); + fidx->ei_block = border; + ext4_idx_store_pblock(fidx, oldblock); + + ext_debug(inode, "int.index at %d (block %llu): %u -> %llu\n", + i, newblock, le32_to_cpu(border), oldblock); + + /* move remainder of path[i] to the new index block */ + if (unlikely(EXT_MAX_INDEX(path[i].p_hdr) != + EXT_LAST_INDEX(path[i].p_hdr))) { + EXT4_ERROR_INODE(inode, + "EXT_MAX_INDEX != EXT_LAST_INDEX ee_block %d!", + le32_to_cpu(path[i].p_ext->ee_block)); + err = -EFSCORRUPTED; + goto cleanup; + } + /* start copy indexes */ + m = EXT_MAX_INDEX(path[i].p_hdr) - path[i].p_idx++; + ext_debug(inode, "cur 0x%p, last 0x%p\n", path[i].p_idx, + EXT_MAX_INDEX(path[i].p_hdr)); + ext4_ext_show_move(inode, path, newblock, i); + if (m) { + memmove(++fidx, path[i].p_idx, + sizeof(struct ext4_extent_idx) * m); + le16_add_cpu(&neh->eh_entries, m); + } + /* zero out unused area in the extent block */ + ext_size = sizeof(struct ext4_extent_header) + + (sizeof(struct ext4_extent) * le16_to_cpu(neh->eh_entries)); + memset(bh->b_data + ext_size, 0, + inode->i_sb->s_blocksize - ext_size); + ext4_extent_block_csum_set(inode, neh); + set_buffer_uptodate(bh); + unlock_buffer(bh); + + err = ext4_handle_dirty_metadata(handle, inode, bh); + if (err) + goto cleanup; + brelse(bh); + bh = NULL; + + /* correct old index */ + if (m) { + err = ext4_ext_get_access(handle, inode, path + i); + if (err) + goto cleanup; + le16_add_cpu(&path[i].p_hdr->eh_entries, -m); + err = ext4_ext_dirty(handle, inode, path + i); + if (err) + goto cleanup; + } + + i--; + } + + /* insert new index */ + err = ext4_ext_insert_index(handle, inode, path + at, + le32_to_cpu(border), newblock); + +cleanup: + if (bh) { + if (buffer_locked(bh)) + unlock_buffer(bh); + brelse(bh); + } + + if (err) { + /* free all allocated blocks in error case */ + for (i = 0; i < depth; i++) { + if (!ablocks[i]) + continue; + ext4_free_blocks(handle, inode, NULL, ablocks[i], 1, + EXT4_FREE_BLOCKS_METADATA); + } + } + kfree(ablocks); + + return err; +} + +/* + * ext4_ext_grow_indepth: + * implements tree growing procedure: + * - allocates new block + * - moves top-level data (index block or leaf) into the new block + * - initializes new top-level, creating index that points to the + * just created block + */ +static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode, + unsigned int flags) +{ + struct ext4_extent_header *neh; + struct buffer_head *bh; + ext4_fsblk_t newblock, goal = 0; + struct ext4_super_block *es = EXT4_SB(inode->i_sb)->s_es; + int err = 0; + size_t ext_size = 0; + + /* Try to prepend new index to old one */ + if (ext_depth(inode)) + goal = ext4_idx_pblock(EXT_FIRST_INDEX(ext_inode_hdr(inode))); + if (goal > le32_to_cpu(es->s_first_data_block)) { + flags |= EXT4_MB_HINT_TRY_GOAL; + goal--; + } else + goal = ext4_inode_to_goal_block(inode); + newblock = ext4_new_meta_blocks(handle, inode, goal, flags, + NULL, &err); + if (newblock == 0) + return err; + + bh = sb_getblk_gfp(inode->i_sb, newblock, __GFP_MOVABLE | GFP_NOFS); + if (unlikely(!bh)) + return -ENOMEM; + lock_buffer(bh); + + err = ext4_journal_get_create_access(handle, inode->i_sb, bh, + EXT4_JTR_NONE); + if (err) { + unlock_buffer(bh); + goto out; + } + + ext_size = sizeof(EXT4_I(inode)->i_data); + /* move top-level index/leaf into new block */ + memmove(bh->b_data, EXT4_I(inode)->i_data, ext_size); + /* zero out unused area in the extent block */ + memset(bh->b_data + ext_size, 0, inode->i_sb->s_blocksize - ext_size); + + /* set size of new block */ + neh = ext_block_hdr(bh); + /* old root could have indexes or leaves + * so calculate e_max right way */ + if (ext_depth(inode)) + neh->eh_max = cpu_to_le16(ext4_ext_space_block_idx(inode, 0)); + else + neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode, 0)); + neh->eh_magic = EXT4_EXT_MAGIC; + ext4_extent_block_csum_set(inode, neh); + set_buffer_uptodate(bh); + set_buffer_verified(bh); + unlock_buffer(bh); + + err = ext4_handle_dirty_metadata(handle, inode, bh); + if (err) + goto out; + + /* Update top-level index: num,max,pointer */ + neh = ext_inode_hdr(inode); + neh->eh_entries = cpu_to_le16(1); + ext4_idx_store_pblock(EXT_FIRST_INDEX(neh), newblock); + if (neh->eh_depth == 0) { + /* Root extent block becomes index block */ + neh->eh_max = cpu_to_le16(ext4_ext_space_root_idx(inode, 0)); + EXT_FIRST_INDEX(neh)->ei_block = + EXT_FIRST_EXTENT(neh)->ee_block; + } + ext_debug(inode, "new root: num %d(%d), lblock %d, ptr %llu\n", + le16_to_cpu(neh->eh_entries), le16_to_cpu(neh->eh_max), + le32_to_cpu(EXT_FIRST_INDEX(neh)->ei_block), + ext4_idx_pblock(EXT_FIRST_INDEX(neh))); + + le16_add_cpu(&neh->eh_depth, 1); + err = ext4_mark_inode_dirty(handle, inode); +out: + brelse(bh); + + return err; +} + +/* + * ext4_ext_create_new_leaf: + * finds empty index and adds new leaf. + * if no free index is found, then it requests in-depth growing. + */ +static struct ext4_ext_path * +ext4_ext_create_new_leaf(handle_t *handle, struct inode *inode, + unsigned int mb_flags, unsigned int gb_flags, + struct ext4_ext_path *path, + struct ext4_extent *newext) +{ + struct ext4_ext_path *curp; + int depth, i, err = 0; + ext4_lblk_t ee_block = le32_to_cpu(newext->ee_block); + +repeat: + i = depth = ext_depth(inode); + + /* walk up to the tree and look for free index entry */ + curp = path + depth; + while (i > 0 && !EXT_HAS_FREE_INDEX(curp)) { + i--; + curp--; + } + + /* we use already allocated block for index block, + * so subsequent data blocks should be contiguous */ + if (EXT_HAS_FREE_INDEX(curp)) { + /* if we found index with free entry, then use that + * entry: create all needed subtree and add new leaf */ + err = ext4_ext_split(handle, inode, mb_flags, path, newext, i); + if (err) + goto errout; + + /* refill path */ + path = ext4_find_extent(inode, ee_block, path, gb_flags); + return path; + } + + /* tree is full, time to grow in depth */ + err = ext4_ext_grow_indepth(handle, inode, mb_flags); + if (err) + goto errout; + + /* refill path */ + path = ext4_find_extent(inode, ee_block, path, gb_flags); + if (IS_ERR(path)) + return path; + + /* + * only first (depth 0 -> 1) produces free space; + * in all other cases we have to split the grown tree + */ + depth = ext_depth(inode); + if (path[depth].p_hdr->eh_entries == path[depth].p_hdr->eh_max) { + /* now we need to split */ + goto repeat; + } + + return path; + +errout: + ext4_free_ext_path(path); + return ERR_PTR(err); +} + +/* + * search the closest allocated block to the left for *logical + * and returns it at @logical + it's physical address at @phys + * if *logical is the smallest allocated block, the function + * returns 0 at @phys + * return value contains 0 (success) or error code + */ +static int ext4_ext_search_left(struct inode *inode, + struct ext4_ext_path *path, + ext4_lblk_t *logical, ext4_fsblk_t *phys) +{ + struct ext4_extent_idx *ix; + struct ext4_extent *ex; + int depth, ee_len; + + if (unlikely(path == NULL)) { + EXT4_ERROR_INODE(inode, "path == NULL *logical %d!", *logical); + return -EFSCORRUPTED; + } + depth = path->p_depth; + *phys = 0; + + if (depth == 0 && path->p_ext == NULL) + return 0; + + /* usually extent in the path covers blocks smaller + * then *logical, but it can be that extent is the + * first one in the file */ + + ex = path[depth].p_ext; + ee_len = ext4_ext_get_actual_len(ex); + if (*logical < le32_to_cpu(ex->ee_block)) { + if (unlikely(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex)) { + EXT4_ERROR_INODE(inode, + "EXT_FIRST_EXTENT != ex *logical %d ee_block %d!", + *logical, le32_to_cpu(ex->ee_block)); + return -EFSCORRUPTED; + } + while (--depth >= 0) { + ix = path[depth].p_idx; + if (unlikely(ix != EXT_FIRST_INDEX(path[depth].p_hdr))) { + EXT4_ERROR_INODE(inode, + "ix (%d) != EXT_FIRST_INDEX (%d) (depth %d)!", + ix != NULL ? le32_to_cpu(ix->ei_block) : 0, + le32_to_cpu(EXT_FIRST_INDEX(path[depth].p_hdr)->ei_block), + depth); + return -EFSCORRUPTED; + } + } + return 0; + } + + if (unlikely(*logical < (le32_to_cpu(ex->ee_block) + ee_len))) { + EXT4_ERROR_INODE(inode, + "logical %d < ee_block %d + ee_len %d!", + *logical, le32_to_cpu(ex->ee_block), ee_len); + return -EFSCORRUPTED; + } + + *logical = le32_to_cpu(ex->ee_block) + ee_len - 1; + *phys = ext4_ext_pblock(ex) + ee_len - 1; + return 0; +} + +/* + * Search the closest allocated block to the right for *logical + * and returns it at @logical + it's physical address at @phys. + * If not exists, return 0 and @phys is set to 0. We will return + * 1 which means we found an allocated block and ret_ex is valid. + * Or return a (< 0) error code. + */ +static int ext4_ext_search_right(struct inode *inode, + struct ext4_ext_path *path, + ext4_lblk_t *logical, ext4_fsblk_t *phys, + struct ext4_extent *ret_ex, int flags) +{ + struct buffer_head *bh = NULL; + struct ext4_extent_header *eh; + struct ext4_extent_idx *ix; + struct ext4_extent *ex; + int depth; /* Note, NOT eh_depth; depth from top of tree */ + int ee_len; + + if (unlikely(path == NULL)) { + EXT4_ERROR_INODE(inode, "path == NULL *logical %d!", *logical); + return -EFSCORRUPTED; + } + depth = path->p_depth; + *phys = 0; + + if (depth == 0 && path->p_ext == NULL) + return 0; + + /* usually extent in the path covers blocks smaller + * then *logical, but it can be that extent is the + * first one in the file */ + + ex = path[depth].p_ext; + ee_len = ext4_ext_get_actual_len(ex); + if (*logical < le32_to_cpu(ex->ee_block)) { + if (unlikely(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex)) { + EXT4_ERROR_INODE(inode, + "first_extent(path[%d].p_hdr) != ex", + depth); + return -EFSCORRUPTED; + } + while (--depth >= 0) { + ix = path[depth].p_idx; + if (unlikely(ix != EXT_FIRST_INDEX(path[depth].p_hdr))) { + EXT4_ERROR_INODE(inode, + "ix != EXT_FIRST_INDEX *logical %d!", + *logical); + return -EFSCORRUPTED; + } + } + goto found_extent; + } + + if (unlikely(*logical < (le32_to_cpu(ex->ee_block) + ee_len))) { + EXT4_ERROR_INODE(inode, + "logical %d < ee_block %d + ee_len %d!", + *logical, le32_to_cpu(ex->ee_block), ee_len); + return -EFSCORRUPTED; + } + + if (ex != EXT_LAST_EXTENT(path[depth].p_hdr)) { + /* next allocated block in this leaf */ + ex++; + goto found_extent; + } + + /* go up and search for index to the right */ + while (--depth >= 0) { + ix = path[depth].p_idx; + if (ix != EXT_LAST_INDEX(path[depth].p_hdr)) + goto got_index; + } + + /* we've gone up to the root and found no index to the right */ + return 0; + +got_index: + /* we've found index to the right, let's + * follow it and find the closest allocated + * block to the right */ + ix++; + while (++depth < path->p_depth) { + /* subtract from p_depth to get proper eh_depth */ + bh = read_extent_tree_block(inode, ix, path->p_depth - depth, + flags); + if (IS_ERR(bh)) + return PTR_ERR(bh); + eh = ext_block_hdr(bh); + ix = EXT_FIRST_INDEX(eh); + put_bh(bh); + } + + bh = read_extent_tree_block(inode, ix, path->p_depth - depth, flags); + if (IS_ERR(bh)) + return PTR_ERR(bh); + eh = ext_block_hdr(bh); + ex = EXT_FIRST_EXTENT(eh); +found_extent: + *logical = le32_to_cpu(ex->ee_block); + *phys = ext4_ext_pblock(ex); + if (ret_ex) + *ret_ex = *ex; + if (bh) + put_bh(bh); + return 1; +} + +/* + * ext4_ext_next_allocated_block: + * returns allocated block in subsequent extent or EXT_MAX_BLOCKS. + * NOTE: it considers block number from index entry as + * allocated block. Thus, index entries have to be consistent + * with leaves. + */ +ext4_lblk_t +ext4_ext_next_allocated_block(struct ext4_ext_path *path) +{ + int depth; + + BUG_ON(path == NULL); + depth = path->p_depth; + + if (depth == 0 && path->p_ext == NULL) + return EXT_MAX_BLOCKS; + + while (depth >= 0) { + struct ext4_ext_path *p = &path[depth]; + + if (depth == path->p_depth) { + /* leaf */ + if (p->p_ext && p->p_ext != EXT_LAST_EXTENT(p->p_hdr)) + return le32_to_cpu(p->p_ext[1].ee_block); + } else { + /* index */ + if (p->p_idx != EXT_LAST_INDEX(p->p_hdr)) + return le32_to_cpu(p->p_idx[1].ei_block); + } + depth--; + } + + return EXT_MAX_BLOCKS; +} + +/* + * ext4_ext_next_leaf_block: + * returns first allocated block from next leaf or EXT_MAX_BLOCKS + */ +static ext4_lblk_t ext4_ext_next_leaf_block(struct ext4_ext_path *path) +{ + int depth; + + BUG_ON(path == NULL); + depth = path->p_depth; + + /* zero-tree has no leaf blocks at all */ + if (depth == 0) + return EXT_MAX_BLOCKS; + + /* go to index block */ + depth--; + + while (depth >= 0) { + if (path[depth].p_idx != + EXT_LAST_INDEX(path[depth].p_hdr)) + return (ext4_lblk_t) + le32_to_cpu(path[depth].p_idx[1].ei_block); + depth--; + } + + return EXT_MAX_BLOCKS; +} + +/* + * ext4_ext_correct_indexes: + * if leaf gets modified and modified extent is first in the leaf, + * then we have to correct all indexes above. + * TODO: do we need to correct tree in all cases? + */ +static int ext4_ext_correct_indexes(handle_t *handle, struct inode *inode, + struct ext4_ext_path *path) +{ + struct ext4_extent_header *eh; + int depth = ext_depth(inode); + struct ext4_extent *ex; + __le32 border; + int k, err = 0; + + eh = path[depth].p_hdr; + ex = path[depth].p_ext; + + if (unlikely(ex == NULL || eh == NULL)) { + EXT4_ERROR_INODE(inode, + "ex %p == NULL or eh %p == NULL", ex, eh); + return -EFSCORRUPTED; + } + + if (depth == 0) { + /* there is no tree at all */ + return 0; + } + + if (ex != EXT_FIRST_EXTENT(eh)) { + /* we correct tree if first leaf got modified only */ + return 0; + } + + /* + * TODO: we need correction if border is smaller than current one + */ + k = depth - 1; + border = path[depth].p_ext->ee_block; + err = ext4_ext_get_access(handle, inode, path + k); + if (err) + return err; + path[k].p_idx->ei_block = border; + err = ext4_ext_dirty(handle, inode, path + k); + if (err) + return err; + + while (k--) { + /* change all left-side indexes */ + if (path[k+1].p_idx != EXT_FIRST_INDEX(path[k+1].p_hdr)) + break; + err = ext4_ext_get_access(handle, inode, path + k); + if (err) + goto clean; + path[k].p_idx->ei_block = border; + err = ext4_ext_dirty(handle, inode, path + k); + if (err) + goto clean; + } + return 0; + +clean: + /* + * The path[k].p_bh is either unmodified or with no verified bit + * set (see ext4_ext_get_access()). So just clear the verified bit + * of the successfully modified extents buffers, which will force + * these extents to be checked to avoid using inconsistent data. + */ + while (++k < depth) + clear_buffer_verified(path[k].p_bh); + + return err; +} + +static int ext4_can_extents_be_merged(struct inode *inode, + struct ext4_extent *ex1, + struct ext4_extent *ex2) +{ + unsigned short ext1_ee_len, ext2_ee_len; + + if (ext4_ext_is_unwritten(ex1) != ext4_ext_is_unwritten(ex2)) + return 0; + + ext1_ee_len = ext4_ext_get_actual_len(ex1); + ext2_ee_len = ext4_ext_get_actual_len(ex2); + + if (le32_to_cpu(ex1->ee_block) + ext1_ee_len != + le32_to_cpu(ex2->ee_block)) + return 0; + + if (ext1_ee_len + ext2_ee_len > EXT_INIT_MAX_LEN) + return 0; + + if (ext4_ext_is_unwritten(ex1) && + ext1_ee_len + ext2_ee_len > EXT_UNWRITTEN_MAX_LEN) + return 0; +#ifdef AGGRESSIVE_TEST + if (ext1_ee_len >= 4) + return 0; +#endif + + if (ext4_ext_pblock(ex1) + ext1_ee_len == ext4_ext_pblock(ex2)) + return 1; + return 0; +} + +/* + * This function tries to merge the "ex" extent to the next extent in the tree. + * It always tries to merge towards right. If you want to merge towards + * left, pass "ex - 1" as argument instead of "ex". + * Returns 0 if the extents (ex and ex+1) were _not_ merged and returns + * 1 if they got merged. + */ +static int ext4_ext_try_to_merge_right(struct inode *inode, + struct ext4_ext_path *path, + struct ext4_extent *ex) +{ + struct ext4_extent_header *eh; + unsigned int depth, len; + int merge_done = 0, unwritten; + + depth = ext_depth(inode); + BUG_ON(path[depth].p_hdr == NULL); + eh = path[depth].p_hdr; + + while (ex < EXT_LAST_EXTENT(eh)) { + if (!ext4_can_extents_be_merged(inode, ex, ex + 1)) + break; + /* merge with next extent! */ + unwritten = ext4_ext_is_unwritten(ex); + ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex) + + ext4_ext_get_actual_len(ex + 1)); + if (unwritten) + ext4_ext_mark_unwritten(ex); + + if (ex + 1 < EXT_LAST_EXTENT(eh)) { + len = (EXT_LAST_EXTENT(eh) - ex - 1) + * sizeof(struct ext4_extent); + memmove(ex + 1, ex + 2, len); + } + le16_add_cpu(&eh->eh_entries, -1); + merge_done = 1; + WARN_ON(eh->eh_entries == 0); + if (!eh->eh_entries) + EXT4_ERROR_INODE(inode, "eh->eh_entries = 0!"); + } + + return merge_done; +} + +/* + * This function does a very simple check to see if we can collapse + * an extent tree with a single extent tree leaf block into the inode. + */ +static void ext4_ext_try_to_merge_up(handle_t *handle, + struct inode *inode, + struct ext4_ext_path *path) +{ + size_t s; + unsigned max_root = ext4_ext_space_root(inode, 0); + ext4_fsblk_t blk; + + if ((path[0].p_depth != 1) || + (le16_to_cpu(path[0].p_hdr->eh_entries) != 1) || + (le16_to_cpu(path[1].p_hdr->eh_entries) > max_root)) + return; + + /* + * We need to modify the block allocation bitmap and the block + * group descriptor to release the extent tree block. If we + * can't get the journal credits, give up. + */ + if (ext4_journal_extend(handle, 2, + ext4_free_metadata_revoke_credits(inode->i_sb, 1))) + return; + + /* + * Copy the extent data up to the inode + */ + blk = ext4_idx_pblock(path[0].p_idx); + s = le16_to_cpu(path[1].p_hdr->eh_entries) * + sizeof(struct ext4_extent_idx); + s += sizeof(struct ext4_extent_header); + + path[1].p_maxdepth = path[0].p_maxdepth; + memcpy(path[0].p_hdr, path[1].p_hdr, s); + path[0].p_depth = 0; + path[0].p_ext = EXT_FIRST_EXTENT(path[0].p_hdr) + + (path[1].p_ext - EXT_FIRST_EXTENT(path[1].p_hdr)); + path[0].p_hdr->eh_max = cpu_to_le16(max_root); + + ext4_ext_path_brelse(path + 1); + ext4_free_blocks(handle, inode, NULL, blk, 1, + EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET); +} + +/* + * This function tries to merge the @ex extent to neighbours in the tree, then + * tries to collapse the extent tree into the inode. + */ +static void ext4_ext_try_to_merge(handle_t *handle, + struct inode *inode, + struct ext4_ext_path *path, + struct ext4_extent *ex) +{ + struct ext4_extent_header *eh; + unsigned int depth; + int merge_done = 0; + + depth = ext_depth(inode); + BUG_ON(path[depth].p_hdr == NULL); + eh = path[depth].p_hdr; + + if (ex > EXT_FIRST_EXTENT(eh)) + merge_done = ext4_ext_try_to_merge_right(inode, path, ex - 1); + + if (!merge_done) + (void) ext4_ext_try_to_merge_right(inode, path, ex); + + ext4_ext_try_to_merge_up(handle, inode, path); +} + +/* + * check if a portion of the "newext" extent overlaps with an + * existing extent. + * + * If there is an overlap discovered, it updates the length of the newext + * such that there will be no overlap, and then returns 1. + * If there is no overlap found, it returns 0. + */ +static unsigned int ext4_ext_check_overlap(struct ext4_sb_info *sbi, + struct inode *inode, + struct ext4_extent *newext, + struct ext4_ext_path *path) +{ + ext4_lblk_t b1, b2; + unsigned int depth, len1; + unsigned int ret = 0; + + b1 = le32_to_cpu(newext->ee_block); + len1 = ext4_ext_get_actual_len(newext); + depth = ext_depth(inode); + if (!path[depth].p_ext) + goto out; + b2 = EXT4_LBLK_CMASK(sbi, le32_to_cpu(path[depth].p_ext->ee_block)); + + /* + * get the next allocated block if the extent in the path + * is before the requested block(s) + */ + if (b2 < b1) { + b2 = ext4_ext_next_allocated_block(path); + if (b2 == EXT_MAX_BLOCKS) + goto out; + b2 = EXT4_LBLK_CMASK(sbi, b2); + } + + /* check for wrap through zero on extent logical start block*/ + if (b1 + len1 < b1) { + len1 = EXT_MAX_BLOCKS - b1; + newext->ee_len = cpu_to_le16(len1); + ret = 1; + } + + /* check for overlap */ + if (b1 + len1 > b2) { + newext->ee_len = cpu_to_le16(b2 - b1); + ret = 1; + } +out: + return ret; +} + +/* + * ext4_ext_insert_extent: + * tries to merge requested extent into the existing extent or + * inserts requested extent as new one into the tree, + * creating new leaf in the no-space case. + */ +struct ext4_ext_path * +ext4_ext_insert_extent(handle_t *handle, struct inode *inode, + struct ext4_ext_path *path, + struct ext4_extent *newext, int gb_flags) +{ + struct ext4_extent_header *eh; + struct ext4_extent *ex, *fex; + struct ext4_extent *nearex; /* nearest extent */ + int depth, len, err = 0; + ext4_lblk_t next; + int mb_flags = 0, unwritten; + + if (gb_flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) + mb_flags |= EXT4_MB_DELALLOC_RESERVED; + if (unlikely(ext4_ext_get_actual_len(newext) == 0)) { + EXT4_ERROR_INODE(inode, "ext4_ext_get_actual_len(newext) == 0"); + err = -EFSCORRUPTED; + goto errout; + } + depth = ext_depth(inode); + ex = path[depth].p_ext; + eh = path[depth].p_hdr; + if (unlikely(path[depth].p_hdr == NULL)) { + EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth); + err = -EFSCORRUPTED; + goto errout; + } + + /* try to insert block into found extent and return */ + if (ex && !(gb_flags & EXT4_GET_BLOCKS_PRE_IO)) { + + /* + * Try to see whether we should rather test the extent on + * right from ex, or from the left of ex. This is because + * ext4_find_extent() can return either extent on the + * left, or on the right from the searched position. This + * will make merging more effective. + */ + if (ex < EXT_LAST_EXTENT(eh) && + (le32_to_cpu(ex->ee_block) + + ext4_ext_get_actual_len(ex) < + le32_to_cpu(newext->ee_block))) { + ex += 1; + goto prepend; + } else if ((ex > EXT_FIRST_EXTENT(eh)) && + (le32_to_cpu(newext->ee_block) + + ext4_ext_get_actual_len(newext) < + le32_to_cpu(ex->ee_block))) + ex -= 1; + + /* Try to append newex to the ex */ + if (ext4_can_extents_be_merged(inode, ex, newext)) { + ext_debug(inode, "append [%d]%d block to %u:[%d]%d" + "(from %llu)\n", + ext4_ext_is_unwritten(newext), + ext4_ext_get_actual_len(newext), + le32_to_cpu(ex->ee_block), + ext4_ext_is_unwritten(ex), + ext4_ext_get_actual_len(ex), + ext4_ext_pblock(ex)); + err = ext4_ext_get_access(handle, inode, + path + depth); + if (err) + goto errout; + unwritten = ext4_ext_is_unwritten(ex); + ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex) + + ext4_ext_get_actual_len(newext)); + if (unwritten) + ext4_ext_mark_unwritten(ex); + nearex = ex; + goto merge; + } + +prepend: + /* Try to prepend newex to the ex */ + if (ext4_can_extents_be_merged(inode, newext, ex)) { + ext_debug(inode, "prepend %u[%d]%d block to %u:[%d]%d" + "(from %llu)\n", + le32_to_cpu(newext->ee_block), + ext4_ext_is_unwritten(newext), + ext4_ext_get_actual_len(newext), + le32_to_cpu(ex->ee_block), + ext4_ext_is_unwritten(ex), + ext4_ext_get_actual_len(ex), + ext4_ext_pblock(ex)); + err = ext4_ext_get_access(handle, inode, + path + depth); + if (err) + goto errout; + + unwritten = ext4_ext_is_unwritten(ex); + ex->ee_block = newext->ee_block; + ext4_ext_store_pblock(ex, ext4_ext_pblock(newext)); + ex->ee_len = cpu_to_le16(ext4_ext_get_actual_len(ex) + + ext4_ext_get_actual_len(newext)); + if (unwritten) + ext4_ext_mark_unwritten(ex); + nearex = ex; + goto merge; + } + } + + depth = ext_depth(inode); + eh = path[depth].p_hdr; + if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max)) + goto has_space; + + /* probably next leaf has space for us? */ + fex = EXT_LAST_EXTENT(eh); + next = EXT_MAX_BLOCKS; + if (le32_to_cpu(newext->ee_block) > le32_to_cpu(fex->ee_block)) + next = ext4_ext_next_leaf_block(path); + if (next != EXT_MAX_BLOCKS) { + struct ext4_ext_path *npath; + + ext_debug(inode, "next leaf block - %u\n", next); + npath = ext4_find_extent(inode, next, NULL, gb_flags); + if (IS_ERR(npath)) { + err = PTR_ERR(npath); + goto errout; + } + BUG_ON(npath->p_depth != path->p_depth); + eh = npath[depth].p_hdr; + if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max)) { + ext_debug(inode, "next leaf isn't full(%d)\n", + le16_to_cpu(eh->eh_entries)); + ext4_free_ext_path(path); + path = npath; + goto has_space; + } + ext_debug(inode, "next leaf has no free space(%d,%d)\n", + le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max)); + ext4_free_ext_path(npath); + } + + /* + * There is no free space in the found leaf. + * We're gonna add a new leaf in the tree. + */ + if (gb_flags & EXT4_GET_BLOCKS_METADATA_NOFAIL) + mb_flags |= EXT4_MB_USE_RESERVED; + path = ext4_ext_create_new_leaf(handle, inode, mb_flags, gb_flags, + path, newext); + if (IS_ERR(path)) + return path; + depth = ext_depth(inode); + eh = path[depth].p_hdr; + +has_space: + nearex = path[depth].p_ext; + + err = ext4_ext_get_access(handle, inode, path + depth); + if (err) + goto errout; + + if (!nearex) { + /* there is no extent in this leaf, create first one */ + ext_debug(inode, "first extent in the leaf: %u:%llu:[%d]%d\n", + le32_to_cpu(newext->ee_block), + ext4_ext_pblock(newext), + ext4_ext_is_unwritten(newext), + ext4_ext_get_actual_len(newext)); + nearex = EXT_FIRST_EXTENT(eh); + } else { + if (le32_to_cpu(newext->ee_block) + > le32_to_cpu(nearex->ee_block)) { + /* Insert after */ + ext_debug(inode, "insert %u:%llu:[%d]%d before: " + "nearest %p\n", + le32_to_cpu(newext->ee_block), + ext4_ext_pblock(newext), + ext4_ext_is_unwritten(newext), + ext4_ext_get_actual_len(newext), + nearex); + nearex++; + } else { + /* Insert before */ + BUG_ON(newext->ee_block == nearex->ee_block); + ext_debug(inode, "insert %u:%llu:[%d]%d after: " + "nearest %p\n", + le32_to_cpu(newext->ee_block), + ext4_ext_pblock(newext), + ext4_ext_is_unwritten(newext), + ext4_ext_get_actual_len(newext), + nearex); + } + len = EXT_LAST_EXTENT(eh) - nearex + 1; + if (len > 0) { + ext_debug(inode, "insert %u:%llu:[%d]%d: " + "move %d extents from 0x%p to 0x%p\n", + le32_to_cpu(newext->ee_block), + ext4_ext_pblock(newext), + ext4_ext_is_unwritten(newext), + ext4_ext_get_actual_len(newext), + len, nearex, nearex + 1); + memmove(nearex + 1, nearex, + len * sizeof(struct ext4_extent)); + } + } + + le16_add_cpu(&eh->eh_entries, 1); + path[depth].p_ext = nearex; + nearex->ee_block = newext->ee_block; + ext4_ext_store_pblock(nearex, ext4_ext_pblock(newext)); + nearex->ee_len = newext->ee_len; + +merge: + /* try to merge extents */ + if (!(gb_flags & EXT4_GET_BLOCKS_PRE_IO)) + ext4_ext_try_to_merge(handle, inode, path, nearex); + + /* time to correct all indexes above */ + err = ext4_ext_correct_indexes(handle, inode, path); + if (err) + goto errout; + + err = ext4_ext_dirty(handle, inode, path + path->p_depth); + if (err) + goto errout; + + return path; + +errout: + ext4_free_ext_path(path); + return ERR_PTR(err); +} + +static int ext4_fill_es_cache_info(struct inode *inode, + ext4_lblk_t block, ext4_lblk_t num, + struct fiemap_extent_info *fieinfo) +{ + ext4_lblk_t next, end = block + num - 1; + struct extent_status es; + unsigned char blksize_bits = inode->i_sb->s_blocksize_bits; + unsigned int flags; + int err; + + while (block <= end) { + next = 0; + flags = 0; + if (!ext4_es_lookup_extent(inode, block, &next, &es)) + break; + if (ext4_es_is_unwritten(&es)) + flags |= FIEMAP_EXTENT_UNWRITTEN; + if (ext4_es_is_delayed(&es)) + flags |= (FIEMAP_EXTENT_DELALLOC | + FIEMAP_EXTENT_UNKNOWN); + if (ext4_es_is_hole(&es)) + flags |= EXT4_FIEMAP_EXTENT_HOLE; + if (next == 0) + flags |= FIEMAP_EXTENT_LAST; + if (flags & (FIEMAP_EXTENT_DELALLOC| + EXT4_FIEMAP_EXTENT_HOLE)) + es.es_pblk = 0; + else + es.es_pblk = ext4_es_pblock(&es); + err = fiemap_fill_next_extent(fieinfo, + (__u64)es.es_lblk << blksize_bits, + (__u64)es.es_pblk << blksize_bits, + (__u64)es.es_len << blksize_bits, + flags); + if (next == 0) + break; + block = next; + if (err < 0) + return err; + if (err == 1) + return 0; + } + return 0; +} + + +/* + * ext4_ext_find_hole - find hole around given block according to the given path + * @inode: inode we lookup in + * @path: path in extent tree to @lblk + * @lblk: pointer to logical block around which we want to determine hole + * + * Determine hole length (and start if easily possible) around given logical + * block. We don't try too hard to find the beginning of the hole but @path + * actually points to extent before @lblk, we provide it. + * + * The function returns the length of a hole starting at @lblk. We update @lblk + * to the beginning of the hole if we managed to find it. + */ +static ext4_lblk_t ext4_ext_find_hole(struct inode *inode, + struct ext4_ext_path *path, + ext4_lblk_t *lblk) +{ + int depth = ext_depth(inode); + struct ext4_extent *ex; + ext4_lblk_t len; + + ex = path[depth].p_ext; + if (ex == NULL) { + /* there is no extent yet, so gap is [0;-] */ + *lblk = 0; + len = EXT_MAX_BLOCKS; + } else if (*lblk < le32_to_cpu(ex->ee_block)) { + len = le32_to_cpu(ex->ee_block) - *lblk; + } else if (*lblk >= le32_to_cpu(ex->ee_block) + + ext4_ext_get_actual_len(ex)) { + ext4_lblk_t next; + + *lblk = le32_to_cpu(ex->ee_block) + ext4_ext_get_actual_len(ex); + next = ext4_ext_next_allocated_block(path); + BUG_ON(next == *lblk); + len = next - *lblk; + } else { + BUG(); + } + return len; +} + +/* + * ext4_ext_rm_idx: + * removes index from the index block. + */ +static int ext4_ext_rm_idx(handle_t *handle, struct inode *inode, + struct ext4_ext_path *path, int depth) +{ + int err; + ext4_fsblk_t leaf; + int k = depth - 1; + + /* free index block */ + leaf = ext4_idx_pblock(path[k].p_idx); + if (unlikely(path[k].p_hdr->eh_entries == 0)) { + EXT4_ERROR_INODE(inode, "path[%d].p_hdr->eh_entries == 0", k); + return -EFSCORRUPTED; + } + err = ext4_ext_get_access(handle, inode, path + k); + if (err) + return err; + + if (path[k].p_idx != EXT_LAST_INDEX(path[k].p_hdr)) { + int len = EXT_LAST_INDEX(path[k].p_hdr) - path[k].p_idx; + len *= sizeof(struct ext4_extent_idx); + memmove(path[k].p_idx, path[k].p_idx + 1, len); + } + + le16_add_cpu(&path[k].p_hdr->eh_entries, -1); + err = ext4_ext_dirty(handle, inode, path + k); + if (err) + return err; + ext_debug(inode, "index is empty, remove it, free block %llu\n", leaf); + trace_ext4_ext_rm_idx(inode, leaf); + + ext4_free_blocks(handle, inode, NULL, leaf, 1, + EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET); + + while (--k >= 0) { + if (path[k + 1].p_idx != EXT_FIRST_INDEX(path[k + 1].p_hdr)) + break; + err = ext4_ext_get_access(handle, inode, path + k); + if (err) + goto clean; + path[k].p_idx->ei_block = path[k + 1].p_idx->ei_block; + err = ext4_ext_dirty(handle, inode, path + k); + if (err) + goto clean; + } + return 0; + +clean: + /* + * The path[k].p_bh is either unmodified or with no verified bit + * set (see ext4_ext_get_access()). So just clear the verified bit + * of the successfully modified extents buffers, which will force + * these extents to be checked to avoid using inconsistent data. + */ + while (++k < depth) + clear_buffer_verified(path[k].p_bh); + + return err; +} + +/* + * ext4_ext_calc_credits_for_single_extent: + * This routine returns max. credits that needed to insert an extent + * to the extent tree. + * When pass the actual path, the caller should calculate credits + * under i_data_sem. + */ +int ext4_ext_calc_credits_for_single_extent(struct inode *inode, int nrblocks, + struct ext4_ext_path *path) +{ + if (path) { + int depth = ext_depth(inode); + int ret = 0; + + /* probably there is space in leaf? */ + if (le16_to_cpu(path[depth].p_hdr->eh_entries) + < le16_to_cpu(path[depth].p_hdr->eh_max)) { + + /* + * There are some space in the leaf tree, no + * need to account for leaf block credit + * + * bitmaps and block group descriptor blocks + * and other metadata blocks still need to be + * accounted. + */ + /* 1 bitmap, 1 block group descriptor */ + ret = 2 + EXT4_META_TRANS_BLOCKS(inode->i_sb); + return ret; + } + } + + return ext4_chunk_trans_blocks(inode, nrblocks); +} + +/* + * How many index/leaf blocks need to change/allocate to add @extents extents? + * + * If we add a single extent, then in the worse case, each tree level + * index/leaf need to be changed in case of the tree split. + * + * If more extents are inserted, they could cause the whole tree split more + * than once, but this is really rare. + */ +int ext4_ext_index_trans_blocks(struct inode *inode, int extents) +{ + int index; + + /* If we are converting the inline data, only one is needed here. */ + if (ext4_has_inline_data(inode)) + return 1; + + /* + * Extent tree can change between the time we estimate credits and + * the time we actually modify the tree. Assume the worst case. + */ + if (extents <= 1) + index = (EXT4_MAX_EXTENT_DEPTH * 2) + extents; + else + index = (EXT4_MAX_EXTENT_DEPTH * 3) + + DIV_ROUND_UP(extents, ext4_ext_space_block(inode, 0)); + + return index; +} + +static inline int get_default_free_blocks_flags(struct inode *inode) +{ + if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode) || + ext4_test_inode_flag(inode, EXT4_INODE_EA_INODE)) + return EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET; + else if (ext4_should_journal_data(inode)) + return EXT4_FREE_BLOCKS_FORGET; + return 0; +} + +/* + * ext4_rereserve_cluster - increment the reserved cluster count when + * freeing a cluster with a pending reservation + * + * @inode - file containing the cluster + * @lblk - logical block in cluster to be reserved + * + * Increments the reserved cluster count and adjusts quota in a bigalloc + * file system when freeing a partial cluster containing at least one + * delayed and unwritten block. A partial cluster meeting that + * requirement will have a pending reservation. If so, the + * RERESERVE_CLUSTER flag is used when calling ext4_free_blocks() to + * defer reserved and allocated space accounting to a subsequent call + * to this function. + */ +static void ext4_rereserve_cluster(struct inode *inode, ext4_lblk_t lblk) +{ + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + struct ext4_inode_info *ei = EXT4_I(inode); + + dquot_reclaim_block(inode, EXT4_C2B(sbi, 1)); + + spin_lock(&ei->i_block_reservation_lock); + ei->i_reserved_data_blocks++; + percpu_counter_add(&sbi->s_dirtyclusters_counter, 1); + spin_unlock(&ei->i_block_reservation_lock); + + percpu_counter_add(&sbi->s_freeclusters_counter, 1); + ext4_remove_pending(inode, lblk); +} + +static int ext4_remove_blocks(handle_t *handle, struct inode *inode, + struct ext4_extent *ex, + struct partial_cluster *partial, + ext4_lblk_t from, ext4_lblk_t to) +{ + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + unsigned short ee_len = ext4_ext_get_actual_len(ex); + ext4_fsblk_t last_pblk, pblk; + ext4_lblk_t num; + int flags; + + /* only extent tail removal is allowed */ + if (from < le32_to_cpu(ex->ee_block) || + to != le32_to_cpu(ex->ee_block) + ee_len - 1) { + ext4_error(sbi->s_sb, + "strange request: removal(2) %u-%u from %u:%u", + from, to, le32_to_cpu(ex->ee_block), ee_len); + return 0; + } + +#ifdef EXTENTS_STATS + spin_lock(&sbi->s_ext_stats_lock); + sbi->s_ext_blocks += ee_len; + sbi->s_ext_extents++; + if (ee_len < sbi->s_ext_min) + sbi->s_ext_min = ee_len; + if (ee_len > sbi->s_ext_max) + sbi->s_ext_max = ee_len; + if (ext_depth(inode) > sbi->s_depth_max) + sbi->s_depth_max = ext_depth(inode); + spin_unlock(&sbi->s_ext_stats_lock); +#endif + + trace_ext4_remove_blocks(inode, ex, from, to, partial); + + /* + * if we have a partial cluster, and it's different from the + * cluster of the last block in the extent, we free it + */ + last_pblk = ext4_ext_pblock(ex) + ee_len - 1; + + if (partial->state != initial && + partial->pclu != EXT4_B2C(sbi, last_pblk)) { + if (partial->state == tofree) { + flags = get_default_free_blocks_flags(inode); + if (ext4_is_pending(inode, partial->lblk)) + flags |= EXT4_FREE_BLOCKS_RERESERVE_CLUSTER; + ext4_free_blocks(handle, inode, NULL, + EXT4_C2B(sbi, partial->pclu), + sbi->s_cluster_ratio, flags); + if (flags & EXT4_FREE_BLOCKS_RERESERVE_CLUSTER) + ext4_rereserve_cluster(inode, partial->lblk); + } + partial->state = initial; + } + + num = le32_to_cpu(ex->ee_block) + ee_len - from; + pblk = ext4_ext_pblock(ex) + ee_len - num; + + /* + * We free the partial cluster at the end of the extent (if any), + * unless the cluster is used by another extent (partial_cluster + * state is nofree). If a partial cluster exists here, it must be + * shared with the last block in the extent. + */ + flags = get_default_free_blocks_flags(inode); + + /* partial, left end cluster aligned, right end unaligned */ + if ((EXT4_LBLK_COFF(sbi, to) != sbi->s_cluster_ratio - 1) && + (EXT4_LBLK_CMASK(sbi, to) >= from) && + (partial->state != nofree)) { + if (ext4_is_pending(inode, to)) + flags |= EXT4_FREE_BLOCKS_RERESERVE_CLUSTER; + ext4_free_blocks(handle, inode, NULL, + EXT4_PBLK_CMASK(sbi, last_pblk), + sbi->s_cluster_ratio, flags); + if (flags & EXT4_FREE_BLOCKS_RERESERVE_CLUSTER) + ext4_rereserve_cluster(inode, to); + partial->state = initial; + flags = get_default_free_blocks_flags(inode); + } + + flags |= EXT4_FREE_BLOCKS_NOFREE_LAST_CLUSTER; + + /* + * For bigalloc file systems, we never free a partial cluster + * at the beginning of the extent. Instead, we check to see if we + * need to free it on a subsequent call to ext4_remove_blocks, + * or at the end of ext4_ext_rm_leaf or ext4_ext_remove_space. + */ + flags |= EXT4_FREE_BLOCKS_NOFREE_FIRST_CLUSTER; + ext4_free_blocks(handle, inode, NULL, pblk, num, flags); + + /* reset the partial cluster if we've freed past it */ + if (partial->state != initial && partial->pclu != EXT4_B2C(sbi, pblk)) + partial->state = initial; + + /* + * If we've freed the entire extent but the beginning is not left + * cluster aligned and is not marked as ineligible for freeing we + * record the partial cluster at the beginning of the extent. It + * wasn't freed by the preceding ext4_free_blocks() call, and we + * need to look farther to the left to determine if it's to be freed + * (not shared with another extent). Else, reset the partial + * cluster - we're either done freeing or the beginning of the + * extent is left cluster aligned. + */ + if (EXT4_LBLK_COFF(sbi, from) && num == ee_len) { + if (partial->state == initial) { + partial->pclu = EXT4_B2C(sbi, pblk); + partial->lblk = from; + partial->state = tofree; + } + } else { + partial->state = initial; + } + + return 0; +} + +/* + * ext4_ext_rm_leaf() Removes the extents associated with the + * blocks appearing between "start" and "end". Both "start" + * and "end" must appear in the same extent or EIO is returned. + * + * @handle: The journal handle + * @inode: The files inode + * @path: The path to the leaf + * @partial_cluster: The cluster which we'll have to free if all extents + * has been released from it. However, if this value is + * negative, it's a cluster just to the right of the + * punched region and it must not be freed. + * @start: The first block to remove + * @end: The last block to remove + */ +static int +ext4_ext_rm_leaf(handle_t *handle, struct inode *inode, + struct ext4_ext_path *path, + struct partial_cluster *partial, + ext4_lblk_t start, ext4_lblk_t end) +{ + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + int err = 0, correct_index = 0; + int depth = ext_depth(inode), credits, revoke_credits; + struct ext4_extent_header *eh; + ext4_lblk_t a, b; + unsigned num; + ext4_lblk_t ex_ee_block; + unsigned short ex_ee_len; + unsigned unwritten = 0; + struct ext4_extent *ex; + ext4_fsblk_t pblk; + + /* the header must be checked already in ext4_ext_remove_space() */ + ext_debug(inode, "truncate since %u in leaf to %u\n", start, end); + if (!path[depth].p_hdr) + path[depth].p_hdr = ext_block_hdr(path[depth].p_bh); + eh = path[depth].p_hdr; + if (unlikely(path[depth].p_hdr == NULL)) { + EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth); + return -EFSCORRUPTED; + } + /* find where to start removing */ + ex = path[depth].p_ext; + if (!ex) + ex = EXT_LAST_EXTENT(eh); + + ex_ee_block = le32_to_cpu(ex->ee_block); + ex_ee_len = ext4_ext_get_actual_len(ex); + + trace_ext4_ext_rm_leaf(inode, start, ex, partial); + + while (ex >= EXT_FIRST_EXTENT(eh) && + ex_ee_block + ex_ee_len > start) { + + if (ext4_ext_is_unwritten(ex)) + unwritten = 1; + else + unwritten = 0; + + ext_debug(inode, "remove ext %u:[%d]%d\n", ex_ee_block, + unwritten, ex_ee_len); + path[depth].p_ext = ex; + + a = max(ex_ee_block, start); + b = min(ex_ee_block + ex_ee_len - 1, end); + + ext_debug(inode, " border %u:%u\n", a, b); + + /* If this extent is beyond the end of the hole, skip it */ + if (end < ex_ee_block) { + /* + * We're going to skip this extent and move to another, + * so note that its first cluster is in use to avoid + * freeing it when removing blocks. Eventually, the + * right edge of the truncated/punched region will + * be just to the left. + */ + if (sbi->s_cluster_ratio > 1) { + pblk = ext4_ext_pblock(ex); + partial->pclu = EXT4_B2C(sbi, pblk); + partial->state = nofree; + } + ex--; + ex_ee_block = le32_to_cpu(ex->ee_block); + ex_ee_len = ext4_ext_get_actual_len(ex); + continue; + } else if (b != ex_ee_block + ex_ee_len - 1) { + EXT4_ERROR_INODE(inode, + "can not handle truncate %u:%u " + "on extent %u:%u", + start, end, ex_ee_block, + ex_ee_block + ex_ee_len - 1); + err = -EFSCORRUPTED; + goto out; + } else if (a != ex_ee_block) { + /* remove tail of the extent */ + num = a - ex_ee_block; + } else { + /* remove whole extent: excellent! */ + num = 0; + } + /* + * 3 for leaf, sb, and inode plus 2 (bmap and group + * descriptor) for each block group; assume two block + * groups plus ex_ee_len/blocks_per_block_group for + * the worst case + */ + credits = 7 + 2*(ex_ee_len/EXT4_BLOCKS_PER_GROUP(inode->i_sb)); + if (ex == EXT_FIRST_EXTENT(eh)) { + correct_index = 1; + credits += (ext_depth(inode)) + 1; + } + credits += EXT4_MAXQUOTAS_TRANS_BLOCKS(inode->i_sb); + /* + * We may end up freeing some index blocks and data from the + * punched range. Note that partial clusters are accounted for + * by ext4_free_data_revoke_credits(). + */ + revoke_credits = + ext4_free_metadata_revoke_credits(inode->i_sb, + ext_depth(inode)) + + ext4_free_data_revoke_credits(inode, b - a + 1); + + err = ext4_datasem_ensure_credits(handle, inode, credits, + credits, revoke_credits); + if (err) { + if (err > 0) + err = -EAGAIN; + goto out; + } + + err = ext4_ext_get_access(handle, inode, path + depth); + if (err) + goto out; + + err = ext4_remove_blocks(handle, inode, ex, partial, a, b); + if (err) + goto out; + + if (num == 0) + /* this extent is removed; mark slot entirely unused */ + ext4_ext_store_pblock(ex, 0); + + ex->ee_len = cpu_to_le16(num); + /* + * Do not mark unwritten if all the blocks in the + * extent have been removed. + */ + if (unwritten && num) + ext4_ext_mark_unwritten(ex); + /* + * If the extent was completely released, + * we need to remove it from the leaf + */ + if (num == 0) { + if (end != EXT_MAX_BLOCKS - 1) { + /* + * For hole punching, we need to scoot all the + * extents up when an extent is removed so that + * we dont have blank extents in the middle + */ + memmove(ex, ex+1, (EXT_LAST_EXTENT(eh) - ex) * + sizeof(struct ext4_extent)); + + /* Now get rid of the one at the end */ + memset(EXT_LAST_EXTENT(eh), 0, + sizeof(struct ext4_extent)); + } + le16_add_cpu(&eh->eh_entries, -1); + } + + err = ext4_ext_dirty(handle, inode, path + depth); + if (err) + goto out; + + ext_debug(inode, "new extent: %u:%u:%llu\n", ex_ee_block, num, + ext4_ext_pblock(ex)); + ex--; + ex_ee_block = le32_to_cpu(ex->ee_block); + ex_ee_len = ext4_ext_get_actual_len(ex); + } + + if (correct_index && eh->eh_entries) + err = ext4_ext_correct_indexes(handle, inode, path); + + /* + * If there's a partial cluster and at least one extent remains in + * the leaf, free the partial cluster if it isn't shared with the + * current extent. If it is shared with the current extent + * we reset the partial cluster because we've reached the start of the + * truncated/punched region and we're done removing blocks. + */ + if (partial->state == tofree && ex >= EXT_FIRST_EXTENT(eh)) { + pblk = ext4_ext_pblock(ex) + ex_ee_len - 1; + if (partial->pclu != EXT4_B2C(sbi, pblk)) { + int flags = get_default_free_blocks_flags(inode); + + if (ext4_is_pending(inode, partial->lblk)) + flags |= EXT4_FREE_BLOCKS_RERESERVE_CLUSTER; + ext4_free_blocks(handle, inode, NULL, + EXT4_C2B(sbi, partial->pclu), + sbi->s_cluster_ratio, flags); + if (flags & EXT4_FREE_BLOCKS_RERESERVE_CLUSTER) + ext4_rereserve_cluster(inode, partial->lblk); + } + partial->state = initial; + } + + /* if this leaf is free, then we should + * remove it from index block above */ + if (err == 0 && eh->eh_entries == 0 && path[depth].p_bh != NULL) + err = ext4_ext_rm_idx(handle, inode, path, depth); + +out: + return err; +} + +/* + * ext4_ext_more_to_rm: + * returns 1 if current index has to be freed (even partial) + */ +static int +ext4_ext_more_to_rm(struct ext4_ext_path *path) +{ + BUG_ON(path->p_idx == NULL); + + if (path->p_idx < EXT_FIRST_INDEX(path->p_hdr)) + return 0; + + /* + * if truncate on deeper level happened, it wasn't partial, + * so we have to consider current index for truncation + */ + if (le16_to_cpu(path->p_hdr->eh_entries) == path->p_block) + return 0; + return 1; +} + +int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start, + ext4_lblk_t end) +{ + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + int depth = ext_depth(inode); + struct ext4_ext_path *path = NULL; + struct partial_cluster partial; + handle_t *handle; + int i = 0, err = 0; + int flags = EXT4_EX_NOCACHE | EXT4_EX_NOFAIL; + + partial.pclu = 0; + partial.lblk = 0; + partial.state = initial; + + ext_debug(inode, "truncate since %u to %u\n", start, end); + + /* probably first extent we're gonna free will be last in block */ + handle = ext4_journal_start_with_revoke(inode, EXT4_HT_TRUNCATE, + depth + 1, + ext4_free_metadata_revoke_credits(inode->i_sb, depth)); + if (IS_ERR(handle)) + return PTR_ERR(handle); + +again: + trace_ext4_ext_remove_space(inode, start, end, depth); + + /* + * Check if we are removing extents inside the extent tree. If that + * is the case, we are going to punch a hole inside the extent tree + * so we have to check whether we need to split the extent covering + * the last block to remove so we can easily remove the part of it + * in ext4_ext_rm_leaf(). + */ + if (end < EXT_MAX_BLOCKS - 1) { + struct ext4_extent *ex; + ext4_lblk_t ee_block, ex_end, lblk; + ext4_fsblk_t pblk; + + /* find extent for or closest extent to this block */ + path = ext4_find_extent(inode, end, NULL, flags); + if (IS_ERR(path)) { + ext4_journal_stop(handle); + return PTR_ERR(path); + } + depth = ext_depth(inode); + /* Leaf not may not exist only if inode has no blocks at all */ + ex = path[depth].p_ext; + if (!ex) { + if (depth) { + EXT4_ERROR_INODE(inode, + "path[%d].p_hdr == NULL", + depth); + err = -EFSCORRUPTED; + } + goto out; + } + + ee_block = le32_to_cpu(ex->ee_block); + ex_end = ee_block + ext4_ext_get_actual_len(ex) - 1; + + /* + * See if the last block is inside the extent, if so split + * the extent at 'end' block so we can easily remove the + * tail of the first part of the split extent in + * ext4_ext_rm_leaf(). + */ + if (end >= ee_block && end < ex_end) { + + /* + * If we're going to split the extent, note that + * the cluster containing the block after 'end' is + * in use to avoid freeing it when removing blocks. + */ + if (sbi->s_cluster_ratio > 1) { + pblk = ext4_ext_pblock(ex) + end - ee_block + 1; + partial.pclu = EXT4_B2C(sbi, pblk); + partial.state = nofree; + } + + /* + * Split the extent in two so that 'end' is the last + * block in the first new extent. Also we should not + * fail removing space due to ENOSPC so try to use + * reserved block if that happens. + */ + path = ext4_force_split_extent_at(handle, inode, path, + end + 1, 1); + if (IS_ERR(path)) { + err = PTR_ERR(path); + goto out; + } + } else if (sbi->s_cluster_ratio > 1 && end >= ex_end && + partial.state == initial) { + /* + * If we're punching, there's an extent to the right. + * If the partial cluster hasn't been set, set it to + * that extent's first cluster and its state to nofree + * so it won't be freed should it contain blocks to be + * removed. If it's already set (tofree/nofree), we're + * retrying and keep the original partial cluster info + * so a cluster marked tofree as a result of earlier + * extent removal is not lost. + */ + lblk = ex_end + 1; + err = ext4_ext_search_right(inode, path, &lblk, &pblk, + NULL, flags); + if (err < 0) + goto out; + if (pblk) { + partial.pclu = EXT4_B2C(sbi, pblk); + partial.state = nofree; + } + } + } + /* + * We start scanning from right side, freeing all the blocks + * after i_size and walking into the tree depth-wise. + */ + depth = ext_depth(inode); + if (path) { + int k = i = depth; + while (--k > 0) + path[k].p_block = + le16_to_cpu(path[k].p_hdr->eh_entries)+1; + } else { + path = kcalloc(depth + 1, sizeof(struct ext4_ext_path), + GFP_NOFS | __GFP_NOFAIL); + if (path == NULL) { + ext4_journal_stop(handle); + return -ENOMEM; + } + path[0].p_maxdepth = path[0].p_depth = depth; + path[0].p_hdr = ext_inode_hdr(inode); + i = 0; + + if (ext4_ext_check(inode, path[0].p_hdr, depth, 0)) { + err = -EFSCORRUPTED; + goto out; + } + } + err = 0; + + while (i >= 0 && err == 0) { + if (i == depth) { + /* this is leaf block */ + err = ext4_ext_rm_leaf(handle, inode, path, + &partial, start, end); + /* root level has p_bh == NULL, brelse() eats this */ + ext4_ext_path_brelse(path + i); + i--; + continue; + } + + /* this is index block */ + if (!path[i].p_hdr) { + ext_debug(inode, "initialize header\n"); + path[i].p_hdr = ext_block_hdr(path[i].p_bh); + } + + if (!path[i].p_idx) { + /* this level hasn't been touched yet */ + path[i].p_idx = EXT_LAST_INDEX(path[i].p_hdr); + path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries)+1; + ext_debug(inode, "init index ptr: hdr 0x%p, num %d\n", + path[i].p_hdr, + le16_to_cpu(path[i].p_hdr->eh_entries)); + } else { + /* we were already here, see at next index */ + path[i].p_idx--; + } + + ext_debug(inode, "level %d - index, first 0x%p, cur 0x%p\n", + i, EXT_FIRST_INDEX(path[i].p_hdr), + path[i].p_idx); + if (ext4_ext_more_to_rm(path + i)) { + struct buffer_head *bh; + /* go to the next level */ + ext_debug(inode, "move to level %d (block %llu)\n", + i + 1, ext4_idx_pblock(path[i].p_idx)); + memset(path + i + 1, 0, sizeof(*path)); + bh = read_extent_tree_block(inode, path[i].p_idx, + depth - i - 1, flags); + if (IS_ERR(bh)) { + /* should we reset i_size? */ + err = PTR_ERR(bh); + break; + } + /* Yield here to deal with large extent trees. + * Should be a no-op if we did IO above. */ + cond_resched(); + if (WARN_ON(i + 1 > depth)) { + err = -EFSCORRUPTED; + break; + } + path[i + 1].p_bh = bh; + + /* save actual number of indexes since this + * number is changed at the next iteration */ + path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries); + i++; + } else { + /* we finished processing this index, go up */ + if (path[i].p_hdr->eh_entries == 0 && i > 0) { + /* index is empty, remove it; + * handle must be already prepared by the + * truncatei_leaf() */ + err = ext4_ext_rm_idx(handle, inode, path, i); + } + /* root level has p_bh == NULL, brelse() eats this */ + ext4_ext_path_brelse(path + i); + i--; + ext_debug(inode, "return to level %d\n", i); + } + } + + trace_ext4_ext_remove_space_done(inode, start, end, depth, &partial, + path->p_hdr->eh_entries); + + /* + * if there's a partial cluster and we have removed the first extent + * in the file, then we also free the partial cluster, if any + */ + if (partial.state == tofree && err == 0) { + int flags = get_default_free_blocks_flags(inode); + + if (ext4_is_pending(inode, partial.lblk)) + flags |= EXT4_FREE_BLOCKS_RERESERVE_CLUSTER; + ext4_free_blocks(handle, inode, NULL, + EXT4_C2B(sbi, partial.pclu), + sbi->s_cluster_ratio, flags); + if (flags & EXT4_FREE_BLOCKS_RERESERVE_CLUSTER) + ext4_rereserve_cluster(inode, partial.lblk); + partial.state = initial; + } + + /* TODO: flexible tree reduction should be here */ + if (path->p_hdr->eh_entries == 0) { + /* + * truncate to zero freed all the tree, + * so we need to correct eh_depth + */ + err = ext4_ext_get_access(handle, inode, path); + if (err == 0) { + ext_inode_hdr(inode)->eh_depth = 0; + ext_inode_hdr(inode)->eh_max = + cpu_to_le16(ext4_ext_space_root(inode, 0)); + err = ext4_ext_dirty(handle, inode, path); + } + } +out: + ext4_free_ext_path(path); + path = NULL; + if (err == -EAGAIN) + goto again; + ext4_journal_stop(handle); + + return err; +} + +/* + * called at mount time + */ +void ext4_ext_init(struct super_block *sb) +{ + /* + * possible initialization would be here + */ + + if (ext4_has_feature_extents(sb)) { +#if defined(AGGRESSIVE_TEST) || defined(CHECK_BINSEARCH) || defined(EXTENTS_STATS) + printk(KERN_INFO "EXT4-fs: file extents enabled" +#ifdef AGGRESSIVE_TEST + ", aggressive tests" +#endif +#ifdef CHECK_BINSEARCH + ", check binsearch" +#endif +#ifdef EXTENTS_STATS + ", stats" +#endif + "\n"); +#endif +#ifdef EXTENTS_STATS + spin_lock_init(&EXT4_SB(sb)->s_ext_stats_lock); + EXT4_SB(sb)->s_ext_min = 1 << 30; + EXT4_SB(sb)->s_ext_max = 0; +#endif + } +} + +/* + * called at umount time + */ +void ext4_ext_release(struct super_block *sb) +{ + if (!ext4_has_feature_extents(sb)) + return; + +#ifdef EXTENTS_STATS + if (EXT4_SB(sb)->s_ext_blocks && EXT4_SB(sb)->s_ext_extents) { + struct ext4_sb_info *sbi = EXT4_SB(sb); + printk(KERN_ERR "EXT4-fs: %lu blocks in %lu extents (%lu ave)\n", + sbi->s_ext_blocks, sbi->s_ext_extents, + sbi->s_ext_blocks / sbi->s_ext_extents); + printk(KERN_ERR "EXT4-fs: extents: %lu min, %lu max, max depth %lu\n", + sbi->s_ext_min, sbi->s_ext_max, sbi->s_depth_max); + } +#endif +} + +static void ext4_zeroout_es(struct inode *inode, struct ext4_extent *ex) +{ + ext4_lblk_t ee_block; + ext4_fsblk_t ee_pblock; + unsigned int ee_len; + + ee_block = le32_to_cpu(ex->ee_block); + ee_len = ext4_ext_get_actual_len(ex); + ee_pblock = ext4_ext_pblock(ex); + + if (ee_len == 0) + return; + + ext4_es_insert_extent(inode, ee_block, ee_len, ee_pblock, + EXTENT_STATUS_WRITTEN, false); +} + +/* FIXME!! we need to try to merge to left or right after zero-out */ +static int ext4_ext_zeroout(struct inode *inode, struct ext4_extent *ex) +{ + ext4_fsblk_t ee_pblock; + unsigned int ee_len; + + ee_len = ext4_ext_get_actual_len(ex); + ee_pblock = ext4_ext_pblock(ex); + return ext4_issue_zeroout(inode, le32_to_cpu(ex->ee_block), ee_pblock, + ee_len); +} + +/* + * ext4_split_extent_at() splits an extent at given block. + * + * @handle: the journal handle + * @inode: the file inode + * @path: the path to the extent + * @split: the logical block where the extent is splitted. + * @split_flags: indicates if the extent could be zeroout if split fails, and + * the states(init or unwritten) of new extents. + * @flags: flags used to insert new extent to extent tree. + * + * + * Splits extent [a, b] into two extents [a, @split) and [@split, b], states + * of which are determined by split_flag. + * + * There are two cases: + * a> the extent are splitted into two extent. + * b> split is not needed, and just mark the extent. + * + * Return an extent path pointer on success, or an error pointer on failure. + */ +static struct ext4_ext_path *ext4_split_extent_at(handle_t *handle, + struct inode *inode, + struct ext4_ext_path *path, + ext4_lblk_t split, + int split_flag, int flags) +{ + ext4_fsblk_t newblock; + ext4_lblk_t ee_block; + struct ext4_extent *ex, newex, orig_ex, zero_ex; + struct ext4_extent *ex2 = NULL; + unsigned int ee_len, depth; + int err = 0; + + BUG_ON((split_flag & (EXT4_EXT_DATA_VALID1 | EXT4_EXT_DATA_VALID2)) == + (EXT4_EXT_DATA_VALID1 | EXT4_EXT_DATA_VALID2)); + + ext_debug(inode, "logical block %llu\n", (unsigned long long)split); + + ext4_ext_show_leaf(inode, path); + + depth = ext_depth(inode); + ex = path[depth].p_ext; + ee_block = le32_to_cpu(ex->ee_block); + ee_len = ext4_ext_get_actual_len(ex); + newblock = split - ee_block + ext4_ext_pblock(ex); + + BUG_ON(split < ee_block || split >= (ee_block + ee_len)); + BUG_ON(!ext4_ext_is_unwritten(ex) && + split_flag & (EXT4_EXT_MAY_ZEROOUT | + EXT4_EXT_MARK_UNWRIT1 | + EXT4_EXT_MARK_UNWRIT2)); + + err = ext4_ext_get_access(handle, inode, path + depth); + if (err) + goto out; + + if (split == ee_block) { + /* + * case b: block @split is the block that the extent begins with + * then we just change the state of the extent, and splitting + * is not needed. + */ + if (split_flag & EXT4_EXT_MARK_UNWRIT2) + ext4_ext_mark_unwritten(ex); + else + ext4_ext_mark_initialized(ex); + + if (!(flags & EXT4_GET_BLOCKS_PRE_IO)) + ext4_ext_try_to_merge(handle, inode, path, ex); + + err = ext4_ext_dirty(handle, inode, path + path->p_depth); + goto out; + } + + /* case a */ + memcpy(&orig_ex, ex, sizeof(orig_ex)); + ex->ee_len = cpu_to_le16(split - ee_block); + if (split_flag & EXT4_EXT_MARK_UNWRIT1) + ext4_ext_mark_unwritten(ex); + + /* + * path may lead to new leaf, not to original leaf any more + * after ext4_ext_insert_extent() returns, + */ + err = ext4_ext_dirty(handle, inode, path + depth); + if (err) + goto fix_extent_len; + + ex2 = &newex; + ex2->ee_block = cpu_to_le32(split); + ex2->ee_len = cpu_to_le16(ee_len - (split - ee_block)); + ext4_ext_store_pblock(ex2, newblock); + if (split_flag & EXT4_EXT_MARK_UNWRIT2) + ext4_ext_mark_unwritten(ex2); + + path = ext4_ext_insert_extent(handle, inode, path, &newex, flags); + if (!IS_ERR(path)) + goto out; + + err = PTR_ERR(path); + if (err != -ENOSPC && err != -EDQUOT && err != -ENOMEM) + return path; + + /* + * Get a new path to try to zeroout or fix the extent length. + * Using EXT4_EX_NOFAIL guarantees that ext4_find_extent() + * will not return -ENOMEM, otherwise -ENOMEM will cause a + * retry in do_writepages(), and a WARN_ON may be triggered + * in ext4_da_update_reserve_space() due to an incorrect + * ee_len causing the i_reserved_data_blocks exception. + */ + path = ext4_find_extent(inode, ee_block, NULL, flags | EXT4_EX_NOFAIL); + if (IS_ERR(path)) { + EXT4_ERROR_INODE(inode, "Failed split extent on %u, err %ld", + split, PTR_ERR(path)); + return path; + } + depth = ext_depth(inode); + ex = path[depth].p_ext; + + if (EXT4_EXT_MAY_ZEROOUT & split_flag) { + if (split_flag & (EXT4_EXT_DATA_VALID1|EXT4_EXT_DATA_VALID2)) { + if (split_flag & EXT4_EXT_DATA_VALID1) { + err = ext4_ext_zeroout(inode, ex2); + zero_ex.ee_block = ex2->ee_block; + zero_ex.ee_len = cpu_to_le16( + ext4_ext_get_actual_len(ex2)); + ext4_ext_store_pblock(&zero_ex, + ext4_ext_pblock(ex2)); + } else { + err = ext4_ext_zeroout(inode, ex); + zero_ex.ee_block = ex->ee_block; + zero_ex.ee_len = cpu_to_le16( + ext4_ext_get_actual_len(ex)); + ext4_ext_store_pblock(&zero_ex, + ext4_ext_pblock(ex)); + } + } else { + err = ext4_ext_zeroout(inode, &orig_ex); + zero_ex.ee_block = orig_ex.ee_block; + zero_ex.ee_len = cpu_to_le16( + ext4_ext_get_actual_len(&orig_ex)); + ext4_ext_store_pblock(&zero_ex, + ext4_ext_pblock(&orig_ex)); + } + + if (!err) { + /* update the extent length and mark as initialized */ + ex->ee_len = cpu_to_le16(ee_len); + ext4_ext_try_to_merge(handle, inode, path, ex); + err = ext4_ext_dirty(handle, inode, path + path->p_depth); + if (!err) + /* update extent status tree */ + ext4_zeroout_es(inode, &zero_ex); + /* If we failed at this point, we don't know in which + * state the extent tree exactly is so don't try to fix + * length of the original extent as it may do even more + * damage. + */ + goto out; + } + } + +fix_extent_len: + ex->ee_len = orig_ex.ee_len; + /* + * Ignore ext4_ext_dirty return value since we are already in error path + * and err is a non-zero error code. + */ + ext4_ext_dirty(handle, inode, path + path->p_depth); +out: + if (err) { + ext4_free_ext_path(path); + path = ERR_PTR(err); + } + ext4_ext_show_leaf(inode, path); + return path; +} + +/* + * ext4_split_extent() splits an extent and mark extent which is covered + * by @map as split_flags indicates + * + * It may result in splitting the extent into multiple extents (up to three) + * There are three possibilities: + * a> There is no split required + * b> Splits in two extents: Split is happening at either end of the extent + * c> Splits in three extents: Somone is splitting in middle of the extent + * + */ +static struct ext4_ext_path *ext4_split_extent(handle_t *handle, + struct inode *inode, + struct ext4_ext_path *path, + struct ext4_map_blocks *map, + int split_flag, int flags, + unsigned int *allocated) +{ + ext4_lblk_t ee_block; + struct ext4_extent *ex; + unsigned int ee_len, depth; + int unwritten; + int split_flag1, flags1; + + depth = ext_depth(inode); + ex = path[depth].p_ext; + ee_block = le32_to_cpu(ex->ee_block); + ee_len = ext4_ext_get_actual_len(ex); + unwritten = ext4_ext_is_unwritten(ex); + + if (map->m_lblk + map->m_len < ee_block + ee_len) { + split_flag1 = split_flag & EXT4_EXT_MAY_ZEROOUT; + flags1 = flags | EXT4_GET_BLOCKS_PRE_IO; + if (unwritten) + split_flag1 |= EXT4_EXT_MARK_UNWRIT1 | + EXT4_EXT_MARK_UNWRIT2; + if (split_flag & EXT4_EXT_DATA_VALID2) + split_flag1 |= EXT4_EXT_DATA_VALID1; + path = ext4_split_extent_at(handle, inode, path, + map->m_lblk + map->m_len, split_flag1, flags1); + if (IS_ERR(path)) + return path; + /* + * Update path is required because previous ext4_split_extent_at + * may result in split of original leaf or extent zeroout. + */ + path = ext4_find_extent(inode, map->m_lblk, path, flags); + if (IS_ERR(path)) + return path; + depth = ext_depth(inode); + ex = path[depth].p_ext; + if (!ex) { + EXT4_ERROR_INODE(inode, "unexpected hole at %lu", + (unsigned long) map->m_lblk); + ext4_free_ext_path(path); + return ERR_PTR(-EFSCORRUPTED); + } + unwritten = ext4_ext_is_unwritten(ex); + } + + if (map->m_lblk >= ee_block) { + split_flag1 = split_flag & EXT4_EXT_DATA_VALID2; + if (unwritten) { + split_flag1 |= EXT4_EXT_MARK_UNWRIT1; + split_flag1 |= split_flag & (EXT4_EXT_MAY_ZEROOUT | + EXT4_EXT_MARK_UNWRIT2); + } + path = ext4_split_extent_at(handle, inode, path, + map->m_lblk, split_flag1, flags); + if (IS_ERR(path)) + return path; + } + + if (allocated) { + if (map->m_lblk + map->m_len > ee_block + ee_len) + *allocated = ee_len - (map->m_lblk - ee_block); + else + *allocated = map->m_len; + } + ext4_ext_show_leaf(inode, path); + return path; +} + +/* + * This function is called by ext4_ext_map_blocks() if someone tries to write + * to an unwritten extent. It may result in splitting the unwritten + * extent into multiple extents (up to three - one initialized and two + * unwritten). + * There are three possibilities: + * a> There is no split required: Entire extent should be initialized + * b> Splits in two extents: Write is happening at either end of the extent + * c> Splits in three extents: Somone is writing in middle of the extent + * + * Pre-conditions: + * - The extent pointed to by 'path' is unwritten. + * - The extent pointed to by 'path' contains a superset + * of the logical span [map->m_lblk, map->m_lblk + map->m_len). + * + * Post-conditions on success: + * - the returned value is the number of blocks beyond map->l_lblk + * that are allocated and initialized. + * It is guaranteed to be >= map->m_len. + */ +static struct ext4_ext_path * +ext4_ext_convert_to_initialized(handle_t *handle, struct inode *inode, + struct ext4_map_blocks *map, struct ext4_ext_path *path, + int flags, unsigned int *allocated) +{ + struct ext4_sb_info *sbi; + struct ext4_extent_header *eh; + struct ext4_map_blocks split_map; + struct ext4_extent zero_ex1, zero_ex2; + struct ext4_extent *ex, *abut_ex; + ext4_lblk_t ee_block, eof_block; + unsigned int ee_len, depth, map_len = map->m_len; + int err = 0; + int split_flag = EXT4_EXT_DATA_VALID2; + unsigned int max_zeroout = 0; + + ext_debug(inode, "logical block %llu, max_blocks %u\n", + (unsigned long long)map->m_lblk, map_len); + + sbi = EXT4_SB(inode->i_sb); + eof_block = (EXT4_I(inode)->i_disksize + inode->i_sb->s_blocksize - 1) + >> inode->i_sb->s_blocksize_bits; + if (eof_block < map->m_lblk + map_len) + eof_block = map->m_lblk + map_len; + + depth = ext_depth(inode); + eh = path[depth].p_hdr; + ex = path[depth].p_ext; + ee_block = le32_to_cpu(ex->ee_block); + ee_len = ext4_ext_get_actual_len(ex); + zero_ex1.ee_len = 0; + zero_ex2.ee_len = 0; + + trace_ext4_ext_convert_to_initialized_enter(inode, map, ex); + + /* Pre-conditions */ + BUG_ON(!ext4_ext_is_unwritten(ex)); + BUG_ON(!in_range(map->m_lblk, ee_block, ee_len)); + + /* + * Attempt to transfer newly initialized blocks from the currently + * unwritten extent to its neighbor. This is much cheaper + * than an insertion followed by a merge as those involve costly + * memmove() calls. Transferring to the left is the common case in + * steady state for workloads doing fallocate(FALLOC_FL_KEEP_SIZE) + * followed by append writes. + * + * Limitations of the current logic: + * - L1: we do not deal with writes covering the whole extent. + * This would require removing the extent if the transfer + * is possible. + * - L2: we only attempt to merge with an extent stored in the + * same extent tree node. + */ + *allocated = 0; + if ((map->m_lblk == ee_block) && + /* See if we can merge left */ + (map_len < ee_len) && /*L1*/ + (ex > EXT_FIRST_EXTENT(eh))) { /*L2*/ + ext4_lblk_t prev_lblk; + ext4_fsblk_t prev_pblk, ee_pblk; + unsigned int prev_len; + + abut_ex = ex - 1; + prev_lblk = le32_to_cpu(abut_ex->ee_block); + prev_len = ext4_ext_get_actual_len(abut_ex); + prev_pblk = ext4_ext_pblock(abut_ex); + ee_pblk = ext4_ext_pblock(ex); + + /* + * A transfer of blocks from 'ex' to 'abut_ex' is allowed + * upon those conditions: + * - C1: abut_ex is initialized, + * - C2: abut_ex is logically abutting ex, + * - C3: abut_ex is physically abutting ex, + * - C4: abut_ex can receive the additional blocks without + * overflowing the (initialized) length limit. + */ + if ((!ext4_ext_is_unwritten(abut_ex)) && /*C1*/ + ((prev_lblk + prev_len) == ee_block) && /*C2*/ + ((prev_pblk + prev_len) == ee_pblk) && /*C3*/ + (prev_len < (EXT_INIT_MAX_LEN - map_len))) { /*C4*/ + err = ext4_ext_get_access(handle, inode, path + depth); + if (err) + goto errout; + + trace_ext4_ext_convert_to_initialized_fastpath(inode, + map, ex, abut_ex); + + /* Shift the start of ex by 'map_len' blocks */ + ex->ee_block = cpu_to_le32(ee_block + map_len); + ext4_ext_store_pblock(ex, ee_pblk + map_len); + ex->ee_len = cpu_to_le16(ee_len - map_len); + ext4_ext_mark_unwritten(ex); /* Restore the flag */ + + /* Extend abut_ex by 'map_len' blocks */ + abut_ex->ee_len = cpu_to_le16(prev_len + map_len); + + /* Result: number of initialized blocks past m_lblk */ + *allocated = map_len; + } + } else if (((map->m_lblk + map_len) == (ee_block + ee_len)) && + (map_len < ee_len) && /*L1*/ + ex < EXT_LAST_EXTENT(eh)) { /*L2*/ + /* See if we can merge right */ + ext4_lblk_t next_lblk; + ext4_fsblk_t next_pblk, ee_pblk; + unsigned int next_len; + + abut_ex = ex + 1; + next_lblk = le32_to_cpu(abut_ex->ee_block); + next_len = ext4_ext_get_actual_len(abut_ex); + next_pblk = ext4_ext_pblock(abut_ex); + ee_pblk = ext4_ext_pblock(ex); + + /* + * A transfer of blocks from 'ex' to 'abut_ex' is allowed + * upon those conditions: + * - C1: abut_ex is initialized, + * - C2: abut_ex is logically abutting ex, + * - C3: abut_ex is physically abutting ex, + * - C4: abut_ex can receive the additional blocks without + * overflowing the (initialized) length limit. + */ + if ((!ext4_ext_is_unwritten(abut_ex)) && /*C1*/ + ((map->m_lblk + map_len) == next_lblk) && /*C2*/ + ((ee_pblk + ee_len) == next_pblk) && /*C3*/ + (next_len < (EXT_INIT_MAX_LEN - map_len))) { /*C4*/ + err = ext4_ext_get_access(handle, inode, path + depth); + if (err) + goto errout; + + trace_ext4_ext_convert_to_initialized_fastpath(inode, + map, ex, abut_ex); + + /* Shift the start of abut_ex by 'map_len' blocks */ + abut_ex->ee_block = cpu_to_le32(next_lblk - map_len); + ext4_ext_store_pblock(abut_ex, next_pblk - map_len); + ex->ee_len = cpu_to_le16(ee_len - map_len); + ext4_ext_mark_unwritten(ex); /* Restore the flag */ + + /* Extend abut_ex by 'map_len' blocks */ + abut_ex->ee_len = cpu_to_le16(next_len + map_len); + + /* Result: number of initialized blocks past m_lblk */ + *allocated = map_len; + } + } + if (*allocated) { + /* Mark the block containing both extents as dirty */ + err = ext4_ext_dirty(handle, inode, path + depth); + + /* Update path to point to the right extent */ + path[depth].p_ext = abut_ex; + if (err) + goto errout; + goto out; + } else + *allocated = ee_len - (map->m_lblk - ee_block); + + WARN_ON(map->m_lblk < ee_block); + /* + * It is safe to convert extent to initialized via explicit + * zeroout only if extent is fully inside i_size or new_size. + */ + split_flag |= ee_block + ee_len <= eof_block ? EXT4_EXT_MAY_ZEROOUT : 0; + + if (EXT4_EXT_MAY_ZEROOUT & split_flag) + max_zeroout = sbi->s_extent_max_zeroout_kb >> + (inode->i_sb->s_blocksize_bits - 10); + + /* + * five cases: + * 1. split the extent into three extents. + * 2. split the extent into two extents, zeroout the head of the first + * extent. + * 3. split the extent into two extents, zeroout the tail of the second + * extent. + * 4. split the extent into two extents with out zeroout. + * 5. no splitting needed, just possibly zeroout the head and / or the + * tail of the extent. + */ + split_map.m_lblk = map->m_lblk; + split_map.m_len = map->m_len; + + if (max_zeroout && (*allocated > split_map.m_len)) { + if (*allocated <= max_zeroout) { + /* case 3 or 5 */ + zero_ex1.ee_block = + cpu_to_le32(split_map.m_lblk + + split_map.m_len); + zero_ex1.ee_len = + cpu_to_le16(*allocated - split_map.m_len); + ext4_ext_store_pblock(&zero_ex1, + ext4_ext_pblock(ex) + split_map.m_lblk + + split_map.m_len - ee_block); + err = ext4_ext_zeroout(inode, &zero_ex1); + if (err) + goto fallback; + split_map.m_len = *allocated; + } + if (split_map.m_lblk - ee_block + split_map.m_len < + max_zeroout) { + /* case 2 or 5 */ + if (split_map.m_lblk != ee_block) { + zero_ex2.ee_block = ex->ee_block; + zero_ex2.ee_len = cpu_to_le16(split_map.m_lblk - + ee_block); + ext4_ext_store_pblock(&zero_ex2, + ext4_ext_pblock(ex)); + err = ext4_ext_zeroout(inode, &zero_ex2); + if (err) + goto fallback; + } + + split_map.m_len += split_map.m_lblk - ee_block; + split_map.m_lblk = ee_block; + *allocated = map->m_len; + } + } + +fallback: + path = ext4_split_extent(handle, inode, path, &split_map, split_flag, + flags, NULL); + if (IS_ERR(path)) + return path; +out: + /* If we have gotten a failure, don't zero out status tree */ + ext4_zeroout_es(inode, &zero_ex1); + ext4_zeroout_es(inode, &zero_ex2); + return path; + +errout: + ext4_free_ext_path(path); + return ERR_PTR(err); +} + +/* + * This function is called by ext4_ext_map_blocks() from + * ext4_get_blocks_dio_write() when DIO to write + * to an unwritten extent. + * + * Writing to an unwritten extent may result in splitting the unwritten + * extent into multiple initialized/unwritten extents (up to three) + * There are three possibilities: + * a> There is no split required: Entire extent should be unwritten + * b> Splits in two extents: Write is happening at either end of the extent + * c> Splits in three extents: Somone is writing in middle of the extent + * + * This works the same way in the case of initialized -> unwritten conversion. + * + * One of more index blocks maybe needed if the extent tree grow after + * the unwritten extent split. To prevent ENOSPC occur at the IO + * complete, we need to split the unwritten extent before DIO submit + * the IO. The unwritten extent called at this time will be split + * into three unwritten extent(at most). After IO complete, the part + * being filled will be convert to initialized by the end_io callback function + * via ext4_convert_unwritten_extents(). + * + * The size of unwritten extent to be written is passed to the caller via the + * allocated pointer. Return an extent path pointer on success, or an error + * pointer on failure. + */ +static struct ext4_ext_path *ext4_split_convert_extents(handle_t *handle, + struct inode *inode, + struct ext4_map_blocks *map, + struct ext4_ext_path *path, + int flags, unsigned int *allocated) +{ + ext4_lblk_t eof_block; + ext4_lblk_t ee_block; + struct ext4_extent *ex; + unsigned int ee_len; + int split_flag = 0, depth; + + ext_debug(inode, "logical block %llu, max_blocks %u\n", + (unsigned long long)map->m_lblk, map->m_len); + + eof_block = (EXT4_I(inode)->i_disksize + inode->i_sb->s_blocksize - 1) + >> inode->i_sb->s_blocksize_bits; + if (eof_block < map->m_lblk + map->m_len) + eof_block = map->m_lblk + map->m_len; + /* + * It is safe to convert extent to initialized via explicit + * zeroout only if extent is fully inside i_size or new_size. + */ + depth = ext_depth(inode); + ex = path[depth].p_ext; + ee_block = le32_to_cpu(ex->ee_block); + ee_len = ext4_ext_get_actual_len(ex); + + /* Convert to unwritten */ + if (flags & EXT4_GET_BLOCKS_CONVERT_UNWRITTEN) { + split_flag |= EXT4_EXT_DATA_VALID1; + /* Convert to initialized */ + } else if (flags & EXT4_GET_BLOCKS_CONVERT) { + split_flag |= ee_block + ee_len <= eof_block ? + EXT4_EXT_MAY_ZEROOUT : 0; + split_flag |= (EXT4_EXT_MARK_UNWRIT2 | EXT4_EXT_DATA_VALID2); + } + flags |= EXT4_GET_BLOCKS_PRE_IO; + return ext4_split_extent(handle, inode, path, map, split_flag, flags, + allocated); +} + +static struct ext4_ext_path * +ext4_convert_unwritten_extents_endio(handle_t *handle, struct inode *inode, + struct ext4_map_blocks *map, + struct ext4_ext_path *path) +{ + struct ext4_extent *ex; + ext4_lblk_t ee_block; + unsigned int ee_len; + int depth; + int err = 0; + + depth = ext_depth(inode); + ex = path[depth].p_ext; + ee_block = le32_to_cpu(ex->ee_block); + ee_len = ext4_ext_get_actual_len(ex); + + ext_debug(inode, "logical block %llu, max_blocks %u\n", + (unsigned long long)ee_block, ee_len); + + /* If extent is larger than requested it is a clear sign that we still + * have some extent state machine issues left. So extent_split is still + * required. + * TODO: Once all related issues will be fixed this situation should be + * illegal. + */ + if (ee_block != map->m_lblk || ee_len > map->m_len) { +#ifdef CONFIG_EXT4_DEBUG + ext4_warning(inode->i_sb, "Inode (%ld) finished: extent logical block %llu," + " len %u; IO logical block %llu, len %u", + inode->i_ino, (unsigned long long)ee_block, ee_len, + (unsigned long long)map->m_lblk, map->m_len); +#endif + path = ext4_split_convert_extents(handle, inode, map, path, + EXT4_GET_BLOCKS_CONVERT, NULL); + if (IS_ERR(path)) + return path; + + path = ext4_find_extent(inode, map->m_lblk, path, 0); + if (IS_ERR(path)) + return path; + depth = ext_depth(inode); + ex = path[depth].p_ext; + } + + err = ext4_ext_get_access(handle, inode, path + depth); + if (err) + goto errout; + /* first mark the extent as initialized */ + ext4_ext_mark_initialized(ex); + + /* note: ext4_ext_correct_indexes() isn't needed here because + * borders are not changed + */ + ext4_ext_try_to_merge(handle, inode, path, ex); + + /* Mark modified extent as dirty */ + err = ext4_ext_dirty(handle, inode, path + path->p_depth); + if (err) + goto errout; + + ext4_ext_show_leaf(inode, path); + return path; + +errout: + ext4_free_ext_path(path); + return ERR_PTR(err); +} + +static struct ext4_ext_path * +convert_initialized_extent(handle_t *handle, struct inode *inode, + struct ext4_map_blocks *map, + struct ext4_ext_path *path, + unsigned int *allocated) +{ + struct ext4_extent *ex; + ext4_lblk_t ee_block; + unsigned int ee_len; + int depth; + int err = 0; + + /* + * Make sure that the extent is no bigger than we support with + * unwritten extent + */ + if (map->m_len > EXT_UNWRITTEN_MAX_LEN) + map->m_len = EXT_UNWRITTEN_MAX_LEN / 2; + + depth = ext_depth(inode); + ex = path[depth].p_ext; + ee_block = le32_to_cpu(ex->ee_block); + ee_len = ext4_ext_get_actual_len(ex); + + ext_debug(inode, "logical block %llu, max_blocks %u\n", + (unsigned long long)ee_block, ee_len); + + if (ee_block != map->m_lblk || ee_len > map->m_len) { + path = ext4_split_convert_extents(handle, inode, map, path, + EXT4_GET_BLOCKS_CONVERT_UNWRITTEN, NULL); + if (IS_ERR(path)) + return path; + + path = ext4_find_extent(inode, map->m_lblk, path, 0); + if (IS_ERR(path)) + return path; + depth = ext_depth(inode); + ex = path[depth].p_ext; + if (!ex) { + EXT4_ERROR_INODE(inode, "unexpected hole at %lu", + (unsigned long) map->m_lblk); + err = -EFSCORRUPTED; + goto errout; + } + } + + err = ext4_ext_get_access(handle, inode, path + depth); + if (err) + goto errout; + /* first mark the extent as unwritten */ + ext4_ext_mark_unwritten(ex); + + /* note: ext4_ext_correct_indexes() isn't needed here because + * borders are not changed + */ + ext4_ext_try_to_merge(handle, inode, path, ex); + + /* Mark modified extent as dirty */ + err = ext4_ext_dirty(handle, inode, path + path->p_depth); + if (err) + goto errout; + ext4_ext_show_leaf(inode, path); + + ext4_update_inode_fsync_trans(handle, inode, 1); + + map->m_flags |= EXT4_MAP_UNWRITTEN; + if (*allocated > map->m_len) + *allocated = map->m_len; + map->m_len = *allocated; + return path; + +errout: + ext4_free_ext_path(path); + return ERR_PTR(err); +} + +static struct ext4_ext_path * +ext4_ext_handle_unwritten_extents(handle_t *handle, struct inode *inode, + struct ext4_map_blocks *map, + struct ext4_ext_path *path, int flags, + unsigned int *allocated, ext4_fsblk_t newblock) +{ + int err = 0; + + ext_debug(inode, "logical block %llu, max_blocks %u, flags 0x%x, allocated %u\n", + (unsigned long long)map->m_lblk, map->m_len, flags, + *allocated); + ext4_ext_show_leaf(inode, path); + + /* + * When writing into unwritten space, we should not fail to + * allocate metadata blocks for the new extent block if needed. + */ + flags |= EXT4_GET_BLOCKS_METADATA_NOFAIL; + + trace_ext4_ext_handle_unwritten_extents(inode, map, flags, + *allocated, newblock); + + /* get_block() before submitting IO, split the extent */ + if (flags & EXT4_GET_BLOCKS_PRE_IO) { + path = ext4_split_convert_extents(handle, inode, map, path, + flags | EXT4_GET_BLOCKS_CONVERT, allocated); + if (IS_ERR(path)) + return path; + /* + * shouldn't get a 0 allocated when splitting an extent unless + * m_len is 0 (bug) or extent has been corrupted + */ + if (unlikely(*allocated == 0)) { + EXT4_ERROR_INODE(inode, + "unexpected allocated == 0, m_len = %u", + map->m_len); + err = -EFSCORRUPTED; + goto errout; + } + map->m_flags |= EXT4_MAP_UNWRITTEN; + goto out; + } + /* IO end_io complete, convert the filled extent to written */ + if (flags & EXT4_GET_BLOCKS_CONVERT) { + path = ext4_convert_unwritten_extents_endio(handle, inode, + map, path); + if (IS_ERR(path)) + return path; + ext4_update_inode_fsync_trans(handle, inode, 1); + goto map_out; + } + /* buffered IO cases */ + /* + * repeat fallocate creation request + * we already have an unwritten extent + */ + if (flags & EXT4_GET_BLOCKS_UNWRIT_EXT) { + map->m_flags |= EXT4_MAP_UNWRITTEN; + goto map_out; + } + + /* buffered READ or buffered write_begin() lookup */ + if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) { + /* + * We have blocks reserved already. We + * return allocated blocks so that delalloc + * won't do block reservation for us. But + * the buffer head will be unmapped so that + * a read from the block returns 0s. + */ + map->m_flags |= EXT4_MAP_UNWRITTEN; + goto out1; + } + + /* + * Default case when (flags & EXT4_GET_BLOCKS_CREATE) == 1. + * For buffered writes, at writepage time, etc. Convert a + * discovered unwritten extent to written. + */ + path = ext4_ext_convert_to_initialized(handle, inode, map, path, + flags, allocated); + if (IS_ERR(path)) + return path; + ext4_update_inode_fsync_trans(handle, inode, 1); + /* + * shouldn't get a 0 allocated when converting an unwritten extent + * unless m_len is 0 (bug) or extent has been corrupted + */ + if (unlikely(*allocated == 0)) { + EXT4_ERROR_INODE(inode, "unexpected allocated == 0, m_len = %u", + map->m_len); + err = -EFSCORRUPTED; + goto errout; + } + +out: + map->m_flags |= EXT4_MAP_NEW; +map_out: + map->m_flags |= EXT4_MAP_MAPPED; +out1: + map->m_pblk = newblock; + if (*allocated > map->m_len) + *allocated = map->m_len; + map->m_len = *allocated; + ext4_ext_show_leaf(inode, path); + return path; + +errout: + ext4_free_ext_path(path); + return ERR_PTR(err); +} + +/* + * get_implied_cluster_alloc - check to see if the requested + * allocation (in the map structure) overlaps with a cluster already + * allocated in an extent. + * @sb The filesystem superblock structure + * @map The requested lblk->pblk mapping + * @ex The extent structure which might contain an implied + * cluster allocation + * + * This function is called by ext4_ext_map_blocks() after we failed to + * find blocks that were already in the inode's extent tree. Hence, + * we know that the beginning of the requested region cannot overlap + * the extent from the inode's extent tree. There are three cases we + * want to catch. The first is this case: + * + * |--- cluster # N--| + * |--- extent ---| |---- requested region ---| + * |==========| + * + * The second case that we need to test for is this one: + * + * |--------- cluster # N ----------------| + * |--- requested region --| |------- extent ----| + * |=======================| + * + * The third case is when the requested region lies between two extents + * within the same cluster: + * |------------- cluster # N-------------| + * |----- ex -----| |---- ex_right ----| + * |------ requested region ------| + * |================| + * + * In each of the above cases, we need to set the map->m_pblk and + * map->m_len so it corresponds to the return the extent labelled as + * "|====|" from cluster #N, since it is already in use for data in + * cluster EXT4_B2C(sbi, map->m_lblk). We will then return 1 to + * signal to ext4_ext_map_blocks() that map->m_pblk should be treated + * as a new "allocated" block region. Otherwise, we will return 0 and + * ext4_ext_map_blocks() will then allocate one or more new clusters + * by calling ext4_mb_new_blocks(). + */ +static int get_implied_cluster_alloc(struct super_block *sb, + struct ext4_map_blocks *map, + struct ext4_extent *ex, + struct ext4_ext_path *path) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + ext4_lblk_t c_offset = EXT4_LBLK_COFF(sbi, map->m_lblk); + ext4_lblk_t ex_cluster_start, ex_cluster_end; + ext4_lblk_t rr_cluster_start; + ext4_lblk_t ee_block = le32_to_cpu(ex->ee_block); + ext4_fsblk_t ee_start = ext4_ext_pblock(ex); + unsigned short ee_len = ext4_ext_get_actual_len(ex); + + /* The extent passed in that we are trying to match */ + ex_cluster_start = EXT4_B2C(sbi, ee_block); + ex_cluster_end = EXT4_B2C(sbi, ee_block + ee_len - 1); + + /* The requested region passed into ext4_map_blocks() */ + rr_cluster_start = EXT4_B2C(sbi, map->m_lblk); + + if ((rr_cluster_start == ex_cluster_end) || + (rr_cluster_start == ex_cluster_start)) { + if (rr_cluster_start == ex_cluster_end) + ee_start += ee_len - 1; + map->m_pblk = EXT4_PBLK_CMASK(sbi, ee_start) + c_offset; + map->m_len = min(map->m_len, + (unsigned) sbi->s_cluster_ratio - c_offset); + /* + * Check for and handle this case: + * + * |--------- cluster # N-------------| + * |------- extent ----| + * |--- requested region ---| + * |===========| + */ + + if (map->m_lblk < ee_block) + map->m_len = min(map->m_len, ee_block - map->m_lblk); + + /* + * Check for the case where there is already another allocated + * block to the right of 'ex' but before the end of the cluster. + * + * |------------- cluster # N-------------| + * |----- ex -----| |---- ex_right ----| + * |------ requested region ------| + * |================| + */ + if (map->m_lblk > ee_block) { + ext4_lblk_t next = ext4_ext_next_allocated_block(path); + map->m_len = min(map->m_len, next - map->m_lblk); + } + + trace_ext4_get_implied_cluster_alloc_exit(sb, map, 1); + return 1; + } + + trace_ext4_get_implied_cluster_alloc_exit(sb, map, 0); + return 0; +} + +/* + * Determine hole length around the given logical block, first try to + * locate and expand the hole from the given @path, and then adjust it + * if it's partially or completely converted to delayed extents, insert + * it into the extent cache tree if it's indeed a hole, finally return + * the length of the determined extent. + */ +static ext4_lblk_t ext4_ext_determine_insert_hole(struct inode *inode, + struct ext4_ext_path *path, + ext4_lblk_t lblk) +{ + ext4_lblk_t hole_start, len; + struct extent_status es; + + hole_start = lblk; + len = ext4_ext_find_hole(inode, path, &hole_start); +again: + ext4_es_find_extent_range(inode, &ext4_es_is_delayed, hole_start, + hole_start + len - 1, &es); + if (!es.es_len) + goto insert_hole; + + /* + * There's a delalloc extent in the hole, handle it if the delalloc + * extent is in front of, behind and straddle the queried range. + */ + if (lblk >= es.es_lblk + es.es_len) { + /* + * The delalloc extent is in front of the queried range, + * find again from the queried start block. + */ + len -= lblk - hole_start; + hole_start = lblk; + goto again; + } else if (in_range(lblk, es.es_lblk, es.es_len)) { + /* + * The delalloc extent containing lblk, it must have been + * added after ext4_map_blocks() checked the extent status + * tree so we are not holding i_rwsem and delalloc info is + * only stabilized by i_data_sem we are going to release + * soon. Don't modify the extent status tree and report + * extent as a hole, just adjust the length to the delalloc + * extent's after lblk. + */ + len = es.es_lblk + es.es_len - lblk; + return len; + } else { + /* + * The delalloc extent is partially or completely behind + * the queried range, update hole length until the + * beginning of the delalloc extent. + */ + len = min(es.es_lblk - hole_start, len); + } + +insert_hole: + /* Put just found gap into cache to speed up subsequent requests */ + ext_debug(inode, " -> %u:%u\n", hole_start, len); + ext4_es_insert_extent(inode, hole_start, len, ~0, + EXTENT_STATUS_HOLE, false); + + /* Update hole_len to reflect hole size after lblk */ + if (hole_start != lblk) + len -= lblk - hole_start; + + return len; +} + +/* + * Block allocation/map/preallocation routine for extents based files + * + * + * Need to be called with + * down_read(&EXT4_I(inode)->i_data_sem) if not allocating file system block + * (ie, flags is zero). Otherwise down_write(&EXT4_I(inode)->i_data_sem) + * + * return > 0, number of blocks already mapped/allocated + * if flags doesn't contain EXT4_GET_BLOCKS_CREATE and these are pre-allocated blocks + * buffer head is unmapped + * otherwise blocks are mapped + * + * return = 0, if plain look up failed (blocks have not been allocated) + * buffer head is unmapped + * + * return < 0, error case. + */ +int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, + struct ext4_map_blocks *map, int flags) +{ + struct ext4_ext_path *path = NULL; + struct ext4_extent newex, *ex, ex2; + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + ext4_fsblk_t newblock = 0, pblk; + int err = 0, depth; + unsigned int allocated = 0, offset = 0; + unsigned int allocated_clusters = 0; + struct ext4_allocation_request ar; + ext4_lblk_t cluster_offset; + + ext_debug(inode, "blocks %u/%u requested\n", map->m_lblk, map->m_len); + trace_ext4_ext_map_blocks_enter(inode, map->m_lblk, map->m_len, flags); + + /* find extent for this block */ + path = ext4_find_extent(inode, map->m_lblk, NULL, flags); + if (IS_ERR(path)) { + err = PTR_ERR(path); + goto out; + } + + depth = ext_depth(inode); + + /* + * consistent leaf must not be empty; + * this situation is possible, though, _during_ tree modification; + * this is why assert can't be put in ext4_find_extent() + */ + if (unlikely(path[depth].p_ext == NULL && depth != 0)) { + EXT4_ERROR_INODE(inode, "bad extent address " + "lblock: %lu, depth: %d pblock %lld", + (unsigned long) map->m_lblk, depth, + path[depth].p_block); + err = -EFSCORRUPTED; + goto out; + } + + ex = path[depth].p_ext; + if (ex) { + ext4_lblk_t ee_block = le32_to_cpu(ex->ee_block); + ext4_fsblk_t ee_start = ext4_ext_pblock(ex); + unsigned short ee_len; + + + /* + * unwritten extents are treated as holes, except that + * we split out initialized portions during a write. + */ + ee_len = ext4_ext_get_actual_len(ex); + + trace_ext4_ext_show_extent(inode, ee_block, ee_start, ee_len); + + /* if found extent covers block, simply return it */ + if (in_range(map->m_lblk, ee_block, ee_len)) { + newblock = map->m_lblk - ee_block + ee_start; + /* number of remaining blocks in the extent */ + allocated = ee_len - (map->m_lblk - ee_block); + ext_debug(inode, "%u fit into %u:%d -> %llu\n", + map->m_lblk, ee_block, ee_len, newblock); + + /* + * If the extent is initialized check whether the + * caller wants to convert it to unwritten. + */ + if ((!ext4_ext_is_unwritten(ex)) && + (flags & EXT4_GET_BLOCKS_CONVERT_UNWRITTEN)) { + path = convert_initialized_extent(handle, + inode, map, path, &allocated); + if (IS_ERR(path)) + err = PTR_ERR(path); + goto out; + } else if (!ext4_ext_is_unwritten(ex)) { + map->m_flags |= EXT4_MAP_MAPPED; + map->m_pblk = newblock; + if (allocated > map->m_len) + allocated = map->m_len; + map->m_len = allocated; + ext4_ext_show_leaf(inode, path); + goto out; + } + + path = ext4_ext_handle_unwritten_extents( + handle, inode, map, path, flags, + &allocated, newblock); + if (IS_ERR(path)) + err = PTR_ERR(path); + goto out; + } + } + + /* + * requested block isn't allocated yet; + * we couldn't try to create block if flags doesn't contain EXT4_GET_BLOCKS_CREATE + */ + if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) { + ext4_lblk_t len; + + len = ext4_ext_determine_insert_hole(inode, path, map->m_lblk); + + map->m_pblk = 0; + map->m_len = min_t(unsigned int, map->m_len, len); + goto out; + } + + /* + * Okay, we need to do block allocation. + */ + newex.ee_block = cpu_to_le32(map->m_lblk); + cluster_offset = EXT4_LBLK_COFF(sbi, map->m_lblk); + + /* + * If we are doing bigalloc, check to see if the extent returned + * by ext4_find_extent() implies a cluster we can use. + */ + if (cluster_offset && ex && + get_implied_cluster_alloc(inode->i_sb, map, ex, path)) { + ar.len = allocated = map->m_len; + newblock = map->m_pblk; + goto got_allocated_blocks; + } + + /* find neighbour allocated blocks */ + ar.lleft = map->m_lblk; + err = ext4_ext_search_left(inode, path, &ar.lleft, &ar.pleft); + if (err) + goto out; + ar.lright = map->m_lblk; + err = ext4_ext_search_right(inode, path, &ar.lright, &ar.pright, + &ex2, flags); + if (err < 0) + goto out; + + /* Check if the extent after searching to the right implies a + * cluster we can use. */ + if ((sbi->s_cluster_ratio > 1) && err && + get_implied_cluster_alloc(inode->i_sb, map, &ex2, path)) { + ar.len = allocated = map->m_len; + newblock = map->m_pblk; + err = 0; + goto got_allocated_blocks; + } + + /* + * See if request is beyond maximum number of blocks we can have in + * a single extent. For an initialized extent this limit is + * EXT_INIT_MAX_LEN and for an unwritten extent this limit is + * EXT_UNWRITTEN_MAX_LEN. + */ + if (map->m_len > EXT_INIT_MAX_LEN && + !(flags & EXT4_GET_BLOCKS_UNWRIT_EXT)) + map->m_len = EXT_INIT_MAX_LEN; + else if (map->m_len > EXT_UNWRITTEN_MAX_LEN && + (flags & EXT4_GET_BLOCKS_UNWRIT_EXT)) + map->m_len = EXT_UNWRITTEN_MAX_LEN; + + /* Check if we can really insert (m_lblk)::(m_lblk + m_len) extent */ + newex.ee_len = cpu_to_le16(map->m_len); + err = ext4_ext_check_overlap(sbi, inode, &newex, path); + if (err) + allocated = ext4_ext_get_actual_len(&newex); + else + allocated = map->m_len; + + /* allocate new block */ + ar.inode = inode; + ar.goal = ext4_ext_find_goal(inode, path, map->m_lblk); + ar.logical = map->m_lblk; + /* + * We calculate the offset from the beginning of the cluster + * for the logical block number, since when we allocate a + * physical cluster, the physical block should start at the + * same offset from the beginning of the cluster. This is + * needed so that future calls to get_implied_cluster_alloc() + * work correctly. + */ + offset = EXT4_LBLK_COFF(sbi, map->m_lblk); + ar.len = EXT4_NUM_B2C(sbi, offset+allocated); + ar.goal -= offset; + ar.logical -= offset; + if (S_ISREG(inode->i_mode)) + ar.flags = EXT4_MB_HINT_DATA; + else + /* disable in-core preallocation for non-regular files */ + ar.flags = 0; + if (flags & EXT4_GET_BLOCKS_NO_NORMALIZE) + ar.flags |= EXT4_MB_HINT_NOPREALLOC; + if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) + ar.flags |= EXT4_MB_DELALLOC_RESERVED; + if (flags & EXT4_GET_BLOCKS_METADATA_NOFAIL) + ar.flags |= EXT4_MB_USE_RESERVED; + newblock = ext4_mb_new_blocks(handle, &ar, &err); + if (!newblock) + goto out; + allocated_clusters = ar.len; + ar.len = EXT4_C2B(sbi, ar.len) - offset; + ext_debug(inode, "allocate new block: goal %llu, found %llu/%u, requested %u\n", + ar.goal, newblock, ar.len, allocated); + if (ar.len > allocated) + ar.len = allocated; + +got_allocated_blocks: + /* try to insert new extent into found leaf and return */ + pblk = newblock + offset; + ext4_ext_store_pblock(&newex, pblk); + newex.ee_len = cpu_to_le16(ar.len); + /* Mark unwritten */ + if (flags & EXT4_GET_BLOCKS_UNWRIT_EXT) { + ext4_ext_mark_unwritten(&newex); + map->m_flags |= EXT4_MAP_UNWRITTEN; + } + + path = ext4_ext_insert_extent(handle, inode, path, &newex, flags); + if (IS_ERR(path)) { + err = PTR_ERR(path); + if (allocated_clusters) { + int fb_flags = 0; + + /* + * free data blocks we just allocated. + * not a good idea to call discard here directly, + * but otherwise we'd need to call it every free(). + */ + ext4_discard_preallocations(inode); + if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) + fb_flags = EXT4_FREE_BLOCKS_NO_QUOT_UPDATE; + ext4_free_blocks(handle, inode, NULL, newblock, + EXT4_C2B(sbi, allocated_clusters), + fb_flags); + } + goto out; + } + + /* + * Cache the extent and update transaction to commit on fdatasync only + * when it is _not_ an unwritten extent. + */ + if ((flags & EXT4_GET_BLOCKS_UNWRIT_EXT) == 0) + ext4_update_inode_fsync_trans(handle, inode, 1); + else + ext4_update_inode_fsync_trans(handle, inode, 0); + + map->m_flags |= (EXT4_MAP_NEW | EXT4_MAP_MAPPED); + map->m_pblk = pblk; + map->m_len = ar.len; + allocated = map->m_len; + ext4_ext_show_leaf(inode, path); +out: + /* + * We never use EXT4_GET_BLOCKS_QUERY_LAST_IN_LEAF with CREATE flag. + * So we know that the depth used here is correct, since there was no + * block allocation done if EXT4_GET_BLOCKS_QUERY_LAST_IN_LEAF is set. + * If tomorrow we start using this QUERY flag with CREATE, then we will + * need to re-calculate the depth as it might have changed due to block + * allocation. + */ + if (flags & EXT4_GET_BLOCKS_QUERY_LAST_IN_LEAF) { + WARN_ON_ONCE(flags & EXT4_GET_BLOCKS_CREATE); + if (!err && ex && (ex == EXT_LAST_EXTENT(path[depth].p_hdr))) + map->m_flags |= EXT4_MAP_QUERY_LAST_IN_LEAF; + } + + ext4_free_ext_path(path); + + trace_ext4_ext_map_blocks_exit(inode, flags, map, + err ? err : allocated); + return err ? err : allocated; +} + +int ext4_ext_truncate(handle_t *handle, struct inode *inode) +{ + struct super_block *sb = inode->i_sb; + ext4_lblk_t last_block; + int err = 0; + + /* + * TODO: optimization is possible here. + * Probably we need not scan at all, + * because page truncation is enough. + */ + + /* we have to know where to truncate from in crash case */ + EXT4_I(inode)->i_disksize = inode->i_size; + err = ext4_mark_inode_dirty(handle, inode); + if (err) + return err; + + last_block = (inode->i_size + sb->s_blocksize - 1) + >> EXT4_BLOCK_SIZE_BITS(sb); + ext4_es_remove_extent(inode, last_block, EXT_MAX_BLOCKS - last_block); + +retry_remove_space: + err = ext4_ext_remove_space(inode, last_block, EXT_MAX_BLOCKS - 1); + if (err == -ENOMEM) { + memalloc_retry_wait(GFP_ATOMIC); + goto retry_remove_space; + } + return err; +} + +static int ext4_alloc_file_blocks(struct file *file, ext4_lblk_t offset, + ext4_lblk_t len, loff_t new_size, + int flags) +{ + struct inode *inode = file_inode(file); + handle_t *handle; + int ret = 0, ret2 = 0, ret3 = 0; + int retries = 0; + int depth = 0; + struct ext4_map_blocks map; + unsigned int credits; + loff_t epos, old_size = i_size_read(inode); + unsigned int blkbits = inode->i_blkbits; + bool alloc_zero = false; + + BUG_ON(!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)); + map.m_lblk = offset; + map.m_len = len; + /* + * Don't normalize the request if it can fit in one extent so + * that it doesn't get unnecessarily split into multiple + * extents. + */ + if (len <= EXT_UNWRITTEN_MAX_LEN) + flags |= EXT4_GET_BLOCKS_NO_NORMALIZE; + + /* + * Do the actual write zero during a running journal transaction + * costs a lot. First allocate an unwritten extent and then + * convert it to written after zeroing it out. + */ + if (flags & EXT4_GET_BLOCKS_ZERO) { + flags &= ~EXT4_GET_BLOCKS_ZERO; + flags |= EXT4_GET_BLOCKS_UNWRIT_EXT; + alloc_zero = true; + } + + /* + * credits to insert 1 extent into extent tree + */ + credits = ext4_chunk_trans_blocks(inode, len); + depth = ext_depth(inode); + +retry: + while (len) { + /* + * Recalculate credits when extent tree depth changes. + */ + if (depth != ext_depth(inode)) { + credits = ext4_chunk_trans_blocks(inode, len); + depth = ext_depth(inode); + } + + handle = ext4_journal_start(inode, EXT4_HT_MAP_BLOCKS, + credits); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + break; + } + ret = ext4_map_blocks(handle, inode, &map, flags); + if (ret <= 0) { + ext4_debug("inode #%lu: block %u: len %u: " + "ext4_ext_map_blocks returned %d", + inode->i_ino, map.m_lblk, + map.m_len, ret); + ext4_mark_inode_dirty(handle, inode); + ext4_journal_stop(handle); + break; + } + /* + * allow a full retry cycle for any remaining allocations + */ + retries = 0; + epos = (loff_t)(map.m_lblk + ret) << blkbits; + inode_set_ctime_current(inode); + if (new_size) { + if (epos > new_size) + epos = new_size; + if (ext4_update_inode_size(inode, epos) & 0x1) + inode_set_mtime_to_ts(inode, + inode_get_ctime(inode)); + if (epos > old_size) { + pagecache_isize_extended(inode, old_size, epos); + ext4_zero_partial_blocks(handle, inode, + old_size, epos - old_size); + } + } + ret2 = ext4_mark_inode_dirty(handle, inode); + ext4_update_inode_fsync_trans(handle, inode, 1); + ret3 = ext4_journal_stop(handle); + ret2 = ret3 ? ret3 : ret2; + if (unlikely(ret2)) + break; + + if (alloc_zero && + (map.m_flags & (EXT4_MAP_MAPPED | EXT4_MAP_UNWRITTEN))) { + ret2 = ext4_issue_zeroout(inode, map.m_lblk, map.m_pblk, + map.m_len); + if (likely(!ret2)) + ret2 = ext4_convert_unwritten_extents(NULL, + inode, (loff_t)map.m_lblk << blkbits, + (loff_t)map.m_len << blkbits); + if (ret2) + break; + } + + map.m_lblk += ret; + map.m_len = len = len - ret; + } + if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries)) + goto retry; + + return ret > 0 ? ret2 : ret; +} + +static int ext4_collapse_range(struct file *file, loff_t offset, loff_t len); + +static int ext4_insert_range(struct file *file, loff_t offset, loff_t len); + +static long ext4_zero_range(struct file *file, loff_t offset, + loff_t len, int mode) +{ + struct inode *inode = file_inode(file); + handle_t *handle = NULL; + loff_t new_size = 0; + loff_t end = offset + len; + ext4_lblk_t start_lblk, end_lblk; + unsigned int blocksize = i_blocksize(inode); + unsigned int blkbits = inode->i_blkbits; + int ret, flags, credits; + + trace_ext4_zero_range(inode, offset, len, mode); + WARN_ON_ONCE(!inode_is_locked(inode)); + + /* Indirect files do not support unwritten extents */ + if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) + return -EOPNOTSUPP; + + if (!(mode & FALLOC_FL_KEEP_SIZE) && + (end > inode->i_size || end > EXT4_I(inode)->i_disksize)) { + new_size = end; + ret = inode_newsize_ok(inode, new_size); + if (ret) + return ret; + } + + flags = EXT4_GET_BLOCKS_CREATE_UNWRIT_EXT; + /* Preallocate the range including the unaligned edges */ + if (!IS_ALIGNED(offset | end, blocksize)) { + ext4_lblk_t alloc_lblk = offset >> blkbits; + ext4_lblk_t len_lblk = EXT4_MAX_BLOCKS(len, offset, blkbits); + + ret = ext4_alloc_file_blocks(file, alloc_lblk, len_lblk, + new_size, flags); + if (ret) + return ret; + } + + ret = ext4_update_disksize_before_punch(inode, offset, len); + if (ret) + return ret; + + /* Now release the pages and zero block aligned part of pages */ + ret = ext4_truncate_page_cache_block_range(inode, offset, end); + if (ret) + return ret; + + /* Zero range excluding the unaligned edges */ + start_lblk = EXT4_B_TO_LBLK(inode, offset); + end_lblk = end >> blkbits; + if (end_lblk > start_lblk) { + ext4_lblk_t zero_blks = end_lblk - start_lblk; + + if (mode & FALLOC_FL_WRITE_ZEROES) + flags = EXT4_GET_BLOCKS_CREATE_ZERO | EXT4_EX_NOCACHE; + else + flags |= (EXT4_GET_BLOCKS_CONVERT_UNWRITTEN | + EXT4_EX_NOCACHE); + ret = ext4_alloc_file_blocks(file, start_lblk, zero_blks, + new_size, flags); + if (ret) + return ret; + } + /* Finish zeroing out if it doesn't contain partial block */ + if (IS_ALIGNED(offset | end, blocksize)) + return ret; + + /* + * In worst case we have to writeout two nonadjacent unwritten + * blocks and update the inode + */ + credits = (2 * ext4_ext_index_trans_blocks(inode, 2)) + 1; + if (ext4_should_journal_data(inode)) + credits += 2; + handle = ext4_journal_start(inode, EXT4_HT_MISC, credits); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + ext4_std_error(inode->i_sb, ret); + return ret; + } + + /* Zero out partial block at the edges of the range */ + ret = ext4_zero_partial_blocks(handle, inode, offset, len); + if (ret) + goto out_handle; + + if (new_size) + ext4_update_inode_size(inode, new_size); + ret = ext4_mark_inode_dirty(handle, inode); + if (unlikely(ret)) + goto out_handle; + + ext4_update_inode_fsync_trans(handle, inode, 1); + if (file->f_flags & O_SYNC) + ext4_handle_sync(handle); + +out_handle: + ext4_journal_stop(handle); + return ret; +} + +static long ext4_do_fallocate(struct file *file, loff_t offset, + loff_t len, int mode) +{ + struct inode *inode = file_inode(file); + loff_t end = offset + len; + loff_t new_size = 0; + ext4_lblk_t start_lblk, len_lblk; + int ret; + + trace_ext4_fallocate_enter(inode, offset, len, mode); + WARN_ON_ONCE(!inode_is_locked(inode)); + + start_lblk = offset >> inode->i_blkbits; + len_lblk = EXT4_MAX_BLOCKS(len, offset, inode->i_blkbits); + + /* We only support preallocation for extent-based files only. */ + if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) { + ret = -EOPNOTSUPP; + goto out; + } + + if (!(mode & FALLOC_FL_KEEP_SIZE) && + (end > inode->i_size || end > EXT4_I(inode)->i_disksize)) { + new_size = end; + ret = inode_newsize_ok(inode, new_size); + if (ret) + goto out; + } + + ret = ext4_alloc_file_blocks(file, start_lblk, len_lblk, new_size, + EXT4_GET_BLOCKS_CREATE_UNWRIT_EXT); + if (ret) + goto out; + + if (file->f_flags & O_SYNC && EXT4_SB(inode->i_sb)->s_journal) { + ret = ext4_fc_commit(EXT4_SB(inode->i_sb)->s_journal, + EXT4_I(inode)->i_sync_tid); + } +out: + trace_ext4_fallocate_exit(inode, offset, len_lblk, ret); + return ret; +} + +/* + * preallocate space for a file. This implements ext4's fallocate file + * operation, which gets called from sys_fallocate system call. + * For block-mapped files, posix_fallocate should fall back to the method + * of writing zeroes to the required new blocks (the same behavior which is + * expected for file systems which do not support fallocate() system call). + */ +long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len) +{ + struct inode *inode = file_inode(file); + struct address_space *mapping = file->f_mapping; + int ret; + + /* + * Encrypted inodes can't handle collapse range or insert + * range since we would need to re-encrypt blocks with a + * different IV or XTS tweak (which are based on the logical + * block number). + */ + if (IS_ENCRYPTED(inode) && + (mode & (FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_INSERT_RANGE))) + return -EOPNOTSUPP; + /* + * Don't allow writing zeroes if the underlying device does not + * enable the unmap write zeroes operation. + */ + if ((mode & FALLOC_FL_WRITE_ZEROES) && + !bdev_write_zeroes_unmap_sectors(inode->i_sb->s_bdev)) + return -EOPNOTSUPP; + + /* Return error if mode is not supported */ + if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE | + FALLOC_FL_ZERO_RANGE | FALLOC_FL_COLLAPSE_RANGE | + FALLOC_FL_INSERT_RANGE | FALLOC_FL_WRITE_ZEROES)) + return -EOPNOTSUPP; + + inode_lock(inode); + ret = ext4_convert_inline_data(inode); + if (ret) + goto out_inode_lock; + + /* Wait all existing dio workers, newcomers will block on i_rwsem */ + inode_dio_wait(inode); + + ret = file_modified(file); + if (ret) + goto out_inode_lock; + + if ((mode & FALLOC_FL_MODE_MASK) == FALLOC_FL_ALLOCATE_RANGE) { + ret = ext4_do_fallocate(file, offset, len, mode); + goto out_inode_lock; + } + + /* + * Follow-up operations will drop page cache, hold invalidate lock + * to prevent page faults from reinstantiating pages we have + * released from page cache. + */ + filemap_invalidate_lock(mapping); + + ret = ext4_break_layouts(inode); + if (ret) + goto out_invalidate_lock; + + switch (mode & FALLOC_FL_MODE_MASK) { + case FALLOC_FL_PUNCH_HOLE: + ret = ext4_punch_hole(file, offset, len); + break; + case FALLOC_FL_COLLAPSE_RANGE: + ret = ext4_collapse_range(file, offset, len); + break; + case FALLOC_FL_INSERT_RANGE: + ret = ext4_insert_range(file, offset, len); + break; + case FALLOC_FL_ZERO_RANGE: + case FALLOC_FL_WRITE_ZEROES: + ret = ext4_zero_range(file, offset, len, mode); + break; + default: + ret = -EOPNOTSUPP; + } + +out_invalidate_lock: + filemap_invalidate_unlock(mapping); +out_inode_lock: + inode_unlock(inode); + return ret; +} + +/* + * This function converts a range of blocks to written extents. The caller of + * this function will pass the start offset and the size. all unwritten extents + * within this range will be converted to written extents. + * + * This function is called from the direct IO end io call back function for + * atomic writes, to convert the unwritten extents after IO is completed. + * + * Note that the requirement for atomic writes is that all conversion should + * happen atomically in a single fs journal transaction. We mainly only allocate + * unwritten extents either on a hole on a pre-exiting unwritten extent range in + * ext4_map_blocks_atomic_write(). The only case where we can have multiple + * unwritten extents in a range [offset, offset+len) is when there is a split + * unwritten extent between two leaf nodes which was cached in extent status + * cache during ext4_iomap_alloc() time. That will allow + * ext4_map_blocks_atomic_write() to return the unwritten extent range w/o going + * into the slow path. That means we might need a loop for conversion of this + * unwritten extent split across leaf block within a single journal transaction. + * Split extents across leaf nodes is a rare case, but let's still handle that + * to meet the requirements of multi-fsblock atomic writes. + * + * Returns 0 on success. + */ +int ext4_convert_unwritten_extents_atomic(handle_t *handle, struct inode *inode, + loff_t offset, ssize_t len) +{ + unsigned int max_blocks; + int ret = 0, ret2 = 0, ret3 = 0; + struct ext4_map_blocks map; + unsigned int blkbits = inode->i_blkbits; + unsigned int credits = 0; + int flags = EXT4_GET_BLOCKS_IO_CONVERT_EXT | EXT4_EX_NOCACHE; + + map.m_lblk = offset >> blkbits; + max_blocks = EXT4_MAX_BLOCKS(len, offset, blkbits); + + if (!handle) { + /* + * TODO: An optimization can be added later by having an extent + * status flag e.g. EXTENT_STATUS_SPLIT_LEAF. If we query that + * it can tell if the extent in the cache is a split extent. + * But for now let's assume pextents as 2 always. + */ + credits = ext4_meta_trans_blocks(inode, max_blocks, 2); + } + + if (credits) { + handle = ext4_journal_start(inode, EXT4_HT_MAP_BLOCKS, credits); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + return ret; + } + } + + while (ret >= 0 && ret < max_blocks) { + map.m_lblk += ret; + map.m_len = (max_blocks -= ret); + ret = ext4_map_blocks(handle, inode, &map, flags); + if (ret != max_blocks) + ext4_msg(inode->i_sb, KERN_INFO, + "inode #%lu: block %u: len %u: " + "split block mapping found for atomic write, " + "ret = %d", + inode->i_ino, map.m_lblk, + map.m_len, ret); + if (ret <= 0) + break; + } + + ret2 = ext4_mark_inode_dirty(handle, inode); + + if (credits) { + ret3 = ext4_journal_stop(handle); + if (unlikely(ret3)) + ret2 = ret3; + } + + if (ret <= 0 || ret2) + ext4_warning(inode->i_sb, + "inode #%lu: block %u: len %u: " + "returned %d or %d", + inode->i_ino, map.m_lblk, + map.m_len, ret, ret2); + + return ret > 0 ? ret2 : ret; +} + +/* + * This function convert a range of blocks to written extents + * The caller of this function will pass the start offset and the size. + * all unwritten extents within this range will be converted to + * written extents. + * + * This function is called from the direct IO end io call back + * function, to convert the fallocated extents after IO is completed. + * Returns 0 on success. + */ +int ext4_convert_unwritten_extents(handle_t *handle, struct inode *inode, + loff_t offset, ssize_t len) +{ + unsigned int max_blocks; + int ret = 0, ret2 = 0, ret3 = 0; + struct ext4_map_blocks map; + unsigned int blkbits = inode->i_blkbits; + unsigned int credits = 0; + + map.m_lblk = offset >> blkbits; + max_blocks = EXT4_MAX_BLOCKS(len, offset, blkbits); + + if (!handle) { + /* + * credits to insert 1 extent into extent tree + */ + credits = ext4_chunk_trans_blocks(inode, max_blocks); + } + while (ret >= 0 && ret < max_blocks) { + map.m_lblk += ret; + map.m_len = (max_blocks -= ret); + if (credits) { + handle = ext4_journal_start(inode, EXT4_HT_MAP_BLOCKS, + credits); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + break; + } + } + /* + * Do not cache any unrelated extents, as it does not hold the + * i_rwsem or invalidate_lock, which could corrupt the extent + * status tree. + */ + ret = ext4_map_blocks(handle, inode, &map, + EXT4_GET_BLOCKS_IO_CONVERT_EXT | + EXT4_EX_NOCACHE); + if (ret <= 0) + ext4_warning(inode->i_sb, + "inode #%lu: block %u: len %u: " + "ext4_ext_map_blocks returned %d", + inode->i_ino, map.m_lblk, + map.m_len, ret); + ret2 = ext4_mark_inode_dirty(handle, inode); + if (credits) { + ret3 = ext4_journal_stop(handle); + if (unlikely(ret3)) + ret2 = ret3; + } + + if (ret <= 0 || ret2) + break; + } + return ret > 0 ? ret2 : ret; +} + +int ext4_convert_unwritten_io_end_vec(handle_t *handle, ext4_io_end_t *io_end) +{ + int ret = 0, err = 0; + struct ext4_io_end_vec *io_end_vec; + + /* + * This is somewhat ugly but the idea is clear: When transaction is + * reserved, everything goes into it. Otherwise we rather start several + * smaller transactions for conversion of each extent separately. + */ + if (handle) { + handle = ext4_journal_start_reserved(handle, + EXT4_HT_EXT_CONVERT); + if (IS_ERR(handle)) + return PTR_ERR(handle); + } + + list_for_each_entry(io_end_vec, &io_end->list_vec, list) { + ret = ext4_convert_unwritten_extents(handle, io_end->inode, + io_end_vec->offset, + io_end_vec->size); + if (ret) + break; + } + + if (handle) + err = ext4_journal_stop(handle); + + return ret < 0 ? ret : err; +} + +static int ext4_iomap_xattr_fiemap(struct inode *inode, struct iomap *iomap) +{ + __u64 physical = 0; + __u64 length = 0; + int blockbits = inode->i_sb->s_blocksize_bits; + int error = 0; + u16 iomap_type; + + /* in-inode? */ + if (ext4_test_inode_state(inode, EXT4_STATE_XATTR)) { + struct ext4_iloc iloc; + int offset; /* offset of xattr in inode */ + + error = ext4_get_inode_loc(inode, &iloc); + if (error) + return error; + physical = (__u64)iloc.bh->b_blocknr << blockbits; + offset = EXT4_GOOD_OLD_INODE_SIZE + + EXT4_I(inode)->i_extra_isize; + physical += offset; + length = EXT4_SB(inode->i_sb)->s_inode_size - offset; + brelse(iloc.bh); + iomap_type = IOMAP_INLINE; + } else if (EXT4_I(inode)->i_file_acl) { /* external block */ + physical = (__u64)EXT4_I(inode)->i_file_acl << blockbits; + length = inode->i_sb->s_blocksize; + iomap_type = IOMAP_MAPPED; + } else { + /* no in-inode or external block for xattr, so return -ENOENT */ + error = -ENOENT; + goto out; + } + + iomap->addr = physical; + iomap->offset = 0; + iomap->length = length; + iomap->type = iomap_type; + iomap->flags = 0; +out: + return error; +} + +static int ext4_iomap_xattr_begin(struct inode *inode, loff_t offset, + loff_t length, unsigned flags, + struct iomap *iomap, struct iomap *srcmap) +{ + int error; + + error = ext4_iomap_xattr_fiemap(inode, iomap); + if (error == 0 && (offset >= iomap->length)) + error = -ENOENT; + return error; +} + +static const struct iomap_ops ext4_iomap_xattr_ops = { + .iomap_begin = ext4_iomap_xattr_begin, +}; + +static int ext4_fiemap_check_ranges(struct inode *inode, u64 start, u64 *len) +{ + u64 maxbytes = ext4_get_maxbytes(inode); + + if (*len == 0) + return -EINVAL; + if (start > maxbytes) + return -EFBIG; + + /* + * Shrink request scope to what the fs can actually handle. + */ + if (*len > maxbytes || (maxbytes - *len) < start) + *len = maxbytes - start; + return 0; +} + +int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, + u64 start, u64 len) +{ + int error = 0; + + inode_lock_shared(inode); + if (fieinfo->fi_flags & FIEMAP_FLAG_CACHE) { + error = ext4_ext_precache(inode); + if (error) + goto unlock; + fieinfo->fi_flags &= ~FIEMAP_FLAG_CACHE; + } + + /* + * For bitmap files the maximum size limit could be smaller than + * s_maxbytes, so check len here manually instead of just relying on the + * generic check. + */ + error = ext4_fiemap_check_ranges(inode, start, &len); + if (error) + goto unlock; + + if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR) { + fieinfo->fi_flags &= ~FIEMAP_FLAG_XATTR; + error = iomap_fiemap(inode, fieinfo, start, len, + &ext4_iomap_xattr_ops); + } else { + error = iomap_fiemap(inode, fieinfo, start, len, + &ext4_iomap_report_ops); + } +unlock: + inode_unlock_shared(inode); + return error; +} + +int ext4_get_es_cache(struct inode *inode, struct fiemap_extent_info *fieinfo, + __u64 start, __u64 len) +{ + ext4_lblk_t start_blk, len_blks; + __u64 last_blk; + int error = 0; + + if (ext4_has_inline_data(inode)) { + int has_inline; + + down_read(&EXT4_I(inode)->xattr_sem); + has_inline = ext4_has_inline_data(inode); + up_read(&EXT4_I(inode)->xattr_sem); + if (has_inline) + return 0; + } + + if (fieinfo->fi_flags & FIEMAP_FLAG_CACHE) { + inode_lock_shared(inode); + error = ext4_ext_precache(inode); + inode_unlock_shared(inode); + if (error) + return error; + fieinfo->fi_flags &= ~FIEMAP_FLAG_CACHE; + } + + error = fiemap_prep(inode, fieinfo, start, &len, 0); + if (error) + return error; + + error = ext4_fiemap_check_ranges(inode, start, &len); + if (error) + return error; + + start_blk = start >> inode->i_sb->s_blocksize_bits; + last_blk = (start + len - 1) >> inode->i_sb->s_blocksize_bits; + if (last_blk >= EXT_MAX_BLOCKS) + last_blk = EXT_MAX_BLOCKS-1; + len_blks = ((ext4_lblk_t) last_blk) - start_blk + 1; + + /* + * Walk the extent tree gathering extent information + * and pushing extents back to the user. + */ + return ext4_fill_es_cache_info(inode, start_blk, len_blks, fieinfo); +} + +/* + * ext4_ext_shift_path_extents: + * Shift the extents of a path structure lying between path[depth].p_ext + * and EXT_LAST_EXTENT(path[depth].p_hdr), by @shift blocks. @SHIFT tells + * if it is right shift or left shift operation. + */ +static int +ext4_ext_shift_path_extents(struct ext4_ext_path *path, ext4_lblk_t shift, + struct inode *inode, handle_t *handle, + enum SHIFT_DIRECTION SHIFT) +{ + int depth, err = 0; + struct ext4_extent *ex_start, *ex_last; + bool update = false; + int credits, restart_credits; + depth = path->p_depth; + + while (depth >= 0) { + if (depth == path->p_depth) { + ex_start = path[depth].p_ext; + if (!ex_start) + return -EFSCORRUPTED; + + ex_last = EXT_LAST_EXTENT(path[depth].p_hdr); + /* leaf + sb + inode */ + credits = 3; + if (ex_start == EXT_FIRST_EXTENT(path[depth].p_hdr)) { + update = true; + /* extent tree + sb + inode */ + credits = depth + 2; + } + + restart_credits = ext4_chunk_trans_extent(inode, 0); + err = ext4_datasem_ensure_credits(handle, inode, credits, + restart_credits, 0); + if (err) { + if (err > 0) + err = -EAGAIN; + goto out; + } + + err = ext4_ext_get_access(handle, inode, path + depth); + if (err) + goto out; + + while (ex_start <= ex_last) { + if (SHIFT == SHIFT_LEFT) { + le32_add_cpu(&ex_start->ee_block, + -shift); + /* Try to merge to the left. */ + if ((ex_start > + EXT_FIRST_EXTENT(path[depth].p_hdr)) + && + ext4_ext_try_to_merge_right(inode, + path, ex_start - 1)) + ex_last--; + else + ex_start++; + } else { + le32_add_cpu(&ex_last->ee_block, shift); + ext4_ext_try_to_merge_right(inode, path, + ex_last); + ex_last--; + } + } + err = ext4_ext_dirty(handle, inode, path + depth); + if (err) + goto out; + + if (--depth < 0 || !update) + break; + } + + /* Update index too */ + err = ext4_ext_get_access(handle, inode, path + depth); + if (err) + goto out; + + if (SHIFT == SHIFT_LEFT) + le32_add_cpu(&path[depth].p_idx->ei_block, -shift); + else + le32_add_cpu(&path[depth].p_idx->ei_block, shift); + err = ext4_ext_dirty(handle, inode, path + depth); + if (err) + goto out; + + /* we are done if current index is not a starting index */ + if (path[depth].p_idx != EXT_FIRST_INDEX(path[depth].p_hdr)) + break; + + depth--; + } + +out: + return err; +} + +/* + * ext4_ext_shift_extents: + * All the extents which lies in the range from @start to the last allocated + * block for the @inode are shifted either towards left or right (depending + * upon @SHIFT) by @shift blocks. + * On success, 0 is returned, error otherwise. + */ +static int +ext4_ext_shift_extents(struct inode *inode, handle_t *handle, + ext4_lblk_t start, ext4_lblk_t shift, + enum SHIFT_DIRECTION SHIFT) +{ + struct ext4_ext_path *path; + int ret = 0, depth; + struct ext4_extent *extent; + ext4_lblk_t stop, *iterator, ex_start, ex_end; + ext4_lblk_t tmp = EXT_MAX_BLOCKS; + + /* Let path point to the last extent */ + path = ext4_find_extent(inode, EXT_MAX_BLOCKS - 1, NULL, + EXT4_EX_NOCACHE); + if (IS_ERR(path)) + return PTR_ERR(path); + + depth = path->p_depth; + extent = path[depth].p_ext; + if (!extent) + goto out; + + stop = le32_to_cpu(extent->ee_block); + + /* + * For left shifts, make sure the hole on the left is big enough to + * accommodate the shift. For right shifts, make sure the last extent + * won't be shifted beyond EXT_MAX_BLOCKS. + */ + if (SHIFT == SHIFT_LEFT) { + path = ext4_find_extent(inode, start - 1, path, + EXT4_EX_NOCACHE); + if (IS_ERR(path)) + return PTR_ERR(path); + depth = path->p_depth; + extent = path[depth].p_ext; + if (extent) { + ex_start = le32_to_cpu(extent->ee_block); + ex_end = le32_to_cpu(extent->ee_block) + + ext4_ext_get_actual_len(extent); + } else { + ex_start = 0; + ex_end = 0; + } + + if ((start == ex_start && shift > ex_start) || + (shift > start - ex_end)) { + ret = -EINVAL; + goto out; + } + } else { + if (shift > EXT_MAX_BLOCKS - + (stop + ext4_ext_get_actual_len(extent))) { + ret = -EINVAL; + goto out; + } + } + + /* + * In case of left shift, iterator points to start and it is increased + * till we reach stop. In case of right shift, iterator points to stop + * and it is decreased till we reach start. + */ +again: + ret = 0; + if (SHIFT == SHIFT_LEFT) + iterator = &start; + else + iterator = &stop; + + if (tmp != EXT_MAX_BLOCKS) + *iterator = tmp; + + /* + * Its safe to start updating extents. Start and stop are unsigned, so + * in case of right shift if extent with 0 block is reached, iterator + * becomes NULL to indicate the end of the loop. + */ + while (iterator && start <= stop) { + path = ext4_find_extent(inode, *iterator, path, + EXT4_EX_NOCACHE); + if (IS_ERR(path)) + return PTR_ERR(path); + depth = path->p_depth; + extent = path[depth].p_ext; + if (!extent) { + EXT4_ERROR_INODE(inode, "unexpected hole at %lu", + (unsigned long) *iterator); + return -EFSCORRUPTED; + } + if (SHIFT == SHIFT_LEFT && *iterator > + le32_to_cpu(extent->ee_block)) { + /* Hole, move to the next extent */ + if (extent < EXT_LAST_EXTENT(path[depth].p_hdr)) { + path[depth].p_ext++; + } else { + *iterator = ext4_ext_next_allocated_block(path); + continue; + } + } + + tmp = *iterator; + if (SHIFT == SHIFT_LEFT) { + extent = EXT_LAST_EXTENT(path[depth].p_hdr); + *iterator = le32_to_cpu(extent->ee_block) + + ext4_ext_get_actual_len(extent); + } else { + extent = EXT_FIRST_EXTENT(path[depth].p_hdr); + if (le32_to_cpu(extent->ee_block) > start) + *iterator = le32_to_cpu(extent->ee_block) - 1; + else if (le32_to_cpu(extent->ee_block) == start) + iterator = NULL; + else { + extent = EXT_LAST_EXTENT(path[depth].p_hdr); + while (le32_to_cpu(extent->ee_block) >= start) + extent--; + + if (extent == EXT_LAST_EXTENT(path[depth].p_hdr)) + break; + + extent++; + iterator = NULL; + } + path[depth].p_ext = extent; + } + ret = ext4_ext_shift_path_extents(path, shift, inode, + handle, SHIFT); + /* iterator can be NULL which means we should break */ + if (ret == -EAGAIN) + goto again; + if (ret) + break; + } +out: + ext4_free_ext_path(path); + return ret; +} + +/* + * ext4_collapse_range: + * This implements the fallocate's collapse range functionality for ext4 + * Returns: 0 and non-zero on error. + */ +static int ext4_collapse_range(struct file *file, loff_t offset, loff_t len) +{ + struct inode *inode = file_inode(file); + struct super_block *sb = inode->i_sb; + struct address_space *mapping = inode->i_mapping; + loff_t end = offset + len; + ext4_lblk_t start_lblk, end_lblk; + handle_t *handle; + unsigned int credits; + loff_t start, new_size; + int ret; + + trace_ext4_collapse_range(inode, offset, len); + WARN_ON_ONCE(!inode_is_locked(inode)); + + /* Currently just for extent based files */ + if (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) + return -EOPNOTSUPP; + /* Collapse range works only on fs cluster size aligned regions. */ + if (!IS_ALIGNED(offset | len, EXT4_CLUSTER_SIZE(sb))) + return -EINVAL; + /* + * There is no need to overlap collapse range with EOF, in which case + * it is effectively a truncate operation + */ + if (end >= inode->i_size) + return -EINVAL; + + /* + * Write tail of the last page before removed range and data that + * will be shifted since they will get removed from the page cache + * below. We are also protected from pages becoming dirty by + * i_rwsem and invalidate_lock. + * Need to round down offset to be aligned with page size boundary + * for page size > block size. + */ + start = round_down(offset, PAGE_SIZE); + ret = filemap_write_and_wait_range(mapping, start, offset); + if (!ret) + ret = filemap_write_and_wait_range(mapping, end, LLONG_MAX); + if (ret) + return ret; + + truncate_pagecache(inode, start); + + credits = ext4_chunk_trans_extent(inode, 0); + handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, credits); + if (IS_ERR(handle)) + return PTR_ERR(handle); + + ext4_fc_mark_ineligible(sb, EXT4_FC_REASON_FALLOC_RANGE, handle); + + start_lblk = offset >> inode->i_blkbits; + end_lblk = (offset + len) >> inode->i_blkbits; + + ext4_check_map_extents_env(inode); + + down_write(&EXT4_I(inode)->i_data_sem); + ext4_discard_preallocations(inode); + ext4_es_remove_extent(inode, start_lblk, EXT_MAX_BLOCKS - start_lblk); + + ret = ext4_ext_remove_space(inode, start_lblk, end_lblk - 1); + if (ret) { + up_write(&EXT4_I(inode)->i_data_sem); + goto out_handle; + } + ext4_discard_preallocations(inode); + + ret = ext4_ext_shift_extents(inode, handle, end_lblk, + end_lblk - start_lblk, SHIFT_LEFT); + if (ret) { + up_write(&EXT4_I(inode)->i_data_sem); + goto out_handle; + } + + new_size = inode->i_size - len; + i_size_write(inode, new_size); + EXT4_I(inode)->i_disksize = new_size; + + up_write(&EXT4_I(inode)->i_data_sem); + ret = ext4_mark_inode_dirty(handle, inode); + if (ret) + goto out_handle; + + ext4_update_inode_fsync_trans(handle, inode, 1); + if (IS_SYNC(inode)) + ext4_handle_sync(handle); + +out_handle: + ext4_journal_stop(handle); + return ret; +} + +/* + * ext4_insert_range: + * This function implements the FALLOC_FL_INSERT_RANGE flag of fallocate. + * The data blocks starting from @offset to the EOF are shifted by @len + * towards right to create a hole in the @inode. Inode size is increased + * by len bytes. + * Returns 0 on success, error otherwise. + */ +static int ext4_insert_range(struct file *file, loff_t offset, loff_t len) +{ + struct inode *inode = file_inode(file); + struct super_block *sb = inode->i_sb; + struct address_space *mapping = inode->i_mapping; + handle_t *handle; + struct ext4_ext_path *path; + struct ext4_extent *extent; + ext4_lblk_t start_lblk, len_lblk, ee_start_lblk = 0; + unsigned int credits, ee_len; + int ret, depth, split_flag = 0; + loff_t start; + + trace_ext4_insert_range(inode, offset, len); + WARN_ON_ONCE(!inode_is_locked(inode)); + + /* Currently just for extent based files */ + if (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) + return -EOPNOTSUPP; + /* Insert range works only on fs cluster size aligned regions. */ + if (!IS_ALIGNED(offset | len, EXT4_CLUSTER_SIZE(sb))) + return -EINVAL; + /* Offset must be less than i_size */ + if (offset >= inode->i_size) + return -EINVAL; + /* Check whether the maximum file size would be exceeded */ + if (len > inode->i_sb->s_maxbytes - inode->i_size) + return -EFBIG; + + /* + * Write out all dirty pages. Need to round down to align start offset + * to page size boundary for page size > block size. + */ + start = round_down(offset, PAGE_SIZE); + ret = filemap_write_and_wait_range(mapping, start, LLONG_MAX); + if (ret) + return ret; + + truncate_pagecache(inode, start); + + credits = ext4_chunk_trans_extent(inode, 0); + handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, credits); + if (IS_ERR(handle)) + return PTR_ERR(handle); + + ext4_fc_mark_ineligible(sb, EXT4_FC_REASON_FALLOC_RANGE, handle); + + /* Expand file to avoid data loss if there is error while shifting */ + inode->i_size += len; + EXT4_I(inode)->i_disksize += len; + ret = ext4_mark_inode_dirty(handle, inode); + if (ret) + goto out_handle; + + start_lblk = offset >> inode->i_blkbits; + len_lblk = len >> inode->i_blkbits; + + ext4_check_map_extents_env(inode); + + down_write(&EXT4_I(inode)->i_data_sem); + ext4_discard_preallocations(inode); + + path = ext4_find_extent(inode, start_lblk, NULL, 0); + if (IS_ERR(path)) { + up_write(&EXT4_I(inode)->i_data_sem); + ret = PTR_ERR(path); + goto out_handle; + } + + depth = ext_depth(inode); + extent = path[depth].p_ext; + if (extent) { + ee_start_lblk = le32_to_cpu(extent->ee_block); + ee_len = ext4_ext_get_actual_len(extent); + + /* + * If start_lblk is not the starting block of extent, split + * the extent @start_lblk + */ + if ((start_lblk > ee_start_lblk) && + (start_lblk < (ee_start_lblk + ee_len))) { + if (ext4_ext_is_unwritten(extent)) + split_flag = EXT4_EXT_MARK_UNWRIT1 | + EXT4_EXT_MARK_UNWRIT2; + path = ext4_split_extent_at(handle, inode, path, + start_lblk, split_flag, + EXT4_EX_NOCACHE | + EXT4_GET_BLOCKS_PRE_IO | + EXT4_GET_BLOCKS_METADATA_NOFAIL); + } + + if (IS_ERR(path)) { + up_write(&EXT4_I(inode)->i_data_sem); + ret = PTR_ERR(path); + goto out_handle; + } + } + + ext4_free_ext_path(path); + ext4_es_remove_extent(inode, start_lblk, EXT_MAX_BLOCKS - start_lblk); + + /* + * if start_lblk lies in a hole which is at start of file, use + * ee_start_lblk to shift extents + */ + ret = ext4_ext_shift_extents(inode, handle, + max(ee_start_lblk, start_lblk), len_lblk, SHIFT_RIGHT); + up_write(&EXT4_I(inode)->i_data_sem); + if (ret) + goto out_handle; + + ext4_update_inode_fsync_trans(handle, inode, 1); + if (IS_SYNC(inode)) + ext4_handle_sync(handle); + +out_handle: + ext4_journal_stop(handle); + return ret; +} + +/** + * ext4_swap_extents() - Swap extents between two inodes + * @handle: handle for this transaction + * @inode1: First inode + * @inode2: Second inode + * @lblk1: Start block for first inode + * @lblk2: Start block for second inode + * @count: Number of blocks to swap + * @unwritten: Mark second inode's extents as unwritten after swap + * @erp: Pointer to save error value + * + * This helper routine does exactly what is promise "swap extents". All other + * stuff such as page-cache locking consistency, bh mapping consistency or + * extent's data copying must be performed by caller. + * Locking: + * i_rwsem is held for both inodes + * i_data_sem is locked for write for both inodes + * Assumptions: + * All pages from requested range are locked for both inodes + */ +int +ext4_swap_extents(handle_t *handle, struct inode *inode1, + struct inode *inode2, ext4_lblk_t lblk1, ext4_lblk_t lblk2, + ext4_lblk_t count, int unwritten, int *erp) +{ + struct ext4_ext_path *path1 = NULL; + struct ext4_ext_path *path2 = NULL; + int replaced_count = 0; + + BUG_ON(!rwsem_is_locked(&EXT4_I(inode1)->i_data_sem)); + BUG_ON(!rwsem_is_locked(&EXT4_I(inode2)->i_data_sem)); + BUG_ON(!inode_is_locked(inode1)); + BUG_ON(!inode_is_locked(inode2)); + + ext4_es_remove_extent(inode1, lblk1, count); + ext4_es_remove_extent(inode2, lblk2, count); + + while (count) { + struct ext4_extent *ex1, *ex2, tmp_ex; + ext4_lblk_t e1_blk, e2_blk; + int e1_len, e2_len, len; + int split = 0; + + path1 = ext4_find_extent(inode1, lblk1, path1, EXT4_EX_NOCACHE); + if (IS_ERR(path1)) { + *erp = PTR_ERR(path1); + goto errout; + } + path2 = ext4_find_extent(inode2, lblk2, path2, EXT4_EX_NOCACHE); + if (IS_ERR(path2)) { + *erp = PTR_ERR(path2); + goto errout; + } + ex1 = path1[path1->p_depth].p_ext; + ex2 = path2[path2->p_depth].p_ext; + /* Do we have something to swap ? */ + if (unlikely(!ex2 || !ex1)) + goto errout; + + e1_blk = le32_to_cpu(ex1->ee_block); + e2_blk = le32_to_cpu(ex2->ee_block); + e1_len = ext4_ext_get_actual_len(ex1); + e2_len = ext4_ext_get_actual_len(ex2); + + /* Hole handling */ + if (!in_range(lblk1, e1_blk, e1_len) || + !in_range(lblk2, e2_blk, e2_len)) { + ext4_lblk_t next1, next2; + + /* if hole after extent, then go to next extent */ + next1 = ext4_ext_next_allocated_block(path1); + next2 = ext4_ext_next_allocated_block(path2); + /* If hole before extent, then shift to that extent */ + if (e1_blk > lblk1) + next1 = e1_blk; + if (e2_blk > lblk2) + next2 = e2_blk; + /* Do we have something to swap */ + if (next1 == EXT_MAX_BLOCKS || next2 == EXT_MAX_BLOCKS) + goto errout; + /* Move to the rightest boundary */ + len = next1 - lblk1; + if (len < next2 - lblk2) + len = next2 - lblk2; + if (len > count) + len = count; + lblk1 += len; + lblk2 += len; + count -= len; + continue; + } + + /* Prepare left boundary */ + if (e1_blk < lblk1) { + split = 1; + path1 = ext4_force_split_extent_at(handle, inode1, + path1, lblk1, 0); + if (IS_ERR(path1)) { + *erp = PTR_ERR(path1); + goto errout; + } + } + if (e2_blk < lblk2) { + split = 1; + path2 = ext4_force_split_extent_at(handle, inode2, + path2, lblk2, 0); + if (IS_ERR(path2)) { + *erp = PTR_ERR(path2); + goto errout; + } + } + /* ext4_split_extent_at() may result in leaf extent split, + * path must to be revalidated. */ + if (split) + continue; + + /* Prepare right boundary */ + len = count; + if (len > e1_blk + e1_len - lblk1) + len = e1_blk + e1_len - lblk1; + if (len > e2_blk + e2_len - lblk2) + len = e2_blk + e2_len - lblk2; + + if (len != e1_len) { + split = 1; + path1 = ext4_force_split_extent_at(handle, inode1, + path1, lblk1 + len, 0); + if (IS_ERR(path1)) { + *erp = PTR_ERR(path1); + goto errout; + } + } + if (len != e2_len) { + split = 1; + path2 = ext4_force_split_extent_at(handle, inode2, + path2, lblk2 + len, 0); + if (IS_ERR(path2)) { + *erp = PTR_ERR(path2); + goto errout; + } + } + /* ext4_split_extent_at() may result in leaf extent split, + * path must to be revalidated. */ + if (split) + continue; + + BUG_ON(e2_len != e1_len); + *erp = ext4_ext_get_access(handle, inode1, path1 + path1->p_depth); + if (unlikely(*erp)) + goto errout; + *erp = ext4_ext_get_access(handle, inode2, path2 + path2->p_depth); + if (unlikely(*erp)) + goto errout; + + /* Both extents are fully inside boundaries. Swap it now */ + tmp_ex = *ex1; + ext4_ext_store_pblock(ex1, ext4_ext_pblock(ex2)); + ext4_ext_store_pblock(ex2, ext4_ext_pblock(&tmp_ex)); + ex1->ee_len = cpu_to_le16(e2_len); + ex2->ee_len = cpu_to_le16(e1_len); + if (unwritten) + ext4_ext_mark_unwritten(ex2); + if (ext4_ext_is_unwritten(&tmp_ex)) + ext4_ext_mark_unwritten(ex1); + + ext4_ext_try_to_merge(handle, inode2, path2, ex2); + ext4_ext_try_to_merge(handle, inode1, path1, ex1); + *erp = ext4_ext_dirty(handle, inode2, path2 + + path2->p_depth); + if (unlikely(*erp)) + goto errout; + *erp = ext4_ext_dirty(handle, inode1, path1 + + path1->p_depth); + /* + * Looks scarry ah..? second inode already points to new blocks, + * and it was successfully dirtied. But luckily error may happen + * only due to journal error, so full transaction will be + * aborted anyway. + */ + if (unlikely(*erp)) + goto errout; + + lblk1 += len; + lblk2 += len; + replaced_count += len; + count -= len; + } + +errout: + ext4_free_ext_path(path1); + ext4_free_ext_path(path2); + return replaced_count; +} + +/* + * ext4_clu_mapped - determine whether any block in a logical cluster has + * been mapped to a physical cluster + * + * @inode - file containing the logical cluster + * @lclu - logical cluster of interest + * + * Returns 1 if any block in the logical cluster is mapped, signifying + * that a physical cluster has been allocated for it. Otherwise, + * returns 0. Can also return negative error codes. Derived from + * ext4_ext_map_blocks(). + */ +int ext4_clu_mapped(struct inode *inode, ext4_lblk_t lclu) +{ + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + struct ext4_ext_path *path; + int depth, mapped = 0, err = 0; + struct ext4_extent *extent; + ext4_lblk_t first_lblk, first_lclu, last_lclu; + + /* + * if data can be stored inline, the logical cluster isn't + * mapped - no physical clusters have been allocated, and the + * file has no extents + */ + if (ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA) || + ext4_has_inline_data(inode)) + return 0; + + /* search for the extent closest to the first block in the cluster */ + path = ext4_find_extent(inode, EXT4_C2B(sbi, lclu), NULL, 0); + if (IS_ERR(path)) + return PTR_ERR(path); + + depth = ext_depth(inode); + + /* + * A consistent leaf must not be empty. This situation is possible, + * though, _during_ tree modification, and it's why an assert can't + * be put in ext4_find_extent(). + */ + if (unlikely(path[depth].p_ext == NULL && depth != 0)) { + EXT4_ERROR_INODE(inode, + "bad extent address - lblock: %lu, depth: %d, pblock: %lld", + (unsigned long) EXT4_C2B(sbi, lclu), + depth, path[depth].p_block); + err = -EFSCORRUPTED; + goto out; + } + + extent = path[depth].p_ext; + + /* can't be mapped if the extent tree is empty */ + if (extent == NULL) + goto out; + + first_lblk = le32_to_cpu(extent->ee_block); + first_lclu = EXT4_B2C(sbi, first_lblk); + + /* + * Three possible outcomes at this point - found extent spanning + * the target cluster, to the left of the target cluster, or to the + * right of the target cluster. The first two cases are handled here. + * The last case indicates the target cluster is not mapped. + */ + if (lclu >= first_lclu) { + last_lclu = EXT4_B2C(sbi, first_lblk + + ext4_ext_get_actual_len(extent) - 1); + if (lclu <= last_lclu) { + mapped = 1; + } else { + first_lblk = ext4_ext_next_allocated_block(path); + first_lclu = EXT4_B2C(sbi, first_lblk); + if (lclu == first_lclu) + mapped = 1; + } + } + +out: + ext4_free_ext_path(path); + + return err ? err : mapped; +} + +/* + * Updates physical block address and unwritten status of extent + * starting at lblk start and of len. If such an extent doesn't exist, + * this function splits the extent tree appropriately to create an + * extent like this. This function is called in the fast commit + * replay path. Returns 0 on success and error on failure. + */ +int ext4_ext_replay_update_ex(struct inode *inode, ext4_lblk_t start, + int len, int unwritten, ext4_fsblk_t pblk) +{ + struct ext4_ext_path *path; + struct ext4_extent *ex; + int ret; + + path = ext4_find_extent(inode, start, NULL, 0); + if (IS_ERR(path)) + return PTR_ERR(path); + ex = path[path->p_depth].p_ext; + if (!ex) { + ret = -EFSCORRUPTED; + goto out; + } + + if (le32_to_cpu(ex->ee_block) != start || + ext4_ext_get_actual_len(ex) != len) { + /* We need to split this extent to match our extent first */ + down_write(&EXT4_I(inode)->i_data_sem); + path = ext4_force_split_extent_at(NULL, inode, path, start, 1); + up_write(&EXT4_I(inode)->i_data_sem); + if (IS_ERR(path)) { + ret = PTR_ERR(path); + goto out; + } + + path = ext4_find_extent(inode, start, path, 0); + if (IS_ERR(path)) + return PTR_ERR(path); + + ex = path[path->p_depth].p_ext; + WARN_ON(le32_to_cpu(ex->ee_block) != start); + + if (ext4_ext_get_actual_len(ex) != len) { + down_write(&EXT4_I(inode)->i_data_sem); + path = ext4_force_split_extent_at(NULL, inode, path, + start + len, 1); + up_write(&EXT4_I(inode)->i_data_sem); + if (IS_ERR(path)) { + ret = PTR_ERR(path); + goto out; + } + + path = ext4_find_extent(inode, start, path, 0); + if (IS_ERR(path)) + return PTR_ERR(path); + ex = path[path->p_depth].p_ext; + } + } + if (unwritten) + ext4_ext_mark_unwritten(ex); + else + ext4_ext_mark_initialized(ex); + ext4_ext_store_pblock(ex, pblk); + down_write(&EXT4_I(inode)->i_data_sem); + ret = ext4_ext_dirty(NULL, inode, &path[path->p_depth]); + up_write(&EXT4_I(inode)->i_data_sem); +out: + ext4_free_ext_path(path); + ext4_mark_inode_dirty(NULL, inode); + return ret; +} + +/* Try to shrink the extent tree */ +void ext4_ext_replay_shrink_inode(struct inode *inode, ext4_lblk_t end) +{ + struct ext4_ext_path *path = NULL; + struct ext4_extent *ex; + ext4_lblk_t old_cur, cur = 0; + + while (cur < end) { + path = ext4_find_extent(inode, cur, NULL, 0); + if (IS_ERR(path)) + return; + ex = path[path->p_depth].p_ext; + if (!ex) { + ext4_free_ext_path(path); + ext4_mark_inode_dirty(NULL, inode); + return; + } + old_cur = cur; + cur = le32_to_cpu(ex->ee_block) + ext4_ext_get_actual_len(ex); + if (cur <= old_cur) + cur = old_cur + 1; + ext4_ext_try_to_merge(NULL, inode, path, ex); + down_write(&EXT4_I(inode)->i_data_sem); + ext4_ext_dirty(NULL, inode, &path[path->p_depth]); + up_write(&EXT4_I(inode)->i_data_sem); + ext4_mark_inode_dirty(NULL, inode); + ext4_free_ext_path(path); + } +} + +/* Check if *cur is a hole and if it is, skip it */ +static int skip_hole(struct inode *inode, ext4_lblk_t *cur) +{ + int ret; + struct ext4_map_blocks map; + + map.m_lblk = *cur; + map.m_len = ((inode->i_size) >> inode->i_sb->s_blocksize_bits) - *cur; + + ret = ext4_map_blocks(NULL, inode, &map, 0); + if (ret < 0) + return ret; + if (ret != 0) + return 0; + *cur = *cur + map.m_len; + return 0; +} + +/* Count number of blocks used by this inode and update i_blocks */ +int ext4_ext_replay_set_iblocks(struct inode *inode) +{ + struct ext4_ext_path *path = NULL, *path2 = NULL; + struct ext4_extent *ex; + ext4_lblk_t cur = 0, end; + int numblks = 0, i, ret = 0; + ext4_fsblk_t cmp1, cmp2; + struct ext4_map_blocks map; + + /* Determin the size of the file first */ + path = ext4_find_extent(inode, EXT_MAX_BLOCKS - 1, NULL, + EXT4_EX_NOCACHE); + if (IS_ERR(path)) + return PTR_ERR(path); + ex = path[path->p_depth].p_ext; + if (!ex) + goto out; + end = le32_to_cpu(ex->ee_block) + ext4_ext_get_actual_len(ex); + + /* Count the number of data blocks */ + cur = 0; + while (cur < end) { + map.m_lblk = cur; + map.m_len = end - cur; + ret = ext4_map_blocks(NULL, inode, &map, 0); + if (ret < 0) + break; + if (ret > 0) + numblks += ret; + cur = cur + map.m_len; + } + + /* + * Count the number of extent tree blocks. We do it by looking up + * two successive extents and determining the difference between + * their paths. When path is different for 2 successive extents + * we compare the blocks in the path at each level and increment + * iblocks by total number of differences found. + */ + cur = 0; + ret = skip_hole(inode, &cur); + if (ret < 0) + goto out; + path = ext4_find_extent(inode, cur, path, 0); + if (IS_ERR(path)) + goto out; + numblks += path->p_depth; + while (cur < end) { + path = ext4_find_extent(inode, cur, path, 0); + if (IS_ERR(path)) + break; + ex = path[path->p_depth].p_ext; + if (!ex) + goto cleanup; + + cur = max(cur + 1, le32_to_cpu(ex->ee_block) + + ext4_ext_get_actual_len(ex)); + ret = skip_hole(inode, &cur); + if (ret < 0) + break; + + path2 = ext4_find_extent(inode, cur, path2, 0); + if (IS_ERR(path2)) + break; + + for (i = 0; i <= max(path->p_depth, path2->p_depth); i++) { + cmp1 = cmp2 = 0; + if (i <= path->p_depth) + cmp1 = path[i].p_bh ? + path[i].p_bh->b_blocknr : 0; + if (i <= path2->p_depth) + cmp2 = path2[i].p_bh ? + path2[i].p_bh->b_blocknr : 0; + if (cmp1 != cmp2 && cmp2 != 0) + numblks++; + } + } + +out: + inode->i_blocks = numblks << (inode->i_sb->s_blocksize_bits - 9); + ext4_mark_inode_dirty(NULL, inode); +cleanup: + ext4_free_ext_path(path); + ext4_free_ext_path(path2); + return 0; +} + +int ext4_ext_clear_bb(struct inode *inode) +{ + struct ext4_ext_path *path = NULL; + struct ext4_extent *ex; + ext4_lblk_t cur = 0, end; + int j, ret = 0; + struct ext4_map_blocks map; + + if (ext4_test_inode_flag(inode, EXT4_INODE_INLINE_DATA)) + return 0; + + /* Determin the size of the file first */ + path = ext4_find_extent(inode, EXT_MAX_BLOCKS - 1, NULL, + EXT4_EX_NOCACHE); + if (IS_ERR(path)) + return PTR_ERR(path); + ex = path[path->p_depth].p_ext; + if (!ex) + goto out; + end = le32_to_cpu(ex->ee_block) + ext4_ext_get_actual_len(ex); + + cur = 0; + while (cur < end) { + map.m_lblk = cur; + map.m_len = end - cur; + ret = ext4_map_blocks(NULL, inode, &map, 0); + if (ret < 0) + break; + if (ret > 0) { + path = ext4_find_extent(inode, map.m_lblk, path, 0); + if (!IS_ERR(path)) { + for (j = 0; j < path->p_depth; j++) { + ext4_mb_mark_bb(inode->i_sb, + path[j].p_block, 1, false); + ext4_fc_record_regions(inode->i_sb, inode->i_ino, + 0, path[j].p_block, 1, 1); + } + } else { + path = NULL; + } + ext4_mb_mark_bb(inode->i_sb, map.m_pblk, map.m_len, false); + ext4_fc_record_regions(inode->i_sb, inode->i_ino, + map.m_lblk, map.m_pblk, map.m_len, 1); + } + cur = cur + map.m_len; + } + +out: + ext4_free_ext_path(path); + return 0; +} -- 2.43.0

From: Simon Glass <simon.glass@canonical.com> Copy mballoc.c from Linux v6.18 fs/ext4 directory. This file implements the multiblock allocator for efficient block allocation. Co-developed-by: Claude Opus 4.5 <noreply@anthropic.com> --- fs/ext4l/mballoc.c | 7175 ++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 7175 insertions(+) create mode 100644 fs/ext4l/mballoc.c diff --git a/fs/ext4l/mballoc.c b/fs/ext4l/mballoc.c new file mode 100644 index 00000000000..9087183602e --- /dev/null +++ b/fs/ext4l/mballoc.c @@ -0,0 +1,7175 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright (c) 2003-2006, Cluster File Systems, Inc, info@clusterfs.com + * Written by Alex Tomas <alex@clusterfs.com> + */ + + +/* + * mballoc.c contains the multiblocks allocation routines + */ + +#include "ext4_jbd2.h" +#include "mballoc.h" +#include <linux/log2.h> +#include <linux/module.h> +#include <linux/slab.h> +#include <linux/nospec.h> +#include <linux/backing-dev.h> +#include <linux/freezer.h> +#include <trace/events/ext4.h> +#include <kunit/static_stub.h> + +/* + * MUSTDO: + * - test ext4_ext_search_left() and ext4_ext_search_right() + * - search for metadata in few groups + * + * TODO v4: + * - normalization should take into account whether file is still open + * - discard preallocations if no free space left (policy?) + * - don't normalize tails + * - quota + * - reservation for superuser + * + * TODO v3: + * - bitmap read-ahead (proposed by Oleg Drokin aka green) + * - track min/max extents in each group for better group selection + * - mb_mark_used() may allocate chunk right after splitting buddy + * - tree of groups sorted by number of free blocks + * - error handling + */ + +/* + * The allocation request involve request for multiple number of blocks + * near to the goal(block) value specified. + * + * During initialization phase of the allocator we decide to use the + * group preallocation or inode preallocation depending on the size of + * the file. The size of the file could be the resulting file size we + * would have after allocation, or the current file size, which ever + * is larger. If the size is less than sbi->s_mb_stream_request we + * select to use the group preallocation. The default value of + * s_mb_stream_request is 16 blocks. This can also be tuned via + * /sys/fs/ext4/<partition>/mb_stream_req. The value is represented in + * terms of number of blocks. + * + * The main motivation for having small file use group preallocation is to + * ensure that we have small files closer together on the disk. + * + * First stage the allocator looks at the inode prealloc list, + * ext4_inode_info->i_prealloc_list, which contains list of prealloc + * spaces for this particular inode. The inode prealloc space is + * represented as: + * + * pa_lstart -> the logical start block for this prealloc space + * pa_pstart -> the physical start block for this prealloc space + * pa_len -> length for this prealloc space (in clusters) + * pa_free -> free space available in this prealloc space (in clusters) + * + * The inode preallocation space is used looking at the _logical_ start + * block. If only the logical file block falls within the range of prealloc + * space we will consume the particular prealloc space. This makes sure that + * we have contiguous physical blocks representing the file blocks + * + * The important thing to be noted in case of inode prealloc space is that + * we don't modify the values associated to inode prealloc space except + * pa_free. + * + * If we are not able to find blocks in the inode prealloc space and if we + * have the group allocation flag set then we look at the locality group + * prealloc space. These are per CPU prealloc list represented as + * + * ext4_sb_info.s_locality_groups[smp_processor_id()] + * + * The reason for having a per cpu locality group is to reduce the contention + * between CPUs. It is possible to get scheduled at this point. + * + * The locality group prealloc space is used looking at whether we have + * enough free space (pa_free) within the prealloc space. + * + * If we can't allocate blocks via inode prealloc or/and locality group + * prealloc then we look at the buddy cache. The buddy cache is represented + * by ext4_sb_info.s_buddy_cache (struct inode) whose file offset gets + * mapped to the buddy and bitmap information regarding different + * groups. The buddy information is attached to buddy cache inode so that + * we can access them through the page cache. The information regarding + * each group is loaded via ext4_mb_load_buddy. The information involve + * block bitmap and buddy information. The information are stored in the + * inode as: + * + * { page } + * [ group 0 bitmap][ group 0 buddy] [group 1][ group 1]... + * + * + * one block each for bitmap and buddy information. So for each group we + * take up 2 blocks. A page can contain blocks_per_page (PAGE_SIZE / + * blocksize) blocks. So it can have information regarding groups_per_page + * which is blocks_per_page/2 + * + * The buddy cache inode is not stored on disk. The inode is thrown + * away when the filesystem is unmounted. + * + * We look for count number of blocks in the buddy cache. If we were able + * to locate that many free blocks we return with additional information + * regarding rest of the contiguous physical block available + * + * Before allocating blocks via buddy cache we normalize the request + * blocks. This ensure we ask for more blocks that we needed. The extra + * blocks that we get after allocation is added to the respective prealloc + * list. In case of inode preallocation we follow a list of heuristics + * based on file size. This can be found in ext4_mb_normalize_request. If + * we are doing a group prealloc we try to normalize the request to + * sbi->s_mb_group_prealloc. The default value of s_mb_group_prealloc is + * dependent on the cluster size; for non-bigalloc file systems, it is + * 512 blocks. This can be tuned via + * /sys/fs/ext4/<partition>/mb_group_prealloc. The value is represented in + * terms of number of blocks. If we have mounted the file system with -O + * stripe=<value> option the group prealloc request is normalized to the + * smallest multiple of the stripe value (sbi->s_stripe) which is + * greater than the default mb_group_prealloc. + * + * If "mb_optimize_scan" mount option is set, we maintain in memory group info + * structures in two data structures: + * + * 1) Array of largest free order xarrays (sbi->s_mb_largest_free_orders) + * + * Locking: Writers use xa_lock, readers use rcu_read_lock. + * + * This is an array of xarrays where the index in the array represents the + * largest free order in the buddy bitmap of the participating group infos of + * that xarray. So, there are exactly MB_NUM_ORDERS(sb) (which means total + * number of buddy bitmap orders possible) number of xarrays. Group-infos are + * placed in appropriate xarrays. + * + * 2) Average fragment size xarrays (sbi->s_mb_avg_fragment_size) + * + * Locking: Writers use xa_lock, readers use rcu_read_lock. + * + * This is an array of xarrays where in the i-th xarray there are groups with + * average fragment size >= 2^i and < 2^(i+1). The average fragment size + * is computed as ext4_group_info->bb_free / ext4_group_info->bb_fragments. + * Note that we don't bother with a special xarray for completely empty + * groups so we only have MB_NUM_ORDERS(sb) xarrays. Group-infos are placed + * in appropriate xarrays. + * + * In xarray, the index is the block group number, the value is the block group + * information, and a non-empty value indicates the block group is present in + * the current xarray. + * + * When "mb_optimize_scan" mount option is set, mballoc consults the above data + * structures to decide the order in which groups are to be traversed for + * fulfilling an allocation request. + * + * At CR_POWER2_ALIGNED , we look for groups which have the largest_free_order + * >= the order of the request. We directly look at the largest free order list + * in the data structure (1) above where largest_free_order = order of the + * request. If that list is empty, we look at remaining list in the increasing + * order of largest_free_order. This allows us to perform CR_POWER2_ALIGNED + * lookup in O(1) time. + * + * At CR_GOAL_LEN_FAST, we only consider groups where + * average fragment size > request size. So, we lookup a group which has average + * fragment size just above or equal to request size using our average fragment + * size group lists (data structure 2) in O(1) time. + * + * At CR_BEST_AVAIL_LEN, we aim to optimize allocations which can't be satisfied + * in CR_GOAL_LEN_FAST. The fact that we couldn't find a group in + * CR_GOAL_LEN_FAST suggests that there is no BG that has avg + * fragment size > goal length. So before falling to the slower + * CR_GOAL_LEN_SLOW, in CR_BEST_AVAIL_LEN we proactively trim goal length and + * then use the same fragment lists as CR_GOAL_LEN_FAST to find a BG with a big + * enough average fragment size. This increases the chances of finding a + * suitable block group in O(1) time and results in faster allocation at the + * cost of reduced size of allocation. + * + * If "mb_optimize_scan" mount option is not set, mballoc traverses groups in + * linear order which requires O(N) search time for each CR_POWER2_ALIGNED and + * CR_GOAL_LEN_FAST phase. + * + * The regular allocator (using the buddy cache) supports a few tunables. + * + * /sys/fs/ext4/<partition>/mb_min_to_scan + * /sys/fs/ext4/<partition>/mb_max_to_scan + * /sys/fs/ext4/<partition>/mb_order2_req + * /sys/fs/ext4/<partition>/mb_max_linear_groups + * + * The regular allocator uses buddy scan only if the request len is power of + * 2 blocks and the order of allocation is >= sbi->s_mb_order2_reqs. The + * value of s_mb_order2_reqs can be tuned via + * /sys/fs/ext4/<partition>/mb_order2_req. If the request len is equal to + * stripe size (sbi->s_stripe), we try to search for contiguous block in + * stripe size. This should result in better allocation on RAID setups. If + * not, we search in the specific group using bitmap for best extents. The + * tunable min_to_scan and max_to_scan control the behaviour here. + * min_to_scan indicate how long the mballoc __must__ look for a best + * extent and max_to_scan indicates how long the mballoc __can__ look for a + * best extent in the found extents. Searching for the blocks starts with + * the group specified as the goal value in allocation context via + * ac_g_ex. Each group is first checked based on the criteria whether it + * can be used for allocation. ext4_mb_good_group explains how the groups are + * checked. + * + * When "mb_optimize_scan" is turned on, as mentioned above, the groups may not + * get traversed linearly. That may result in subsequent allocations being not + * close to each other. And so, the underlying device may get filled up in a + * non-linear fashion. While that may not matter on non-rotational devices, for + * rotational devices that may result in higher seek times. "mb_max_linear_groups" + * tells mballoc how many groups mballoc should search linearly before + * performing consulting above data structures for more efficient lookups. For + * non rotational devices, this value defaults to 0 and for rotational devices + * this is set to MB_DEFAULT_LINEAR_LIMIT. + * + * Both the prealloc space are getting populated as above. So for the first + * request we will hit the buddy cache which will result in this prealloc + * space getting filled. The prealloc space is then later used for the + * subsequent request. + */ + +/* + * mballoc operates on the following data: + * - on-disk bitmap + * - in-core buddy (actually includes buddy and bitmap) + * - preallocation descriptors (PAs) + * + * there are two types of preallocations: + * - inode + * assiged to specific inode and can be used for this inode only. + * it describes part of inode's space preallocated to specific + * physical blocks. any block from that preallocated can be used + * independent. the descriptor just tracks number of blocks left + * unused. so, before taking some block from descriptor, one must + * make sure corresponded logical block isn't allocated yet. this + * also means that freeing any block within descriptor's range + * must discard all preallocated blocks. + * - locality group + * assigned to specific locality group which does not translate to + * permanent set of inodes: inode can join and leave group. space + * from this type of preallocation can be used for any inode. thus + * it's consumed from the beginning to the end. + * + * relation between them can be expressed as: + * in-core buddy = on-disk bitmap + preallocation descriptors + * + * this mean blocks mballoc considers used are: + * - allocated blocks (persistent) + * - preallocated blocks (non-persistent) + * + * consistency in mballoc world means that at any time a block is either + * free or used in ALL structures. notice: "any time" should not be read + * literally -- time is discrete and delimited by locks. + * + * to keep it simple, we don't use block numbers, instead we count number of + * blocks: how many blocks marked used/free in on-disk bitmap, buddy and PA. + * + * all operations can be expressed as: + * - init buddy: buddy = on-disk + PAs + * - new PA: buddy += N; PA = N + * - use inode PA: on-disk += N; PA -= N + * - discard inode PA buddy -= on-disk - PA; PA = 0 + * - use locality group PA on-disk += N; PA -= N + * - discard locality group PA buddy -= PA; PA = 0 + * note: 'buddy -= on-disk - PA' is used to show that on-disk bitmap + * is used in real operation because we can't know actual used + * bits from PA, only from on-disk bitmap + * + * if we follow this strict logic, then all operations above should be atomic. + * given some of them can block, we'd have to use something like semaphores + * killing performance on high-end SMP hardware. let's try to relax it using + * the following knowledge: + * 1) if buddy is referenced, it's already initialized + * 2) while block is used in buddy and the buddy is referenced, + * nobody can re-allocate that block + * 3) we work on bitmaps and '+' actually means 'set bits'. if on-disk has + * bit set and PA claims same block, it's OK. IOW, one can set bit in + * on-disk bitmap if buddy has same bit set or/and PA covers corresponded + * block + * + * so, now we're building a concurrency table: + * - init buddy vs. + * - new PA + * blocks for PA are allocated in the buddy, buddy must be referenced + * until PA is linked to allocation group to avoid concurrent buddy init + * - use inode PA + * we need to make sure that either on-disk bitmap or PA has uptodate data + * given (3) we care that PA-=N operation doesn't interfere with init + * - discard inode PA + * the simplest way would be to have buddy initialized by the discard + * - use locality group PA + * again PA-=N must be serialized with init + * - discard locality group PA + * the simplest way would be to have buddy initialized by the discard + * - new PA vs. + * - use inode PA + * i_data_sem serializes them + * - discard inode PA + * discard process must wait until PA isn't used by another process + * - use locality group PA + * some mutex should serialize them + * - discard locality group PA + * discard process must wait until PA isn't used by another process + * - use inode PA + * - use inode PA + * i_data_sem or another mutex should serializes them + * - discard inode PA + * discard process must wait until PA isn't used by another process + * - use locality group PA + * nothing wrong here -- they're different PAs covering different blocks + * - discard locality group PA + * discard process must wait until PA isn't used by another process + * + * now we're ready to make few consequences: + * - PA is referenced and while it is no discard is possible + * - PA is referenced until block isn't marked in on-disk bitmap + * - PA changes only after on-disk bitmap + * - discard must not compete with init. either init is done before + * any discard or they're serialized somehow + * - buddy init as sum of on-disk bitmap and PAs is done atomically + * + * a special case when we've used PA to emptiness. no need to modify buddy + * in this case, but we should care about concurrent init + * + */ + + /* + * Logic in few words: + * + * - allocation: + * load group + * find blocks + * mark bits in on-disk bitmap + * release group + * + * - use preallocation: + * find proper PA (per-inode or group) + * load group + * mark bits in on-disk bitmap + * release group + * release PA + * + * - free: + * load group + * mark bits in on-disk bitmap + * release group + * + * - discard preallocations in group: + * mark PAs deleted + * move them onto local list + * load on-disk bitmap + * load group + * remove PA from object (inode or locality group) + * mark free blocks in-core + * + * - discard inode's preallocations: + */ + +/* + * Locking rules + * + * Locks: + * - bitlock on a group (group) + * - object (inode/locality) (object) + * - per-pa lock (pa) + * - cr_power2_aligned lists lock (cr_power2_aligned) + * - cr_goal_len_fast lists lock (cr_goal_len_fast) + * + * Paths: + * - new pa + * object + * group + * + * - find and use pa: + * pa + * + * - release consumed pa: + * pa + * group + * object + * + * - generate in-core bitmap: + * group + * pa + * + * - discard all for given object (inode, locality group): + * object + * pa + * group + * + * - discard all for given group: + * group + * pa + * group + * object + * + * - allocation path (ext4_mb_regular_allocator) + * group + * cr_power2_aligned/cr_goal_len_fast + */ +static struct kmem_cache *ext4_pspace_cachep; +static struct kmem_cache *ext4_ac_cachep; +static struct kmem_cache *ext4_free_data_cachep; + +/* We create slab caches for groupinfo data structures based on the + * superblock block size. There will be one per mounted filesystem for + * each unique s_blocksize_bits */ +#define NR_GRPINFO_CACHES 8 +static struct kmem_cache *ext4_groupinfo_caches[NR_GRPINFO_CACHES]; + +static const char * const ext4_groupinfo_slab_names[NR_GRPINFO_CACHES] = { + "ext4_groupinfo_1k", "ext4_groupinfo_2k", "ext4_groupinfo_4k", + "ext4_groupinfo_8k", "ext4_groupinfo_16k", "ext4_groupinfo_32k", + "ext4_groupinfo_64k", "ext4_groupinfo_128k" +}; + +static void ext4_mb_generate_from_pa(struct super_block *sb, void *bitmap, + ext4_group_t group); +static void ext4_mb_new_preallocation(struct ext4_allocation_context *ac); + +static int ext4_mb_scan_group(struct ext4_allocation_context *ac, + ext4_group_t group); + +static int ext4_try_to_trim_range(struct super_block *sb, + struct ext4_buddy *e4b, ext4_grpblk_t start, + ext4_grpblk_t max, ext4_grpblk_t minblocks); + +/* + * The algorithm using this percpu seq counter goes below: + * 1. We sample the percpu discard_pa_seq counter before trying for block + * allocation in ext4_mb_new_blocks(). + * 2. We increment this percpu discard_pa_seq counter when we either allocate + * or free these blocks i.e. while marking those blocks as used/free in + * mb_mark_used()/mb_free_blocks(). + * 3. We also increment this percpu seq counter when we successfully identify + * that the bb_prealloc_list is not empty and hence proceed for discarding + * of those PAs inside ext4_mb_discard_group_preallocations(). + * + * Now to make sure that the regular fast path of block allocation is not + * affected, as a small optimization we only sample the percpu seq counter + * on that cpu. Only when the block allocation fails and when freed blocks + * found were 0, that is when we sample percpu seq counter for all cpus using + * below function ext4_get_discard_pa_seq_sum(). This happens after making + * sure that all the PAs on grp->bb_prealloc_list got freed or if it's empty. + */ +static DEFINE_PER_CPU(u64, discard_pa_seq); +static inline u64 ext4_get_discard_pa_seq_sum(void) +{ + int __cpu; + u64 __seq = 0; + + for_each_possible_cpu(__cpu) + __seq += per_cpu(discard_pa_seq, __cpu); + return __seq; +} + +static inline void *mb_correct_addr_and_bit(int *bit, void *addr) +{ +#if BITS_PER_LONG == 64 + *bit += ((unsigned long) addr & 7UL) << 3; + addr = (void *) ((unsigned long) addr & ~7UL); +#elif BITS_PER_LONG == 32 + *bit += ((unsigned long) addr & 3UL) << 3; + addr = (void *) ((unsigned long) addr & ~3UL); +#else +#error "how many bits you are?!" +#endif + return addr; +} + +static inline int mb_test_bit(int bit, void *addr) +{ + /* + * ext4_test_bit on architecture like powerpc + * needs unsigned long aligned address + */ + addr = mb_correct_addr_and_bit(&bit, addr); + return ext4_test_bit(bit, addr); +} + +static inline void mb_set_bit(int bit, void *addr) +{ + addr = mb_correct_addr_and_bit(&bit, addr); + ext4_set_bit(bit, addr); +} + +static inline void mb_clear_bit(int bit, void *addr) +{ + addr = mb_correct_addr_and_bit(&bit, addr); + ext4_clear_bit(bit, addr); +} + +static inline int mb_test_and_clear_bit(int bit, void *addr) +{ + addr = mb_correct_addr_and_bit(&bit, addr); + return ext4_test_and_clear_bit(bit, addr); +} + +static inline int mb_find_next_zero_bit(void *addr, int max, int start) +{ + int fix = 0, ret, tmpmax; + addr = mb_correct_addr_and_bit(&fix, addr); + tmpmax = max + fix; + start += fix; + + ret = ext4_find_next_zero_bit(addr, tmpmax, start) - fix; + if (ret > max) + return max; + return ret; +} + +static inline int mb_find_next_bit(void *addr, int max, int start) +{ + int fix = 0, ret, tmpmax; + addr = mb_correct_addr_and_bit(&fix, addr); + tmpmax = max + fix; + start += fix; + + ret = ext4_find_next_bit(addr, tmpmax, start) - fix; + if (ret > max) + return max; + return ret; +} + +static void *mb_find_buddy(struct ext4_buddy *e4b, int order, int *max) +{ + char *bb; + + BUG_ON(e4b->bd_bitmap == e4b->bd_buddy); + BUG_ON(max == NULL); + + if (order > e4b->bd_blkbits + 1) { + *max = 0; + return NULL; + } + + /* at order 0 we see each particular block */ + if (order == 0) { + *max = 1 << (e4b->bd_blkbits + 3); + return e4b->bd_bitmap; + } + + bb = e4b->bd_buddy + EXT4_SB(e4b->bd_sb)->s_mb_offsets[order]; + *max = EXT4_SB(e4b->bd_sb)->s_mb_maxs[order]; + + return bb; +} + +#ifdef DOUBLE_CHECK +static void mb_free_blocks_double(struct inode *inode, struct ext4_buddy *e4b, + int first, int count) +{ + int i; + struct super_block *sb = e4b->bd_sb; + + if (unlikely(e4b->bd_info->bb_bitmap == NULL)) + return; + assert_spin_locked(ext4_group_lock_ptr(sb, e4b->bd_group)); + for (i = 0; i < count; i++) { + if (!mb_test_bit(first + i, e4b->bd_info->bb_bitmap)) { + ext4_fsblk_t blocknr; + + blocknr = ext4_group_first_block_no(sb, e4b->bd_group); + blocknr += EXT4_C2B(EXT4_SB(sb), first + i); + ext4_mark_group_bitmap_corrupted(sb, e4b->bd_group, + EXT4_GROUP_INFO_BBITMAP_CORRUPT); + ext4_grp_locked_error(sb, e4b->bd_group, + inode ? inode->i_ino : 0, + blocknr, + "freeing block already freed " + "(bit %u)", + first + i); + } + mb_clear_bit(first + i, e4b->bd_info->bb_bitmap); + } +} + +static void mb_mark_used_double(struct ext4_buddy *e4b, int first, int count) +{ + int i; + + if (unlikely(e4b->bd_info->bb_bitmap == NULL)) + return; + assert_spin_locked(ext4_group_lock_ptr(e4b->bd_sb, e4b->bd_group)); + for (i = 0; i < count; i++) { + BUG_ON(mb_test_bit(first + i, e4b->bd_info->bb_bitmap)); + mb_set_bit(first + i, e4b->bd_info->bb_bitmap); + } +} + +static void mb_cmp_bitmaps(struct ext4_buddy *e4b, void *bitmap) +{ + if (unlikely(e4b->bd_info->bb_bitmap == NULL)) + return; + if (memcmp(e4b->bd_info->bb_bitmap, bitmap, e4b->bd_sb->s_blocksize)) { + unsigned char *b1, *b2; + int i; + b1 = (unsigned char *) e4b->bd_info->bb_bitmap; + b2 = (unsigned char *) bitmap; + for (i = 0; i < e4b->bd_sb->s_blocksize; i++) { + if (b1[i] != b2[i]) { + ext4_msg(e4b->bd_sb, KERN_ERR, + "corruption in group %u " + "at byte %u(%u): %x in copy != %x " + "on disk/prealloc", + e4b->bd_group, i, i * 8, b1[i], b2[i]); + BUG(); + } + } + } +} + +static void mb_group_bb_bitmap_alloc(struct super_block *sb, + struct ext4_group_info *grp, ext4_group_t group) +{ + struct buffer_head *bh; + + grp->bb_bitmap = kmalloc(sb->s_blocksize, GFP_NOFS); + if (!grp->bb_bitmap) + return; + + bh = ext4_read_block_bitmap(sb, group); + if (IS_ERR_OR_NULL(bh)) { + kfree(grp->bb_bitmap); + grp->bb_bitmap = NULL; + return; + } + + memcpy(grp->bb_bitmap, bh->b_data, sb->s_blocksize); + put_bh(bh); +} + +static void mb_group_bb_bitmap_free(struct ext4_group_info *grp) +{ + kfree(grp->bb_bitmap); +} + +#else +static inline void mb_free_blocks_double(struct inode *inode, + struct ext4_buddy *e4b, int first, int count) +{ + return; +} +static inline void mb_mark_used_double(struct ext4_buddy *e4b, + int first, int count) +{ + return; +} +static inline void mb_cmp_bitmaps(struct ext4_buddy *e4b, void *bitmap) +{ + return; +} + +static inline void mb_group_bb_bitmap_alloc(struct super_block *sb, + struct ext4_group_info *grp, ext4_group_t group) +{ + return; +} + +static inline void mb_group_bb_bitmap_free(struct ext4_group_info *grp) +{ + return; +} +#endif + +#ifdef AGGRESSIVE_CHECK + +#define MB_CHECK_ASSERT(assert) \ +do { \ + if (!(assert)) { \ + printk(KERN_EMERG \ + "Assertion failure in %s() at %s:%d: \"%s\"\n", \ + function, file, line, # assert); \ + BUG(); \ + } \ +} while (0) + +static void __mb_check_buddy(struct ext4_buddy *e4b, char *file, + const char *function, int line) +{ + struct super_block *sb = e4b->bd_sb; + int order = e4b->bd_blkbits + 1; + int max; + int max2; + int i; + int j; + int k; + int count; + struct ext4_group_info *grp; + int fragments = 0; + int fstart; + struct list_head *cur; + void *buddy; + void *buddy2; + + if (e4b->bd_info->bb_check_counter++ % 10) + return; + + while (order > 1) { + buddy = mb_find_buddy(e4b, order, &max); + MB_CHECK_ASSERT(buddy); + buddy2 = mb_find_buddy(e4b, order - 1, &max2); + MB_CHECK_ASSERT(buddy2); + MB_CHECK_ASSERT(buddy != buddy2); + MB_CHECK_ASSERT(max * 2 == max2); + + count = 0; + for (i = 0; i < max; i++) { + + if (mb_test_bit(i, buddy)) { + /* only single bit in buddy2 may be 0 */ + if (!mb_test_bit(i << 1, buddy2)) { + MB_CHECK_ASSERT( + mb_test_bit((i<<1)+1, buddy2)); + } + continue; + } + + /* both bits in buddy2 must be 1 */ + MB_CHECK_ASSERT(mb_test_bit(i << 1, buddy2)); + MB_CHECK_ASSERT(mb_test_bit((i << 1) + 1, buddy2)); + + for (j = 0; j < (1 << order); j++) { + k = (i * (1 << order)) + j; + MB_CHECK_ASSERT( + !mb_test_bit(k, e4b->bd_bitmap)); + } + count++; + } + MB_CHECK_ASSERT(e4b->bd_info->bb_counters[order] == count); + order--; + } + + fstart = -1; + buddy = mb_find_buddy(e4b, 0, &max); + for (i = 0; i < max; i++) { + if (!mb_test_bit(i, buddy)) { + MB_CHECK_ASSERT(i >= e4b->bd_info->bb_first_free); + if (fstart == -1) { + fragments++; + fstart = i; + } + continue; + } + fstart = -1; + /* check used bits only */ + for (j = 0; j < e4b->bd_blkbits + 1; j++) { + buddy2 = mb_find_buddy(e4b, j, &max2); + k = i >> j; + MB_CHECK_ASSERT(k < max2); + MB_CHECK_ASSERT(mb_test_bit(k, buddy2)); + } + } + MB_CHECK_ASSERT(!EXT4_MB_GRP_NEED_INIT(e4b->bd_info)); + MB_CHECK_ASSERT(e4b->bd_info->bb_fragments == fragments); + + grp = ext4_get_group_info(sb, e4b->bd_group); + if (!grp) + return; + list_for_each(cur, &grp->bb_prealloc_list) { + ext4_group_t groupnr; + struct ext4_prealloc_space *pa; + pa = list_entry(cur, struct ext4_prealloc_space, pa_group_list); + ext4_get_group_no_and_offset(sb, pa->pa_pstart, &groupnr, &k); + MB_CHECK_ASSERT(groupnr == e4b->bd_group); + for (i = 0; i < pa->pa_len; i++) + MB_CHECK_ASSERT(mb_test_bit(k + i, buddy)); + } +} +#undef MB_CHECK_ASSERT +#define mb_check_buddy(e4b) __mb_check_buddy(e4b, \ + __FILE__, __func__, __LINE__) +#else +#define mb_check_buddy(e4b) +#endif + +/* + * Divide blocks started from @first with length @len into + * smaller chunks with power of 2 blocks. + * Clear the bits in bitmap which the blocks of the chunk(s) covered, + * then increase bb_counters[] for corresponded chunk size. + */ +static void ext4_mb_mark_free_simple(struct super_block *sb, + void *buddy, ext4_grpblk_t first, ext4_grpblk_t len, + struct ext4_group_info *grp) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + ext4_grpblk_t min; + ext4_grpblk_t max; + ext4_grpblk_t chunk; + unsigned int border; + + BUG_ON(len > EXT4_CLUSTERS_PER_GROUP(sb)); + + border = 2 << sb->s_blocksize_bits; + + while (len > 0) { + /* find how many blocks can be covered since this position */ + max = ffs(first | border) - 1; + + /* find how many blocks of power 2 we need to mark */ + min = fls(len) - 1; + + if (max < min) + min = max; + chunk = 1 << min; + + /* mark multiblock chunks only */ + grp->bb_counters[min]++; + if (min > 0) + mb_clear_bit(first >> min, + buddy + sbi->s_mb_offsets[min]); + + len -= chunk; + first += chunk; + } +} + +static int mb_avg_fragment_size_order(struct super_block *sb, ext4_grpblk_t len) +{ + int order; + + /* + * We don't bother with a special lists groups with only 1 block free + * extents and for completely empty groups. + */ + order = fls(len) - 2; + if (order < 0) + return 0; + if (order == MB_NUM_ORDERS(sb)) + order--; + if (WARN_ON_ONCE(order > MB_NUM_ORDERS(sb))) + order = MB_NUM_ORDERS(sb) - 1; + return order; +} + +/* Move group to appropriate avg_fragment_size list */ +static void +mb_update_avg_fragment_size(struct super_block *sb, struct ext4_group_info *grp) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + int new, old; + + if (!test_opt2(sb, MB_OPTIMIZE_SCAN)) + return; + + old = grp->bb_avg_fragment_size_order; + new = grp->bb_fragments == 0 ? -1 : + mb_avg_fragment_size_order(sb, grp->bb_free / grp->bb_fragments); + if (new == old) + return; + + if (old >= 0) + xa_erase(&sbi->s_mb_avg_fragment_size[old], grp->bb_group); + + grp->bb_avg_fragment_size_order = new; + if (new >= 0) { + /* + * Cannot use __GFP_NOFAIL because we hold the group lock. + * Although allocation for insertion may fails, it's not fatal + * as we have linear traversal to fall back on. + */ + int err = xa_insert(&sbi->s_mb_avg_fragment_size[new], + grp->bb_group, grp, GFP_ATOMIC); + if (err) + mb_debug(sb, "insert group: %u to s_mb_avg_fragment_size[%d] failed, err %d", + grp->bb_group, new, err); + } +} + +static int ext4_mb_scan_groups_xa_range(struct ext4_allocation_context *ac, + struct xarray *xa, + ext4_group_t start, ext4_group_t end) +{ + struct super_block *sb = ac->ac_sb; + struct ext4_sb_info *sbi = EXT4_SB(sb); + enum criteria cr = ac->ac_criteria; + ext4_group_t ngroups = ext4_get_groups_count(sb); + unsigned long group = start; + struct ext4_group_info *grp; + + if (WARN_ON_ONCE(end > ngroups || start >= end)) + return 0; + + xa_for_each_range(xa, group, grp, start, end - 1) { + int err; + + if (sbi->s_mb_stats) + atomic64_inc(&sbi->s_bal_cX_groups_considered[cr]); + + err = ext4_mb_scan_group(ac, grp->bb_group); + if (err || ac->ac_status != AC_STATUS_CONTINUE) + return err; + + cond_resched(); + } + + return 0; +} + +/* + * Find a suitable group of given order from the largest free orders xarray. + */ +static inline int +ext4_mb_scan_groups_largest_free_order_range(struct ext4_allocation_context *ac, + int order, ext4_group_t start, + ext4_group_t end) +{ + struct xarray *xa = &EXT4_SB(ac->ac_sb)->s_mb_largest_free_orders[order]; + + if (xa_empty(xa)) + return 0; + + return ext4_mb_scan_groups_xa_range(ac, xa, start, end); +} + +/* + * Choose next group by traversing largest_free_order lists. Updates *new_cr if + * cr level needs an update. + */ +static int ext4_mb_scan_groups_p2_aligned(struct ext4_allocation_context *ac, + ext4_group_t group) +{ + struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); + int i; + int ret = 0; + ext4_group_t start, end; + + start = group; + end = ext4_get_groups_count(ac->ac_sb); +wrap_around: + for (i = ac->ac_2order; i < MB_NUM_ORDERS(ac->ac_sb); i++) { + ret = ext4_mb_scan_groups_largest_free_order_range(ac, i, + start, end); + if (ret || ac->ac_status != AC_STATUS_CONTINUE) + return ret; + } + if (start) { + end = start; + start = 0; + goto wrap_around; + } + + if (sbi->s_mb_stats) + atomic64_inc(&sbi->s_bal_cX_failed[ac->ac_criteria]); + + /* Increment cr and search again if no group is found */ + ac->ac_criteria = CR_GOAL_LEN_FAST; + return ret; +} + +/* + * Find a suitable group of given order from the average fragments xarray. + */ +static int +ext4_mb_scan_groups_avg_frag_order_range(struct ext4_allocation_context *ac, + int order, ext4_group_t start, + ext4_group_t end) +{ + struct xarray *xa = &EXT4_SB(ac->ac_sb)->s_mb_avg_fragment_size[order]; + + if (xa_empty(xa)) + return 0; + + return ext4_mb_scan_groups_xa_range(ac, xa, start, end); +} + +/* + * Choose next group by traversing average fragment size list of suitable + * order. Updates *new_cr if cr level needs an update. + */ +static int ext4_mb_scan_groups_goal_fast(struct ext4_allocation_context *ac, + ext4_group_t group) +{ + struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); + int i, ret = 0; + ext4_group_t start, end; + + start = group; + end = ext4_get_groups_count(ac->ac_sb); +wrap_around: + i = mb_avg_fragment_size_order(ac->ac_sb, ac->ac_g_ex.fe_len); + for (; i < MB_NUM_ORDERS(ac->ac_sb); i++) { + ret = ext4_mb_scan_groups_avg_frag_order_range(ac, i, + start, end); + if (ret || ac->ac_status != AC_STATUS_CONTINUE) + return ret; + } + if (start) { + end = start; + start = 0; + goto wrap_around; + } + + if (sbi->s_mb_stats) + atomic64_inc(&sbi->s_bal_cX_failed[ac->ac_criteria]); + /* + * CR_BEST_AVAIL_LEN works based on the concept that we have + * a larger normalized goal len request which can be trimmed to + * a smaller goal len such that it can still satisfy original + * request len. However, allocation request for non-regular + * files never gets normalized. + * See function ext4_mb_normalize_request() (EXT4_MB_HINT_DATA). + */ + if (ac->ac_flags & EXT4_MB_HINT_DATA) + ac->ac_criteria = CR_BEST_AVAIL_LEN; + else + ac->ac_criteria = CR_GOAL_LEN_SLOW; + + return ret; +} + +/* + * We couldn't find a group in CR_GOAL_LEN_FAST so try to find the highest free fragment + * order we have and proactively trim the goal request length to that order to + * find a suitable group faster. + * + * This optimizes allocation speed at the cost of slightly reduced + * preallocations. However, we make sure that we don't trim the request too + * much and fall to CR_GOAL_LEN_SLOW in that case. + */ +static int ext4_mb_scan_groups_best_avail(struct ext4_allocation_context *ac, + ext4_group_t group) +{ + int ret = 0; + struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); + int i, order, min_order; + unsigned long num_stripe_clusters = 0; + ext4_group_t start, end; + + /* + * mb_avg_fragment_size_order() returns order in a way that makes + * retrieving back the length using (1 << order) inaccurate. Hence, use + * fls() instead since we need to know the actual length while modifying + * goal length. + */ + order = fls(ac->ac_g_ex.fe_len) - 1; + if (WARN_ON_ONCE(order - 1 > MB_NUM_ORDERS(ac->ac_sb))) + order = MB_NUM_ORDERS(ac->ac_sb); + min_order = order - sbi->s_mb_best_avail_max_trim_order; + if (min_order < 0) + min_order = 0; + + if (sbi->s_stripe > 0) { + /* + * We are assuming that stripe size is always a multiple of + * cluster ratio otherwise __ext4_fill_super exists early. + */ + num_stripe_clusters = EXT4_NUM_B2C(sbi, sbi->s_stripe); + if (1 << min_order < num_stripe_clusters) + /* + * We consider 1 order less because later we round + * up the goal len to num_stripe_clusters + */ + min_order = fls(num_stripe_clusters) - 1; + } + + if (1 << min_order < ac->ac_o_ex.fe_len) + min_order = fls(ac->ac_o_ex.fe_len); + + start = group; + end = ext4_get_groups_count(ac->ac_sb); +wrap_around: + for (i = order; i >= min_order; i--) { + int frag_order; + /* + * Scale down goal len to make sure we find something + * in the free fragments list. Basically, reduce + * preallocations. + */ + ac->ac_g_ex.fe_len = 1 << i; + + if (num_stripe_clusters > 0) { + /* + * Try to round up the adjusted goal length to + * stripe size (in cluster units) multiple for + * efficiency. + */ + ac->ac_g_ex.fe_len = roundup(ac->ac_g_ex.fe_len, + num_stripe_clusters); + } + + frag_order = mb_avg_fragment_size_order(ac->ac_sb, + ac->ac_g_ex.fe_len); + + ret = ext4_mb_scan_groups_avg_frag_order_range(ac, frag_order, + start, end); + if (ret || ac->ac_status != AC_STATUS_CONTINUE) + return ret; + } + if (start) { + end = start; + start = 0; + goto wrap_around; + } + + /* Reset goal length to original goal length before falling into CR_GOAL_LEN_SLOW */ + ac->ac_g_ex.fe_len = ac->ac_orig_goal_len; + if (sbi->s_mb_stats) + atomic64_inc(&sbi->s_bal_cX_failed[ac->ac_criteria]); + ac->ac_criteria = CR_GOAL_LEN_SLOW; + + return ret; +} + +static inline int should_optimize_scan(struct ext4_allocation_context *ac) +{ + if (unlikely(!test_opt2(ac->ac_sb, MB_OPTIMIZE_SCAN))) + return 0; + if (ac->ac_criteria >= CR_GOAL_LEN_SLOW) + return 0; + if (!ext4_test_inode_flag(ac->ac_inode, EXT4_INODE_EXTENTS)) + return 0; + return 1; +} + +/* + * next linear group for allocation. + */ +static void next_linear_group(ext4_group_t *group, ext4_group_t ngroups) +{ + /* + * Artificially restricted ngroups for non-extent + * files makes group > ngroups possible on first loop. + */ + *group = *group + 1 >= ngroups ? 0 : *group + 1; +} + +static int ext4_mb_scan_groups_linear(struct ext4_allocation_context *ac, + ext4_group_t ngroups, ext4_group_t *start, ext4_group_t count) +{ + int ret, i; + enum criteria cr = ac->ac_criteria; + struct super_block *sb = ac->ac_sb; + struct ext4_sb_info *sbi = EXT4_SB(sb); + ext4_group_t group = *start; + + for (i = 0; i < count; i++, next_linear_group(&group, ngroups)) { + ret = ext4_mb_scan_group(ac, group); + if (ret || ac->ac_status != AC_STATUS_CONTINUE) + return ret; + cond_resched(); + } + + *start = group; + if (count == ngroups) + ac->ac_criteria++; + + /* Processed all groups and haven't found blocks */ + if (sbi->s_mb_stats && i == ngroups) + atomic64_inc(&sbi->s_bal_cX_failed[cr]); + + return 0; +} + +static int ext4_mb_scan_groups(struct ext4_allocation_context *ac) +{ + int ret = 0; + ext4_group_t start; + struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); + ext4_group_t ngroups = ext4_get_groups_count(ac->ac_sb); + + /* non-extent files are limited to low blocks/groups */ + if (!(ext4_test_inode_flag(ac->ac_inode, EXT4_INODE_EXTENTS))) + ngroups = sbi->s_blockfile_groups; + + /* searching for the right group start from the goal value specified */ + start = ac->ac_g_ex.fe_group; + ac->ac_prefetch_grp = start; + ac->ac_prefetch_nr = 0; + + if (!should_optimize_scan(ac)) + return ext4_mb_scan_groups_linear(ac, ngroups, &start, ngroups); + + /* + * Optimized scanning can return non adjacent groups which can cause + * seek overhead for rotational disks. So try few linear groups before + * trying optimized scan. + */ + if (sbi->s_mb_max_linear_groups) + ret = ext4_mb_scan_groups_linear(ac, ngroups, &start, + sbi->s_mb_max_linear_groups); + if (ret || ac->ac_status != AC_STATUS_CONTINUE) + return ret; + + switch (ac->ac_criteria) { + case CR_POWER2_ALIGNED: + return ext4_mb_scan_groups_p2_aligned(ac, start); + case CR_GOAL_LEN_FAST: + return ext4_mb_scan_groups_goal_fast(ac, start); + case CR_BEST_AVAIL_LEN: + return ext4_mb_scan_groups_best_avail(ac, start); + default: + /* + * TODO: For CR_GOAL_LEN_SLOW, we can arrange groups in an + * rb tree sorted by bb_free. But until that happens, we should + * never come here. + */ + WARN_ON(1); + } + + return 0; +} + +/* + * Cache the order of the largest free extent we have available in this block + * group. + */ +static void +mb_set_largest_free_order(struct super_block *sb, struct ext4_group_info *grp) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + int new, old = grp->bb_largest_free_order; + + for (new = MB_NUM_ORDERS(sb) - 1; new >= 0; new--) + if (grp->bb_counters[new] > 0) + break; + + /* No need to move between order lists? */ + if (new == old) + return; + + if (old >= 0) { + struct xarray *xa = &sbi->s_mb_largest_free_orders[old]; + + if (!xa_empty(xa) && xa_load(xa, grp->bb_group)) + xa_erase(xa, grp->bb_group); + } + + grp->bb_largest_free_order = new; + if (test_opt2(sb, MB_OPTIMIZE_SCAN) && new >= 0 && grp->bb_free) { + /* + * Cannot use __GFP_NOFAIL because we hold the group lock. + * Although allocation for insertion may fails, it's not fatal + * as we have linear traversal to fall back on. + */ + int err = xa_insert(&sbi->s_mb_largest_free_orders[new], + grp->bb_group, grp, GFP_ATOMIC); + if (err) + mb_debug(sb, "insert group: %u to s_mb_largest_free_orders[%d] failed, err %d", + grp->bb_group, new, err); + } +} + +static noinline_for_stack +void ext4_mb_generate_buddy(struct super_block *sb, + void *buddy, void *bitmap, ext4_group_t group, + struct ext4_group_info *grp) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + ext4_grpblk_t max = EXT4_CLUSTERS_PER_GROUP(sb); + ext4_grpblk_t i = 0; + ext4_grpblk_t first; + ext4_grpblk_t len; + unsigned free = 0; + unsigned fragments = 0; + unsigned long long period = get_cycles(); + + /* initialize buddy from bitmap which is aggregation + * of on-disk bitmap and preallocations */ + i = mb_find_next_zero_bit(bitmap, max, 0); + grp->bb_first_free = i; + while (i < max) { + fragments++; + first = i; + i = mb_find_next_bit(bitmap, max, i); + len = i - first; + free += len; + if (len > 1) + ext4_mb_mark_free_simple(sb, buddy, first, len, grp); + else + grp->bb_counters[0]++; + if (i < max) + i = mb_find_next_zero_bit(bitmap, max, i); + } + grp->bb_fragments = fragments; + + if (free != grp->bb_free) { + ext4_grp_locked_error(sb, group, 0, 0, + "block bitmap and bg descriptor " + "inconsistent: %u vs %u free clusters", + free, grp->bb_free); + /* + * If we intend to continue, we consider group descriptor + * corrupt and update bb_free using bitmap value + */ + grp->bb_free = free; + ext4_mark_group_bitmap_corrupted(sb, group, + EXT4_GROUP_INFO_BBITMAP_CORRUPT); + } + mb_set_largest_free_order(sb, grp); + mb_update_avg_fragment_size(sb, grp); + + clear_bit(EXT4_GROUP_INFO_NEED_INIT_BIT, &(grp->bb_state)); + + period = get_cycles() - period; + atomic_inc(&sbi->s_mb_buddies_generated); + atomic64_add(period, &sbi->s_mb_generation_time); +} + +static void mb_regenerate_buddy(struct ext4_buddy *e4b) +{ + int count; + int order = 1; + void *buddy; + + while ((buddy = mb_find_buddy(e4b, order++, &count))) + mb_set_bits(buddy, 0, count); + + e4b->bd_info->bb_fragments = 0; + memset(e4b->bd_info->bb_counters, 0, + sizeof(*e4b->bd_info->bb_counters) * + (e4b->bd_sb->s_blocksize_bits + 2)); + + ext4_mb_generate_buddy(e4b->bd_sb, e4b->bd_buddy, + e4b->bd_bitmap, e4b->bd_group, e4b->bd_info); +} + +/* The buddy information is attached the buddy cache inode + * for convenience. The information regarding each group + * is loaded via ext4_mb_load_buddy. The information involve + * block bitmap and buddy information. The information are + * stored in the inode as + * + * { page } + * [ group 0 bitmap][ group 0 buddy] [group 1][ group 1]... + * + * + * one block each for bitmap and buddy information. + * So for each group we take up 2 blocks. A page can + * contain blocks_per_page (PAGE_SIZE / blocksize) blocks. + * So it can have information regarding groups_per_page which + * is blocks_per_page/2 + * + * Locking note: This routine takes the block group lock of all groups + * for this page; do not hold this lock when calling this routine! + */ + +static int ext4_mb_init_cache(struct folio *folio, char *incore, gfp_t gfp) +{ + ext4_group_t ngroups; + unsigned int blocksize; + int blocks_per_page; + int groups_per_page; + int err = 0; + int i; + ext4_group_t first_group, group; + int first_block; + struct super_block *sb; + struct buffer_head *bhs; + struct buffer_head **bh = NULL; + struct inode *inode; + char *data; + char *bitmap; + struct ext4_group_info *grinfo; + + inode = folio->mapping->host; + sb = inode->i_sb; + ngroups = ext4_get_groups_count(sb); + blocksize = i_blocksize(inode); + blocks_per_page = PAGE_SIZE / blocksize; + + mb_debug(sb, "init folio %lu\n", folio->index); + + groups_per_page = blocks_per_page >> 1; + if (groups_per_page == 0) + groups_per_page = 1; + + /* allocate buffer_heads to read bitmaps */ + if (groups_per_page > 1) { + i = sizeof(struct buffer_head *) * groups_per_page; + bh = kzalloc(i, gfp); + if (bh == NULL) + return -ENOMEM; + } else + bh = &bhs; + + first_group = folio->index * blocks_per_page / 2; + + /* read all groups the folio covers into the cache */ + for (i = 0, group = first_group; i < groups_per_page; i++, group++) { + if (group >= ngroups) + break; + + grinfo = ext4_get_group_info(sb, group); + if (!grinfo) + continue; + /* + * If page is uptodate then we came here after online resize + * which added some new uninitialized group info structs, so + * we must skip all initialized uptodate buddies on the folio, + * which may be currently in use by an allocating task. + */ + if (folio_test_uptodate(folio) && + !EXT4_MB_GRP_NEED_INIT(grinfo)) { + bh[i] = NULL; + continue; + } + bh[i] = ext4_read_block_bitmap_nowait(sb, group, false); + if (IS_ERR(bh[i])) { + err = PTR_ERR(bh[i]); + bh[i] = NULL; + goto out; + } + mb_debug(sb, "read bitmap for group %u\n", group); + } + + /* wait for I/O completion */ + for (i = 0, group = first_group; i < groups_per_page; i++, group++) { + int err2; + + if (!bh[i]) + continue; + err2 = ext4_wait_block_bitmap(sb, group, bh[i]); + if (!err) + err = err2; + } + + first_block = folio->index * blocks_per_page; + for (i = 0; i < blocks_per_page; i++) { + group = (first_block + i) >> 1; + if (group >= ngroups) + break; + + if (!bh[group - first_group]) + /* skip initialized uptodate buddy */ + continue; + + if (!buffer_verified(bh[group - first_group])) + /* Skip faulty bitmaps */ + continue; + err = 0; + + /* + * data carry information regarding this + * particular group in the format specified + * above + * + */ + data = folio_address(folio) + (i * blocksize); + bitmap = bh[group - first_group]->b_data; + + /* + * We place the buddy block and bitmap block + * close together + */ + grinfo = ext4_get_group_info(sb, group); + if (!grinfo) { + err = -EFSCORRUPTED; + goto out; + } + if ((first_block + i) & 1) { + /* this is block of buddy */ + BUG_ON(incore == NULL); + mb_debug(sb, "put buddy for group %u in folio %lu/%x\n", + group, folio->index, i * blocksize); + trace_ext4_mb_buddy_bitmap_load(sb, group); + grinfo->bb_fragments = 0; + memset(grinfo->bb_counters, 0, + sizeof(*grinfo->bb_counters) * + (MB_NUM_ORDERS(sb))); + /* + * incore got set to the group block bitmap below + */ + ext4_lock_group(sb, group); + /* init the buddy */ + memset(data, 0xff, blocksize); + ext4_mb_generate_buddy(sb, data, incore, group, grinfo); + ext4_unlock_group(sb, group); + incore = NULL; + } else { + /* this is block of bitmap */ + BUG_ON(incore != NULL); + mb_debug(sb, "put bitmap for group %u in folio %lu/%x\n", + group, folio->index, i * blocksize); + trace_ext4_mb_bitmap_load(sb, group); + + /* see comments in ext4_mb_put_pa() */ + ext4_lock_group(sb, group); + memcpy(data, bitmap, blocksize); + + /* mark all preallocated blks used in in-core bitmap */ + ext4_mb_generate_from_pa(sb, data, group); + WARN_ON_ONCE(!RB_EMPTY_ROOT(&grinfo->bb_free_root)); + ext4_unlock_group(sb, group); + + /* set incore so that the buddy information can be + * generated using this + */ + incore = data; + } + } + folio_mark_uptodate(folio); + +out: + if (bh) { + for (i = 0; i < groups_per_page; i++) + brelse(bh[i]); + if (bh != &bhs) + kfree(bh); + } + return err; +} + +/* + * Lock the buddy and bitmap pages. This make sure other parallel init_group + * on the same buddy page doesn't happen whild holding the buddy page lock. + * Return locked buddy and bitmap pages on e4b struct. If buddy and bitmap + * are on the same page e4b->bd_buddy_folio is NULL and return value is 0. + */ +static int ext4_mb_get_buddy_page_lock(struct super_block *sb, + ext4_group_t group, struct ext4_buddy *e4b, gfp_t gfp) +{ + struct inode *inode = EXT4_SB(sb)->s_buddy_cache; + int block, pnum, poff; + int blocks_per_page; + struct folio *folio; + + e4b->bd_buddy_folio = NULL; + e4b->bd_bitmap_folio = NULL; + + blocks_per_page = PAGE_SIZE / sb->s_blocksize; + /* + * the buddy cache inode stores the block bitmap + * and buddy information in consecutive blocks. + * So for each group we need two blocks. + */ + block = group * 2; + pnum = block / blocks_per_page; + poff = block % blocks_per_page; + folio = __filemap_get_folio(inode->i_mapping, pnum, + FGP_LOCK | FGP_ACCESSED | FGP_CREAT, gfp); + if (IS_ERR(folio)) + return PTR_ERR(folio); + BUG_ON(folio->mapping != inode->i_mapping); + e4b->bd_bitmap_folio = folio; + e4b->bd_bitmap = folio_address(folio) + (poff * sb->s_blocksize); + + if (blocks_per_page >= 2) { + /* buddy and bitmap are on the same page */ + return 0; + } + + /* blocks_per_page == 1, hence we need another page for the buddy */ + folio = __filemap_get_folio(inode->i_mapping, block + 1, + FGP_LOCK | FGP_ACCESSED | FGP_CREAT, gfp); + if (IS_ERR(folio)) + return PTR_ERR(folio); + BUG_ON(folio->mapping != inode->i_mapping); + e4b->bd_buddy_folio = folio; + return 0; +} + +static void ext4_mb_put_buddy_page_lock(struct ext4_buddy *e4b) +{ + if (e4b->bd_bitmap_folio) { + folio_unlock(e4b->bd_bitmap_folio); + folio_put(e4b->bd_bitmap_folio); + } + if (e4b->bd_buddy_folio) { + folio_unlock(e4b->bd_buddy_folio); + folio_put(e4b->bd_buddy_folio); + } +} + +/* + * Locking note: This routine calls ext4_mb_init_cache(), which takes the + * block group lock of all groups for this page; do not hold the BG lock when + * calling this routine! + */ +static noinline_for_stack +int ext4_mb_init_group(struct super_block *sb, ext4_group_t group, gfp_t gfp) +{ + + struct ext4_group_info *this_grp; + struct ext4_buddy e4b; + struct folio *folio; + int ret = 0; + + might_sleep(); + mb_debug(sb, "init group %u\n", group); + this_grp = ext4_get_group_info(sb, group); + if (!this_grp) + return -EFSCORRUPTED; + + /* + * This ensures that we don't reinit the buddy cache + * page which map to the group from which we are already + * allocating. If we are looking at the buddy cache we would + * have taken a reference using ext4_mb_load_buddy and that + * would have pinned buddy page to page cache. + * The call to ext4_mb_get_buddy_page_lock will mark the + * page accessed. + */ + ret = ext4_mb_get_buddy_page_lock(sb, group, &e4b, gfp); + if (ret || !EXT4_MB_GRP_NEED_INIT(this_grp)) { + /* + * somebody initialized the group + * return without doing anything + */ + goto err; + } + + folio = e4b.bd_bitmap_folio; + ret = ext4_mb_init_cache(folio, NULL, gfp); + if (ret) + goto err; + if (!folio_test_uptodate(folio)) { + ret = -EIO; + goto err; + } + + if (e4b.bd_buddy_folio == NULL) { + /* + * If both the bitmap and buddy are in + * the same page we don't need to force + * init the buddy + */ + ret = 0; + goto err; + } + /* init buddy cache */ + folio = e4b.bd_buddy_folio; + ret = ext4_mb_init_cache(folio, e4b.bd_bitmap, gfp); + if (ret) + goto err; + if (!folio_test_uptodate(folio)) { + ret = -EIO; + goto err; + } +err: + ext4_mb_put_buddy_page_lock(&e4b); + return ret; +} + +/* + * Locking note: This routine calls ext4_mb_init_cache(), which takes the + * block group lock of all groups for this page; do not hold the BG lock when + * calling this routine! + */ +static noinline_for_stack int +ext4_mb_load_buddy_gfp(struct super_block *sb, ext4_group_t group, + struct ext4_buddy *e4b, gfp_t gfp) +{ + int blocks_per_page; + int block; + int pnum; + int poff; + struct folio *folio; + int ret; + struct ext4_group_info *grp; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct inode *inode = sbi->s_buddy_cache; + + might_sleep(); + mb_debug(sb, "load group %u\n", group); + + blocks_per_page = PAGE_SIZE / sb->s_blocksize; + grp = ext4_get_group_info(sb, group); + if (!grp) + return -EFSCORRUPTED; + + e4b->bd_blkbits = sb->s_blocksize_bits; + e4b->bd_info = grp; + e4b->bd_sb = sb; + e4b->bd_group = group; + e4b->bd_buddy_folio = NULL; + e4b->bd_bitmap_folio = NULL; + + if (unlikely(EXT4_MB_GRP_NEED_INIT(grp))) { + /* + * we need full data about the group + * to make a good selection + */ + ret = ext4_mb_init_group(sb, group, gfp); + if (ret) + return ret; + } + + /* + * the buddy cache inode stores the block bitmap + * and buddy information in consecutive blocks. + * So for each group we need two blocks. + */ + block = group * 2; + pnum = block / blocks_per_page; + poff = block % blocks_per_page; + + /* Avoid locking the folio in the fast path ... */ + folio = __filemap_get_folio(inode->i_mapping, pnum, FGP_ACCESSED, 0); + if (IS_ERR(folio) || !folio_test_uptodate(folio)) { + if (!IS_ERR(folio)) + /* + * drop the folio reference and try + * to get the folio with lock. If we + * are not uptodate that implies + * somebody just created the folio but + * is yet to initialize it. So + * wait for it to initialize. + */ + folio_put(folio); + folio = __filemap_get_folio(inode->i_mapping, pnum, + FGP_LOCK | FGP_ACCESSED | FGP_CREAT, gfp); + if (!IS_ERR(folio)) { + if (WARN_RATELIMIT(folio->mapping != inode->i_mapping, + "ext4: bitmap's mapping != inode->i_mapping\n")) { + /* should never happen */ + folio_unlock(folio); + ret = -EINVAL; + goto err; + } + if (!folio_test_uptodate(folio)) { + ret = ext4_mb_init_cache(folio, NULL, gfp); + if (ret) { + folio_unlock(folio); + goto err; + } + mb_cmp_bitmaps(e4b, folio_address(folio) + + (poff * sb->s_blocksize)); + } + folio_unlock(folio); + } + } + if (IS_ERR(folio)) { + ret = PTR_ERR(folio); + goto err; + } + if (!folio_test_uptodate(folio)) { + ret = -EIO; + goto err; + } + + /* Folios marked accessed already */ + e4b->bd_bitmap_folio = folio; + e4b->bd_bitmap = folio_address(folio) + (poff * sb->s_blocksize); + + block++; + pnum = block / blocks_per_page; + poff = block % blocks_per_page; + + folio = __filemap_get_folio(inode->i_mapping, pnum, FGP_ACCESSED, 0); + if (IS_ERR(folio) || !folio_test_uptodate(folio)) { + if (!IS_ERR(folio)) + folio_put(folio); + folio = __filemap_get_folio(inode->i_mapping, pnum, + FGP_LOCK | FGP_ACCESSED | FGP_CREAT, gfp); + if (!IS_ERR(folio)) { + if (WARN_RATELIMIT(folio->mapping != inode->i_mapping, + "ext4: buddy bitmap's mapping != inode->i_mapping\n")) { + /* should never happen */ + folio_unlock(folio); + ret = -EINVAL; + goto err; + } + if (!folio_test_uptodate(folio)) { + ret = ext4_mb_init_cache(folio, e4b->bd_bitmap, + gfp); + if (ret) { + folio_unlock(folio); + goto err; + } + } + folio_unlock(folio); + } + } + if (IS_ERR(folio)) { + ret = PTR_ERR(folio); + goto err; + } + if (!folio_test_uptodate(folio)) { + ret = -EIO; + goto err; + } + + /* Folios marked accessed already */ + e4b->bd_buddy_folio = folio; + e4b->bd_buddy = folio_address(folio) + (poff * sb->s_blocksize); + + return 0; + +err: + if (!IS_ERR_OR_NULL(folio)) + folio_put(folio); + if (e4b->bd_bitmap_folio) + folio_put(e4b->bd_bitmap_folio); + + e4b->bd_buddy = NULL; + e4b->bd_bitmap = NULL; + return ret; +} + +static int ext4_mb_load_buddy(struct super_block *sb, ext4_group_t group, + struct ext4_buddy *e4b) +{ + return ext4_mb_load_buddy_gfp(sb, group, e4b, GFP_NOFS); +} + +static void ext4_mb_unload_buddy(struct ext4_buddy *e4b) +{ + if (e4b->bd_bitmap_folio) + folio_put(e4b->bd_bitmap_folio); + if (e4b->bd_buddy_folio) + folio_put(e4b->bd_buddy_folio); +} + + +static int mb_find_order_for_block(struct ext4_buddy *e4b, int block) +{ + int order = 1, max; + void *bb; + + BUG_ON(e4b->bd_bitmap == e4b->bd_buddy); + BUG_ON(block >= (1 << (e4b->bd_blkbits + 3))); + + while (order <= e4b->bd_blkbits + 1) { + bb = mb_find_buddy(e4b, order, &max); + if (!mb_test_bit(block >> order, bb)) { + /* this block is part of buddy of order 'order' */ + return order; + } + order++; + } + return 0; +} + +static void mb_clear_bits(void *bm, int cur, int len) +{ + __u32 *addr; + + len = cur + len; + while (cur < len) { + if ((cur & 31) == 0 && (len - cur) >= 32) { + /* fast path: clear whole word at once */ + addr = bm + (cur >> 3); + *addr = 0; + cur += 32; + continue; + } + mb_clear_bit(cur, bm); + cur++; + } +} + +/* clear bits in given range + * will return first found zero bit if any, -1 otherwise + */ +static int mb_test_and_clear_bits(void *bm, int cur, int len) +{ + __u32 *addr; + int zero_bit = -1; + + len = cur + len; + while (cur < len) { + if ((cur & 31) == 0 && (len - cur) >= 32) { + /* fast path: clear whole word at once */ + addr = bm + (cur >> 3); + if (*addr != (__u32)(-1) && zero_bit == -1) + zero_bit = cur + mb_find_next_zero_bit(addr, 32, 0); + *addr = 0; + cur += 32; + continue; + } + if (!mb_test_and_clear_bit(cur, bm) && zero_bit == -1) + zero_bit = cur; + cur++; + } + + return zero_bit; +} + +void mb_set_bits(void *bm, int cur, int len) +{ + __u32 *addr; + + len = cur + len; + while (cur < len) { + if ((cur & 31) == 0 && (len - cur) >= 32) { + /* fast path: set whole word at once */ + addr = bm + (cur >> 3); + *addr = 0xffffffff; + cur += 32; + continue; + } + mb_set_bit(cur, bm); + cur++; + } +} + +static inline int mb_buddy_adjust_border(int* bit, void* bitmap, int side) +{ + if (mb_test_bit(*bit + side, bitmap)) { + mb_clear_bit(*bit, bitmap); + (*bit) -= side; + return 1; + } + else { + (*bit) += side; + mb_set_bit(*bit, bitmap); + return -1; + } +} + +static void mb_buddy_mark_free(struct ext4_buddy *e4b, int first, int last) +{ + int max; + int order = 1; + void *buddy = mb_find_buddy(e4b, order, &max); + + while (buddy) { + void *buddy2; + + /* Bits in range [first; last] are known to be set since + * corresponding blocks were allocated. Bits in range + * (first; last) will stay set because they form buddies on + * upper layer. We just deal with borders if they don't + * align with upper layer and then go up. + * Releasing entire group is all about clearing + * single bit of highest order buddy. + */ + + /* Example: + * --------------------------------- + * | 1 | 1 | 1 | 1 | + * --------------------------------- + * | 0 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | + * --------------------------------- + * 0 1 2 3 4 5 6 7 + * \_____________________/ + * + * Neither [1] nor [6] is aligned to above layer. + * Left neighbour [0] is free, so mark it busy, + * decrease bb_counters and extend range to + * [0; 6] + * Right neighbour [7] is busy. It can't be coaleasced with [6], so + * mark [6] free, increase bb_counters and shrink range to + * [0; 5]. + * Then shift range to [0; 2], go up and do the same. + */ + + + if (first & 1) + e4b->bd_info->bb_counters[order] += mb_buddy_adjust_border(&first, buddy, -1); + if (!(last & 1)) + e4b->bd_info->bb_counters[order] += mb_buddy_adjust_border(&last, buddy, 1); + if (first > last) + break; + order++; + + buddy2 = mb_find_buddy(e4b, order, &max); + if (!buddy2) { + mb_clear_bits(buddy, first, last - first + 1); + e4b->bd_info->bb_counters[order - 1] += last - first + 1; + break; + } + first >>= 1; + last >>= 1; + buddy = buddy2; + } +} + +static void mb_free_blocks(struct inode *inode, struct ext4_buddy *e4b, + int first, int count) +{ + int left_is_free = 0; + int right_is_free = 0; + int block; + int last = first + count - 1; + struct super_block *sb = e4b->bd_sb; + + if (WARN_ON(count == 0)) + return; + BUG_ON(last >= (sb->s_blocksize << 3)); + assert_spin_locked(ext4_group_lock_ptr(sb, e4b->bd_group)); + /* Don't bother if the block group is corrupt. */ + if (unlikely(EXT4_MB_GRP_BBITMAP_CORRUPT(e4b->bd_info))) + return; + + mb_check_buddy(e4b); + mb_free_blocks_double(inode, e4b, first, count); + + /* access memory sequentially: check left neighbour, + * clear range and then check right neighbour + */ + if (first != 0) + left_is_free = !mb_test_bit(first - 1, e4b->bd_bitmap); + block = mb_test_and_clear_bits(e4b->bd_bitmap, first, count); + if (last + 1 < EXT4_SB(sb)->s_mb_maxs[0]) + right_is_free = !mb_test_bit(last + 1, e4b->bd_bitmap); + + if (unlikely(block != -1)) { + struct ext4_sb_info *sbi = EXT4_SB(sb); + ext4_fsblk_t blocknr; + + /* + * Fastcommit replay can free already freed blocks which + * corrupts allocation info. Regenerate it. + */ + if (sbi->s_mount_state & EXT4_FC_REPLAY) { + mb_regenerate_buddy(e4b); + goto check; + } + + blocknr = ext4_group_first_block_no(sb, e4b->bd_group); + blocknr += EXT4_C2B(sbi, block); + ext4_mark_group_bitmap_corrupted(sb, e4b->bd_group, + EXT4_GROUP_INFO_BBITMAP_CORRUPT); + ext4_grp_locked_error(sb, e4b->bd_group, + inode ? inode->i_ino : 0, blocknr, + "freeing already freed block (bit %u); block bitmap corrupt.", + block); + return; + } + + this_cpu_inc(discard_pa_seq); + e4b->bd_info->bb_free += count; + if (first < e4b->bd_info->bb_first_free) + e4b->bd_info->bb_first_free = first; + + /* let's maintain fragments counter */ + if (left_is_free && right_is_free) + e4b->bd_info->bb_fragments--; + else if (!left_is_free && !right_is_free) + e4b->bd_info->bb_fragments++; + + /* buddy[0] == bd_bitmap is a special case, so handle + * it right away and let mb_buddy_mark_free stay free of + * zero order checks. + * Check if neighbours are to be coaleasced, + * adjust bitmap bb_counters and borders appropriately. + */ + if (first & 1) { + first += !left_is_free; + e4b->bd_info->bb_counters[0] += left_is_free ? -1 : 1; + } + if (!(last & 1)) { + last -= !right_is_free; + e4b->bd_info->bb_counters[0] += right_is_free ? -1 : 1; + } + + if (first <= last) + mb_buddy_mark_free(e4b, first >> 1, last >> 1); + + mb_set_largest_free_order(sb, e4b->bd_info); + mb_update_avg_fragment_size(sb, e4b->bd_info); +check: + mb_check_buddy(e4b); +} + +static int mb_find_extent(struct ext4_buddy *e4b, int block, + int needed, struct ext4_free_extent *ex) +{ + int max, order, next; + void *buddy; + + assert_spin_locked(ext4_group_lock_ptr(e4b->bd_sb, e4b->bd_group)); + BUG_ON(ex == NULL); + + buddy = mb_find_buddy(e4b, 0, &max); + BUG_ON(buddy == NULL); + BUG_ON(block >= max); + if (mb_test_bit(block, buddy)) { + ex->fe_len = 0; + ex->fe_start = 0; + ex->fe_group = 0; + return 0; + } + + /* find actual order */ + order = mb_find_order_for_block(e4b, block); + + ex->fe_len = (1 << order) - (block & ((1 << order) - 1)); + ex->fe_start = block; + ex->fe_group = e4b->bd_group; + + block = block >> order; + + while (needed > ex->fe_len && + mb_find_buddy(e4b, order, &max)) { + + if (block + 1 >= max) + break; + + next = (block + 1) * (1 << order); + if (mb_test_bit(next, e4b->bd_bitmap)) + break; + + order = mb_find_order_for_block(e4b, next); + + block = next >> order; + ex->fe_len += 1 << order; + } + + if (ex->fe_start + ex->fe_len > EXT4_CLUSTERS_PER_GROUP(e4b->bd_sb)) { + /* Should never happen! (but apparently sometimes does?!?) */ + WARN_ON(1); + ext4_grp_locked_error(e4b->bd_sb, e4b->bd_group, 0, 0, + "corruption or bug in mb_find_extent " + "block=%d, order=%d needed=%d ex=%u/%d/%d@%u", + block, order, needed, ex->fe_group, ex->fe_start, + ex->fe_len, ex->fe_logical); + ex->fe_len = 0; + ex->fe_start = 0; + ex->fe_group = 0; + } + return ex->fe_len; +} + +static int mb_mark_used(struct ext4_buddy *e4b, struct ext4_free_extent *ex) +{ + int ord; + int mlen = 0; + int max = 0; + int start = ex->fe_start; + int len = ex->fe_len; + unsigned ret = 0; + int len0 = len; + void *buddy; + int ord_start, ord_end; + + BUG_ON(start + len > (e4b->bd_sb->s_blocksize << 3)); + BUG_ON(e4b->bd_group != ex->fe_group); + assert_spin_locked(ext4_group_lock_ptr(e4b->bd_sb, e4b->bd_group)); + mb_check_buddy(e4b); + mb_mark_used_double(e4b, start, len); + + this_cpu_inc(discard_pa_seq); + e4b->bd_info->bb_free -= len; + if (e4b->bd_info->bb_first_free == start) + e4b->bd_info->bb_first_free += len; + + /* let's maintain fragments counter */ + if (start != 0) + mlen = !mb_test_bit(start - 1, e4b->bd_bitmap); + if (start + len < EXT4_SB(e4b->bd_sb)->s_mb_maxs[0]) + max = !mb_test_bit(start + len, e4b->bd_bitmap); + if (mlen && max) + e4b->bd_info->bb_fragments++; + else if (!mlen && !max) + e4b->bd_info->bb_fragments--; + + /* let's maintain buddy itself */ + while (len) { + ord = mb_find_order_for_block(e4b, start); + + if (((start >> ord) << ord) == start && len >= (1 << ord)) { + /* the whole chunk may be allocated at once! */ + mlen = 1 << ord; + buddy = mb_find_buddy(e4b, ord, &max); + BUG_ON((start >> ord) >= max); + mb_set_bit(start >> ord, buddy); + e4b->bd_info->bb_counters[ord]--; + start += mlen; + len -= mlen; + BUG_ON(len < 0); + continue; + } + + /* store for history */ + if (ret == 0) + ret = len | (ord << 16); + + BUG_ON(ord <= 0); + buddy = mb_find_buddy(e4b, ord, &max); + mb_set_bit(start >> ord, buddy); + e4b->bd_info->bb_counters[ord]--; + + ord_start = (start >> ord) << ord; + ord_end = ord_start + (1 << ord); + /* first chunk */ + if (start > ord_start) + ext4_mb_mark_free_simple(e4b->bd_sb, e4b->bd_buddy, + ord_start, start - ord_start, + e4b->bd_info); + + /* last chunk */ + if (start + len < ord_end) { + ext4_mb_mark_free_simple(e4b->bd_sb, e4b->bd_buddy, + start + len, + ord_end - (start + len), + e4b->bd_info); + break; + } + len = start + len - ord_end; + start = ord_end; + } + mb_set_largest_free_order(e4b->bd_sb, e4b->bd_info); + + mb_update_avg_fragment_size(e4b->bd_sb, e4b->bd_info); + mb_set_bits(e4b->bd_bitmap, ex->fe_start, len0); + mb_check_buddy(e4b); + + return ret; +} + +/* + * Must be called under group lock! + */ +static void ext4_mb_use_best_found(struct ext4_allocation_context *ac, + struct ext4_buddy *e4b) +{ + struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); + int ret; + + BUG_ON(ac->ac_b_ex.fe_group != e4b->bd_group); + BUG_ON(ac->ac_status == AC_STATUS_FOUND); + + ac->ac_b_ex.fe_len = min(ac->ac_b_ex.fe_len, ac->ac_g_ex.fe_len); + ac->ac_b_ex.fe_logical = ac->ac_g_ex.fe_logical; + ret = mb_mark_used(e4b, &ac->ac_b_ex); + + /* preallocation can change ac_b_ex, thus we store actually + * allocated blocks for history */ + ac->ac_f_ex = ac->ac_b_ex; + + ac->ac_status = AC_STATUS_FOUND; + ac->ac_tail = ret & 0xffff; + ac->ac_buddy = ret >> 16; + + /* + * take the page reference. We want the page to be pinned + * so that we don't get a ext4_mb_init_cache_call for this + * group until we update the bitmap. That would mean we + * double allocate blocks. The reference is dropped + * in ext4_mb_release_context + */ + ac->ac_bitmap_folio = e4b->bd_bitmap_folio; + folio_get(ac->ac_bitmap_folio); + ac->ac_buddy_folio = e4b->bd_buddy_folio; + folio_get(ac->ac_buddy_folio); + /* store last allocated for subsequent stream allocation */ + if (ac->ac_flags & EXT4_MB_STREAM_ALLOC) { + int hash = ac->ac_inode->i_ino % sbi->s_mb_nr_global_goals; + + WRITE_ONCE(sbi->s_mb_last_groups[hash], ac->ac_f_ex.fe_group); + } + + /* + * As we've just preallocated more space than + * user requested originally, we store allocated + * space in a special descriptor. + */ + if (ac->ac_o_ex.fe_len < ac->ac_b_ex.fe_len) + ext4_mb_new_preallocation(ac); + +} + +static void ext4_mb_check_limits(struct ext4_allocation_context *ac, + struct ext4_buddy *e4b, + int finish_group) +{ + struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); + struct ext4_free_extent *bex = &ac->ac_b_ex; + struct ext4_free_extent *gex = &ac->ac_g_ex; + + if (ac->ac_status == AC_STATUS_FOUND) + return; + /* + * We don't want to scan for a whole year + */ + if (ac->ac_found > sbi->s_mb_max_to_scan && + !(ac->ac_flags & EXT4_MB_HINT_FIRST)) { + ac->ac_status = AC_STATUS_BREAK; + return; + } + + /* + * Haven't found good chunk so far, let's continue + */ + if (bex->fe_len < gex->fe_len) + return; + + if (finish_group || ac->ac_found > sbi->s_mb_min_to_scan) + ext4_mb_use_best_found(ac, e4b); +} + +/* + * The routine checks whether found extent is good enough. If it is, + * then the extent gets marked used and flag is set to the context + * to stop scanning. Otherwise, the extent is compared with the + * previous found extent and if new one is better, then it's stored + * in the context. Later, the best found extent will be used, if + * mballoc can't find good enough extent. + * + * The algorithm used is roughly as follows: + * + * * If free extent found is exactly as big as goal, then + * stop the scan and use it immediately + * + * * If free extent found is smaller than goal, then keep retrying + * upto a max of sbi->s_mb_max_to_scan times (default 200). After + * that stop scanning and use whatever we have. + * + * * If free extent found is bigger than goal, then keep retrying + * upto a max of sbi->s_mb_min_to_scan times (default 10) before + * stopping the scan and using the extent. + * + * + * FIXME: real allocation policy is to be designed yet! + */ +static void ext4_mb_measure_extent(struct ext4_allocation_context *ac, + struct ext4_free_extent *ex, + struct ext4_buddy *e4b) +{ + struct ext4_free_extent *bex = &ac->ac_b_ex; + struct ext4_free_extent *gex = &ac->ac_g_ex; + + BUG_ON(ex->fe_len <= 0); + BUG_ON(ex->fe_len > EXT4_CLUSTERS_PER_GROUP(ac->ac_sb)); + BUG_ON(ex->fe_start >= EXT4_CLUSTERS_PER_GROUP(ac->ac_sb)); + BUG_ON(ac->ac_status != AC_STATUS_CONTINUE); + + ac->ac_found++; + ac->ac_cX_found[ac->ac_criteria]++; + + /* + * The special case - take what you catch first + */ + if (unlikely(ac->ac_flags & EXT4_MB_HINT_FIRST)) { + *bex = *ex; + ext4_mb_use_best_found(ac, e4b); + return; + } + + /* + * Let's check whether the chuck is good enough + */ + if (ex->fe_len == gex->fe_len) { + *bex = *ex; + ext4_mb_use_best_found(ac, e4b); + return; + } + + /* + * If this is first found extent, just store it in the context + */ + if (bex->fe_len == 0) { + *bex = *ex; + return; + } + + /* + * If new found extent is better, store it in the context + */ + if (bex->fe_len < gex->fe_len) { + /* if the request isn't satisfied, any found extent + * larger than previous best one is better */ + if (ex->fe_len > bex->fe_len) + *bex = *ex; + } else if (ex->fe_len > gex->fe_len) { + /* if the request is satisfied, then we try to find + * an extent that still satisfy the request, but is + * smaller than previous one */ + if (ex->fe_len < bex->fe_len) + *bex = *ex; + } + + ext4_mb_check_limits(ac, e4b, 0); +} + +static noinline_for_stack +void ext4_mb_try_best_found(struct ext4_allocation_context *ac, + struct ext4_buddy *e4b) +{ + struct ext4_free_extent ex = ac->ac_b_ex; + ext4_group_t group = ex.fe_group; + int max; + int err; + + BUG_ON(ex.fe_len <= 0); + err = ext4_mb_load_buddy(ac->ac_sb, group, e4b); + if (err) + return; + + ext4_lock_group(ac->ac_sb, group); + if (unlikely(EXT4_MB_GRP_BBITMAP_CORRUPT(e4b->bd_info))) + goto out; + + max = mb_find_extent(e4b, ex.fe_start, ex.fe_len, &ex); + + if (max > 0) { + ac->ac_b_ex = ex; + ext4_mb_use_best_found(ac, e4b); + } + +out: + ext4_unlock_group(ac->ac_sb, group); + ext4_mb_unload_buddy(e4b); +} + +static noinline_for_stack +int ext4_mb_find_by_goal(struct ext4_allocation_context *ac, + struct ext4_buddy *e4b) +{ + ext4_group_t group = ac->ac_g_ex.fe_group; + int max; + int err; + struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); + struct ext4_group_info *grp = ext4_get_group_info(ac->ac_sb, group); + struct ext4_free_extent ex; + + if (!grp) + return -EFSCORRUPTED; + if (!(ac->ac_flags & (EXT4_MB_HINT_TRY_GOAL | EXT4_MB_HINT_GOAL_ONLY))) + return 0; + if (grp->bb_free == 0) + return 0; + + err = ext4_mb_load_buddy(ac->ac_sb, group, e4b); + if (err) + return err; + + ext4_lock_group(ac->ac_sb, group); + if (unlikely(EXT4_MB_GRP_BBITMAP_CORRUPT(e4b->bd_info))) + goto out; + + max = mb_find_extent(e4b, ac->ac_g_ex.fe_start, + ac->ac_g_ex.fe_len, &ex); + ex.fe_logical = 0xDEADFA11; /* debug value */ + + if (max >= ac->ac_g_ex.fe_len && + ac->ac_g_ex.fe_len == EXT4_NUM_B2C(sbi, sbi->s_stripe)) { + ext4_fsblk_t start; + + start = ext4_grp_offs_to_block(ac->ac_sb, &ex); + /* use do_div to get remainder (would be 64-bit modulo) */ + if (do_div(start, sbi->s_stripe) == 0) { + ac->ac_found++; + ac->ac_b_ex = ex; + ext4_mb_use_best_found(ac, e4b); + } + } else if (max >= ac->ac_g_ex.fe_len) { + BUG_ON(ex.fe_len <= 0); + BUG_ON(ex.fe_group != ac->ac_g_ex.fe_group); + BUG_ON(ex.fe_start != ac->ac_g_ex.fe_start); + ac->ac_found++; + ac->ac_b_ex = ex; + ext4_mb_use_best_found(ac, e4b); + } else if (max > 0 && (ac->ac_flags & EXT4_MB_HINT_MERGE)) { + /* Sometimes, caller may want to merge even small + * number of blocks to an existing extent */ + BUG_ON(ex.fe_len <= 0); + BUG_ON(ex.fe_group != ac->ac_g_ex.fe_group); + BUG_ON(ex.fe_start != ac->ac_g_ex.fe_start); + ac->ac_found++; + ac->ac_b_ex = ex; + ext4_mb_use_best_found(ac, e4b); + } +out: + ext4_unlock_group(ac->ac_sb, group); + ext4_mb_unload_buddy(e4b); + + return 0; +} + +/* + * The routine scans buddy structures (not bitmap!) from given order + * to max order and tries to find big enough chunk to satisfy the req + */ +static noinline_for_stack +void ext4_mb_simple_scan_group(struct ext4_allocation_context *ac, + struct ext4_buddy *e4b) +{ + struct super_block *sb = ac->ac_sb; + struct ext4_group_info *grp = e4b->bd_info; + void *buddy; + int i; + int k; + int max; + + BUG_ON(ac->ac_2order <= 0); + for (i = ac->ac_2order; i < MB_NUM_ORDERS(sb); i++) { + if (grp->bb_counters[i] == 0) + continue; + + buddy = mb_find_buddy(e4b, i, &max); + if (WARN_RATELIMIT(buddy == NULL, + "ext4: mb_simple_scan_group: mb_find_buddy failed, (%d)\n", i)) + continue; + + k = mb_find_next_zero_bit(buddy, max, 0); + if (k >= max) { + ext4_mark_group_bitmap_corrupted(ac->ac_sb, + e4b->bd_group, + EXT4_GROUP_INFO_BBITMAP_CORRUPT); + ext4_grp_locked_error(ac->ac_sb, e4b->bd_group, 0, 0, + "%d free clusters of order %d. But found 0", + grp->bb_counters[i], i); + break; + } + ac->ac_found++; + ac->ac_cX_found[ac->ac_criteria]++; + + ac->ac_b_ex.fe_len = 1 << i; + ac->ac_b_ex.fe_start = k << i; + ac->ac_b_ex.fe_group = e4b->bd_group; + + ext4_mb_use_best_found(ac, e4b); + + BUG_ON(ac->ac_f_ex.fe_len != ac->ac_g_ex.fe_len); + + if (EXT4_SB(sb)->s_mb_stats) + atomic_inc(&EXT4_SB(sb)->s_bal_2orders); + + break; + } +} + +/* + * The routine scans the group and measures all found extents. + * In order to optimize scanning, caller must pass number of + * free blocks in the group, so the routine can know upper limit. + */ +static noinline_for_stack +void ext4_mb_complex_scan_group(struct ext4_allocation_context *ac, + struct ext4_buddy *e4b) +{ + struct super_block *sb = ac->ac_sb; + void *bitmap = e4b->bd_bitmap; + struct ext4_free_extent ex; + int i, j, freelen; + int free; + + free = e4b->bd_info->bb_free; + if (WARN_ON(free <= 0)) + return; + + i = e4b->bd_info->bb_first_free; + + while (free && ac->ac_status == AC_STATUS_CONTINUE) { + i = mb_find_next_zero_bit(bitmap, + EXT4_CLUSTERS_PER_GROUP(sb), i); + if (i >= EXT4_CLUSTERS_PER_GROUP(sb)) { + /* + * IF we have corrupt bitmap, we won't find any + * free blocks even though group info says we + * have free blocks + */ + ext4_mark_group_bitmap_corrupted(sb, e4b->bd_group, + EXT4_GROUP_INFO_BBITMAP_CORRUPT); + ext4_grp_locked_error(sb, e4b->bd_group, 0, 0, + "%d free clusters as per " + "group info. But bitmap says 0", + free); + break; + } + + if (!ext4_mb_cr_expensive(ac->ac_criteria)) { + /* + * In CR_GOAL_LEN_FAST and CR_BEST_AVAIL_LEN, we are + * sure that this group will have a large enough + * continuous free extent, so skip over the smaller free + * extents + */ + j = mb_find_next_bit(bitmap, + EXT4_CLUSTERS_PER_GROUP(sb), i); + freelen = j - i; + + if (freelen < ac->ac_g_ex.fe_len) { + i = j; + free -= freelen; + continue; + } + } + + mb_find_extent(e4b, i, ac->ac_g_ex.fe_len, &ex); + if (WARN_ON(ex.fe_len <= 0)) + break; + if (free < ex.fe_len) { + ext4_mark_group_bitmap_corrupted(sb, e4b->bd_group, + EXT4_GROUP_INFO_BBITMAP_CORRUPT); + ext4_grp_locked_error(sb, e4b->bd_group, 0, 0, + "%d free clusters as per " + "group info. But got %d blocks", + free, ex.fe_len); + /* + * The number of free blocks differs. This mostly + * indicate that the bitmap is corrupt. So exit + * without claiming the space. + */ + break; + } + ex.fe_logical = 0xDEADC0DE; /* debug value */ + ext4_mb_measure_extent(ac, &ex, e4b); + + i += ex.fe_len; + free -= ex.fe_len; + } + + ext4_mb_check_limits(ac, e4b, 1); +} + +/* + * This is a special case for storages like raid5 + * we try to find stripe-aligned chunks for stripe-size-multiple requests + */ +static noinline_for_stack +void ext4_mb_scan_aligned(struct ext4_allocation_context *ac, + struct ext4_buddy *e4b) +{ + struct super_block *sb = ac->ac_sb; + struct ext4_sb_info *sbi = EXT4_SB(sb); + void *bitmap = e4b->bd_bitmap; + struct ext4_free_extent ex; + ext4_fsblk_t first_group_block; + ext4_fsblk_t a; + ext4_grpblk_t i, stripe; + int max; + + BUG_ON(sbi->s_stripe == 0); + + /* find first stripe-aligned block in group */ + first_group_block = ext4_group_first_block_no(sb, e4b->bd_group); + + a = first_group_block + sbi->s_stripe - 1; + do_div(a, sbi->s_stripe); + i = (a * sbi->s_stripe) - first_group_block; + + stripe = EXT4_NUM_B2C(sbi, sbi->s_stripe); + i = EXT4_B2C(sbi, i); + while (i < EXT4_CLUSTERS_PER_GROUP(sb)) { + if (!mb_test_bit(i, bitmap)) { + max = mb_find_extent(e4b, i, stripe, &ex); + if (max >= stripe) { + ac->ac_found++; + ac->ac_cX_found[ac->ac_criteria]++; + ex.fe_logical = 0xDEADF00D; /* debug value */ + ac->ac_b_ex = ex; + ext4_mb_use_best_found(ac, e4b); + break; + } + } + i += stripe; + } +} + +static void __ext4_mb_scan_group(struct ext4_allocation_context *ac) +{ + bool is_stripe_aligned; + struct ext4_sb_info *sbi; + enum criteria cr = ac->ac_criteria; + + ac->ac_groups_scanned++; + if (cr == CR_POWER2_ALIGNED) + return ext4_mb_simple_scan_group(ac, ac->ac_e4b); + + sbi = EXT4_SB(ac->ac_sb); + is_stripe_aligned = false; + if ((sbi->s_stripe >= sbi->s_cluster_ratio) && + !(ac->ac_g_ex.fe_len % EXT4_NUM_B2C(sbi, sbi->s_stripe))) + is_stripe_aligned = true; + + if ((cr == CR_GOAL_LEN_FAST || cr == CR_BEST_AVAIL_LEN) && + is_stripe_aligned) + ext4_mb_scan_aligned(ac, ac->ac_e4b); + + if (ac->ac_status == AC_STATUS_CONTINUE) + ext4_mb_complex_scan_group(ac, ac->ac_e4b); +} + +/* + * This is also called BEFORE we load the buddy bitmap. + * Returns either 1 or 0 indicating that the group is either suitable + * for the allocation or not. + */ +static bool ext4_mb_good_group(struct ext4_allocation_context *ac, + ext4_group_t group, enum criteria cr) +{ + ext4_grpblk_t free, fragments; + int flex_size = ext4_flex_bg_size(EXT4_SB(ac->ac_sb)); + struct ext4_group_info *grp = ext4_get_group_info(ac->ac_sb, group); + + BUG_ON(cr < CR_POWER2_ALIGNED || cr >= EXT4_MB_NUM_CRS); + + if (unlikely(!grp || EXT4_MB_GRP_BBITMAP_CORRUPT(grp))) + return false; + + free = grp->bb_free; + if (free == 0) + return false; + + fragments = grp->bb_fragments; + if (fragments == 0) + return false; + + switch (cr) { + case CR_POWER2_ALIGNED: + BUG_ON(ac->ac_2order == 0); + + /* Avoid using the first bg of a flexgroup for data files */ + if ((ac->ac_flags & EXT4_MB_HINT_DATA) && + (flex_size >= EXT4_FLEX_SIZE_DIR_ALLOC_SCHEME) && + ((group % flex_size) == 0)) + return false; + + if (free < ac->ac_g_ex.fe_len) + return false; + + if (ac->ac_2order >= MB_NUM_ORDERS(ac->ac_sb)) + return true; + + if (grp->bb_largest_free_order < ac->ac_2order) + return false; + + return true; + case CR_GOAL_LEN_FAST: + case CR_BEST_AVAIL_LEN: + if ((free / fragments) >= ac->ac_g_ex.fe_len) + return true; + break; + case CR_GOAL_LEN_SLOW: + if (free >= ac->ac_g_ex.fe_len) + return true; + break; + case CR_ANY_FREE: + return true; + default: + BUG(); + } + + return false; +} + +/* + * This could return negative error code if something goes wrong + * during ext4_mb_init_group(). This should not be called with + * ext4_lock_group() held. + * + * Note: because we are conditionally operating with the group lock in + * the EXT4_MB_STRICT_CHECK case, we need to fake out sparse in this + * function using __acquire and __release. This means we need to be + * super careful before messing with the error path handling via "goto + * out"! + */ +static int ext4_mb_good_group_nolock(struct ext4_allocation_context *ac, + ext4_group_t group, enum criteria cr) +{ + struct ext4_group_info *grp = ext4_get_group_info(ac->ac_sb, group); + struct super_block *sb = ac->ac_sb; + struct ext4_sb_info *sbi = EXT4_SB(sb); + bool should_lock = ac->ac_flags & EXT4_MB_STRICT_CHECK; + ext4_grpblk_t free; + int ret = 0; + + if (!grp) + return -EFSCORRUPTED; + if (sbi->s_mb_stats) + atomic64_inc(&sbi->s_bal_cX_groups_considered[ac->ac_criteria]); + if (should_lock) { + ext4_lock_group(sb, group); + __release(ext4_group_lock_ptr(sb, group)); + } + free = grp->bb_free; + if (free == 0) + goto out; + /* + * In all criterias except CR_ANY_FREE we try to avoid groups that + * can't possibly satisfy the full goal request due to insufficient + * free blocks. + */ + if (cr < CR_ANY_FREE && free < ac->ac_g_ex.fe_len) + goto out; + if (unlikely(EXT4_MB_GRP_BBITMAP_CORRUPT(grp))) + goto out; + if (should_lock) { + __acquire(ext4_group_lock_ptr(sb, group)); + ext4_unlock_group(sb, group); + } + + /* We only do this if the grp has never been initialized */ + if (unlikely(EXT4_MB_GRP_NEED_INIT(grp))) { + struct ext4_group_desc *gdp = + ext4_get_group_desc(sb, group, NULL); + int ret; + + /* + * CR_POWER2_ALIGNED/CR_GOAL_LEN_FAST is a very optimistic + * search to find large good chunks almost for free. If buddy + * data is not ready, then this optimization makes no sense. But + * we never skip the first block group in a flex_bg, since this + * gets used for metadata block allocation, and we want to make + * sure we locate metadata blocks in the first block group in + * the flex_bg if possible. + */ + if (!ext4_mb_cr_expensive(cr) && + (!sbi->s_log_groups_per_flex || + ((group & ((1 << sbi->s_log_groups_per_flex) - 1)) != 0)) && + !(ext4_has_group_desc_csum(sb) && + (gdp->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)))) + return 0; + ret = ext4_mb_init_group(sb, group, GFP_NOFS); + if (ret) + return ret; + } + + if (should_lock) { + ext4_lock_group(sb, group); + __release(ext4_group_lock_ptr(sb, group)); + } + ret = ext4_mb_good_group(ac, group, cr); +out: + if (should_lock) { + __acquire(ext4_group_lock_ptr(sb, group)); + ext4_unlock_group(sb, group); + } + return ret; +} + +/* + * Start prefetching @nr block bitmaps starting at @group. + * Return the next group which needs to be prefetched. + */ +ext4_group_t ext4_mb_prefetch(struct super_block *sb, ext4_group_t group, + unsigned int nr, int *cnt) +{ + ext4_group_t ngroups = ext4_get_groups_count(sb); + struct buffer_head *bh; + struct blk_plug plug; + + blk_start_plug(&plug); + while (nr-- > 0) { + struct ext4_group_desc *gdp = ext4_get_group_desc(sb, group, + NULL); + struct ext4_group_info *grp = ext4_get_group_info(sb, group); + + /* + * Prefetch block groups with free blocks; but don't + * bother if it is marked uninitialized on disk, since + * it won't require I/O to read. Also only try to + * prefetch once, so we avoid getblk() call, which can + * be expensive. + */ + if (gdp && grp && !EXT4_MB_GRP_TEST_AND_SET_READ(grp) && + EXT4_MB_GRP_NEED_INIT(grp) && + ext4_free_group_clusters(sb, gdp) > 0 ) { + bh = ext4_read_block_bitmap_nowait(sb, group, true); + if (bh && !IS_ERR(bh)) { + if (!buffer_uptodate(bh) && cnt) + (*cnt)++; + brelse(bh); + } + } + if (++group >= ngroups) + group = 0; + } + blk_finish_plug(&plug); + return group; +} + +/* + * Batch reads of the block allocation bitmaps to get + * multiple READs in flight; limit prefetching at inexpensive + * CR, otherwise mballoc can spend a lot of time loading + * imperfect groups + */ +static void ext4_mb_might_prefetch(struct ext4_allocation_context *ac, + ext4_group_t group) +{ + struct ext4_sb_info *sbi; + + if (ac->ac_prefetch_grp != group) + return; + + sbi = EXT4_SB(ac->ac_sb); + if (ext4_mb_cr_expensive(ac->ac_criteria) || + ac->ac_prefetch_ios < sbi->s_mb_prefetch_limit) { + unsigned int nr = sbi->s_mb_prefetch; + + if (ext4_has_feature_flex_bg(ac->ac_sb)) { + nr = 1 << sbi->s_log_groups_per_flex; + nr -= group & (nr - 1); + nr = umin(nr, sbi->s_mb_prefetch); + } + + ac->ac_prefetch_nr = nr; + ac->ac_prefetch_grp = ext4_mb_prefetch(ac->ac_sb, group, nr, + &ac->ac_prefetch_ios); + } +} + +/* + * Prefetching reads the block bitmap into the buffer cache; but we + * need to make sure that the buddy bitmap in the page cache has been + * initialized. Note that ext4_mb_init_group() will block if the I/O + * is not yet completed, or indeed if it was not initiated by + * ext4_mb_prefetch did not start the I/O. + * + * TODO: We should actually kick off the buddy bitmap setup in a work + * queue when the buffer I/O is completed, so that we don't block + * waiting for the block allocation bitmap read to finish when + * ext4_mb_prefetch_fini is called from ext4_mb_regular_allocator(). + */ +void ext4_mb_prefetch_fini(struct super_block *sb, ext4_group_t group, + unsigned int nr) +{ + struct ext4_group_desc *gdp; + struct ext4_group_info *grp; + + while (nr-- > 0) { + if (!group) + group = ext4_get_groups_count(sb); + group--; + gdp = ext4_get_group_desc(sb, group, NULL); + grp = ext4_get_group_info(sb, group); + + if (grp && gdp && EXT4_MB_GRP_NEED_INIT(grp) && + ext4_free_group_clusters(sb, gdp) > 0) { + if (ext4_mb_init_group(sb, group, GFP_NOFS)) + break; + } + } +} + +static int ext4_mb_scan_group(struct ext4_allocation_context *ac, + ext4_group_t group) +{ + int ret; + struct super_block *sb = ac->ac_sb; + enum criteria cr = ac->ac_criteria; + + ext4_mb_might_prefetch(ac, group); + + /* prevent unnecessary buddy loading. */ + if (cr < CR_ANY_FREE && spin_is_locked(ext4_group_lock_ptr(sb, group))) + return 0; + + /* This now checks without needing the buddy page */ + ret = ext4_mb_good_group_nolock(ac, group, cr); + if (ret <= 0) { + if (!ac->ac_first_err) + ac->ac_first_err = ret; + return 0; + } + + ret = ext4_mb_load_buddy(sb, group, ac->ac_e4b); + if (ret) + return ret; + + /* skip busy group */ + if (cr >= CR_ANY_FREE) + ext4_lock_group(sb, group); + else if (!ext4_try_lock_group(sb, group)) + goto out_unload; + + /* We need to check again after locking the block group. */ + if (unlikely(!ext4_mb_good_group(ac, group, cr))) + goto out_unlock; + + __ext4_mb_scan_group(ac); + +out_unlock: + ext4_unlock_group(sb, group); +out_unload: + ext4_mb_unload_buddy(ac->ac_e4b); + return ret; +} + +static noinline_for_stack int +ext4_mb_regular_allocator(struct ext4_allocation_context *ac) +{ + ext4_group_t i; + int err = 0; + struct super_block *sb = ac->ac_sb; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_buddy e4b; + + BUG_ON(ac->ac_status == AC_STATUS_FOUND); + + /* first, try the goal */ + err = ext4_mb_find_by_goal(ac, &e4b); + if (err || ac->ac_status == AC_STATUS_FOUND) + goto out; + + if (unlikely(ac->ac_flags & EXT4_MB_HINT_GOAL_ONLY)) + goto out; + + /* + * ac->ac_2order is set only if the fe_len is a power of 2 + * if ac->ac_2order is set we also set criteria to CR_POWER2_ALIGNED + * so that we try exact allocation using buddy. + */ + i = fls(ac->ac_g_ex.fe_len); + ac->ac_2order = 0; + /* + * We search using buddy data only if the order of the request + * is greater than equal to the sbi_s_mb_order2_reqs + * You can tune it via /sys/fs/ext4/<partition>/mb_order2_req + * We also support searching for power-of-two requests only for + * requests upto maximum buddy size we have constructed. + */ + if (i >= sbi->s_mb_order2_reqs && i <= MB_NUM_ORDERS(sb)) { + if (is_power_of_2(ac->ac_g_ex.fe_len)) + ac->ac_2order = array_index_nospec(i - 1, + MB_NUM_ORDERS(sb)); + } + + /* if stream allocation is enabled, use global goal */ + if (ac->ac_flags & EXT4_MB_STREAM_ALLOC) { + int hash = ac->ac_inode->i_ino % sbi->s_mb_nr_global_goals; + + ac->ac_g_ex.fe_group = READ_ONCE(sbi->s_mb_last_groups[hash]); + ac->ac_g_ex.fe_start = -1; + ac->ac_flags &= ~EXT4_MB_HINT_TRY_GOAL; + } + + /* + * Let's just scan groups to find more-less suitable blocks We + * start with CR_GOAL_LEN_FAST, unless it is power of 2 + * aligned, in which case let's do that faster approach first. + */ + ac->ac_criteria = CR_GOAL_LEN_FAST; + if (ac->ac_2order) + ac->ac_criteria = CR_POWER2_ALIGNED; + + ac->ac_e4b = &e4b; + ac->ac_prefetch_ios = 0; + ac->ac_first_err = 0; +repeat: + while (ac->ac_criteria < EXT4_MB_NUM_CRS) { + err = ext4_mb_scan_groups(ac); + if (err) + goto out; + + if (ac->ac_status != AC_STATUS_CONTINUE) + break; + } + + if (ac->ac_b_ex.fe_len > 0 && ac->ac_status != AC_STATUS_FOUND && + !(ac->ac_flags & EXT4_MB_HINT_FIRST)) { + /* + * We've been searching too long. Let's try to allocate + * the best chunk we've found so far + */ + ext4_mb_try_best_found(ac, &e4b); + if (ac->ac_status != AC_STATUS_FOUND) { + int lost; + + /* + * Someone more lucky has already allocated it. + * The only thing we can do is just take first + * found block(s) + */ + lost = atomic_inc_return(&sbi->s_mb_lost_chunks); + mb_debug(sb, "lost chunk, group: %u, start: %d, len: %d, lost: %d\n", + ac->ac_b_ex.fe_group, ac->ac_b_ex.fe_start, + ac->ac_b_ex.fe_len, lost); + + ac->ac_b_ex.fe_group = 0; + ac->ac_b_ex.fe_start = 0; + ac->ac_b_ex.fe_len = 0; + ac->ac_status = AC_STATUS_CONTINUE; + ac->ac_flags |= EXT4_MB_HINT_FIRST; + ac->ac_criteria = CR_ANY_FREE; + goto repeat; + } + } + + if (sbi->s_mb_stats && ac->ac_status == AC_STATUS_FOUND) { + atomic64_inc(&sbi->s_bal_cX_hits[ac->ac_criteria]); + if (ac->ac_flags & EXT4_MB_STREAM_ALLOC && + ac->ac_b_ex.fe_group == ac->ac_g_ex.fe_group) + atomic_inc(&sbi->s_bal_stream_goals); + } +out: + if (!err && ac->ac_status != AC_STATUS_FOUND && ac->ac_first_err) + err = ac->ac_first_err; + + mb_debug(sb, "Best len %d, origin len %d, ac_status %u, ac_flags 0x%x, cr %d ret %d\n", + ac->ac_b_ex.fe_len, ac->ac_o_ex.fe_len, ac->ac_status, + ac->ac_flags, ac->ac_criteria, err); + + if (ac->ac_prefetch_nr) + ext4_mb_prefetch_fini(sb, ac->ac_prefetch_grp, ac->ac_prefetch_nr); + + return err; +} + +static void *ext4_mb_seq_groups_start(struct seq_file *seq, loff_t *pos) +{ + struct super_block *sb = pde_data(file_inode(seq->file)); + ext4_group_t group; + + if (*pos < 0 || *pos >= ext4_get_groups_count(sb)) + return NULL; + group = *pos + 1; + return (void *) ((unsigned long) group); +} + +static void *ext4_mb_seq_groups_next(struct seq_file *seq, void *v, loff_t *pos) +{ + struct super_block *sb = pde_data(file_inode(seq->file)); + ext4_group_t group; + + ++*pos; + if (*pos < 0 || *pos >= ext4_get_groups_count(sb)) + return NULL; + group = *pos + 1; + return (void *) ((unsigned long) group); +} + +static int ext4_mb_seq_groups_show(struct seq_file *seq, void *v) +{ + struct super_block *sb = pde_data(file_inode(seq->file)); + ext4_group_t group = (ext4_group_t) ((unsigned long) v); + int i, err; + char nbuf[16]; + struct ext4_buddy e4b; + struct ext4_group_info *grinfo; + unsigned char blocksize_bits = min_t(unsigned char, + sb->s_blocksize_bits, + EXT4_MAX_BLOCK_LOG_SIZE); + DEFINE_RAW_FLEX(struct ext4_group_info, sg, bb_counters, + EXT4_MAX_BLOCK_LOG_SIZE + 2); + + group--; + if (group == 0) + seq_puts(seq, "#group: free frags first [" + " 2^0 2^1 2^2 2^3 2^4 2^5 2^6 " + " 2^7 2^8 2^9 2^10 2^11 2^12 2^13 ]\n"); + + i = (blocksize_bits + 2) * sizeof(sg->bb_counters[0]) + + sizeof(struct ext4_group_info); + + grinfo = ext4_get_group_info(sb, group); + if (!grinfo) + return 0; + /* Load the group info in memory only if not already loaded. */ + if (unlikely(EXT4_MB_GRP_NEED_INIT(grinfo))) { + err = ext4_mb_load_buddy(sb, group, &e4b); + if (err) { + seq_printf(seq, "#%-5u: %s\n", group, ext4_decode_error(NULL, err, nbuf)); + return 0; + } + ext4_mb_unload_buddy(&e4b); + } + + /* + * We care only about free space counters in the group info and + * these are safe to access even after the buddy has been unloaded + */ + memcpy(sg, grinfo, i); + seq_printf(seq, "#%-5u: %-5u %-5u %-5u [", group, sg->bb_free, + sg->bb_fragments, sg->bb_first_free); + for (i = 0; i <= 13; i++) + seq_printf(seq, " %-5u", i <= blocksize_bits + 1 ? + sg->bb_counters[i] : 0); + seq_puts(seq, " ]"); + if (EXT4_MB_GRP_BBITMAP_CORRUPT(sg)) + seq_puts(seq, " Block bitmap corrupted!"); + seq_putc(seq, '\n'); + return 0; +} + +static void ext4_mb_seq_groups_stop(struct seq_file *seq, void *v) +{ +} + +const struct seq_operations ext4_mb_seq_groups_ops = { + .start = ext4_mb_seq_groups_start, + .next = ext4_mb_seq_groups_next, + .stop = ext4_mb_seq_groups_stop, + .show = ext4_mb_seq_groups_show, +}; + +int ext4_seq_mb_stats_show(struct seq_file *seq, void *offset) +{ + struct super_block *sb = seq->private; + struct ext4_sb_info *sbi = EXT4_SB(sb); + + seq_puts(seq, "mballoc:\n"); + if (!sbi->s_mb_stats) { + seq_puts(seq, "\tmb stats collection turned off.\n"); + seq_puts( + seq, + "\tTo enable, please write \"1\" to sysfs file mb_stats.\n"); + return 0; + } + seq_printf(seq, "\treqs: %u\n", atomic_read(&sbi->s_bal_reqs)); + seq_printf(seq, "\tsuccess: %u\n", atomic_read(&sbi->s_bal_success)); + + seq_printf(seq, "\tgroups_scanned: %u\n", + atomic_read(&sbi->s_bal_groups_scanned)); + + /* CR_POWER2_ALIGNED stats */ + seq_puts(seq, "\tcr_p2_aligned_stats:\n"); + seq_printf(seq, "\t\thits: %llu\n", + atomic64_read(&sbi->s_bal_cX_hits[CR_POWER2_ALIGNED])); + seq_printf( + seq, "\t\tgroups_considered: %llu\n", + atomic64_read( + &sbi->s_bal_cX_groups_considered[CR_POWER2_ALIGNED])); + seq_printf(seq, "\t\textents_scanned: %u\n", + atomic_read(&sbi->s_bal_cX_ex_scanned[CR_POWER2_ALIGNED])); + seq_printf(seq, "\t\tuseless_loops: %llu\n", + atomic64_read(&sbi->s_bal_cX_failed[CR_POWER2_ALIGNED])); + + /* CR_GOAL_LEN_FAST stats */ + seq_puts(seq, "\tcr_goal_fast_stats:\n"); + seq_printf(seq, "\t\thits: %llu\n", + atomic64_read(&sbi->s_bal_cX_hits[CR_GOAL_LEN_FAST])); + seq_printf(seq, "\t\tgroups_considered: %llu\n", + atomic64_read( + &sbi->s_bal_cX_groups_considered[CR_GOAL_LEN_FAST])); + seq_printf(seq, "\t\textents_scanned: %u\n", + atomic_read(&sbi->s_bal_cX_ex_scanned[CR_GOAL_LEN_FAST])); + seq_printf(seq, "\t\tuseless_loops: %llu\n", + atomic64_read(&sbi->s_bal_cX_failed[CR_GOAL_LEN_FAST])); + + /* CR_BEST_AVAIL_LEN stats */ + seq_puts(seq, "\tcr_best_avail_stats:\n"); + seq_printf(seq, "\t\thits: %llu\n", + atomic64_read(&sbi->s_bal_cX_hits[CR_BEST_AVAIL_LEN])); + seq_printf( + seq, "\t\tgroups_considered: %llu\n", + atomic64_read( + &sbi->s_bal_cX_groups_considered[CR_BEST_AVAIL_LEN])); + seq_printf(seq, "\t\textents_scanned: %u\n", + atomic_read(&sbi->s_bal_cX_ex_scanned[CR_BEST_AVAIL_LEN])); + seq_printf(seq, "\t\tuseless_loops: %llu\n", + atomic64_read(&sbi->s_bal_cX_failed[CR_BEST_AVAIL_LEN])); + + /* CR_GOAL_LEN_SLOW stats */ + seq_puts(seq, "\tcr_goal_slow_stats:\n"); + seq_printf(seq, "\t\thits: %llu\n", + atomic64_read(&sbi->s_bal_cX_hits[CR_GOAL_LEN_SLOW])); + seq_printf(seq, "\t\tgroups_considered: %llu\n", + atomic64_read( + &sbi->s_bal_cX_groups_considered[CR_GOAL_LEN_SLOW])); + seq_printf(seq, "\t\textents_scanned: %u\n", + atomic_read(&sbi->s_bal_cX_ex_scanned[CR_GOAL_LEN_SLOW])); + seq_printf(seq, "\t\tuseless_loops: %llu\n", + atomic64_read(&sbi->s_bal_cX_failed[CR_GOAL_LEN_SLOW])); + + /* CR_ANY_FREE stats */ + seq_puts(seq, "\tcr_any_free_stats:\n"); + seq_printf(seq, "\t\thits: %llu\n", + atomic64_read(&sbi->s_bal_cX_hits[CR_ANY_FREE])); + seq_printf( + seq, "\t\tgroups_considered: %llu\n", + atomic64_read(&sbi->s_bal_cX_groups_considered[CR_ANY_FREE])); + seq_printf(seq, "\t\textents_scanned: %u\n", + atomic_read(&sbi->s_bal_cX_ex_scanned[CR_ANY_FREE])); + seq_printf(seq, "\t\tuseless_loops: %llu\n", + atomic64_read(&sbi->s_bal_cX_failed[CR_ANY_FREE])); + + /* Aggregates */ + seq_printf(seq, "\textents_scanned: %u\n", + atomic_read(&sbi->s_bal_ex_scanned)); + seq_printf(seq, "\t\tgoal_hits: %u\n", atomic_read(&sbi->s_bal_goals)); + seq_printf(seq, "\t\tstream_goal_hits: %u\n", + atomic_read(&sbi->s_bal_stream_goals)); + seq_printf(seq, "\t\tlen_goal_hits: %u\n", + atomic_read(&sbi->s_bal_len_goals)); + seq_printf(seq, "\t\t2^n_hits: %u\n", atomic_read(&sbi->s_bal_2orders)); + seq_printf(seq, "\t\tbreaks: %u\n", atomic_read(&sbi->s_bal_breaks)); + seq_printf(seq, "\t\tlost: %u\n", atomic_read(&sbi->s_mb_lost_chunks)); + seq_printf(seq, "\tbuddies_generated: %u/%u\n", + atomic_read(&sbi->s_mb_buddies_generated), + ext4_get_groups_count(sb)); + seq_printf(seq, "\tbuddies_time_used: %llu\n", + atomic64_read(&sbi->s_mb_generation_time)); + seq_printf(seq, "\tpreallocated: %u\n", + atomic_read(&sbi->s_mb_preallocated)); + seq_printf(seq, "\tdiscarded: %u\n", atomic_read(&sbi->s_mb_discarded)); + return 0; +} + +static void *ext4_mb_seq_structs_summary_start(struct seq_file *seq, loff_t *pos) +{ + struct super_block *sb = pde_data(file_inode(seq->file)); + unsigned long position; + + if (*pos < 0 || *pos >= 2*MB_NUM_ORDERS(sb)) + return NULL; + position = *pos + 1; + return (void *) ((unsigned long) position); +} + +static void *ext4_mb_seq_structs_summary_next(struct seq_file *seq, void *v, loff_t *pos) +{ + struct super_block *sb = pde_data(file_inode(seq->file)); + unsigned long position; + + ++*pos; + if (*pos < 0 || *pos >= 2*MB_NUM_ORDERS(sb)) + return NULL; + position = *pos + 1; + return (void *) ((unsigned long) position); +} + +static int ext4_mb_seq_structs_summary_show(struct seq_file *seq, void *v) +{ + struct super_block *sb = pde_data(file_inode(seq->file)); + struct ext4_sb_info *sbi = EXT4_SB(sb); + unsigned long position = ((unsigned long) v); + struct ext4_group_info *grp; + unsigned int count; + unsigned long idx; + + position--; + if (position >= MB_NUM_ORDERS(sb)) { + position -= MB_NUM_ORDERS(sb); + if (position == 0) + seq_puts(seq, "avg_fragment_size_lists:\n"); + + count = 0; + xa_for_each(&sbi->s_mb_avg_fragment_size[position], idx, grp) + count++; + seq_printf(seq, "\tlist_order_%u_groups: %u\n", + (unsigned int)position, count); + return 0; + } + + if (position == 0) { + seq_printf(seq, "optimize_scan: %d\n", + test_opt2(sb, MB_OPTIMIZE_SCAN) ? 1 : 0); + seq_puts(seq, "max_free_order_lists:\n"); + } + count = 0; + xa_for_each(&sbi->s_mb_largest_free_orders[position], idx, grp) + count++; + seq_printf(seq, "\tlist_order_%u_groups: %u\n", + (unsigned int)position, count); + + return 0; +} + +static void ext4_mb_seq_structs_summary_stop(struct seq_file *seq, void *v) +{ +} + +const struct seq_operations ext4_mb_seq_structs_summary_ops = { + .start = ext4_mb_seq_structs_summary_start, + .next = ext4_mb_seq_structs_summary_next, + .stop = ext4_mb_seq_structs_summary_stop, + .show = ext4_mb_seq_structs_summary_show, +}; + +static struct kmem_cache *get_groupinfo_cache(int blocksize_bits) +{ + int cache_index = blocksize_bits - EXT4_MIN_BLOCK_LOG_SIZE; + struct kmem_cache *cachep = ext4_groupinfo_caches[cache_index]; + + BUG_ON(!cachep); + return cachep; +} + +/* + * Allocate the top-level s_group_info array for the specified number + * of groups + */ +int ext4_mb_alloc_groupinfo(struct super_block *sb, ext4_group_t ngroups) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + unsigned size; + struct ext4_group_info ***old_groupinfo, ***new_groupinfo; + + size = (ngroups + EXT4_DESC_PER_BLOCK(sb) - 1) >> + EXT4_DESC_PER_BLOCK_BITS(sb); + if (size <= sbi->s_group_info_size) + return 0; + + size = roundup_pow_of_two(sizeof(*sbi->s_group_info) * size); + new_groupinfo = kvzalloc(size, GFP_KERNEL); + if (!new_groupinfo) { + ext4_msg(sb, KERN_ERR, "can't allocate buddy meta group"); + return -ENOMEM; + } + rcu_read_lock(); + old_groupinfo = rcu_dereference(sbi->s_group_info); + if (old_groupinfo) + memcpy(new_groupinfo, old_groupinfo, + sbi->s_group_info_size * sizeof(*sbi->s_group_info)); + rcu_read_unlock(); + rcu_assign_pointer(sbi->s_group_info, new_groupinfo); + sbi->s_group_info_size = size / sizeof(*sbi->s_group_info); + if (old_groupinfo) + ext4_kvfree_array_rcu(old_groupinfo); + ext4_debug("allocated s_groupinfo array for %d meta_bg's\n", + sbi->s_group_info_size); + return 0; +} + +/* Create and initialize ext4_group_info data for the given group. */ +int ext4_mb_add_groupinfo(struct super_block *sb, ext4_group_t group, + struct ext4_group_desc *desc) +{ + int i; + int metalen = 0; + int idx = group >> EXT4_DESC_PER_BLOCK_BITS(sb); + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_group_info **meta_group_info; + struct kmem_cache *cachep = get_groupinfo_cache(sb->s_blocksize_bits); + + /* + * First check if this group is the first of a reserved block. + * If it's true, we have to allocate a new table of pointers + * to ext4_group_info structures + */ + if (group % EXT4_DESC_PER_BLOCK(sb) == 0) { + metalen = sizeof(*meta_group_info) << + EXT4_DESC_PER_BLOCK_BITS(sb); + meta_group_info = kmalloc(metalen, GFP_NOFS); + if (meta_group_info == NULL) { + ext4_msg(sb, KERN_ERR, "can't allocate mem " + "for a buddy group"); + return -ENOMEM; + } + rcu_read_lock(); + rcu_dereference(sbi->s_group_info)[idx] = meta_group_info; + rcu_read_unlock(); + } + + meta_group_info = sbi_array_rcu_deref(sbi, s_group_info, idx); + i = group & (EXT4_DESC_PER_BLOCK(sb) - 1); + + meta_group_info[i] = kmem_cache_zalloc(cachep, GFP_NOFS); + if (meta_group_info[i] == NULL) { + ext4_msg(sb, KERN_ERR, "can't allocate buddy mem"); + goto exit_group_info; + } + set_bit(EXT4_GROUP_INFO_NEED_INIT_BIT, + &(meta_group_info[i]->bb_state)); + + /* + * initialize bb_free to be able to skip + * empty groups without initialization + */ + if (ext4_has_group_desc_csum(sb) && + (desc->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT))) { + meta_group_info[i]->bb_free = + ext4_free_clusters_after_init(sb, group, desc); + } else { + meta_group_info[i]->bb_free = + ext4_free_group_clusters(sb, desc); + } + + INIT_LIST_HEAD(&meta_group_info[i]->bb_prealloc_list); + init_rwsem(&meta_group_info[i]->alloc_sem); + meta_group_info[i]->bb_free_root = RB_ROOT; + meta_group_info[i]->bb_largest_free_order = -1; /* uninit */ + meta_group_info[i]->bb_avg_fragment_size_order = -1; /* uninit */ + meta_group_info[i]->bb_group = group; + + mb_group_bb_bitmap_alloc(sb, meta_group_info[i], group); + return 0; + +exit_group_info: + /* If a meta_group_info table has been allocated, release it now */ + if (group % EXT4_DESC_PER_BLOCK(sb) == 0) { + struct ext4_group_info ***group_info; + + rcu_read_lock(); + group_info = rcu_dereference(sbi->s_group_info); + kfree(group_info[idx]); + group_info[idx] = NULL; + rcu_read_unlock(); + } + return -ENOMEM; +} /* ext4_mb_add_groupinfo */ + +static int ext4_mb_init_backend(struct super_block *sb) +{ + ext4_group_t ngroups = ext4_get_groups_count(sb); + ext4_group_t i; + struct ext4_sb_info *sbi = EXT4_SB(sb); + int err; + struct ext4_group_desc *desc; + struct ext4_group_info ***group_info; + struct kmem_cache *cachep; + + err = ext4_mb_alloc_groupinfo(sb, ngroups); + if (err) + return err; + + sbi->s_buddy_cache = new_inode(sb); + if (sbi->s_buddy_cache == NULL) { + ext4_msg(sb, KERN_ERR, "can't get new inode"); + goto err_freesgi; + } + /* To avoid potentially colliding with an valid on-disk inode number, + * use EXT4_BAD_INO for the buddy cache inode number. This inode is + * not in the inode hash, so it should never be found by iget(), but + * this will avoid confusion if it ever shows up during debugging. */ + sbi->s_buddy_cache->i_ino = EXT4_BAD_INO; + EXT4_I(sbi->s_buddy_cache)->i_disksize = 0; + for (i = 0; i < ngroups; i++) { + cond_resched(); + desc = ext4_get_group_desc(sb, i, NULL); + if (desc == NULL) { + ext4_msg(sb, KERN_ERR, "can't read descriptor %u", i); + goto err_freebuddy; + } + if (ext4_mb_add_groupinfo(sb, i, desc) != 0) + goto err_freebuddy; + } + + if (ext4_has_feature_flex_bg(sb)) { + /* a single flex group is supposed to be read by a single IO. + * 2 ^ s_log_groups_per_flex != UINT_MAX as s_mb_prefetch is + * unsigned integer, so the maximum shift is 32. + */ + if (sbi->s_es->s_log_groups_per_flex >= 32) { + ext4_msg(sb, KERN_ERR, "too many log groups per flexible block group"); + goto err_freebuddy; + } + sbi->s_mb_prefetch = min_t(uint, 1 << sbi->s_es->s_log_groups_per_flex, + BLK_MAX_SEGMENT_SIZE >> (sb->s_blocksize_bits - 9)); + sbi->s_mb_prefetch *= 8; /* 8 prefetch IOs in flight at most */ + } else { + sbi->s_mb_prefetch = 32; + } + if (sbi->s_mb_prefetch > ext4_get_groups_count(sb)) + sbi->s_mb_prefetch = ext4_get_groups_count(sb); + /* + * now many real IOs to prefetch within a single allocation at + * CR_POWER2_ALIGNED. Given CR_POWER2_ALIGNED is an CPU-related + * optimization we shouldn't try to load too many groups, at some point + * we should start to use what we've got in memory. + * with an average random access time 5ms, it'd take a second to get + * 200 groups (* N with flex_bg), so let's make this limit 4 + */ + sbi->s_mb_prefetch_limit = sbi->s_mb_prefetch * 4; + if (sbi->s_mb_prefetch_limit > ext4_get_groups_count(sb)) + sbi->s_mb_prefetch_limit = ext4_get_groups_count(sb); + + return 0; + +err_freebuddy: + cachep = get_groupinfo_cache(sb->s_blocksize_bits); + while (i-- > 0) { + struct ext4_group_info *grp = ext4_get_group_info(sb, i); + + if (grp) + kmem_cache_free(cachep, grp); + } + i = sbi->s_group_info_size; + rcu_read_lock(); + group_info = rcu_dereference(sbi->s_group_info); + while (i-- > 0) + kfree(group_info[i]); + rcu_read_unlock(); + iput(sbi->s_buddy_cache); +err_freesgi: + rcu_read_lock(); + kvfree(rcu_dereference(sbi->s_group_info)); + rcu_read_unlock(); + return -ENOMEM; +} + +static void ext4_groupinfo_destroy_slabs(void) +{ + int i; + + for (i = 0; i < NR_GRPINFO_CACHES; i++) { + kmem_cache_destroy(ext4_groupinfo_caches[i]); + ext4_groupinfo_caches[i] = NULL; + } +} + +static int ext4_groupinfo_create_slab(size_t size) +{ + static DEFINE_MUTEX(ext4_grpinfo_slab_create_mutex); + int slab_size; + int blocksize_bits = order_base_2(size); + int cache_index = blocksize_bits - EXT4_MIN_BLOCK_LOG_SIZE; + struct kmem_cache *cachep; + + if (cache_index >= NR_GRPINFO_CACHES) + return -EINVAL; + + if (unlikely(cache_index < 0)) + cache_index = 0; + + mutex_lock(&ext4_grpinfo_slab_create_mutex); + if (ext4_groupinfo_caches[cache_index]) { + mutex_unlock(&ext4_grpinfo_slab_create_mutex); + return 0; /* Already created */ + } + + slab_size = offsetof(struct ext4_group_info, + bb_counters[blocksize_bits + 2]); + + cachep = kmem_cache_create(ext4_groupinfo_slab_names[cache_index], + slab_size, 0, SLAB_RECLAIM_ACCOUNT, + NULL); + + ext4_groupinfo_caches[cache_index] = cachep; + + mutex_unlock(&ext4_grpinfo_slab_create_mutex); + if (!cachep) { + printk(KERN_EMERG + "EXT4-fs: no memory for groupinfo slab cache\n"); + return -ENOMEM; + } + + return 0; +} + +static void ext4_discard_work(struct work_struct *work) +{ + struct ext4_sb_info *sbi = container_of(work, + struct ext4_sb_info, s_discard_work); + struct super_block *sb = sbi->s_sb; + struct ext4_free_data *fd, *nfd; + struct ext4_buddy e4b; + LIST_HEAD(discard_list); + ext4_group_t grp, load_grp; + int err = 0; + + spin_lock(&sbi->s_md_lock); + list_splice_init(&sbi->s_discard_list, &discard_list); + spin_unlock(&sbi->s_md_lock); + + load_grp = UINT_MAX; + list_for_each_entry_safe(fd, nfd, &discard_list, efd_list) { + /* + * If filesystem is umounting or no memory or suffering + * from no space, give up the discard + */ + if ((sb->s_flags & SB_ACTIVE) && !err && + !atomic_read(&sbi->s_retry_alloc_pending)) { + grp = fd->efd_group; + if (grp != load_grp) { + if (load_grp != UINT_MAX) + ext4_mb_unload_buddy(&e4b); + + err = ext4_mb_load_buddy(sb, grp, &e4b); + if (err) { + kmem_cache_free(ext4_free_data_cachep, fd); + load_grp = UINT_MAX; + continue; + } else { + load_grp = grp; + } + } + + ext4_lock_group(sb, grp); + ext4_try_to_trim_range(sb, &e4b, fd->efd_start_cluster, + fd->efd_start_cluster + fd->efd_count - 1, 1); + ext4_unlock_group(sb, grp); + } + kmem_cache_free(ext4_free_data_cachep, fd); + } + + if (load_grp != UINT_MAX) + ext4_mb_unload_buddy(&e4b); +} + +static inline void ext4_mb_avg_fragment_size_destroy(struct ext4_sb_info *sbi) +{ + if (!sbi->s_mb_avg_fragment_size) + return; + + for (int i = 0; i < MB_NUM_ORDERS(sbi->s_sb); i++) + xa_destroy(&sbi->s_mb_avg_fragment_size[i]); + + kfree(sbi->s_mb_avg_fragment_size); + sbi->s_mb_avg_fragment_size = NULL; +} + +static inline void ext4_mb_largest_free_orders_destroy(struct ext4_sb_info *sbi) +{ + if (!sbi->s_mb_largest_free_orders) + return; + + for (int i = 0; i < MB_NUM_ORDERS(sbi->s_sb); i++) + xa_destroy(&sbi->s_mb_largest_free_orders[i]); + + kfree(sbi->s_mb_largest_free_orders); + sbi->s_mb_largest_free_orders = NULL; +} + +int ext4_mb_init(struct super_block *sb) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + unsigned i, j; + unsigned offset, offset_incr; + unsigned max; + int ret; + + i = MB_NUM_ORDERS(sb) * sizeof(*sbi->s_mb_offsets); + + sbi->s_mb_offsets = kmalloc(i, GFP_KERNEL); + if (sbi->s_mb_offsets == NULL) { + ret = -ENOMEM; + goto out; + } + + i = MB_NUM_ORDERS(sb) * sizeof(*sbi->s_mb_maxs); + sbi->s_mb_maxs = kmalloc(i, GFP_KERNEL); + if (sbi->s_mb_maxs == NULL) { + ret = -ENOMEM; + goto out; + } + + ret = ext4_groupinfo_create_slab(sb->s_blocksize); + if (ret < 0) + goto out; + + /* order 0 is regular bitmap */ + sbi->s_mb_maxs[0] = sb->s_blocksize << 3; + sbi->s_mb_offsets[0] = 0; + + i = 1; + offset = 0; + offset_incr = 1 << (sb->s_blocksize_bits - 1); + max = sb->s_blocksize << 2; + do { + sbi->s_mb_offsets[i] = offset; + sbi->s_mb_maxs[i] = max; + offset += offset_incr; + offset_incr = offset_incr >> 1; + max = max >> 1; + i++; + } while (i < MB_NUM_ORDERS(sb)); + + sbi->s_mb_avg_fragment_size = + kmalloc_array(MB_NUM_ORDERS(sb), sizeof(struct xarray), + GFP_KERNEL); + if (!sbi->s_mb_avg_fragment_size) { + ret = -ENOMEM; + goto out; + } + for (i = 0; i < MB_NUM_ORDERS(sb); i++) + xa_init(&sbi->s_mb_avg_fragment_size[i]); + + sbi->s_mb_largest_free_orders = + kmalloc_array(MB_NUM_ORDERS(sb), sizeof(struct xarray), + GFP_KERNEL); + if (!sbi->s_mb_largest_free_orders) { + ret = -ENOMEM; + goto out; + } + for (i = 0; i < MB_NUM_ORDERS(sb); i++) + xa_init(&sbi->s_mb_largest_free_orders[i]); + + spin_lock_init(&sbi->s_md_lock); + atomic_set(&sbi->s_mb_free_pending, 0); + INIT_LIST_HEAD(&sbi->s_freed_data_list[0]); + INIT_LIST_HEAD(&sbi->s_freed_data_list[1]); + INIT_LIST_HEAD(&sbi->s_discard_list); + INIT_WORK(&sbi->s_discard_work, ext4_discard_work); + atomic_set(&sbi->s_retry_alloc_pending, 0); + + sbi->s_mb_max_to_scan = MB_DEFAULT_MAX_TO_SCAN; + sbi->s_mb_min_to_scan = MB_DEFAULT_MIN_TO_SCAN; + sbi->s_mb_stats = MB_DEFAULT_STATS; + sbi->s_mb_stream_request = MB_DEFAULT_STREAM_THRESHOLD; + sbi->s_mb_order2_reqs = MB_DEFAULT_ORDER2_REQS; + sbi->s_mb_best_avail_max_trim_order = MB_DEFAULT_BEST_AVAIL_TRIM_ORDER; + + /* + * The default group preallocation is 512, which for 4k block + * sizes translates to 2 megabytes. However for bigalloc file + * systems, this is probably too big (i.e, if the cluster size + * is 1 megabyte, then group preallocation size becomes half a + * gigabyte!). As a default, we will keep a two megabyte + * group pralloc size for cluster sizes up to 64k, and after + * that, we will force a minimum group preallocation size of + * 32 clusters. This translates to 8 megs when the cluster + * size is 256k, and 32 megs when the cluster size is 1 meg, + * which seems reasonable as a default. + */ + sbi->s_mb_group_prealloc = max(MB_DEFAULT_GROUP_PREALLOC >> + sbi->s_cluster_bits, 32); + /* + * If there is a s_stripe > 1, then we set the s_mb_group_prealloc + * to the lowest multiple of s_stripe which is bigger than + * the s_mb_group_prealloc as determined above. We want + * the preallocation size to be an exact multiple of the + * RAID stripe size so that preallocations don't fragment + * the stripes. + */ + if (sbi->s_stripe > 1) { + sbi->s_mb_group_prealloc = roundup( + sbi->s_mb_group_prealloc, EXT4_NUM_B2C(sbi, sbi->s_stripe)); + } + + sbi->s_mb_nr_global_goals = umin(num_possible_cpus(), + DIV_ROUND_UP(sbi->s_groups_count, 4)); + sbi->s_mb_last_groups = kcalloc(sbi->s_mb_nr_global_goals, + sizeof(ext4_group_t), GFP_KERNEL); + if (sbi->s_mb_last_groups == NULL) { + ret = -ENOMEM; + goto out; + } + + sbi->s_locality_groups = alloc_percpu(struct ext4_locality_group); + if (sbi->s_locality_groups == NULL) { + ret = -ENOMEM; + goto out_free_last_groups; + } + for_each_possible_cpu(i) { + struct ext4_locality_group *lg; + lg = per_cpu_ptr(sbi->s_locality_groups, i); + mutex_init(&lg->lg_mutex); + for (j = 0; j < PREALLOC_TB_SIZE; j++) + INIT_LIST_HEAD(&lg->lg_prealloc_list[j]); + spin_lock_init(&lg->lg_prealloc_lock); + } + + if (bdev_nonrot(sb->s_bdev)) + sbi->s_mb_max_linear_groups = 0; + else + sbi->s_mb_max_linear_groups = MB_DEFAULT_LINEAR_LIMIT; + /* init file for buddy data */ + ret = ext4_mb_init_backend(sb); + if (ret != 0) + goto out_free_locality_groups; + + return 0; + +out_free_locality_groups: + free_percpu(sbi->s_locality_groups); + sbi->s_locality_groups = NULL; +out_free_last_groups: + kfree(sbi->s_mb_last_groups); + sbi->s_mb_last_groups = NULL; +out: + ext4_mb_avg_fragment_size_destroy(sbi); + ext4_mb_largest_free_orders_destroy(sbi); + kfree(sbi->s_mb_offsets); + sbi->s_mb_offsets = NULL; + kfree(sbi->s_mb_maxs); + sbi->s_mb_maxs = NULL; + return ret; +} + +/* need to called with the ext4 group lock held */ +static int ext4_mb_cleanup_pa(struct ext4_group_info *grp) +{ + struct ext4_prealloc_space *pa; + struct list_head *cur, *tmp; + int count = 0; + + list_for_each_safe(cur, tmp, &grp->bb_prealloc_list) { + pa = list_entry(cur, struct ext4_prealloc_space, pa_group_list); + list_del(&pa->pa_group_list); + count++; + kmem_cache_free(ext4_pspace_cachep, pa); + } + return count; +} + +void ext4_mb_release(struct super_block *sb) +{ + ext4_group_t ngroups = ext4_get_groups_count(sb); + ext4_group_t i; + int num_meta_group_infos; + struct ext4_group_info *grinfo, ***group_info; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct kmem_cache *cachep = get_groupinfo_cache(sb->s_blocksize_bits); + int count; + + if (test_opt(sb, DISCARD)) { + /* + * wait the discard work to drain all of ext4_free_data + */ + flush_work(&sbi->s_discard_work); + WARN_ON_ONCE(!list_empty(&sbi->s_discard_list)); + } + + if (sbi->s_group_info) { + for (i = 0; i < ngroups; i++) { + cond_resched(); + grinfo = ext4_get_group_info(sb, i); + if (!grinfo) + continue; + mb_group_bb_bitmap_free(grinfo); + ext4_lock_group(sb, i); + count = ext4_mb_cleanup_pa(grinfo); + if (count) + mb_debug(sb, "mballoc: %d PAs left\n", + count); + ext4_unlock_group(sb, i); + kmem_cache_free(cachep, grinfo); + } + num_meta_group_infos = (ngroups + + EXT4_DESC_PER_BLOCK(sb) - 1) >> + EXT4_DESC_PER_BLOCK_BITS(sb); + rcu_read_lock(); + group_info = rcu_dereference(sbi->s_group_info); + for (i = 0; i < num_meta_group_infos; i++) + kfree(group_info[i]); + kvfree(group_info); + rcu_read_unlock(); + } + ext4_mb_avg_fragment_size_destroy(sbi); + ext4_mb_largest_free_orders_destroy(sbi); + kfree(sbi->s_mb_offsets); + kfree(sbi->s_mb_maxs); + iput(sbi->s_buddy_cache); + if (sbi->s_mb_stats) { + ext4_msg(sb, KERN_INFO, + "mballoc: %u blocks %u reqs (%u success)", + atomic_read(&sbi->s_bal_allocated), + atomic_read(&sbi->s_bal_reqs), + atomic_read(&sbi->s_bal_success)); + ext4_msg(sb, KERN_INFO, + "mballoc: %u extents scanned, %u groups scanned, %u goal hits, " + "%u 2^N hits, %u breaks, %u lost", + atomic_read(&sbi->s_bal_ex_scanned), + atomic_read(&sbi->s_bal_groups_scanned), + atomic_read(&sbi->s_bal_goals), + atomic_read(&sbi->s_bal_2orders), + atomic_read(&sbi->s_bal_breaks), + atomic_read(&sbi->s_mb_lost_chunks)); + ext4_msg(sb, KERN_INFO, + "mballoc: %u generated and it took %llu", + atomic_read(&sbi->s_mb_buddies_generated), + atomic64_read(&sbi->s_mb_generation_time)); + ext4_msg(sb, KERN_INFO, + "mballoc: %u preallocated, %u discarded", + atomic_read(&sbi->s_mb_preallocated), + atomic_read(&sbi->s_mb_discarded)); + } + + free_percpu(sbi->s_locality_groups); + kfree(sbi->s_mb_last_groups); +} + +static inline int ext4_issue_discard(struct super_block *sb, + ext4_group_t block_group, ext4_grpblk_t cluster, int count) +{ + ext4_fsblk_t discard_block; + + discard_block = (EXT4_C2B(EXT4_SB(sb), cluster) + + ext4_group_first_block_no(sb, block_group)); + count = EXT4_C2B(EXT4_SB(sb), count); + trace_ext4_discard_blocks(sb, + (unsigned long long) discard_block, count); + + return sb_issue_discard(sb, discard_block, count, GFP_NOFS, 0); +} + +static void ext4_free_data_in_buddy(struct super_block *sb, + struct ext4_free_data *entry) +{ + struct ext4_buddy e4b; + struct ext4_group_info *db; + int err, count = 0; + + mb_debug(sb, "gonna free %u blocks in group %u (0x%p):", + entry->efd_count, entry->efd_group, entry); + + err = ext4_mb_load_buddy(sb, entry->efd_group, &e4b); + /* we expect to find existing buddy because it's pinned */ + BUG_ON(err != 0); + + atomic_sub(entry->efd_count, &EXT4_SB(sb)->s_mb_free_pending); + db = e4b.bd_info; + /* there are blocks to put in buddy to make them really free */ + count += entry->efd_count; + ext4_lock_group(sb, entry->efd_group); + /* Take it out of per group rb tree */ + rb_erase(&entry->efd_node, &(db->bb_free_root)); + mb_free_blocks(NULL, &e4b, entry->efd_start_cluster, entry->efd_count); + + /* + * Clear the trimmed flag for the group so that the next + * ext4_trim_fs can trim it. + */ + EXT4_MB_GRP_CLEAR_TRIMMED(db); + + if (!db->bb_free_root.rb_node) { + /* No more items in the per group rb tree + * balance refcounts from ext4_mb_free_metadata() + */ + folio_put(e4b.bd_buddy_folio); + folio_put(e4b.bd_bitmap_folio); + } + ext4_unlock_group(sb, entry->efd_group); + ext4_mb_unload_buddy(&e4b); + + mb_debug(sb, "freed %d blocks in 1 structures\n", count); +} + +/* + * This function is called by the jbd2 layer once the commit has finished, + * so we know we can free the blocks that were released with that commit. + */ +void ext4_process_freed_data(struct super_block *sb, tid_t commit_tid) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_free_data *entry, *tmp; + LIST_HEAD(freed_data_list); + struct list_head *s_freed_head = &sbi->s_freed_data_list[commit_tid & 1]; + bool wake; + + list_replace_init(s_freed_head, &freed_data_list); + + list_for_each_entry(entry, &freed_data_list, efd_list) + ext4_free_data_in_buddy(sb, entry); + + if (test_opt(sb, DISCARD)) { + spin_lock(&sbi->s_md_lock); + wake = list_empty(&sbi->s_discard_list); + list_splice_tail(&freed_data_list, &sbi->s_discard_list); + spin_unlock(&sbi->s_md_lock); + if (wake) + queue_work(system_dfl_wq, &sbi->s_discard_work); + } else { + list_for_each_entry_safe(entry, tmp, &freed_data_list, efd_list) + kmem_cache_free(ext4_free_data_cachep, entry); + } +} + +int __init ext4_init_mballoc(void) +{ + ext4_pspace_cachep = KMEM_CACHE(ext4_prealloc_space, + SLAB_RECLAIM_ACCOUNT); + if (ext4_pspace_cachep == NULL) + goto out; + + ext4_ac_cachep = KMEM_CACHE(ext4_allocation_context, + SLAB_RECLAIM_ACCOUNT); + if (ext4_ac_cachep == NULL) + goto out_pa_free; + + ext4_free_data_cachep = KMEM_CACHE(ext4_free_data, + SLAB_RECLAIM_ACCOUNT); + if (ext4_free_data_cachep == NULL) + goto out_ac_free; + + return 0; + +out_ac_free: + kmem_cache_destroy(ext4_ac_cachep); +out_pa_free: + kmem_cache_destroy(ext4_pspace_cachep); +out: + return -ENOMEM; +} + +void ext4_exit_mballoc(void) +{ + /* + * Wait for completion of call_rcu()'s on ext4_pspace_cachep + * before destroying the slab cache. + */ + rcu_barrier(); + kmem_cache_destroy(ext4_pspace_cachep); + kmem_cache_destroy(ext4_ac_cachep); + kmem_cache_destroy(ext4_free_data_cachep); + ext4_groupinfo_destroy_slabs(); +} + +#define EXT4_MB_BITMAP_MARKED_CHECK 0x0001 +#define EXT4_MB_SYNC_UPDATE 0x0002 +static int +ext4_mb_mark_context(handle_t *handle, struct super_block *sb, bool state, + ext4_group_t group, ext4_grpblk_t blkoff, + ext4_grpblk_t len, int flags, ext4_grpblk_t *ret_changed) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct buffer_head *bitmap_bh = NULL; + struct ext4_group_desc *gdp; + struct buffer_head *gdp_bh; + int err; + unsigned int i, already, changed = len; + + KUNIT_STATIC_STUB_REDIRECT(ext4_mb_mark_context, + handle, sb, state, group, blkoff, len, + flags, ret_changed); + + if (ret_changed) + *ret_changed = 0; + bitmap_bh = ext4_read_block_bitmap(sb, group); + if (IS_ERR(bitmap_bh)) + return PTR_ERR(bitmap_bh); + + if (handle) { + BUFFER_TRACE(bitmap_bh, "getting write access"); + err = ext4_journal_get_write_access(handle, sb, bitmap_bh, + EXT4_JTR_NONE); + if (err) + goto out_err; + } + + err = -EIO; + gdp = ext4_get_group_desc(sb, group, &gdp_bh); + if (!gdp) + goto out_err; + + if (handle) { + BUFFER_TRACE(gdp_bh, "get_write_access"); + err = ext4_journal_get_write_access(handle, sb, gdp_bh, + EXT4_JTR_NONE); + if (err) + goto out_err; + } + + ext4_lock_group(sb, group); + if (ext4_has_group_desc_csum(sb) && + (gdp->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT))) { + gdp->bg_flags &= cpu_to_le16(~EXT4_BG_BLOCK_UNINIT); + ext4_free_group_clusters_set(sb, gdp, + ext4_free_clusters_after_init(sb, group, gdp)); + } + + if (flags & EXT4_MB_BITMAP_MARKED_CHECK) { + already = 0; + for (i = 0; i < len; i++) + if (mb_test_bit(blkoff + i, bitmap_bh->b_data) == + state) + already++; + changed = len - already; + } + + if (state) { + mb_set_bits(bitmap_bh->b_data, blkoff, len); + ext4_free_group_clusters_set(sb, gdp, + ext4_free_group_clusters(sb, gdp) - changed); + } else { + mb_clear_bits(bitmap_bh->b_data, blkoff, len); + ext4_free_group_clusters_set(sb, gdp, + ext4_free_group_clusters(sb, gdp) + changed); + } + + ext4_block_bitmap_csum_set(sb, gdp, bitmap_bh); + ext4_group_desc_csum_set(sb, group, gdp); + ext4_unlock_group(sb, group); + if (ret_changed) + *ret_changed = changed; + + if (sbi->s_log_groups_per_flex) { + ext4_group_t flex_group = ext4_flex_group(sbi, group); + struct flex_groups *fg = sbi_array_rcu_deref(sbi, + s_flex_groups, flex_group); + + if (state) + atomic64_sub(changed, &fg->free_clusters); + else + atomic64_add(changed, &fg->free_clusters); + } + + err = ext4_handle_dirty_metadata(handle, NULL, bitmap_bh); + if (err) + goto out_err; + err = ext4_handle_dirty_metadata(handle, NULL, gdp_bh); + if (err) + goto out_err; + + if (flags & EXT4_MB_SYNC_UPDATE) { + sync_dirty_buffer(bitmap_bh); + sync_dirty_buffer(gdp_bh); + } + +out_err: + brelse(bitmap_bh); + return err; +} + +/* + * Check quota and mark chosen space (ac->ac_b_ex) non-free in bitmaps + * Returns 0 if success or error code + */ +static noinline_for_stack int +ext4_mb_mark_diskspace_used(struct ext4_allocation_context *ac, + handle_t *handle, unsigned int reserv_clstrs) +{ + struct ext4_group_desc *gdp; + struct ext4_sb_info *sbi; + struct super_block *sb; + ext4_fsblk_t block; + int err, len; + int flags = 0; + ext4_grpblk_t changed; + + BUG_ON(ac->ac_status != AC_STATUS_FOUND); + BUG_ON(ac->ac_b_ex.fe_len <= 0); + + sb = ac->ac_sb; + sbi = EXT4_SB(sb); + + gdp = ext4_get_group_desc(sb, ac->ac_b_ex.fe_group, NULL); + if (!gdp) + return -EIO; + ext4_debug("using block group %u(%d)\n", ac->ac_b_ex.fe_group, + ext4_free_group_clusters(sb, gdp)); + + block = ext4_grp_offs_to_block(sb, &ac->ac_b_ex); + len = EXT4_C2B(sbi, ac->ac_b_ex.fe_len); + if (!ext4_inode_block_valid(ac->ac_inode, block, len)) { + ext4_error(sb, "Allocating blocks %llu-%llu which overlap " + "fs metadata", block, block+len); + /* File system mounted not to panic on error + * Fix the bitmap and return EFSCORRUPTED + * We leak some of the blocks here. + */ + err = ext4_mb_mark_context(handle, sb, true, + ac->ac_b_ex.fe_group, + ac->ac_b_ex.fe_start, + ac->ac_b_ex.fe_len, + 0, NULL); + if (!err) + err = -EFSCORRUPTED; + return err; + } + +#ifdef AGGRESSIVE_CHECK + flags |= EXT4_MB_BITMAP_MARKED_CHECK; +#endif + err = ext4_mb_mark_context(handle, sb, true, ac->ac_b_ex.fe_group, + ac->ac_b_ex.fe_start, ac->ac_b_ex.fe_len, + flags, &changed); + + if (err && changed == 0) + return err; + +#ifdef AGGRESSIVE_CHECK + BUG_ON(changed != ac->ac_b_ex.fe_len); +#endif + percpu_counter_sub(&sbi->s_freeclusters_counter, ac->ac_b_ex.fe_len); + /* + * Now reduce the dirty block count also. Should not go negative + */ + if (!(ac->ac_flags & EXT4_MB_DELALLOC_RESERVED)) + /* release all the reserved blocks if non delalloc */ + percpu_counter_sub(&sbi->s_dirtyclusters_counter, + reserv_clstrs); + + return err; +} + +/* + * Idempotent helper for Ext4 fast commit replay path to set the state of + * blocks in bitmaps and update counters. + */ +void ext4_mb_mark_bb(struct super_block *sb, ext4_fsblk_t block, + int len, bool state) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + ext4_group_t group; + ext4_grpblk_t blkoff; + int err = 0; + unsigned int clen, thisgrp_len; + + while (len > 0) { + ext4_get_group_no_and_offset(sb, block, &group, &blkoff); + + /* + * Check to see if we are freeing blocks across a group + * boundary. + * In case of flex_bg, this can happen that (block, len) may + * span across more than one group. In that case we need to + * get the corresponding group metadata to work with. + * For this we have goto again loop. + */ + thisgrp_len = min_t(unsigned int, (unsigned int)len, + EXT4_BLOCKS_PER_GROUP(sb) - EXT4_C2B(sbi, blkoff)); + clen = EXT4_NUM_B2C(sbi, thisgrp_len); + + if (!ext4_sb_block_valid(sb, NULL, block, thisgrp_len)) { + ext4_error(sb, "Marking blocks in system zone - " + "Block = %llu, len = %u", + block, thisgrp_len); + break; + } + + err = ext4_mb_mark_context(NULL, sb, state, + group, blkoff, clen, + EXT4_MB_BITMAP_MARKED_CHECK | + EXT4_MB_SYNC_UPDATE, + NULL); + if (err) + break; + + block += thisgrp_len; + len -= thisgrp_len; + BUG_ON(len < 0); + } +} + +/* + * here we normalize request for locality group + * Group request are normalized to s_mb_group_prealloc, which goes to + * s_strip if we set the same via mount option. + * s_mb_group_prealloc can be configured via + * /sys/fs/ext4/<partition>/mb_group_prealloc + * + * XXX: should we try to preallocate more than the group has now? + */ +static void ext4_mb_normalize_group_request(struct ext4_allocation_context *ac) +{ + struct super_block *sb = ac->ac_sb; + struct ext4_locality_group *lg = ac->ac_lg; + + BUG_ON(lg == NULL); + ac->ac_g_ex.fe_len = EXT4_SB(sb)->s_mb_group_prealloc; + mb_debug(sb, "goal %u blocks for locality group\n", ac->ac_g_ex.fe_len); +} + +/* + * This function returns the next element to look at during inode + * PA rbtree walk. We assume that we have held the inode PA rbtree lock + * (ei->i_prealloc_lock) + * + * new_start The start of the range we want to compare + * cur_start The existing start that we are comparing against + * node The node of the rb_tree + */ +static inline struct rb_node* +ext4_mb_pa_rb_next_iter(ext4_lblk_t new_start, ext4_lblk_t cur_start, struct rb_node *node) +{ + if (new_start < cur_start) + return node->rb_left; + else + return node->rb_right; +} + +static inline void +ext4_mb_pa_assert_overlap(struct ext4_allocation_context *ac, + ext4_lblk_t start, loff_t end) +{ + struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); + struct ext4_inode_info *ei = EXT4_I(ac->ac_inode); + struct ext4_prealloc_space *tmp_pa; + ext4_lblk_t tmp_pa_start; + loff_t tmp_pa_end; + struct rb_node *iter; + + read_lock(&ei->i_prealloc_lock); + for (iter = ei->i_prealloc_node.rb_node; iter; + iter = ext4_mb_pa_rb_next_iter(start, tmp_pa_start, iter)) { + tmp_pa = rb_entry(iter, struct ext4_prealloc_space, + pa_node.inode_node); + tmp_pa_start = tmp_pa->pa_lstart; + tmp_pa_end = pa_logical_end(sbi, tmp_pa); + + spin_lock(&tmp_pa->pa_lock); + if (tmp_pa->pa_deleted == 0) + BUG_ON(!(start >= tmp_pa_end || end <= tmp_pa_start)); + spin_unlock(&tmp_pa->pa_lock); + } + read_unlock(&ei->i_prealloc_lock); +} + +/* + * Given an allocation context "ac" and a range "start", "end", check + * and adjust boundaries if the range overlaps with any of the existing + * preallocatoins stored in the corresponding inode of the allocation context. + * + * Parameters: + * ac allocation context + * start start of the new range + * end end of the new range + */ +static inline void +ext4_mb_pa_adjust_overlap(struct ext4_allocation_context *ac, + ext4_lblk_t *start, loff_t *end) +{ + struct ext4_inode_info *ei = EXT4_I(ac->ac_inode); + struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); + struct ext4_prealloc_space *tmp_pa = NULL, *left_pa = NULL, *right_pa = NULL; + struct rb_node *iter; + ext4_lblk_t new_start, tmp_pa_start, right_pa_start = -1; + loff_t new_end, tmp_pa_end, left_pa_end = -1; + + new_start = *start; + new_end = *end; + + /* + * Adjust the normalized range so that it doesn't overlap with any + * existing preallocated blocks(PAs). Make sure to hold the rbtree lock + * so it doesn't change underneath us. + */ + read_lock(&ei->i_prealloc_lock); + + /* Step 1: find any one immediate neighboring PA of the normalized range */ + for (iter = ei->i_prealloc_node.rb_node; iter; + iter = ext4_mb_pa_rb_next_iter(ac->ac_o_ex.fe_logical, + tmp_pa_start, iter)) { + tmp_pa = rb_entry(iter, struct ext4_prealloc_space, + pa_node.inode_node); + tmp_pa_start = tmp_pa->pa_lstart; + tmp_pa_end = pa_logical_end(sbi, tmp_pa); + + /* PA must not overlap original request */ + spin_lock(&tmp_pa->pa_lock); + if (tmp_pa->pa_deleted == 0) + BUG_ON(!(ac->ac_o_ex.fe_logical >= tmp_pa_end || + ac->ac_o_ex.fe_logical < tmp_pa_start)); + spin_unlock(&tmp_pa->pa_lock); + } + + /* + * Step 2: check if the found PA is left or right neighbor and + * get the other neighbor + */ + if (tmp_pa) { + if (tmp_pa->pa_lstart < ac->ac_o_ex.fe_logical) { + struct rb_node *tmp; + + left_pa = tmp_pa; + tmp = rb_next(&left_pa->pa_node.inode_node); + if (tmp) { + right_pa = rb_entry(tmp, + struct ext4_prealloc_space, + pa_node.inode_node); + } + } else { + struct rb_node *tmp; + + right_pa = tmp_pa; + tmp = rb_prev(&right_pa->pa_node.inode_node); + if (tmp) { + left_pa = rb_entry(tmp, + struct ext4_prealloc_space, + pa_node.inode_node); + } + } + } + + /* Step 3: get the non deleted neighbors */ + if (left_pa) { + for (iter = &left_pa->pa_node.inode_node;; + iter = rb_prev(iter)) { + if (!iter) { + left_pa = NULL; + break; + } + + tmp_pa = rb_entry(iter, struct ext4_prealloc_space, + pa_node.inode_node); + left_pa = tmp_pa; + spin_lock(&tmp_pa->pa_lock); + if (tmp_pa->pa_deleted == 0) { + spin_unlock(&tmp_pa->pa_lock); + break; + } + spin_unlock(&tmp_pa->pa_lock); + } + } + + if (right_pa) { + for (iter = &right_pa->pa_node.inode_node;; + iter = rb_next(iter)) { + if (!iter) { + right_pa = NULL; + break; + } + + tmp_pa = rb_entry(iter, struct ext4_prealloc_space, + pa_node.inode_node); + right_pa = tmp_pa; + spin_lock(&tmp_pa->pa_lock); + if (tmp_pa->pa_deleted == 0) { + spin_unlock(&tmp_pa->pa_lock); + break; + } + spin_unlock(&tmp_pa->pa_lock); + } + } + + if (left_pa) { + left_pa_end = pa_logical_end(sbi, left_pa); + BUG_ON(left_pa_end > ac->ac_o_ex.fe_logical); + } + + if (right_pa) { + right_pa_start = right_pa->pa_lstart; + BUG_ON(right_pa_start <= ac->ac_o_ex.fe_logical); + } + + /* Step 4: trim our normalized range to not overlap with the neighbors */ + if (left_pa) { + if (left_pa_end > new_start) + new_start = left_pa_end; + } + + if (right_pa) { + if (right_pa_start < new_end) + new_end = right_pa_start; + } + read_unlock(&ei->i_prealloc_lock); + + /* XXX: extra loop to check we really don't overlap preallocations */ + ext4_mb_pa_assert_overlap(ac, new_start, new_end); + + *start = new_start; + *end = new_end; +} + +/* + * Normalization means making request better in terms of + * size and alignment + */ +static noinline_for_stack void +ext4_mb_normalize_request(struct ext4_allocation_context *ac, + struct ext4_allocation_request *ar) +{ + struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); + struct ext4_super_block *es = sbi->s_es; + int bsbits, max; + loff_t size, start_off, end; + loff_t orig_size __maybe_unused; + ext4_lblk_t start; + + /* do normalize only data requests, metadata requests + do not need preallocation */ + if (!(ac->ac_flags & EXT4_MB_HINT_DATA)) + return; + + /* sometime caller may want exact blocks */ + if (unlikely(ac->ac_flags & EXT4_MB_HINT_GOAL_ONLY)) + return; + + /* caller may indicate that preallocation isn't + * required (it's a tail, for example) */ + if (ac->ac_flags & EXT4_MB_HINT_NOPREALLOC) + return; + + if (ac->ac_flags & EXT4_MB_HINT_GROUP_ALLOC) { + ext4_mb_normalize_group_request(ac); + return ; + } + + bsbits = ac->ac_sb->s_blocksize_bits; + + /* first, let's learn actual file size + * given current request is allocated */ + size = extent_logical_end(sbi, &ac->ac_o_ex); + size = size << bsbits; + if (size < i_size_read(ac->ac_inode)) + size = i_size_read(ac->ac_inode); + orig_size = size; + + /* max size of free chunks */ + max = 2 << bsbits; + +#define NRL_CHECK_SIZE(req, size, max, chunk_size) \ + (req <= (size) || max <= (chunk_size)) + + /* first, try to predict filesize */ + /* XXX: should this table be tunable? */ + start_off = 0; + if (size <= 16 * 1024) { + size = 16 * 1024; + } else if (size <= 32 * 1024) { + size = 32 * 1024; + } else if (size <= 64 * 1024) { + size = 64 * 1024; + } else if (size <= 128 * 1024) { + size = 128 * 1024; + } else if (size <= 256 * 1024) { + size = 256 * 1024; + } else if (size <= 512 * 1024) { + size = 512 * 1024; + } else if (size <= 1024 * 1024) { + size = 1024 * 1024; + } else if (NRL_CHECK_SIZE(size, 4 * 1024 * 1024, max, 2 * 1024)) { + start_off = ((loff_t)ac->ac_o_ex.fe_logical >> + (21 - bsbits)) << 21; + size = 2 * 1024 * 1024; + } else if (NRL_CHECK_SIZE(size, 8 * 1024 * 1024, max, 4 * 1024)) { + start_off = ((loff_t)ac->ac_o_ex.fe_logical >> + (22 - bsbits)) << 22; + size = 4 * 1024 * 1024; + } else if (NRL_CHECK_SIZE(EXT4_C2B(sbi, ac->ac_o_ex.fe_len), + (8<<20)>>bsbits, max, 8 * 1024)) { + start_off = ((loff_t)ac->ac_o_ex.fe_logical >> + (23 - bsbits)) << 23; + size = 8 * 1024 * 1024; + } else { + start_off = (loff_t) ac->ac_o_ex.fe_logical << bsbits; + size = (loff_t) EXT4_C2B(sbi, + ac->ac_o_ex.fe_len) << bsbits; + } + size = size >> bsbits; + start = start_off >> bsbits; + + /* + * For tiny groups (smaller than 8MB) the chosen allocation + * alignment may be larger than group size. Make sure the + * alignment does not move allocation to a different group which + * makes mballoc fail assertions later. + */ + start = max(start, rounddown(ac->ac_o_ex.fe_logical, + (ext4_lblk_t)EXT4_BLOCKS_PER_GROUP(ac->ac_sb))); + + /* avoid unnecessary preallocation that may trigger assertions */ + if (start + size > EXT_MAX_BLOCKS) + size = EXT_MAX_BLOCKS - start; + + /* don't cover already allocated blocks in selected range */ + if (ar->pleft && start <= ar->lleft) { + size -= ar->lleft + 1 - start; + start = ar->lleft + 1; + } + if (ar->pright && start + size - 1 >= ar->lright) + size -= start + size - ar->lright; + + /* + * Trim allocation request for filesystems with artificially small + * groups. + */ + if (size > EXT4_BLOCKS_PER_GROUP(ac->ac_sb)) + size = EXT4_BLOCKS_PER_GROUP(ac->ac_sb); + + end = start + size; + + ext4_mb_pa_adjust_overlap(ac, &start, &end); + + size = end - start; + + /* + * In this function "start" and "size" are normalized for better + * alignment and length such that we could preallocate more blocks. + * This normalization is done such that original request of + * ac->ac_o_ex.fe_logical & fe_len should always lie within "start" and + * "size" boundaries. + * (Note fe_len can be relaxed since FS block allocation API does not + * provide gurantee on number of contiguous blocks allocation since that + * depends upon free space left, etc). + * In case of inode pa, later we use the allocated blocks + * [pa_pstart + fe_logical - pa_lstart, fe_len/size] from the preallocated + * range of goal/best blocks [start, size] to put it at the + * ac_o_ex.fe_logical extent of this inode. + * (See ext4_mb_use_inode_pa() for more details) + */ + if (start + size <= ac->ac_o_ex.fe_logical || + start > ac->ac_o_ex.fe_logical) { + ext4_msg(ac->ac_sb, KERN_ERR, + "start %lu, size %lu, fe_logical %lu", + (unsigned long) start, (unsigned long) size, + (unsigned long) ac->ac_o_ex.fe_logical); + BUG(); + } + BUG_ON(size <= 0 || size > EXT4_BLOCKS_PER_GROUP(ac->ac_sb)); + + /* now prepare goal request */ + + /* XXX: is it better to align blocks WRT to logical + * placement or satisfy big request as is */ + ac->ac_g_ex.fe_logical = start; + ac->ac_g_ex.fe_len = EXT4_NUM_B2C(sbi, size); + ac->ac_orig_goal_len = ac->ac_g_ex.fe_len; + + /* define goal start in order to merge */ + if (ar->pright && (ar->lright == (start + size)) && + ar->pright >= size && + ar->pright - size >= le32_to_cpu(es->s_first_data_block)) { + /* merge to the right */ + ext4_get_group_no_and_offset(ac->ac_sb, ar->pright - size, + &ac->ac_g_ex.fe_group, + &ac->ac_g_ex.fe_start); + ac->ac_flags |= EXT4_MB_HINT_TRY_GOAL; + } + if (ar->pleft && (ar->lleft + 1 == start) && + ar->pleft + 1 < ext4_blocks_count(es)) { + /* merge to the left */ + ext4_get_group_no_and_offset(ac->ac_sb, ar->pleft + 1, + &ac->ac_g_ex.fe_group, + &ac->ac_g_ex.fe_start); + ac->ac_flags |= EXT4_MB_HINT_TRY_GOAL; + } + + mb_debug(ac->ac_sb, "goal: %lld(was %lld) blocks at %u\n", size, + orig_size, start); +} + +static void ext4_mb_collect_stats(struct ext4_allocation_context *ac) +{ + struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); + + if (sbi->s_mb_stats && ac->ac_g_ex.fe_len >= 1) { + atomic_inc(&sbi->s_bal_reqs); + atomic_add(ac->ac_b_ex.fe_len, &sbi->s_bal_allocated); + if (ac->ac_b_ex.fe_len >= ac->ac_o_ex.fe_len) + atomic_inc(&sbi->s_bal_success); + + atomic_add(ac->ac_found, &sbi->s_bal_ex_scanned); + for (int i=0; i<EXT4_MB_NUM_CRS; i++) { + atomic_add(ac->ac_cX_found[i], &sbi->s_bal_cX_ex_scanned[i]); + } + + atomic_add(ac->ac_groups_scanned, &sbi->s_bal_groups_scanned); + if (ac->ac_g_ex.fe_start == ac->ac_b_ex.fe_start && + ac->ac_g_ex.fe_group == ac->ac_b_ex.fe_group) + atomic_inc(&sbi->s_bal_goals); + /* did we allocate as much as normalizer originally wanted? */ + if (ac->ac_f_ex.fe_len == ac->ac_orig_goal_len) + atomic_inc(&sbi->s_bal_len_goals); + + if (ac->ac_found > sbi->s_mb_max_to_scan) + atomic_inc(&sbi->s_bal_breaks); + } + + if (ac->ac_op == EXT4_MB_HISTORY_ALLOC) + trace_ext4_mballoc_alloc(ac); + else + trace_ext4_mballoc_prealloc(ac); +} + +/* + * Called on failure; free up any blocks from the inode PA for this + * context. We don't need this for MB_GROUP_PA because we only change + * pa_free in ext4_mb_release_context(), but on failure, we've already + * zeroed out ac->ac_b_ex.fe_len, so group_pa->pa_free is not changed. + */ +static void ext4_discard_allocated_blocks(struct ext4_allocation_context *ac) +{ + struct ext4_prealloc_space *pa = ac->ac_pa; + struct ext4_buddy e4b; + int err; + + if (pa == NULL) { + if (ac->ac_f_ex.fe_len == 0) + return; + err = ext4_mb_load_buddy(ac->ac_sb, ac->ac_f_ex.fe_group, &e4b); + if (WARN_RATELIMIT(err, + "ext4: mb_load_buddy failed (%d)", err)) + /* + * This should never happen since we pin the + * pages in the ext4_allocation_context so + * ext4_mb_load_buddy() should never fail. + */ + return; + ext4_lock_group(ac->ac_sb, ac->ac_f_ex.fe_group); + mb_free_blocks(ac->ac_inode, &e4b, ac->ac_f_ex.fe_start, + ac->ac_f_ex.fe_len); + ext4_unlock_group(ac->ac_sb, ac->ac_f_ex.fe_group); + ext4_mb_unload_buddy(&e4b); + return; + } + if (pa->pa_type == MB_INODE_PA) { + spin_lock(&pa->pa_lock); + pa->pa_free += ac->ac_b_ex.fe_len; + spin_unlock(&pa->pa_lock); + } +} + +/* + * use blocks preallocated to inode + */ +static void ext4_mb_use_inode_pa(struct ext4_allocation_context *ac, + struct ext4_prealloc_space *pa) +{ + struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); + ext4_fsblk_t start; + ext4_fsblk_t end; + int len; + + /* found preallocated blocks, use them */ + start = pa->pa_pstart + (ac->ac_o_ex.fe_logical - pa->pa_lstart); + end = min(pa->pa_pstart + EXT4_C2B(sbi, pa->pa_len), + start + EXT4_C2B(sbi, ac->ac_o_ex.fe_len)); + len = EXT4_NUM_B2C(sbi, end - start); + ext4_get_group_no_and_offset(ac->ac_sb, start, &ac->ac_b_ex.fe_group, + &ac->ac_b_ex.fe_start); + ac->ac_b_ex.fe_len = len; + ac->ac_status = AC_STATUS_FOUND; + ac->ac_pa = pa; + + BUG_ON(start < pa->pa_pstart); + BUG_ON(end > pa->pa_pstart + EXT4_C2B(sbi, pa->pa_len)); + BUG_ON(pa->pa_free < len); + BUG_ON(ac->ac_b_ex.fe_len <= 0); + pa->pa_free -= len; + + mb_debug(ac->ac_sb, "use %llu/%d from inode pa %p\n", start, len, pa); +} + +/* + * use blocks preallocated to locality group + */ +static void ext4_mb_use_group_pa(struct ext4_allocation_context *ac, + struct ext4_prealloc_space *pa) +{ + unsigned int len = ac->ac_o_ex.fe_len; + + ext4_get_group_no_and_offset(ac->ac_sb, pa->pa_pstart, + &ac->ac_b_ex.fe_group, + &ac->ac_b_ex.fe_start); + ac->ac_b_ex.fe_len = len; + ac->ac_status = AC_STATUS_FOUND; + ac->ac_pa = pa; + + /* we don't correct pa_pstart or pa_len here to avoid + * possible race when the group is being loaded concurrently + * instead we correct pa later, after blocks are marked + * in on-disk bitmap -- see ext4_mb_release_context() + * Other CPUs are prevented from allocating from this pa by lg_mutex + */ + mb_debug(ac->ac_sb, "use %u/%u from group pa %p\n", + pa->pa_lstart, len, pa); +} + +/* + * Return the prealloc space that have minimal distance + * from the goal block. @cpa is the prealloc + * space that is having currently known minimal distance + * from the goal block. + */ +static struct ext4_prealloc_space * +ext4_mb_check_group_pa(ext4_fsblk_t goal_block, + struct ext4_prealloc_space *pa, + struct ext4_prealloc_space *cpa) +{ + ext4_fsblk_t cur_distance, new_distance; + + if (cpa == NULL) { + atomic_inc(&pa->pa_count); + return pa; + } + cur_distance = abs(goal_block - cpa->pa_pstart); + new_distance = abs(goal_block - pa->pa_pstart); + + if (cur_distance <= new_distance) + return cpa; + + /* drop the previous reference */ + atomic_dec(&cpa->pa_count); + atomic_inc(&pa->pa_count); + return pa; +} + +/* + * check if found pa meets EXT4_MB_HINT_GOAL_ONLY + */ +static bool +ext4_mb_pa_goal_check(struct ext4_allocation_context *ac, + struct ext4_prealloc_space *pa) +{ + struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); + ext4_fsblk_t start; + + if (likely(!(ac->ac_flags & EXT4_MB_HINT_GOAL_ONLY))) + return true; + + /* + * If EXT4_MB_HINT_GOAL_ONLY is set, ac_g_ex will not be adjusted + * in ext4_mb_normalize_request and will keep same with ac_o_ex + * from ext4_mb_initialize_context. Choose ac_g_ex here to keep + * consistent with ext4_mb_find_by_goal. + */ + start = pa->pa_pstart + + (ac->ac_g_ex.fe_logical - pa->pa_lstart); + if (ext4_grp_offs_to_block(ac->ac_sb, &ac->ac_g_ex) != start) + return false; + + if (ac->ac_g_ex.fe_len > pa->pa_len - + EXT4_B2C(sbi, ac->ac_g_ex.fe_logical - pa->pa_lstart)) + return false; + + return true; +} + +/* + * search goal blocks in preallocated space + */ +static noinline_for_stack bool +ext4_mb_use_preallocated(struct ext4_allocation_context *ac) +{ + struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); + int order, i; + struct ext4_inode_info *ei = EXT4_I(ac->ac_inode); + struct ext4_locality_group *lg; + struct ext4_prealloc_space *tmp_pa = NULL, *cpa = NULL; + struct rb_node *iter; + ext4_fsblk_t goal_block; + + /* only data can be preallocated */ + if (!(ac->ac_flags & EXT4_MB_HINT_DATA)) + return false; + + /* + * first, try per-file preallocation by searching the inode pa rbtree. + * + * Here, we can't do a direct traversal of the tree because + * ext4_mb_discard_group_preallocation() can paralelly mark the pa + * deleted and that can cause direct traversal to skip some entries. + */ + read_lock(&ei->i_prealloc_lock); + + if (RB_EMPTY_ROOT(&ei->i_prealloc_node)) { + goto try_group_pa; + } + + /* + * Step 1: Find a pa with logical start immediately adjacent to the + * original logical start. This could be on the left or right. + * + * (tmp_pa->pa_lstart never changes so we can skip locking for it). + */ + for (iter = ei->i_prealloc_node.rb_node; iter; + iter = ext4_mb_pa_rb_next_iter(ac->ac_o_ex.fe_logical, + tmp_pa->pa_lstart, iter)) { + tmp_pa = rb_entry(iter, struct ext4_prealloc_space, + pa_node.inode_node); + } + + /* + * Step 2: The adjacent pa might be to the right of logical start, find + * the left adjacent pa. After this step we'd have a valid tmp_pa whose + * logical start is towards the left of original request's logical start + */ + if (tmp_pa->pa_lstart > ac->ac_o_ex.fe_logical) { + struct rb_node *tmp; + tmp = rb_prev(&tmp_pa->pa_node.inode_node); + + if (tmp) { + tmp_pa = rb_entry(tmp, struct ext4_prealloc_space, + pa_node.inode_node); + } else { + /* + * If there is no adjacent pa to the left then finding + * an overlapping pa is not possible hence stop searching + * inode pa tree + */ + goto try_group_pa; + } + } + + BUG_ON(!(tmp_pa && tmp_pa->pa_lstart <= ac->ac_o_ex.fe_logical)); + + /* + * Step 3: If the left adjacent pa is deleted, keep moving left to find + * the first non deleted adjacent pa. After this step we should have a + * valid tmp_pa which is guaranteed to be non deleted. + */ + for (iter = &tmp_pa->pa_node.inode_node;; iter = rb_prev(iter)) { + if (!iter) { + /* + * no non deleted left adjacent pa, so stop searching + * inode pa tree + */ + goto try_group_pa; + } + tmp_pa = rb_entry(iter, struct ext4_prealloc_space, + pa_node.inode_node); + spin_lock(&tmp_pa->pa_lock); + if (tmp_pa->pa_deleted == 0) { + /* + * We will keep holding the pa_lock from + * this point on because we don't want group discard + * to delete this pa underneath us. Since group + * discard is anyways an ENOSPC operation it + * should be okay for it to wait a few more cycles. + */ + break; + } else { + spin_unlock(&tmp_pa->pa_lock); + } + } + + BUG_ON(!(tmp_pa && tmp_pa->pa_lstart <= ac->ac_o_ex.fe_logical)); + BUG_ON(tmp_pa->pa_deleted == 1); + + /* + * Step 4: We now have the non deleted left adjacent pa. Only this + * pa can possibly satisfy the request hence check if it overlaps + * original logical start and stop searching if it doesn't. + */ + if (ac->ac_o_ex.fe_logical >= pa_logical_end(sbi, tmp_pa)) { + spin_unlock(&tmp_pa->pa_lock); + goto try_group_pa; + } + + /* non-extent files can't have physical blocks past 2^32 */ + if (!(ext4_test_inode_flag(ac->ac_inode, EXT4_INODE_EXTENTS)) && + (tmp_pa->pa_pstart + EXT4_C2B(sbi, tmp_pa->pa_len) > + EXT4_MAX_BLOCK_FILE_PHYS)) { + /* + * Since PAs don't overlap, we won't find any other PA to + * satisfy this. + */ + spin_unlock(&tmp_pa->pa_lock); + goto try_group_pa; + } + + if (tmp_pa->pa_free && likely(ext4_mb_pa_goal_check(ac, tmp_pa))) { + atomic_inc(&tmp_pa->pa_count); + ext4_mb_use_inode_pa(ac, tmp_pa); + spin_unlock(&tmp_pa->pa_lock); + read_unlock(&ei->i_prealloc_lock); + return true; + } else { + /* + * We found a valid overlapping pa but couldn't use it because + * it had no free blocks. This should ideally never happen + * because: + * + * 1. When a new inode pa is added to rbtree it must have + * pa_free > 0 since otherwise we won't actually need + * preallocation. + * + * 2. An inode pa that is in the rbtree can only have it's + * pa_free become zero when another thread calls: + * ext4_mb_new_blocks + * ext4_mb_use_preallocated + * ext4_mb_use_inode_pa + * + * 3. Further, after the above calls make pa_free == 0, we will + * immediately remove it from the rbtree in: + * ext4_mb_new_blocks + * ext4_mb_release_context + * ext4_mb_put_pa + * + * 4. Since the pa_free becoming 0 and pa_free getting removed + * from tree both happen in ext4_mb_new_blocks, which is always + * called with i_data_sem held for data allocations, we can be + * sure that another process will never see a pa in rbtree with + * pa_free == 0. + */ + WARN_ON_ONCE(tmp_pa->pa_free == 0); + } + spin_unlock(&tmp_pa->pa_lock); +try_group_pa: + read_unlock(&ei->i_prealloc_lock); + + /* can we use group allocation? */ + if (!(ac->ac_flags & EXT4_MB_HINT_GROUP_ALLOC)) + return false; + + /* inode may have no locality group for some reason */ + lg = ac->ac_lg; + if (lg == NULL) + return false; + order = fls(ac->ac_o_ex.fe_len) - 1; + if (order > PREALLOC_TB_SIZE - 1) + /* The max size of hash table is PREALLOC_TB_SIZE */ + order = PREALLOC_TB_SIZE - 1; + + goal_block = ext4_grp_offs_to_block(ac->ac_sb, &ac->ac_g_ex); + /* + * search for the prealloc space that is having + * minimal distance from the goal block. + */ + for (i = order; i < PREALLOC_TB_SIZE; i++) { + rcu_read_lock(); + list_for_each_entry_rcu(tmp_pa, &lg->lg_prealloc_list[i], + pa_node.lg_list) { + spin_lock(&tmp_pa->pa_lock); + if (tmp_pa->pa_deleted == 0 && + tmp_pa->pa_free >= ac->ac_o_ex.fe_len) { + + cpa = ext4_mb_check_group_pa(goal_block, + tmp_pa, cpa); + } + spin_unlock(&tmp_pa->pa_lock); + } + rcu_read_unlock(); + } + if (cpa) { + ext4_mb_use_group_pa(ac, cpa); + return true; + } + return false; +} + +/* + * the function goes through all preallocation in this group and marks them + * used in in-core bitmap. buddy must be generated from this bitmap + * Need to be called with ext4 group lock held + */ +static noinline_for_stack +void ext4_mb_generate_from_pa(struct super_block *sb, void *bitmap, + ext4_group_t group) +{ + struct ext4_group_info *grp = ext4_get_group_info(sb, group); + struct ext4_prealloc_space *pa; + struct list_head *cur; + ext4_group_t groupnr; + ext4_grpblk_t start; + int preallocated = 0; + int len; + + if (!grp) + return; + + /* all form of preallocation discards first load group, + * so the only competing code is preallocation use. + * we don't need any locking here + * notice we do NOT ignore preallocations with pa_deleted + * otherwise we could leave used blocks available for + * allocation in buddy when concurrent ext4_mb_put_pa() + * is dropping preallocation + */ + list_for_each(cur, &grp->bb_prealloc_list) { + pa = list_entry(cur, struct ext4_prealloc_space, pa_group_list); + spin_lock(&pa->pa_lock); + ext4_get_group_no_and_offset(sb, pa->pa_pstart, + &groupnr, &start); + len = pa->pa_len; + spin_unlock(&pa->pa_lock); + if (unlikely(len == 0)) + continue; + BUG_ON(groupnr != group); + mb_set_bits(bitmap, start, len); + preallocated += len; + } + mb_debug(sb, "preallocated %d for group %u\n", preallocated, group); +} + +static void ext4_mb_mark_pa_deleted(struct super_block *sb, + struct ext4_prealloc_space *pa) +{ + struct ext4_inode_info *ei; + + if (pa->pa_deleted) { + ext4_warning(sb, "deleted pa, type:%d, pblk:%llu, lblk:%u, len:%d\n", + pa->pa_type, pa->pa_pstart, pa->pa_lstart, + pa->pa_len); + return; + } + + pa->pa_deleted = 1; + + if (pa->pa_type == MB_INODE_PA) { + ei = EXT4_I(pa->pa_inode); + atomic_dec(&ei->i_prealloc_active); + } +} + +static inline void ext4_mb_pa_free(struct ext4_prealloc_space *pa) +{ + BUG_ON(!pa); + BUG_ON(atomic_read(&pa->pa_count)); + BUG_ON(pa->pa_deleted == 0); + kmem_cache_free(ext4_pspace_cachep, pa); +} + +static void ext4_mb_pa_callback(struct rcu_head *head) +{ + struct ext4_prealloc_space *pa; + + pa = container_of(head, struct ext4_prealloc_space, u.pa_rcu); + ext4_mb_pa_free(pa); +} + +/* + * drops a reference to preallocated space descriptor + * if this was the last reference and the space is consumed + */ +static void ext4_mb_put_pa(struct ext4_allocation_context *ac, + struct super_block *sb, struct ext4_prealloc_space *pa) +{ + ext4_group_t grp; + ext4_fsblk_t grp_blk; + struct ext4_inode_info *ei = EXT4_I(ac->ac_inode); + + /* in this short window concurrent discard can set pa_deleted */ + spin_lock(&pa->pa_lock); + if (!atomic_dec_and_test(&pa->pa_count) || pa->pa_free != 0) { + spin_unlock(&pa->pa_lock); + return; + } + + if (pa->pa_deleted == 1) { + spin_unlock(&pa->pa_lock); + return; + } + + ext4_mb_mark_pa_deleted(sb, pa); + spin_unlock(&pa->pa_lock); + + grp_blk = pa->pa_pstart; + /* + * If doing group-based preallocation, pa_pstart may be in the + * next group when pa is used up + */ + if (pa->pa_type == MB_GROUP_PA) + grp_blk--; + + grp = ext4_get_group_number(sb, grp_blk); + + /* + * possible race: + * + * P1 (buddy init) P2 (regular allocation) + * find block B in PA + * copy on-disk bitmap to buddy + * mark B in on-disk bitmap + * drop PA from group + * mark all PAs in buddy + * + * thus, P1 initializes buddy with B available. to prevent this + * we make "copy" and "mark all PAs" atomic and serialize "drop PA" + * against that pair + */ + ext4_lock_group(sb, grp); + list_del(&pa->pa_group_list); + ext4_unlock_group(sb, grp); + + if (pa->pa_type == MB_INODE_PA) { + write_lock(pa->pa_node_lock.inode_lock); + rb_erase(&pa->pa_node.inode_node, &ei->i_prealloc_node); + write_unlock(pa->pa_node_lock.inode_lock); + ext4_mb_pa_free(pa); + } else { + spin_lock(pa->pa_node_lock.lg_lock); + list_del_rcu(&pa->pa_node.lg_list); + spin_unlock(pa->pa_node_lock.lg_lock); + call_rcu(&(pa)->u.pa_rcu, ext4_mb_pa_callback); + } +} + +static void ext4_mb_pa_rb_insert(struct rb_root *root, struct rb_node *new) +{ + struct rb_node **iter = &root->rb_node, *parent = NULL; + struct ext4_prealloc_space *iter_pa, *new_pa; + ext4_lblk_t iter_start, new_start; + + while (*iter) { + iter_pa = rb_entry(*iter, struct ext4_prealloc_space, + pa_node.inode_node); + new_pa = rb_entry(new, struct ext4_prealloc_space, + pa_node.inode_node); + iter_start = iter_pa->pa_lstart; + new_start = new_pa->pa_lstart; + + parent = *iter; + if (new_start < iter_start) + iter = &((*iter)->rb_left); + else + iter = &((*iter)->rb_right); + } + + rb_link_node(new, parent, iter); + rb_insert_color(new, root); +} + +/* + * creates new preallocated space for given inode + */ +static noinline_for_stack void +ext4_mb_new_inode_pa(struct ext4_allocation_context *ac) +{ + struct super_block *sb = ac->ac_sb; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_prealloc_space *pa; + struct ext4_group_info *grp; + struct ext4_inode_info *ei; + + /* preallocate only when found space is larger then requested */ + BUG_ON(ac->ac_o_ex.fe_len >= ac->ac_b_ex.fe_len); + BUG_ON(ac->ac_status != AC_STATUS_FOUND); + BUG_ON(!S_ISREG(ac->ac_inode->i_mode)); + BUG_ON(ac->ac_pa == NULL); + + pa = ac->ac_pa; + + if (ac->ac_b_ex.fe_len < ac->ac_orig_goal_len) { + struct ext4_free_extent ex = { + .fe_logical = ac->ac_g_ex.fe_logical, + .fe_len = ac->ac_orig_goal_len, + }; + loff_t orig_goal_end = extent_logical_end(sbi, &ex); + loff_t o_ex_end = extent_logical_end(sbi, &ac->ac_o_ex); + + /* + * We can't allocate as much as normalizer wants, so we try + * to get proper lstart to cover the original request, except + * when the goal doesn't cover the original request as below: + * + * orig_ex:2045/2055(10), isize:8417280 -> normalized:0/2048 + * best_ex:0/200(200) -> adjusted: 1848/2048(200) + */ + BUG_ON(ac->ac_g_ex.fe_logical > ac->ac_o_ex.fe_logical); + BUG_ON(ac->ac_g_ex.fe_len < ac->ac_o_ex.fe_len); + + /* + * Use the below logic for adjusting best extent as it keeps + * fragmentation in check while ensuring logical range of best + * extent doesn't overflow out of goal extent: + * + * 1. Check if best ex can be kept at end of goal (before + * cr_best_avail trimmed it) and still cover original start + * 2. Else, check if best ex can be kept at start of goal and + * still cover original end + * 3. Else, keep the best ex at start of original request. + */ + ex.fe_len = ac->ac_b_ex.fe_len; + + ex.fe_logical = orig_goal_end - EXT4_C2B(sbi, ex.fe_len); + if (ac->ac_o_ex.fe_logical >= ex.fe_logical) + goto adjust_bex; + + ex.fe_logical = ac->ac_g_ex.fe_logical; + if (o_ex_end <= extent_logical_end(sbi, &ex)) + goto adjust_bex; + + ex.fe_logical = ac->ac_o_ex.fe_logical; +adjust_bex: + ac->ac_b_ex.fe_logical = ex.fe_logical; + + BUG_ON(ac->ac_o_ex.fe_logical < ac->ac_b_ex.fe_logical); + BUG_ON(extent_logical_end(sbi, &ex) > orig_goal_end); + } + + pa->pa_lstart = ac->ac_b_ex.fe_logical; + pa->pa_pstart = ext4_grp_offs_to_block(sb, &ac->ac_b_ex); + pa->pa_len = ac->ac_b_ex.fe_len; + pa->pa_free = pa->pa_len; + spin_lock_init(&pa->pa_lock); + INIT_LIST_HEAD(&pa->pa_group_list); + pa->pa_deleted = 0; + pa->pa_type = MB_INODE_PA; + + mb_debug(sb, "new inode pa %p: %llu/%d for %u\n", pa, pa->pa_pstart, + pa->pa_len, pa->pa_lstart); + trace_ext4_mb_new_inode_pa(ac, pa); + + atomic_add(pa->pa_free, &sbi->s_mb_preallocated); + ext4_mb_use_inode_pa(ac, pa); + + ei = EXT4_I(ac->ac_inode); + grp = ext4_get_group_info(sb, ac->ac_b_ex.fe_group); + if (!grp) + return; + + pa->pa_node_lock.inode_lock = &ei->i_prealloc_lock; + pa->pa_inode = ac->ac_inode; + + list_add(&pa->pa_group_list, &grp->bb_prealloc_list); + + write_lock(pa->pa_node_lock.inode_lock); + ext4_mb_pa_rb_insert(&ei->i_prealloc_node, &pa->pa_node.inode_node); + write_unlock(pa->pa_node_lock.inode_lock); + atomic_inc(&ei->i_prealloc_active); +} + +/* + * creates new preallocated space for locality group inodes belongs to + */ +static noinline_for_stack void +ext4_mb_new_group_pa(struct ext4_allocation_context *ac) +{ + struct super_block *sb = ac->ac_sb; + struct ext4_locality_group *lg; + struct ext4_prealloc_space *pa; + struct ext4_group_info *grp; + + /* preallocate only when found space is larger then requested */ + BUG_ON(ac->ac_o_ex.fe_len >= ac->ac_b_ex.fe_len); + BUG_ON(ac->ac_status != AC_STATUS_FOUND); + BUG_ON(!S_ISREG(ac->ac_inode->i_mode)); + BUG_ON(ac->ac_pa == NULL); + + pa = ac->ac_pa; + + pa->pa_pstart = ext4_grp_offs_to_block(sb, &ac->ac_b_ex); + pa->pa_lstart = pa->pa_pstart; + pa->pa_len = ac->ac_b_ex.fe_len; + pa->pa_free = pa->pa_len; + spin_lock_init(&pa->pa_lock); + INIT_LIST_HEAD(&pa->pa_node.lg_list); + INIT_LIST_HEAD(&pa->pa_group_list); + pa->pa_deleted = 0; + pa->pa_type = MB_GROUP_PA; + + mb_debug(sb, "new group pa %p: %llu/%d for %u\n", pa, pa->pa_pstart, + pa->pa_len, pa->pa_lstart); + trace_ext4_mb_new_group_pa(ac, pa); + + ext4_mb_use_group_pa(ac, pa); + atomic_add(pa->pa_free, &EXT4_SB(sb)->s_mb_preallocated); + + grp = ext4_get_group_info(sb, ac->ac_b_ex.fe_group); + if (!grp) + return; + lg = ac->ac_lg; + BUG_ON(lg == NULL); + + pa->pa_node_lock.lg_lock = &lg->lg_prealloc_lock; + pa->pa_inode = NULL; + + list_add(&pa->pa_group_list, &grp->bb_prealloc_list); + + /* + * We will later add the new pa to the right bucket + * after updating the pa_free in ext4_mb_release_context + */ +} + +static void ext4_mb_new_preallocation(struct ext4_allocation_context *ac) +{ + if (ac->ac_flags & EXT4_MB_HINT_GROUP_ALLOC) + ext4_mb_new_group_pa(ac); + else + ext4_mb_new_inode_pa(ac); +} + +/* + * finds all unused blocks in on-disk bitmap, frees them in + * in-core bitmap and buddy. + * @pa must be unlinked from inode and group lists, so that + * nobody else can find/use it. + * the caller MUST hold group/inode locks. + * TODO: optimize the case when there are no in-core structures yet + */ +static noinline_for_stack void +ext4_mb_release_inode_pa(struct ext4_buddy *e4b, struct buffer_head *bitmap_bh, + struct ext4_prealloc_space *pa) +{ + struct super_block *sb = e4b->bd_sb; + struct ext4_sb_info *sbi = EXT4_SB(sb); + unsigned int end; + unsigned int next; + ext4_group_t group; + ext4_grpblk_t bit; + unsigned long long grp_blk_start; + int free = 0; + + BUG_ON(pa->pa_deleted == 0); + ext4_get_group_no_and_offset(sb, pa->pa_pstart, &group, &bit); + grp_blk_start = pa->pa_pstart - EXT4_C2B(sbi, bit); + BUG_ON(group != e4b->bd_group && pa->pa_len != 0); + end = bit + pa->pa_len; + + while (bit < end) { + bit = mb_find_next_zero_bit(bitmap_bh->b_data, end, bit); + if (bit >= end) + break; + next = mb_find_next_bit(bitmap_bh->b_data, end, bit); + mb_debug(sb, "free preallocated %u/%u in group %u\n", + (unsigned) ext4_group_first_block_no(sb, group) + bit, + (unsigned) next - bit, (unsigned) group); + free += next - bit; + + trace_ext4_mballoc_discard(sb, NULL, group, bit, next - bit); + trace_ext4_mb_release_inode_pa(pa, (grp_blk_start + + EXT4_C2B(sbi, bit)), + next - bit); + mb_free_blocks(pa->pa_inode, e4b, bit, next - bit); + bit = next + 1; + } + if (free != pa->pa_free) { + ext4_msg(e4b->bd_sb, KERN_CRIT, + "pa %p: logic %lu, phys. %lu, len %d", + pa, (unsigned long) pa->pa_lstart, + (unsigned long) pa->pa_pstart, + pa->pa_len); + ext4_grp_locked_error(sb, group, 0, 0, "free %u, pa_free %u", + free, pa->pa_free); + /* + * pa is already deleted so we use the value obtained + * from the bitmap and continue. + */ + } + atomic_add(free, &sbi->s_mb_discarded); +} + +static noinline_for_stack void +ext4_mb_release_group_pa(struct ext4_buddy *e4b, + struct ext4_prealloc_space *pa) +{ + struct super_block *sb = e4b->bd_sb; + ext4_group_t group; + ext4_grpblk_t bit; + + trace_ext4_mb_release_group_pa(sb, pa); + BUG_ON(pa->pa_deleted == 0); + ext4_get_group_no_and_offset(sb, pa->pa_pstart, &group, &bit); + if (unlikely(group != e4b->bd_group && pa->pa_len != 0)) { + ext4_warning(sb, "bad group: expected %u, group %u, pa_start %llu", + e4b->bd_group, group, pa->pa_pstart); + return; + } + mb_free_blocks(pa->pa_inode, e4b, bit, pa->pa_len); + atomic_add(pa->pa_len, &EXT4_SB(sb)->s_mb_discarded); + trace_ext4_mballoc_discard(sb, NULL, group, bit, pa->pa_len); +} + +/* + * releases all preallocations in given group + * + * first, we need to decide discard policy: + * - when do we discard + * 1) ENOSPC + * - how many do we discard + * 1) how many requested + */ +static noinline_for_stack int +ext4_mb_discard_group_preallocations(struct super_block *sb, + ext4_group_t group, int *busy) +{ + struct ext4_group_info *grp = ext4_get_group_info(sb, group); + struct buffer_head *bitmap_bh = NULL; + struct ext4_prealloc_space *pa, *tmp; + LIST_HEAD(list); + struct ext4_buddy e4b; + struct ext4_inode_info *ei; + int err; + int free = 0; + + if (!grp) + return 0; + mb_debug(sb, "discard preallocation for group %u\n", group); + if (list_empty(&grp->bb_prealloc_list)) + goto out_dbg; + + bitmap_bh = ext4_read_block_bitmap(sb, group); + if (IS_ERR(bitmap_bh)) { + err = PTR_ERR(bitmap_bh); + ext4_error_err(sb, -err, + "Error %d reading block bitmap for %u", + err, group); + goto out_dbg; + } + + err = ext4_mb_load_buddy(sb, group, &e4b); + if (err) { + ext4_warning(sb, "Error %d loading buddy information for %u", + err, group); + put_bh(bitmap_bh); + goto out_dbg; + } + + ext4_lock_group(sb, group); + list_for_each_entry_safe(pa, tmp, + &grp->bb_prealloc_list, pa_group_list) { + spin_lock(&pa->pa_lock); + if (atomic_read(&pa->pa_count)) { + spin_unlock(&pa->pa_lock); + *busy = 1; + continue; + } + if (pa->pa_deleted) { + spin_unlock(&pa->pa_lock); + continue; + } + + /* seems this one can be freed ... */ + ext4_mb_mark_pa_deleted(sb, pa); + + if (!free) + this_cpu_inc(discard_pa_seq); + + /* we can trust pa_free ... */ + free += pa->pa_free; + + spin_unlock(&pa->pa_lock); + + list_del(&pa->pa_group_list); + list_add(&pa->u.pa_tmp_list, &list); + } + + /* now free all selected PAs */ + list_for_each_entry_safe(pa, tmp, &list, u.pa_tmp_list) { + + /* remove from object (inode or locality group) */ + if (pa->pa_type == MB_GROUP_PA) { + spin_lock(pa->pa_node_lock.lg_lock); + list_del_rcu(&pa->pa_node.lg_list); + spin_unlock(pa->pa_node_lock.lg_lock); + } else { + write_lock(pa->pa_node_lock.inode_lock); + ei = EXT4_I(pa->pa_inode); + rb_erase(&pa->pa_node.inode_node, &ei->i_prealloc_node); + write_unlock(pa->pa_node_lock.inode_lock); + } + + list_del(&pa->u.pa_tmp_list); + + if (pa->pa_type == MB_GROUP_PA) { + ext4_mb_release_group_pa(&e4b, pa); + call_rcu(&(pa)->u.pa_rcu, ext4_mb_pa_callback); + } else { + ext4_mb_release_inode_pa(&e4b, bitmap_bh, pa); + ext4_mb_pa_free(pa); + } + } + + ext4_unlock_group(sb, group); + ext4_mb_unload_buddy(&e4b); + put_bh(bitmap_bh); +out_dbg: + mb_debug(sb, "discarded (%d) blocks preallocated for group %u bb_free (%d)\n", + free, group, grp->bb_free); + return free; +} + +/* + * releases all non-used preallocated blocks for given inode + * + * It's important to discard preallocations under i_data_sem + * We don't want another block to be served from the prealloc + * space when we are discarding the inode prealloc space. + * + * FIXME!! Make sure it is valid at all the call sites + */ +void ext4_discard_preallocations(struct inode *inode) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + struct super_block *sb = inode->i_sb; + struct buffer_head *bitmap_bh = NULL; + struct ext4_prealloc_space *pa, *tmp; + ext4_group_t group = 0; + LIST_HEAD(list); + struct ext4_buddy e4b; + struct rb_node *iter; + int err; + + if (!S_ISREG(inode->i_mode)) + return; + + if (EXT4_SB(sb)->s_mount_state & EXT4_FC_REPLAY) + return; + + mb_debug(sb, "discard preallocation for inode %lu\n", + inode->i_ino); + trace_ext4_discard_preallocations(inode, + atomic_read(&ei->i_prealloc_active)); + +repeat: + /* first, collect all pa's in the inode */ + write_lock(&ei->i_prealloc_lock); + for (iter = rb_first(&ei->i_prealloc_node); iter; + iter = rb_next(iter)) { + pa = rb_entry(iter, struct ext4_prealloc_space, + pa_node.inode_node); + BUG_ON(pa->pa_node_lock.inode_lock != &ei->i_prealloc_lock); + + spin_lock(&pa->pa_lock); + if (atomic_read(&pa->pa_count)) { + /* this shouldn't happen often - nobody should + * use preallocation while we're discarding it */ + spin_unlock(&pa->pa_lock); + write_unlock(&ei->i_prealloc_lock); + ext4_msg(sb, KERN_ERR, + "uh-oh! used pa while discarding"); + WARN_ON(1); + schedule_timeout_uninterruptible(HZ); + goto repeat; + + } + if (pa->pa_deleted == 0) { + ext4_mb_mark_pa_deleted(sb, pa); + spin_unlock(&pa->pa_lock); + rb_erase(&pa->pa_node.inode_node, &ei->i_prealloc_node); + list_add(&pa->u.pa_tmp_list, &list); + continue; + } + + /* someone is deleting pa right now */ + spin_unlock(&pa->pa_lock); + write_unlock(&ei->i_prealloc_lock); + + /* we have to wait here because pa_deleted + * doesn't mean pa is already unlinked from + * the list. as we might be called from + * ->clear_inode() the inode will get freed + * and concurrent thread which is unlinking + * pa from inode's list may access already + * freed memory, bad-bad-bad */ + + /* XXX: if this happens too often, we can + * add a flag to force wait only in case + * of ->clear_inode(), but not in case of + * regular truncate */ + schedule_timeout_uninterruptible(HZ); + goto repeat; + } + write_unlock(&ei->i_prealloc_lock); + + list_for_each_entry_safe(pa, tmp, &list, u.pa_tmp_list) { + BUG_ON(pa->pa_type != MB_INODE_PA); + group = ext4_get_group_number(sb, pa->pa_pstart); + + err = ext4_mb_load_buddy_gfp(sb, group, &e4b, + GFP_NOFS|__GFP_NOFAIL); + if (err) { + ext4_error_err(sb, -err, "Error %d loading buddy information for %u", + err, group); + continue; + } + + bitmap_bh = ext4_read_block_bitmap(sb, group); + if (IS_ERR(bitmap_bh)) { + err = PTR_ERR(bitmap_bh); + ext4_error_err(sb, -err, "Error %d reading block bitmap for %u", + err, group); + ext4_mb_unload_buddy(&e4b); + continue; + } + + ext4_lock_group(sb, group); + list_del(&pa->pa_group_list); + ext4_mb_release_inode_pa(&e4b, bitmap_bh, pa); + ext4_unlock_group(sb, group); + + ext4_mb_unload_buddy(&e4b); + put_bh(bitmap_bh); + + list_del(&pa->u.pa_tmp_list); + ext4_mb_pa_free(pa); + } +} + +static int ext4_mb_pa_alloc(struct ext4_allocation_context *ac) +{ + struct ext4_prealloc_space *pa; + + BUG_ON(ext4_pspace_cachep == NULL); + pa = kmem_cache_zalloc(ext4_pspace_cachep, GFP_NOFS); + if (!pa) + return -ENOMEM; + atomic_set(&pa->pa_count, 1); + ac->ac_pa = pa; + return 0; +} + +static void ext4_mb_pa_put_free(struct ext4_allocation_context *ac) +{ + struct ext4_prealloc_space *pa = ac->ac_pa; + + BUG_ON(!pa); + ac->ac_pa = NULL; + WARN_ON(!atomic_dec_and_test(&pa->pa_count)); + /* + * current function is only called due to an error or due to + * len of found blocks < len of requested blocks hence the PA has not + * been added to grp->bb_prealloc_list. So we don't need to lock it + */ + pa->pa_deleted = 1; + ext4_mb_pa_free(pa); +} + +#ifdef CONFIG_EXT4_DEBUG +static inline void ext4_mb_show_pa(struct super_block *sb) +{ + ext4_group_t i, ngroups; + + if (ext4_emergency_state(sb)) + return; + + ngroups = ext4_get_groups_count(sb); + mb_debug(sb, "groups: "); + for (i = 0; i < ngroups; i++) { + struct ext4_group_info *grp = ext4_get_group_info(sb, i); + struct ext4_prealloc_space *pa; + ext4_grpblk_t start; + struct list_head *cur; + + if (!grp) + continue; + ext4_lock_group(sb, i); + list_for_each(cur, &grp->bb_prealloc_list) { + pa = list_entry(cur, struct ext4_prealloc_space, + pa_group_list); + spin_lock(&pa->pa_lock); + ext4_get_group_no_and_offset(sb, pa->pa_pstart, + NULL, &start); + spin_unlock(&pa->pa_lock); + mb_debug(sb, "PA:%u:%d:%d\n", i, start, + pa->pa_len); + } + ext4_unlock_group(sb, i); + mb_debug(sb, "%u: %d/%d\n", i, grp->bb_free, + grp->bb_fragments); + } +} + +static void ext4_mb_show_ac(struct ext4_allocation_context *ac) +{ + struct super_block *sb = ac->ac_sb; + + if (ext4_emergency_state(sb)) + return; + + mb_debug(sb, "Can't allocate:" + " Allocation context details:"); + mb_debug(sb, "status %u flags 0x%x", + ac->ac_status, ac->ac_flags); + mb_debug(sb, "orig %lu/%lu/%lu@%lu, " + "goal %lu/%lu/%lu@%lu, " + "best %lu/%lu/%lu@%lu cr %d", + (unsigned long)ac->ac_o_ex.fe_group, + (unsigned long)ac->ac_o_ex.fe_start, + (unsigned long)ac->ac_o_ex.fe_len, + (unsigned long)ac->ac_o_ex.fe_logical, + (unsigned long)ac->ac_g_ex.fe_group, + (unsigned long)ac->ac_g_ex.fe_start, + (unsigned long)ac->ac_g_ex.fe_len, + (unsigned long)ac->ac_g_ex.fe_logical, + (unsigned long)ac->ac_b_ex.fe_group, + (unsigned long)ac->ac_b_ex.fe_start, + (unsigned long)ac->ac_b_ex.fe_len, + (unsigned long)ac->ac_b_ex.fe_logical, + (int)ac->ac_criteria); + mb_debug(sb, "%u found", ac->ac_found); + mb_debug(sb, "used pa: %s, ", str_yes_no(ac->ac_pa)); + if (ac->ac_pa) + mb_debug(sb, "pa_type %s\n", ac->ac_pa->pa_type == MB_GROUP_PA ? + "group pa" : "inode pa"); + ext4_mb_show_pa(sb); +} +#else +static inline void ext4_mb_show_pa(struct super_block *sb) +{ +} +static inline void ext4_mb_show_ac(struct ext4_allocation_context *ac) +{ + ext4_mb_show_pa(ac->ac_sb); +} +#endif + +/* + * We use locality group preallocation for small size file. The size of the + * file is determined by the current size or the resulting size after + * allocation which ever is larger + * + * One can tune this size via /sys/fs/ext4/<partition>/mb_stream_req + */ +static void ext4_mb_group_or_file(struct ext4_allocation_context *ac) +{ + struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); + int bsbits = ac->ac_sb->s_blocksize_bits; + loff_t size, isize; + bool inode_pa_eligible, group_pa_eligible; + + if (!(ac->ac_flags & EXT4_MB_HINT_DATA)) + return; + + if (unlikely(ac->ac_flags & EXT4_MB_HINT_GOAL_ONLY)) + return; + + group_pa_eligible = sbi->s_mb_group_prealloc > 0; + inode_pa_eligible = true; + size = extent_logical_end(sbi, &ac->ac_o_ex); + isize = (i_size_read(ac->ac_inode) + ac->ac_sb->s_blocksize - 1) + >> bsbits; + + /* No point in using inode preallocation for closed files */ + if ((size == isize) && !ext4_fs_is_busy(sbi) && + !inode_is_open_for_write(ac->ac_inode)) + inode_pa_eligible = false; + + size = max(size, isize); + /* Don't use group allocation for large files */ + if (size > sbi->s_mb_stream_request) + group_pa_eligible = false; + + if (!group_pa_eligible) { + if (inode_pa_eligible) + ac->ac_flags |= EXT4_MB_STREAM_ALLOC; + else + ac->ac_flags |= EXT4_MB_HINT_NOPREALLOC; + return; + } + + BUG_ON(ac->ac_lg != NULL); + /* + * locality group prealloc space are per cpu. The reason for having + * per cpu locality group is to reduce the contention between block + * request from multiple CPUs. + */ + ac->ac_lg = raw_cpu_ptr(sbi->s_locality_groups); + + /* we're going to use group allocation */ + ac->ac_flags |= EXT4_MB_HINT_GROUP_ALLOC; + + /* serialize all allocations in the group */ + mutex_lock(&ac->ac_lg->lg_mutex); +} + +static noinline_for_stack void +ext4_mb_initialize_context(struct ext4_allocation_context *ac, + struct ext4_allocation_request *ar) +{ + struct super_block *sb = ar->inode->i_sb; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_super_block *es = sbi->s_es; + ext4_group_t group; + unsigned int len; + ext4_fsblk_t goal; + ext4_grpblk_t block; + + /* we can't allocate > group size */ + len = ar->len; + + /* just a dirty hack to filter too big requests */ + if (len >= EXT4_CLUSTERS_PER_GROUP(sb)) + len = EXT4_CLUSTERS_PER_GROUP(sb); + + /* start searching from the goal */ + goal = ar->goal; + if (goal < le32_to_cpu(es->s_first_data_block) || + goal >= ext4_blocks_count(es)) + goal = le32_to_cpu(es->s_first_data_block); + ext4_get_group_no_and_offset(sb, goal, &group, &block); + + /* set up allocation goals */ + ac->ac_b_ex.fe_logical = EXT4_LBLK_CMASK(sbi, ar->logical); + ac->ac_status = AC_STATUS_CONTINUE; + ac->ac_sb = sb; + ac->ac_inode = ar->inode; + ac->ac_o_ex.fe_logical = ac->ac_b_ex.fe_logical; + ac->ac_o_ex.fe_group = group; + ac->ac_o_ex.fe_start = block; + ac->ac_o_ex.fe_len = len; + ac->ac_g_ex = ac->ac_o_ex; + ac->ac_orig_goal_len = ac->ac_g_ex.fe_len; + ac->ac_flags = ar->flags; + + /* we have to define context: we'll work with a file or + * locality group. this is a policy, actually */ + ext4_mb_group_or_file(ac); + + mb_debug(sb, "init ac: %u blocks @ %u, goal %u, flags 0x%x, 2^%d, " + "left: %u/%u, right %u/%u to %swritable\n", + (unsigned) ar->len, (unsigned) ar->logical, + (unsigned) ar->goal, ac->ac_flags, ac->ac_2order, + (unsigned) ar->lleft, (unsigned) ar->pleft, + (unsigned) ar->lright, (unsigned) ar->pright, + inode_is_open_for_write(ar->inode) ? "" : "non-"); +} + +static noinline_for_stack void +ext4_mb_discard_lg_preallocations(struct super_block *sb, + struct ext4_locality_group *lg, + int order, int total_entries) +{ + ext4_group_t group = 0; + struct ext4_buddy e4b; + LIST_HEAD(discard_list); + struct ext4_prealloc_space *pa, *tmp; + + mb_debug(sb, "discard locality group preallocation\n"); + + spin_lock(&lg->lg_prealloc_lock); + list_for_each_entry_rcu(pa, &lg->lg_prealloc_list[order], + pa_node.lg_list, + lockdep_is_held(&lg->lg_prealloc_lock)) { + spin_lock(&pa->pa_lock); + if (atomic_read(&pa->pa_count)) { + /* + * This is the pa that we just used + * for block allocation. So don't + * free that + */ + spin_unlock(&pa->pa_lock); + continue; + } + if (pa->pa_deleted) { + spin_unlock(&pa->pa_lock); + continue; + } + /* only lg prealloc space */ + BUG_ON(pa->pa_type != MB_GROUP_PA); + + /* seems this one can be freed ... */ + ext4_mb_mark_pa_deleted(sb, pa); + spin_unlock(&pa->pa_lock); + + list_del_rcu(&pa->pa_node.lg_list); + list_add(&pa->u.pa_tmp_list, &discard_list); + + total_entries--; + if (total_entries <= 5) { + /* + * we want to keep only 5 entries + * allowing it to grow to 8. This + * mak sure we don't call discard + * soon for this list. + */ + break; + } + } + spin_unlock(&lg->lg_prealloc_lock); + + list_for_each_entry_safe(pa, tmp, &discard_list, u.pa_tmp_list) { + int err; + + group = ext4_get_group_number(sb, pa->pa_pstart); + err = ext4_mb_load_buddy_gfp(sb, group, &e4b, + GFP_NOFS|__GFP_NOFAIL); + if (err) { + ext4_error_err(sb, -err, "Error %d loading buddy information for %u", + err, group); + continue; + } + ext4_lock_group(sb, group); + list_del(&pa->pa_group_list); + ext4_mb_release_group_pa(&e4b, pa); + ext4_unlock_group(sb, group); + + ext4_mb_unload_buddy(&e4b); + list_del(&pa->u.pa_tmp_list); + call_rcu(&(pa)->u.pa_rcu, ext4_mb_pa_callback); + } +} + +/* + * We have incremented pa_count. So it cannot be freed at this + * point. Also we hold lg_mutex. So no parallel allocation is + * possible from this lg. That means pa_free cannot be updated. + * + * A parallel ext4_mb_discard_group_preallocations is possible. + * which can cause the lg_prealloc_list to be updated. + */ + +static void ext4_mb_add_n_trim(struct ext4_allocation_context *ac) +{ + int order, added = 0, lg_prealloc_count = 1; + struct super_block *sb = ac->ac_sb; + struct ext4_locality_group *lg = ac->ac_lg; + struct ext4_prealloc_space *tmp_pa, *pa = ac->ac_pa; + + order = fls(pa->pa_free) - 1; + if (order > PREALLOC_TB_SIZE - 1) + /* The max size of hash table is PREALLOC_TB_SIZE */ + order = PREALLOC_TB_SIZE - 1; + /* Add the prealloc space to lg */ + spin_lock(&lg->lg_prealloc_lock); + list_for_each_entry_rcu(tmp_pa, &lg->lg_prealloc_list[order], + pa_node.lg_list, + lockdep_is_held(&lg->lg_prealloc_lock)) { + spin_lock(&tmp_pa->pa_lock); + if (tmp_pa->pa_deleted) { + spin_unlock(&tmp_pa->pa_lock); + continue; + } + if (!added && pa->pa_free < tmp_pa->pa_free) { + /* Add to the tail of the previous entry */ + list_add_tail_rcu(&pa->pa_node.lg_list, + &tmp_pa->pa_node.lg_list); + added = 1; + /* + * we want to count the total + * number of entries in the list + */ + } + spin_unlock(&tmp_pa->pa_lock); + lg_prealloc_count++; + } + if (!added) + list_add_tail_rcu(&pa->pa_node.lg_list, + &lg->lg_prealloc_list[order]); + spin_unlock(&lg->lg_prealloc_lock); + + /* Now trim the list to be not more than 8 elements */ + if (lg_prealloc_count > 8) + ext4_mb_discard_lg_preallocations(sb, lg, + order, lg_prealloc_count); +} + +/* + * release all resource we used in allocation + */ +static void ext4_mb_release_context(struct ext4_allocation_context *ac) +{ + struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); + struct ext4_prealloc_space *pa = ac->ac_pa; + if (pa) { + if (pa->pa_type == MB_GROUP_PA) { + /* see comment in ext4_mb_use_group_pa() */ + spin_lock(&pa->pa_lock); + pa->pa_pstart += EXT4_C2B(sbi, ac->ac_b_ex.fe_len); + pa->pa_lstart += EXT4_C2B(sbi, ac->ac_b_ex.fe_len); + pa->pa_free -= ac->ac_b_ex.fe_len; + pa->pa_len -= ac->ac_b_ex.fe_len; + spin_unlock(&pa->pa_lock); + + /* + * We want to add the pa to the right bucket. + * Remove it from the list and while adding + * make sure the list to which we are adding + * doesn't grow big. + */ + if (likely(pa->pa_free)) { + spin_lock(pa->pa_node_lock.lg_lock); + list_del_rcu(&pa->pa_node.lg_list); + spin_unlock(pa->pa_node_lock.lg_lock); + ext4_mb_add_n_trim(ac); + } + } + + ext4_mb_put_pa(ac, ac->ac_sb, pa); + } + if (ac->ac_bitmap_folio) + folio_put(ac->ac_bitmap_folio); + if (ac->ac_buddy_folio) + folio_put(ac->ac_buddy_folio); + if (ac->ac_flags & EXT4_MB_HINT_GROUP_ALLOC) + mutex_unlock(&ac->ac_lg->lg_mutex); + ext4_mb_collect_stats(ac); +} + +static int ext4_mb_discard_preallocations(struct super_block *sb, int needed) +{ + ext4_group_t i, ngroups = ext4_get_groups_count(sb); + int ret; + int freed = 0, busy = 0; + int retry = 0; + + trace_ext4_mb_discard_preallocations(sb, needed); + + if (needed == 0) + needed = EXT4_CLUSTERS_PER_GROUP(sb) + 1; + repeat: + for (i = 0; i < ngroups && needed > 0; i++) { + ret = ext4_mb_discard_group_preallocations(sb, i, &busy); + freed += ret; + needed -= ret; + cond_resched(); + } + + if (needed > 0 && busy && ++retry < 3) { + busy = 0; + goto repeat; + } + + return freed; +} + +static bool ext4_mb_discard_preallocations_should_retry(struct super_block *sb, + struct ext4_allocation_context *ac, u64 *seq) +{ + int freed; + u64 seq_retry = 0; + bool ret = false; + + freed = ext4_mb_discard_preallocations(sb, ac->ac_o_ex.fe_len); + if (freed) { + ret = true; + goto out_dbg; + } + seq_retry = ext4_get_discard_pa_seq_sum(); + if (!(ac->ac_flags & EXT4_MB_STRICT_CHECK) || seq_retry != *seq) { + ac->ac_flags |= EXT4_MB_STRICT_CHECK; + *seq = seq_retry; + ret = true; + } + +out_dbg: + mb_debug(sb, "freed %d, retry ? %s\n", freed, str_yes_no(ret)); + return ret; +} + +/* + * Simple allocator for Ext4 fast commit replay path. It searches for blocks + * linearly starting at the goal block and also excludes the blocks which + * are going to be in use after fast commit replay. + */ +static ext4_fsblk_t +ext4_mb_new_blocks_simple(struct ext4_allocation_request *ar, int *errp) +{ + struct buffer_head *bitmap_bh; + struct super_block *sb = ar->inode->i_sb; + struct ext4_sb_info *sbi = EXT4_SB(sb); + ext4_group_t group, nr; + ext4_grpblk_t blkoff; + ext4_grpblk_t max = EXT4_CLUSTERS_PER_GROUP(sb); + ext4_grpblk_t i = 0; + ext4_fsblk_t goal, block; + struct ext4_super_block *es = sbi->s_es; + + goal = ar->goal; + if (goal < le32_to_cpu(es->s_first_data_block) || + goal >= ext4_blocks_count(es)) + goal = le32_to_cpu(es->s_first_data_block); + + ar->len = 0; + ext4_get_group_no_and_offset(sb, goal, &group, &blkoff); + for (nr = ext4_get_groups_count(sb); nr > 0; nr--) { + bitmap_bh = ext4_read_block_bitmap(sb, group); + if (IS_ERR(bitmap_bh)) { + *errp = PTR_ERR(bitmap_bh); + pr_warn("Failed to read block bitmap\n"); + return 0; + } + + while (1) { + i = mb_find_next_zero_bit(bitmap_bh->b_data, max, + blkoff); + if (i >= max) + break; + if (ext4_fc_replay_check_excluded(sb, + ext4_group_first_block_no(sb, group) + + EXT4_C2B(sbi, i))) { + blkoff = i + 1; + } else + break; + } + brelse(bitmap_bh); + if (i < max) + break; + + if (++group >= ext4_get_groups_count(sb)) + group = 0; + + blkoff = 0; + } + + if (i >= max) { + *errp = -ENOSPC; + return 0; + } + + block = ext4_group_first_block_no(sb, group) + EXT4_C2B(sbi, i); + ext4_mb_mark_bb(sb, block, 1, true); + ar->len = 1; + + *errp = 0; + return block; +} + +/* + * Main entry point into mballoc to allocate blocks + * it tries to use preallocation first, then falls back + * to usual allocation + */ +ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle, + struct ext4_allocation_request *ar, int *errp) +{ + struct ext4_allocation_context *ac = NULL; + struct ext4_sb_info *sbi; + struct super_block *sb; + ext4_fsblk_t block = 0; + unsigned int inquota = 0; + unsigned int reserv_clstrs = 0; + int retries = 0; + u64 seq; + + might_sleep(); + sb = ar->inode->i_sb; + sbi = EXT4_SB(sb); + + trace_ext4_request_blocks(ar); + if (sbi->s_mount_state & EXT4_FC_REPLAY) + return ext4_mb_new_blocks_simple(ar, errp); + + /* Allow to use superuser reservation for quota file */ + if (ext4_is_quota_file(ar->inode)) + ar->flags |= EXT4_MB_USE_ROOT_BLOCKS; + + if ((ar->flags & EXT4_MB_DELALLOC_RESERVED) == 0) { + /* Without delayed allocation we need to verify + * there is enough free blocks to do block allocation + * and verify allocation doesn't exceed the quota limits. + */ + while (ar->len && + ext4_claim_free_clusters(sbi, ar->len, ar->flags)) { + + /* let others to free the space */ + cond_resched(); + ar->len = ar->len >> 1; + } + if (!ar->len) { + ext4_mb_show_pa(sb); + *errp = -ENOSPC; + return 0; + } + reserv_clstrs = ar->len; + if (ar->flags & EXT4_MB_USE_ROOT_BLOCKS) { + dquot_alloc_block_nofail(ar->inode, + EXT4_C2B(sbi, ar->len)); + } else { + while (ar->len && + dquot_alloc_block(ar->inode, + EXT4_C2B(sbi, ar->len))) { + + ar->flags |= EXT4_MB_HINT_NOPREALLOC; + ar->len--; + } + } + inquota = ar->len; + if (ar->len == 0) { + *errp = -EDQUOT; + goto out; + } + } + + ac = kmem_cache_zalloc(ext4_ac_cachep, GFP_NOFS); + if (!ac) { + ar->len = 0; + *errp = -ENOMEM; + goto out; + } + + ext4_mb_initialize_context(ac, ar); + + ac->ac_op = EXT4_MB_HISTORY_PREALLOC; + seq = this_cpu_read(discard_pa_seq); + if (!ext4_mb_use_preallocated(ac)) { + ac->ac_op = EXT4_MB_HISTORY_ALLOC; + ext4_mb_normalize_request(ac, ar); + + *errp = ext4_mb_pa_alloc(ac); + if (*errp) + goto errout; +repeat: + /* allocate space in core */ + *errp = ext4_mb_regular_allocator(ac); + /* + * pa allocated above is added to grp->bb_prealloc_list only + * when we were able to allocate some block i.e. when + * ac->ac_status == AC_STATUS_FOUND. + * And error from above mean ac->ac_status != AC_STATUS_FOUND + * So we have to free this pa here itself. + */ + if (*errp) { + ext4_mb_pa_put_free(ac); + ext4_discard_allocated_blocks(ac); + goto errout; + } + if (ac->ac_status == AC_STATUS_FOUND && + ac->ac_o_ex.fe_len >= ac->ac_f_ex.fe_len) + ext4_mb_pa_put_free(ac); + } + if (likely(ac->ac_status == AC_STATUS_FOUND)) { + *errp = ext4_mb_mark_diskspace_used(ac, handle, reserv_clstrs); + if (*errp) { + ext4_discard_allocated_blocks(ac); + goto errout; + } else { + block = ext4_grp_offs_to_block(sb, &ac->ac_b_ex); + ar->len = ac->ac_b_ex.fe_len; + } + } else { + if (++retries < 3 && + ext4_mb_discard_preallocations_should_retry(sb, ac, &seq)) + goto repeat; + /* + * If block allocation fails then the pa allocated above + * needs to be freed here itself. + */ + ext4_mb_pa_put_free(ac); + *errp = -ENOSPC; + } + + if (*errp) { +errout: + ac->ac_b_ex.fe_len = 0; + ar->len = 0; + ext4_mb_show_ac(ac); + } + ext4_mb_release_context(ac); + kmem_cache_free(ext4_ac_cachep, ac); +out: + if (inquota && ar->len < inquota) + dquot_free_block(ar->inode, EXT4_C2B(sbi, inquota - ar->len)); + if (!ar->len) { + if ((ar->flags & EXT4_MB_DELALLOC_RESERVED) == 0) + /* release all the reserved blocks if non delalloc */ + percpu_counter_sub(&sbi->s_dirtyclusters_counter, + reserv_clstrs); + } + + trace_ext4_allocate_blocks(ar, (unsigned long long)block); + + return block; +} + +/* + * We can merge two free data extents only if the physical blocks + * are contiguous, AND the extents were freed by the same transaction, + * AND the blocks are associated with the same group. + */ +static inline bool +ext4_freed_extents_can_be_merged(struct ext4_free_data *entry1, + struct ext4_free_data *entry2) +{ + if (entry1->efd_tid != entry2->efd_tid) + return false; + if (entry1->efd_start_cluster + entry1->efd_count != + entry2->efd_start_cluster) + return false; + if (WARN_ON_ONCE(entry1->efd_group != entry2->efd_group)) + return false; + return true; +} + +static inline void +ext4_merge_freed_extents(struct ext4_sb_info *sbi, struct rb_root *root, + struct ext4_free_data *entry1, + struct ext4_free_data *entry2) +{ + entry1->efd_count += entry2->efd_count; + spin_lock(&sbi->s_md_lock); + list_del(&entry2->efd_list); + spin_unlock(&sbi->s_md_lock); + rb_erase(&entry2->efd_node, root); + kmem_cache_free(ext4_free_data_cachep, entry2); +} + +static inline void +ext4_try_merge_freed_extent_prev(struct ext4_sb_info *sbi, struct rb_root *root, + struct ext4_free_data *entry) +{ + struct ext4_free_data *prev; + struct rb_node *node; + + node = rb_prev(&entry->efd_node); + if (!node) + return; + + prev = rb_entry(node, struct ext4_free_data, efd_node); + if (ext4_freed_extents_can_be_merged(prev, entry)) + ext4_merge_freed_extents(sbi, root, prev, entry); +} + +static inline void +ext4_try_merge_freed_extent_next(struct ext4_sb_info *sbi, struct rb_root *root, + struct ext4_free_data *entry) +{ + struct ext4_free_data *next; + struct rb_node *node; + + node = rb_next(&entry->efd_node); + if (!node) + return; + + next = rb_entry(node, struct ext4_free_data, efd_node); + if (ext4_freed_extents_can_be_merged(entry, next)) + ext4_merge_freed_extents(sbi, root, entry, next); +} + +static noinline_for_stack void +ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b, + struct ext4_free_data *new_entry) +{ + ext4_group_t group = e4b->bd_group; + ext4_grpblk_t cluster; + ext4_grpblk_t clusters = new_entry->efd_count; + struct ext4_free_data *entry = NULL; + struct ext4_group_info *db = e4b->bd_info; + struct super_block *sb = e4b->bd_sb; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct rb_root *root = &db->bb_free_root; + struct rb_node **n = &root->rb_node; + struct rb_node *parent = NULL, *new_node; + + BUG_ON(!ext4_handle_valid(handle)); + BUG_ON(e4b->bd_bitmap_folio == NULL); + BUG_ON(e4b->bd_buddy_folio == NULL); + + new_node = &new_entry->efd_node; + cluster = new_entry->efd_start_cluster; + + if (!*n) { + /* first free block exent. We need to + protect buddy cache from being freed, + * otherwise we'll refresh it from + * on-disk bitmap and lose not-yet-available + * blocks */ + folio_get(e4b->bd_buddy_folio); + folio_get(e4b->bd_bitmap_folio); + } + while (*n) { + parent = *n; + entry = rb_entry(parent, struct ext4_free_data, efd_node); + if (cluster < entry->efd_start_cluster) + n = &(*n)->rb_left; + else if (cluster >= (entry->efd_start_cluster + entry->efd_count)) + n = &(*n)->rb_right; + else { + ext4_grp_locked_error(sb, group, 0, + ext4_group_first_block_no(sb, group) + + EXT4_C2B(sbi, cluster), + "Block already on to-be-freed list"); + kmem_cache_free(ext4_free_data_cachep, new_entry); + return; + } + } + + atomic_add(clusters, &sbi->s_mb_free_pending); + if (!entry) + goto insert; + + /* Now try to see the extent can be merged to prev and next */ + if (ext4_freed_extents_can_be_merged(new_entry, entry)) { + entry->efd_start_cluster = cluster; + entry->efd_count += new_entry->efd_count; + kmem_cache_free(ext4_free_data_cachep, new_entry); + ext4_try_merge_freed_extent_prev(sbi, root, entry); + return; + } + if (ext4_freed_extents_can_be_merged(entry, new_entry)) { + entry->efd_count += new_entry->efd_count; + kmem_cache_free(ext4_free_data_cachep, new_entry); + ext4_try_merge_freed_extent_next(sbi, root, entry); + return; + } +insert: + rb_link_node(new_node, parent, n); + rb_insert_color(new_node, root); + + spin_lock(&sbi->s_md_lock); + list_add_tail(&new_entry->efd_list, &sbi->s_freed_data_list[new_entry->efd_tid & 1]); + spin_unlock(&sbi->s_md_lock); +} + +static void ext4_free_blocks_simple(struct inode *inode, ext4_fsblk_t block, + unsigned long count) +{ + struct super_block *sb = inode->i_sb; + ext4_group_t group; + ext4_grpblk_t blkoff; + + ext4_get_group_no_and_offset(sb, block, &group, &blkoff); + ext4_mb_mark_context(NULL, sb, false, group, blkoff, count, + EXT4_MB_BITMAP_MARKED_CHECK | + EXT4_MB_SYNC_UPDATE, + NULL); +} + +/** + * ext4_mb_clear_bb() -- helper function for freeing blocks. + * Used by ext4_free_blocks() + * @handle: handle for this transaction + * @inode: inode + * @block: starting physical block to be freed + * @count: number of blocks to be freed + * @flags: flags used by ext4_free_blocks + */ +static void ext4_mb_clear_bb(handle_t *handle, struct inode *inode, + ext4_fsblk_t block, unsigned long count, + int flags) +{ + struct super_block *sb = inode->i_sb; + struct ext4_group_info *grp; + unsigned int overflow; + ext4_grpblk_t bit; + ext4_group_t block_group; + struct ext4_sb_info *sbi; + struct ext4_buddy e4b; + unsigned int count_clusters; + int err = 0; + int mark_flags = 0; + ext4_grpblk_t changed; + + sbi = EXT4_SB(sb); + + if (!(flags & EXT4_FREE_BLOCKS_VALIDATED) && + !ext4_inode_block_valid(inode, block, count)) { + ext4_error(sb, "Freeing blocks in system zone - " + "Block = %llu, count = %lu", block, count); + /* err = 0. ext4_std_error should be a no op */ + goto error_out; + } + flags |= EXT4_FREE_BLOCKS_VALIDATED; + +do_more: + overflow = 0; + ext4_get_group_no_and_offset(sb, block, &block_group, &bit); + + grp = ext4_get_group_info(sb, block_group); + if (unlikely(!grp || EXT4_MB_GRP_BBITMAP_CORRUPT(grp))) + return; + + /* + * Check to see if we are freeing blocks across a group + * boundary. + */ + if (EXT4_C2B(sbi, bit) + count > EXT4_BLOCKS_PER_GROUP(sb)) { + overflow = EXT4_C2B(sbi, bit) + count - + EXT4_BLOCKS_PER_GROUP(sb); + count -= overflow; + /* The range changed so it's no longer validated */ + flags &= ~EXT4_FREE_BLOCKS_VALIDATED; + } + count_clusters = EXT4_NUM_B2C(sbi, count); + trace_ext4_mballoc_free(sb, inode, block_group, bit, count_clusters); + + /* __GFP_NOFAIL: retry infinitely, ignore TIF_MEMDIE and memcg limit. */ + err = ext4_mb_load_buddy_gfp(sb, block_group, &e4b, + GFP_NOFS|__GFP_NOFAIL); + if (err) + goto error_out; + + if (!(flags & EXT4_FREE_BLOCKS_VALIDATED) && + !ext4_inode_block_valid(inode, block, count)) { + ext4_error(sb, "Freeing blocks in system zone - " + "Block = %llu, count = %lu", block, count); + /* err = 0. ext4_std_error should be a no op */ + goto error_clean; + } + +#ifdef AGGRESSIVE_CHECK + mark_flags |= EXT4_MB_BITMAP_MARKED_CHECK; +#endif + err = ext4_mb_mark_context(handle, sb, false, block_group, bit, + count_clusters, mark_flags, &changed); + + + if (err && changed == 0) + goto error_clean; + +#ifdef AGGRESSIVE_CHECK + BUG_ON(changed != count_clusters); +#endif + + /* + * We need to make sure we don't reuse the freed block until after the + * transaction is committed. We make an exception if the inode is to be + * written in writeback mode since writeback mode has weak data + * consistency guarantees. + */ + if (ext4_handle_valid(handle) && + ((flags & EXT4_FREE_BLOCKS_METADATA) || + !ext4_should_writeback_data(inode))) { + struct ext4_free_data *new_entry; + /* + * We use __GFP_NOFAIL because ext4_free_blocks() is not allowed + * to fail. + */ + new_entry = kmem_cache_alloc(ext4_free_data_cachep, + GFP_NOFS|__GFP_NOFAIL); + new_entry->efd_start_cluster = bit; + new_entry->efd_group = block_group; + new_entry->efd_count = count_clusters; + new_entry->efd_tid = handle->h_transaction->t_tid; + + ext4_lock_group(sb, block_group); + ext4_mb_free_metadata(handle, &e4b, new_entry); + } else { + if (test_opt(sb, DISCARD)) { + err = ext4_issue_discard(sb, block_group, bit, + count_clusters); + /* + * Ignore EOPNOTSUPP error. This is consistent with + * what happens when using journal. + */ + if (err == -EOPNOTSUPP) + err = 0; + if (err) + ext4_msg(sb, KERN_WARNING, "discard request in" + " group:%u block:%d count:%lu failed" + " with %d", block_group, bit, count, + err); + } + + EXT4_MB_GRP_CLEAR_TRIMMED(e4b.bd_info); + + ext4_lock_group(sb, block_group); + mb_free_blocks(inode, &e4b, bit, count_clusters); + } + + ext4_unlock_group(sb, block_group); + + /* + * on a bigalloc file system, defer the s_freeclusters_counter + * update to the caller (ext4_remove_space and friends) so they + * can determine if a cluster freed here should be rereserved + */ + if (!(flags & EXT4_FREE_BLOCKS_RERESERVE_CLUSTER)) { + if (!(flags & EXT4_FREE_BLOCKS_NO_QUOT_UPDATE)) + dquot_free_block(inode, EXT4_C2B(sbi, count_clusters)); + percpu_counter_add(&sbi->s_freeclusters_counter, + count_clusters); + } + + if (overflow && !err) { + block += count; + count = overflow; + ext4_mb_unload_buddy(&e4b); + /* The range changed so it's no longer validated */ + flags &= ~EXT4_FREE_BLOCKS_VALIDATED; + goto do_more; + } + +error_clean: + ext4_mb_unload_buddy(&e4b); +error_out: + ext4_std_error(sb, err); +} + +/** + * ext4_free_blocks() -- Free given blocks and update quota + * @handle: handle for this transaction + * @inode: inode + * @bh: optional buffer of the block to be freed + * @block: starting physical block to be freed + * @count: number of blocks to be freed + * @flags: flags used by ext4_free_blocks + */ +void ext4_free_blocks(handle_t *handle, struct inode *inode, + struct buffer_head *bh, ext4_fsblk_t block, + unsigned long count, int flags) +{ + struct super_block *sb = inode->i_sb; + unsigned int overflow; + struct ext4_sb_info *sbi; + + sbi = EXT4_SB(sb); + + if (bh) { + if (block) + BUG_ON(block != bh->b_blocknr); + else + block = bh->b_blocknr; + } + + if (sbi->s_mount_state & EXT4_FC_REPLAY) { + ext4_free_blocks_simple(inode, block, EXT4_NUM_B2C(sbi, count)); + return; + } + + might_sleep(); + + if (!(flags & EXT4_FREE_BLOCKS_VALIDATED) && + !ext4_inode_block_valid(inode, block, count)) { + ext4_error(sb, "Freeing blocks not in datazone - " + "block = %llu, count = %lu", block, count); + return; + } + flags |= EXT4_FREE_BLOCKS_VALIDATED; + + ext4_debug("freeing block %llu\n", block); + trace_ext4_free_blocks(inode, block, count, flags); + + if (bh && (flags & EXT4_FREE_BLOCKS_FORGET)) { + BUG_ON(count > 1); + + ext4_forget(handle, flags & EXT4_FREE_BLOCKS_METADATA, + inode, bh, block); + } + + /* + * If the extent to be freed does not begin on a cluster + * boundary, we need to deal with partial clusters at the + * beginning and end of the extent. Normally we will free + * blocks at the beginning or the end unless we are explicitly + * requested to avoid doing so. + */ + overflow = EXT4_PBLK_COFF(sbi, block); + if (overflow) { + if (flags & EXT4_FREE_BLOCKS_NOFREE_FIRST_CLUSTER) { + overflow = sbi->s_cluster_ratio - overflow; + block += overflow; + if (count > overflow) + count -= overflow; + else + return; + } else { + block -= overflow; + count += overflow; + } + /* The range changed so it's no longer validated */ + flags &= ~EXT4_FREE_BLOCKS_VALIDATED; + } + overflow = EXT4_LBLK_COFF(sbi, count); + if (overflow) { + if (flags & EXT4_FREE_BLOCKS_NOFREE_LAST_CLUSTER) { + if (count > overflow) + count -= overflow; + else + return; + } else + count += sbi->s_cluster_ratio - overflow; + /* The range changed so it's no longer validated */ + flags &= ~EXT4_FREE_BLOCKS_VALIDATED; + } + + if (!bh && (flags & EXT4_FREE_BLOCKS_FORGET)) { + int i; + int is_metadata = flags & EXT4_FREE_BLOCKS_METADATA; + + for (i = 0; i < count; i++) { + cond_resched(); + if (is_metadata) + bh = sb_find_get_block_nonatomic(inode->i_sb, + block + i); + ext4_forget(handle, is_metadata, inode, bh, block + i); + } + } + + ext4_mb_clear_bb(handle, inode, block, count, flags); +} + +/** + * ext4_group_add_blocks() -- Add given blocks to an existing group + * @handle: handle to this transaction + * @sb: super block + * @block: start physical block to add to the block group + * @count: number of blocks to free + * + * This marks the blocks as free in the bitmap and buddy. + */ +int ext4_group_add_blocks(handle_t *handle, struct super_block *sb, + ext4_fsblk_t block, unsigned long count) +{ + ext4_group_t block_group; + ext4_grpblk_t bit; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_buddy e4b; + int err = 0; + ext4_fsblk_t first_cluster = EXT4_B2C(sbi, block); + ext4_fsblk_t last_cluster = EXT4_B2C(sbi, block + count - 1); + unsigned long cluster_count = last_cluster - first_cluster + 1; + ext4_grpblk_t changed; + + ext4_debug("Adding block(s) %llu-%llu\n", block, block + count - 1); + + if (cluster_count == 0) + return 0; + + ext4_get_group_no_and_offset(sb, block, &block_group, &bit); + /* + * Check to see if we are freeing blocks across a group + * boundary. + */ + if (bit + cluster_count > EXT4_CLUSTERS_PER_GROUP(sb)) { + ext4_warning(sb, "too many blocks added to group %u", + block_group); + err = -EINVAL; + goto error_out; + } + + err = ext4_mb_load_buddy(sb, block_group, &e4b); + if (err) + goto error_out; + + if (!ext4_sb_block_valid(sb, NULL, block, count)) { + ext4_error(sb, "Adding blocks in system zones - " + "Block = %llu, count = %lu", + block, count); + err = -EINVAL; + goto error_clean; + } + + err = ext4_mb_mark_context(handle, sb, false, block_group, bit, + cluster_count, EXT4_MB_BITMAP_MARKED_CHECK, + &changed); + if (err && changed == 0) + goto error_clean; + + if (changed != cluster_count) + ext4_error(sb, "bit already cleared in group %u", block_group); + + ext4_lock_group(sb, block_group); + mb_free_blocks(NULL, &e4b, bit, cluster_count); + ext4_unlock_group(sb, block_group); + percpu_counter_add(&sbi->s_freeclusters_counter, + changed); + +error_clean: + ext4_mb_unload_buddy(&e4b); +error_out: + ext4_std_error(sb, err); + return err; +} + +/** + * ext4_trim_extent -- function to TRIM one single free extent in the group + * @sb: super block for the file system + * @start: starting block of the free extent in the alloc. group + * @count: number of blocks to TRIM + * @e4b: ext4 buddy for the group + * + * Trim "count" blocks starting at "start" in the "group". To assure that no + * one will allocate those blocks, mark it as used in buddy bitmap. This must + * be called with under the group lock. + */ +static int ext4_trim_extent(struct super_block *sb, + int start, int count, struct ext4_buddy *e4b) +__releases(bitlock) +__acquires(bitlock) +{ + struct ext4_free_extent ex; + ext4_group_t group = e4b->bd_group; + int ret = 0; + + trace_ext4_trim_extent(sb, group, start, count); + + assert_spin_locked(ext4_group_lock_ptr(sb, group)); + + ex.fe_start = start; + ex.fe_group = group; + ex.fe_len = count; + + /* + * Mark blocks used, so no one can reuse them while + * being trimmed. + */ + mb_mark_used(e4b, &ex); + ext4_unlock_group(sb, group); + ret = ext4_issue_discard(sb, group, start, count); + ext4_lock_group(sb, group); + mb_free_blocks(NULL, e4b, start, ex.fe_len); + return ret; +} + +static ext4_grpblk_t ext4_last_grp_cluster(struct super_block *sb, + ext4_group_t grp) +{ + unsigned long nr_clusters_in_group; + + if (grp < (ext4_get_groups_count(sb) - 1)) + nr_clusters_in_group = EXT4_CLUSTERS_PER_GROUP(sb); + else + nr_clusters_in_group = (ext4_blocks_count(EXT4_SB(sb)->s_es) - + ext4_group_first_block_no(sb, grp)) + >> EXT4_CLUSTER_BITS(sb); + + return nr_clusters_in_group - 1; +} + +static bool ext4_trim_interrupted(void) +{ + return fatal_signal_pending(current) || freezing(current); +} + +static int ext4_try_to_trim_range(struct super_block *sb, + struct ext4_buddy *e4b, ext4_grpblk_t start, + ext4_grpblk_t max, ext4_grpblk_t minblocks) +__acquires(ext4_group_lock_ptr(sb, e4b->bd_group)) +__releases(ext4_group_lock_ptr(sb, e4b->bd_group)) +{ + ext4_grpblk_t next, count, free_count, last, origin_start; + bool set_trimmed = false; + void *bitmap; + + if (unlikely(EXT4_MB_GRP_BBITMAP_CORRUPT(e4b->bd_info))) + return 0; + + last = ext4_last_grp_cluster(sb, e4b->bd_group); + bitmap = e4b->bd_bitmap; + if (start == 0 && max >= last) + set_trimmed = true; + origin_start = start; + start = max(e4b->bd_info->bb_first_free, start); + count = 0; + free_count = 0; + + while (start <= max) { + start = mb_find_next_zero_bit(bitmap, max + 1, start); + if (start > max) + break; + + next = mb_find_next_bit(bitmap, last + 1, start); + if (origin_start == 0 && next >= last) + set_trimmed = true; + + if ((next - start) >= minblocks) { + int ret = ext4_trim_extent(sb, start, next - start, e4b); + + if (ret && ret != -EOPNOTSUPP) + return count; + count += next - start; + } + free_count += next - start; + start = next + 1; + + if (ext4_trim_interrupted()) + return count; + + if (need_resched()) { + ext4_unlock_group(sb, e4b->bd_group); + cond_resched(); + ext4_lock_group(sb, e4b->bd_group); + } + + if ((e4b->bd_info->bb_free - free_count) < minblocks) + break; + } + + if (set_trimmed) + EXT4_MB_GRP_SET_TRIMMED(e4b->bd_info); + + return count; +} + +/** + * ext4_trim_all_free -- function to trim all free space in alloc. group + * @sb: super block for file system + * @group: group to be trimmed + * @start: first group block to examine + * @max: last group block to examine + * @minblocks: minimum extent block count + * + * ext4_trim_all_free walks through group's block bitmap searching for free + * extents. When the free extent is found, mark it as used in group buddy + * bitmap. Then issue a TRIM command on this extent and free the extent in + * the group buddy bitmap. + */ +static ext4_grpblk_t +ext4_trim_all_free(struct super_block *sb, ext4_group_t group, + ext4_grpblk_t start, ext4_grpblk_t max, + ext4_grpblk_t minblocks) +{ + struct ext4_buddy e4b; + int ret; + + trace_ext4_trim_all_free(sb, group, start, max); + + ret = ext4_mb_load_buddy(sb, group, &e4b); + if (ret) { + ext4_warning(sb, "Error %d loading buddy information for %u", + ret, group); + return ret; + } + + ext4_lock_group(sb, group); + + if (!EXT4_MB_GRP_WAS_TRIMMED(e4b.bd_info) || + minblocks < EXT4_SB(sb)->s_last_trim_minblks) + ret = ext4_try_to_trim_range(sb, &e4b, start, max, minblocks); + else + ret = 0; + + ext4_unlock_group(sb, group); + ext4_mb_unload_buddy(&e4b); + + ext4_debug("trimmed %d blocks in the group %d\n", + ret, group); + + return ret; +} + +/** + * ext4_trim_fs() -- trim ioctl handle function + * @sb: superblock for filesystem + * @range: fstrim_range structure + * + * start: First Byte to trim + * len: number of Bytes to trim from start + * minlen: minimum extent length in Bytes + * ext4_trim_fs goes through all allocation groups containing Bytes from + * start to start+len. For each such a group ext4_trim_all_free function + * is invoked to trim all free space. + */ +int ext4_trim_fs(struct super_block *sb, struct fstrim_range *range) +{ + unsigned int discard_granularity = bdev_discard_granularity(sb->s_bdev); + struct ext4_group_info *grp; + ext4_group_t group, first_group, last_group; + ext4_grpblk_t cnt = 0, first_cluster, last_cluster; + uint64_t start, end, minlen, trimmed = 0; + ext4_fsblk_t first_data_blk = + le32_to_cpu(EXT4_SB(sb)->s_es->s_first_data_block); + ext4_fsblk_t max_blks = ext4_blocks_count(EXT4_SB(sb)->s_es); + int ret = 0; + + start = range->start >> sb->s_blocksize_bits; + end = start + (range->len >> sb->s_blocksize_bits) - 1; + minlen = EXT4_NUM_B2C(EXT4_SB(sb), + range->minlen >> sb->s_blocksize_bits); + + if (minlen > EXT4_CLUSTERS_PER_GROUP(sb) || + start >= max_blks || + range->len < sb->s_blocksize) + return -EINVAL; + /* No point to try to trim less than discard granularity */ + if (range->minlen < discard_granularity) { + minlen = EXT4_NUM_B2C(EXT4_SB(sb), + discard_granularity >> sb->s_blocksize_bits); + if (minlen > EXT4_CLUSTERS_PER_GROUP(sb)) + goto out; + } + if (end >= max_blks - 1) + end = max_blks - 1; + if (end <= first_data_blk) + goto out; + if (start < first_data_blk) + start = first_data_blk; + + /* Determine first and last group to examine based on start and end */ + ext4_get_group_no_and_offset(sb, (ext4_fsblk_t) start, + &first_group, &first_cluster); + ext4_get_group_no_and_offset(sb, (ext4_fsblk_t) end, + &last_group, &last_cluster); + + /* end now represents the last cluster to discard in this group */ + end = EXT4_CLUSTERS_PER_GROUP(sb) - 1; + + for (group = first_group; group <= last_group; group++) { + if (ext4_trim_interrupted()) + break; + grp = ext4_get_group_info(sb, group); + if (!grp) + continue; + /* We only do this if the grp has never been initialized */ + if (unlikely(EXT4_MB_GRP_NEED_INIT(grp))) { + ret = ext4_mb_init_group(sb, group, GFP_NOFS); + if (ret) + break; + } + + /* + * For all the groups except the last one, last cluster will + * always be EXT4_CLUSTERS_PER_GROUP(sb)-1, so we only need to + * change it for the last group, note that last_cluster is + * already computed earlier by ext4_get_group_no_and_offset() + */ + if (group == last_group) + end = last_cluster; + if (grp->bb_free >= minlen) { + cnt = ext4_trim_all_free(sb, group, first_cluster, + end, minlen); + if (cnt < 0) { + ret = cnt; + break; + } + trimmed += cnt; + } + + /* + * For every group except the first one, we are sure + * that the first cluster to discard will be cluster #0. + */ + first_cluster = 0; + } + + if (!ret) + EXT4_SB(sb)->s_last_trim_minblks = minlen; + +out: + range->len = EXT4_C2B(EXT4_SB(sb), trimmed) << sb->s_blocksize_bits; + return ret; +} + +/* Iterate all the free extents in the group. */ +int +ext4_mballoc_query_range( + struct super_block *sb, + ext4_group_t group, + ext4_grpblk_t first, + ext4_grpblk_t end, + ext4_mballoc_query_range_fn meta_formatter, + ext4_mballoc_query_range_fn formatter, + void *priv) +{ + void *bitmap; + ext4_grpblk_t start, next; + struct ext4_buddy e4b; + int error; + + error = ext4_mb_load_buddy(sb, group, &e4b); + if (error) + return error; + bitmap = e4b.bd_bitmap; + + ext4_lock_group(sb, group); + + start = max(e4b.bd_info->bb_first_free, first); + if (end >= EXT4_CLUSTERS_PER_GROUP(sb)) + end = EXT4_CLUSTERS_PER_GROUP(sb) - 1; + if (meta_formatter && start != first) { + if (start > end) + start = end; + ext4_unlock_group(sb, group); + error = meta_formatter(sb, group, first, start - first, + priv); + if (error) + goto out_unload; + ext4_lock_group(sb, group); + } + while (start <= end) { + start = mb_find_next_zero_bit(bitmap, end + 1, start); + if (start > end) + break; + next = mb_find_next_bit(bitmap, end + 1, start); + + ext4_unlock_group(sb, group); + error = formatter(sb, group, start, next - start, priv); + if (error) + goto out_unload; + ext4_lock_group(sb, group); + + start = next + 1; + } + + ext4_unlock_group(sb, group); +out_unload: + ext4_mb_unload_buddy(&e4b); + + return error; +} + +#ifdef CONFIG_EXT4_KUNIT_TESTS +#include "mballoc-test.c" +#endif -- 2.43.0

From: Simon Glass <simon.glass@canonical.com> Copy super.c from Linux v6.18 fs/ext4 directory. This file implements superblock operations including mount, unmount, and filesystem options. Co-developed-by: Claude Opus 4.5 <noreply@anthropic.com> --- fs/ext4l/super.c | 7516 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 7516 insertions(+) create mode 100644 fs/ext4l/super.c diff --git a/fs/ext4l/super.c b/fs/ext4l/super.c new file mode 100644 index 00000000000..33e7c08c952 --- /dev/null +++ b/fs/ext4l/super.c @@ -0,0 +1,7516 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * linux/fs/ext4/super.c + * + * Copyright (C) 1992, 1993, 1994, 1995 + * Remy Card (card@masi.ibp.fr) + * Laboratoire MASI - Institut Blaise Pascal + * Universite Pierre et Marie Curie (Paris VI) + * + * from + * + * linux/fs/minix/inode.c + * + * Copyright (C) 1991, 1992 Linus Torvalds + * + * Big-endian to little-endian byte-swapping/bitmaps by + * David S. Miller (davem@caip.rutgers.edu), 1995 + */ + +#include <linux/module.h> +#include <linux/string.h> +#include <linux/fs.h> +#include <linux/time.h> +#include <linux/vmalloc.h> +#include <linux/slab.h> +#include <linux/init.h> +#include <linux/blkdev.h> +#include <linux/backing-dev.h> +#include <linux/parser.h> +#include <linux/buffer_head.h> +#include <linux/exportfs.h> +#include <linux/vfs.h> +#include <linux/random.h> +#include <linux/mount.h> +#include <linux/namei.h> +#include <linux/quotaops.h> +#include <linux/seq_file.h> +#include <linux/ctype.h> +#include <linux/log2.h> +#include <linux/crc16.h> +#include <linux/dax.h> +#include <linux/uaccess.h> +#include <linux/iversion.h> +#include <linux/unicode.h> +#include <linux/part_stat.h> +#include <linux/kthread.h> +#include <linux/freezer.h> +#include <linux/fsnotify.h> +#include <linux/fs_context.h> +#include <linux/fs_parser.h> + +#include "ext4.h" +#include "ext4_extents.h" /* Needed for trace points definition */ +#include "ext4_jbd2.h" +#include "xattr.h" +#include "acl.h" +#include "mballoc.h" +#include "fsmap.h" + +#define CREATE_TRACE_POINTS +#include <trace/events/ext4.h> + +static struct ext4_lazy_init *ext4_li_info; +static DEFINE_MUTEX(ext4_li_mtx); +static struct ratelimit_state ext4_mount_msg_ratelimit; + +static int ext4_load_journal(struct super_block *, struct ext4_super_block *, + unsigned long journal_devnum); +static int ext4_show_options(struct seq_file *seq, struct dentry *root); +static void ext4_update_super(struct super_block *sb); +static int ext4_commit_super(struct super_block *sb); +static int ext4_mark_recovery_complete(struct super_block *sb, + struct ext4_super_block *es); +static int ext4_clear_journal_err(struct super_block *sb, + struct ext4_super_block *es); +static int ext4_sync_fs(struct super_block *sb, int wait); +static int ext4_statfs(struct dentry *dentry, struct kstatfs *buf); +static int ext4_unfreeze(struct super_block *sb); +static int ext4_freeze(struct super_block *sb); +static inline int ext2_feature_set_ok(struct super_block *sb); +static inline int ext3_feature_set_ok(struct super_block *sb); +static void ext4_unregister_li_request(struct super_block *sb); +static void ext4_clear_request_list(void); +static struct inode *ext4_get_journal_inode(struct super_block *sb, + unsigned int journal_inum); +static int ext4_validate_options(struct fs_context *fc); +static int ext4_check_opt_consistency(struct fs_context *fc, + struct super_block *sb); +static void ext4_apply_options(struct fs_context *fc, struct super_block *sb); +static int ext4_parse_param(struct fs_context *fc, struct fs_parameter *param); +static int ext4_get_tree(struct fs_context *fc); +static int ext4_reconfigure(struct fs_context *fc); +static void ext4_fc_free(struct fs_context *fc); +static int ext4_init_fs_context(struct fs_context *fc); +static void ext4_kill_sb(struct super_block *sb); +static const struct fs_parameter_spec ext4_param_specs[]; + +/* + * Lock ordering + * + * page fault path: + * mmap_lock -> sb_start_pagefault -> invalidate_lock (r) -> transaction start + * -> page lock -> i_data_sem (rw) + * + * buffered write path: + * sb_start_write -> i_mutex -> mmap_lock + * sb_start_write -> i_mutex -> transaction start -> page lock -> + * i_data_sem (rw) + * + * truncate: + * sb_start_write -> i_mutex -> invalidate_lock (w) -> i_mmap_rwsem (w) -> + * page lock + * sb_start_write -> i_mutex -> invalidate_lock (w) -> transaction start -> + * i_data_sem (rw) + * + * direct IO: + * sb_start_write -> i_mutex -> mmap_lock + * sb_start_write -> i_mutex -> transaction start -> i_data_sem (rw) + * + * writepages: + * transaction start -> page lock(s) -> i_data_sem (rw) + */ + +static const struct fs_context_operations ext4_context_ops = { + .parse_param = ext4_parse_param, + .get_tree = ext4_get_tree, + .reconfigure = ext4_reconfigure, + .free = ext4_fc_free, +}; + + +#if !defined(CONFIG_EXT2_FS) && !defined(CONFIG_EXT2_FS_MODULE) && defined(CONFIG_EXT4_USE_FOR_EXT2) +static struct file_system_type ext2_fs_type = { + .owner = THIS_MODULE, + .name = "ext2", + .init_fs_context = ext4_init_fs_context, + .parameters = ext4_param_specs, + .kill_sb = ext4_kill_sb, + .fs_flags = FS_REQUIRES_DEV, +}; +MODULE_ALIAS_FS("ext2"); +MODULE_ALIAS("ext2"); +#define IS_EXT2_SB(sb) ((sb)->s_type == &ext2_fs_type) +#else +#define IS_EXT2_SB(sb) (0) +#endif + + +static struct file_system_type ext3_fs_type = { + .owner = THIS_MODULE, + .name = "ext3", + .init_fs_context = ext4_init_fs_context, + .parameters = ext4_param_specs, + .kill_sb = ext4_kill_sb, + .fs_flags = FS_REQUIRES_DEV, +}; +MODULE_ALIAS_FS("ext3"); +MODULE_ALIAS("ext3"); +#define IS_EXT3_SB(sb) ((sb)->s_type == &ext3_fs_type) + + +static inline void __ext4_read_bh(struct buffer_head *bh, blk_opf_t op_flags, + bh_end_io_t *end_io, bool simu_fail) +{ + if (simu_fail) { + clear_buffer_uptodate(bh); + unlock_buffer(bh); + return; + } + + /* + * buffer's verified bit is no longer valid after reading from + * disk again due to write out error, clear it to make sure we + * recheck the buffer contents. + */ + clear_buffer_verified(bh); + + bh->b_end_io = end_io ? end_io : end_buffer_read_sync; + get_bh(bh); + submit_bh(REQ_OP_READ | op_flags, bh); +} + +void ext4_read_bh_nowait(struct buffer_head *bh, blk_opf_t op_flags, + bh_end_io_t *end_io, bool simu_fail) +{ + BUG_ON(!buffer_locked(bh)); + + if (ext4_buffer_uptodate(bh)) { + unlock_buffer(bh); + return; + } + __ext4_read_bh(bh, op_flags, end_io, simu_fail); +} + +int ext4_read_bh(struct buffer_head *bh, blk_opf_t op_flags, + bh_end_io_t *end_io, bool simu_fail) +{ + BUG_ON(!buffer_locked(bh)); + + if (ext4_buffer_uptodate(bh)) { + unlock_buffer(bh); + return 0; + } + + __ext4_read_bh(bh, op_flags, end_io, simu_fail); + + wait_on_buffer(bh); + if (buffer_uptodate(bh)) + return 0; + return -EIO; +} + +int ext4_read_bh_lock(struct buffer_head *bh, blk_opf_t op_flags, bool wait) +{ + lock_buffer(bh); + if (!wait) { + ext4_read_bh_nowait(bh, op_flags, NULL, false); + return 0; + } + return ext4_read_bh(bh, op_flags, NULL, false); +} + +/* + * This works like __bread_gfp() except it uses ERR_PTR for error + * returns. Currently with sb_bread it's impossible to distinguish + * between ENOMEM and EIO situations (since both result in a NULL + * return. + */ +static struct buffer_head *__ext4_sb_bread_gfp(struct super_block *sb, + sector_t block, + blk_opf_t op_flags, gfp_t gfp) +{ + struct buffer_head *bh; + int ret; + + bh = sb_getblk_gfp(sb, block, gfp); + if (bh == NULL) + return ERR_PTR(-ENOMEM); + if (ext4_buffer_uptodate(bh)) + return bh; + + ret = ext4_read_bh_lock(bh, REQ_META | op_flags, true); + if (ret) { + put_bh(bh); + return ERR_PTR(ret); + } + return bh; +} + +struct buffer_head *ext4_sb_bread(struct super_block *sb, sector_t block, + blk_opf_t op_flags) +{ + gfp_t gfp = mapping_gfp_constraint(sb->s_bdev->bd_mapping, + ~__GFP_FS) | __GFP_MOVABLE; + + return __ext4_sb_bread_gfp(sb, block, op_flags, gfp); +} + +struct buffer_head *ext4_sb_bread_unmovable(struct super_block *sb, + sector_t block) +{ + gfp_t gfp = mapping_gfp_constraint(sb->s_bdev->bd_mapping, + ~__GFP_FS); + + return __ext4_sb_bread_gfp(sb, block, 0, gfp); +} + +struct buffer_head *ext4_sb_bread_nofail(struct super_block *sb, + sector_t block) +{ + gfp_t gfp = mapping_gfp_constraint(sb->s_bdev->bd_mapping, + ~__GFP_FS) | __GFP_MOVABLE | __GFP_NOFAIL; + + return __ext4_sb_bread_gfp(sb, block, 0, gfp); +} + +void ext4_sb_breadahead_unmovable(struct super_block *sb, sector_t block) +{ + struct buffer_head *bh = bdev_getblk(sb->s_bdev, block, + sb->s_blocksize, GFP_NOWAIT); + + if (likely(bh)) { + if (trylock_buffer(bh)) + ext4_read_bh_nowait(bh, REQ_RAHEAD, NULL, false); + brelse(bh); + } +} + +static int ext4_verify_csum_type(struct super_block *sb, + struct ext4_super_block *es) +{ + if (!ext4_has_feature_metadata_csum(sb)) + return 1; + + return es->s_checksum_type == EXT4_CRC32C_CHKSUM; +} + +__le32 ext4_superblock_csum(struct ext4_super_block *es) +{ + int offset = offsetof(struct ext4_super_block, s_checksum); + __u32 csum; + + csum = ext4_chksum(~0, (char *)es, offset); + + return cpu_to_le32(csum); +} + +static int ext4_superblock_csum_verify(struct super_block *sb, + struct ext4_super_block *es) +{ + if (!ext4_has_feature_metadata_csum(sb)) + return 1; + + return es->s_checksum == ext4_superblock_csum(es); +} + +void ext4_superblock_csum_set(struct super_block *sb) +{ + struct ext4_super_block *es = EXT4_SB(sb)->s_es; + + if (!ext4_has_feature_metadata_csum(sb)) + return; + + es->s_checksum = ext4_superblock_csum(es); +} + +ext4_fsblk_t ext4_block_bitmap(struct super_block *sb, + struct ext4_group_desc *bg) +{ + return le32_to_cpu(bg->bg_block_bitmap_lo) | + (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT ? + (ext4_fsblk_t)le32_to_cpu(bg->bg_block_bitmap_hi) << 32 : 0); +} + +ext4_fsblk_t ext4_inode_bitmap(struct super_block *sb, + struct ext4_group_desc *bg) +{ + return le32_to_cpu(bg->bg_inode_bitmap_lo) | + (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT ? + (ext4_fsblk_t)le32_to_cpu(bg->bg_inode_bitmap_hi) << 32 : 0); +} + +ext4_fsblk_t ext4_inode_table(struct super_block *sb, + struct ext4_group_desc *bg) +{ + return le32_to_cpu(bg->bg_inode_table_lo) | + (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT ? + (ext4_fsblk_t)le32_to_cpu(bg->bg_inode_table_hi) << 32 : 0); +} + +__u32 ext4_free_group_clusters(struct super_block *sb, + struct ext4_group_desc *bg) +{ + return le16_to_cpu(bg->bg_free_blocks_count_lo) | + (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT ? + (__u32)le16_to_cpu(bg->bg_free_blocks_count_hi) << 16 : 0); +} + +__u32 ext4_free_inodes_count(struct super_block *sb, + struct ext4_group_desc *bg) +{ + return le16_to_cpu(READ_ONCE(bg->bg_free_inodes_count_lo)) | + (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT ? + (__u32)le16_to_cpu(READ_ONCE(bg->bg_free_inodes_count_hi)) << 16 : 0); +} + +__u32 ext4_used_dirs_count(struct super_block *sb, + struct ext4_group_desc *bg) +{ + return le16_to_cpu(bg->bg_used_dirs_count_lo) | + (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT ? + (__u32)le16_to_cpu(bg->bg_used_dirs_count_hi) << 16 : 0); +} + +__u32 ext4_itable_unused_count(struct super_block *sb, + struct ext4_group_desc *bg) +{ + return le16_to_cpu(bg->bg_itable_unused_lo) | + (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT ? + (__u32)le16_to_cpu(bg->bg_itable_unused_hi) << 16 : 0); +} + +void ext4_block_bitmap_set(struct super_block *sb, + struct ext4_group_desc *bg, ext4_fsblk_t blk) +{ + bg->bg_block_bitmap_lo = cpu_to_le32((u32)blk); + if (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT) + bg->bg_block_bitmap_hi = cpu_to_le32(blk >> 32); +} + +void ext4_inode_bitmap_set(struct super_block *sb, + struct ext4_group_desc *bg, ext4_fsblk_t blk) +{ + bg->bg_inode_bitmap_lo = cpu_to_le32((u32)blk); + if (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT) + bg->bg_inode_bitmap_hi = cpu_to_le32(blk >> 32); +} + +void ext4_inode_table_set(struct super_block *sb, + struct ext4_group_desc *bg, ext4_fsblk_t blk) +{ + bg->bg_inode_table_lo = cpu_to_le32((u32)blk); + if (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT) + bg->bg_inode_table_hi = cpu_to_le32(blk >> 32); +} + +void ext4_free_group_clusters_set(struct super_block *sb, + struct ext4_group_desc *bg, __u32 count) +{ + bg->bg_free_blocks_count_lo = cpu_to_le16((__u16)count); + if (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT) + bg->bg_free_blocks_count_hi = cpu_to_le16(count >> 16); +} + +void ext4_free_inodes_set(struct super_block *sb, + struct ext4_group_desc *bg, __u32 count) +{ + WRITE_ONCE(bg->bg_free_inodes_count_lo, cpu_to_le16((__u16)count)); + if (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT) + WRITE_ONCE(bg->bg_free_inodes_count_hi, cpu_to_le16(count >> 16)); +} + +void ext4_used_dirs_set(struct super_block *sb, + struct ext4_group_desc *bg, __u32 count) +{ + bg->bg_used_dirs_count_lo = cpu_to_le16((__u16)count); + if (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT) + bg->bg_used_dirs_count_hi = cpu_to_le16(count >> 16); +} + +void ext4_itable_unused_set(struct super_block *sb, + struct ext4_group_desc *bg, __u32 count) +{ + bg->bg_itable_unused_lo = cpu_to_le16((__u16)count); + if (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT) + bg->bg_itable_unused_hi = cpu_to_le16(count >> 16); +} + +static void __ext4_update_tstamp(__le32 *lo, __u8 *hi, time64_t now) +{ + now = clamp_val(now, 0, (1ull << 40) - 1); + + *lo = cpu_to_le32(lower_32_bits(now)); + *hi = upper_32_bits(now); +} + +static time64_t __ext4_get_tstamp(__le32 *lo, __u8 *hi) +{ + return ((time64_t)(*hi) << 32) + le32_to_cpu(*lo); +} +#define ext4_update_tstamp(es, tstamp) \ + __ext4_update_tstamp(&(es)->tstamp, &(es)->tstamp ## _hi, \ + ktime_get_real_seconds()) +#define ext4_get_tstamp(es, tstamp) \ + __ext4_get_tstamp(&(es)->tstamp, &(es)->tstamp ## _hi) + +/* + * The ext4_maybe_update_superblock() function checks and updates the + * superblock if needed. + * + * This function is designed to update the on-disk superblock only under + * certain conditions to prevent excessive disk writes and unnecessary + * waking of the disk from sleep. The superblock will be updated if: + * 1. More than sbi->s_sb_update_sec (def: 1 hour) has passed since the last + * superblock update + * 2. More than sbi->s_sb_update_kb (def: 16MB) kbs have been written since the + * last superblock update. + * + * @sb: The superblock + */ +static void ext4_maybe_update_superblock(struct super_block *sb) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_super_block *es = sbi->s_es; + journal_t *journal = sbi->s_journal; + time64_t now; + __u64 last_update; + __u64 lifetime_write_kbytes; + __u64 diff_size; + + if (ext4_emergency_state(sb) || sb_rdonly(sb) || + !(sb->s_flags & SB_ACTIVE) || !journal || + journal->j_flags & JBD2_UNMOUNT) + return; + + now = ktime_get_real_seconds(); + last_update = ext4_get_tstamp(es, s_wtime); + + if (likely(now - last_update < sbi->s_sb_update_sec)) + return; + + lifetime_write_kbytes = sbi->s_kbytes_written + + ((part_stat_read(sb->s_bdev, sectors[STAT_WRITE]) - + sbi->s_sectors_written_start) >> 1); + + /* Get the number of kilobytes not written to disk to account + * for statistics and compare with a multiple of 16 MB. This + * is used to determine when the next superblock commit should + * occur (i.e. not more often than once per 16MB if there was + * less written in an hour). + */ + diff_size = lifetime_write_kbytes - le64_to_cpu(es->s_kbytes_written); + + if (diff_size > sbi->s_sb_update_kb) + schedule_work(&EXT4_SB(sb)->s_sb_upd_work); +} + +static void ext4_journal_commit_callback(journal_t *journal, transaction_t *txn) +{ + struct super_block *sb = journal->j_private; + + BUG_ON(txn->t_state == T_FINISHED); + + ext4_process_freed_data(sb, txn->t_tid); + ext4_maybe_update_superblock(sb); +} + +static bool ext4_journalled_writepage_needs_redirty(struct jbd2_inode *jinode, + struct folio *folio) +{ + struct buffer_head *bh, *head; + struct journal_head *jh; + + bh = head = folio_buffers(folio); + do { + /* + * We have to redirty a page in these cases: + * 1) If buffer is dirty, it means the page was dirty because it + * contains a buffer that needs checkpointing. So the dirty bit + * needs to be preserved so that checkpointing writes the buffer + * properly. + * 2) If buffer is not part of the committing transaction + * (we may have just accidentally come across this buffer because + * inode range tracking is not exact) or if the currently running + * transaction already contains this buffer as well, dirty bit + * needs to be preserved so that the buffer gets writeprotected + * properly on running transaction's commit. + */ + jh = bh2jh(bh); + if (buffer_dirty(bh) || + (jh && (jh->b_transaction != jinode->i_transaction || + jh->b_next_transaction))) + return true; + } while ((bh = bh->b_this_page) != head); + + return false; +} + +static int ext4_journalled_submit_inode_data_buffers(struct jbd2_inode *jinode) +{ + struct address_space *mapping = jinode->i_vfs_inode->i_mapping; + struct writeback_control wbc = { + .sync_mode = WB_SYNC_ALL, + .nr_to_write = LONG_MAX, + .range_start = jinode->i_dirty_start, + .range_end = jinode->i_dirty_end, + }; + struct folio *folio = NULL; + int error; + + /* + * writeback_iter() already checks for dirty pages and calls + * folio_clear_dirty_for_io(), which we want to write protect the + * folios. + * + * However, we may have to redirty a folio sometimes. + */ + while ((folio = writeback_iter(mapping, &wbc, folio, &error))) { + if (ext4_journalled_writepage_needs_redirty(jinode, folio)) + folio_redirty_for_writepage(&wbc, folio); + folio_unlock(folio); + } + + return error; +} + +static int ext4_journal_submit_inode_data_buffers(struct jbd2_inode *jinode) +{ + int ret; + + if (ext4_should_journal_data(jinode->i_vfs_inode)) + ret = ext4_journalled_submit_inode_data_buffers(jinode); + else + ret = ext4_normal_submit_inode_data_buffers(jinode); + return ret; +} + +static int ext4_journal_finish_inode_data_buffers(struct jbd2_inode *jinode) +{ + int ret = 0; + + if (!ext4_should_journal_data(jinode->i_vfs_inode)) + ret = jbd2_journal_finish_inode_data_buffers(jinode); + + return ret; +} + +static bool system_going_down(void) +{ + return system_state == SYSTEM_HALT || system_state == SYSTEM_POWER_OFF + || system_state == SYSTEM_RESTART; +} + +struct ext4_err_translation { + int code; + int errno; +}; + +#define EXT4_ERR_TRANSLATE(err) { .code = EXT4_ERR_##err, .errno = err } + +static struct ext4_err_translation err_translation[] = { + EXT4_ERR_TRANSLATE(EIO), + EXT4_ERR_TRANSLATE(ENOMEM), + EXT4_ERR_TRANSLATE(EFSBADCRC), + EXT4_ERR_TRANSLATE(EFSCORRUPTED), + EXT4_ERR_TRANSLATE(ENOSPC), + EXT4_ERR_TRANSLATE(ENOKEY), + EXT4_ERR_TRANSLATE(EROFS), + EXT4_ERR_TRANSLATE(EFBIG), + EXT4_ERR_TRANSLATE(EEXIST), + EXT4_ERR_TRANSLATE(ERANGE), + EXT4_ERR_TRANSLATE(EOVERFLOW), + EXT4_ERR_TRANSLATE(EBUSY), + EXT4_ERR_TRANSLATE(ENOTDIR), + EXT4_ERR_TRANSLATE(ENOTEMPTY), + EXT4_ERR_TRANSLATE(ESHUTDOWN), + EXT4_ERR_TRANSLATE(EFAULT), +}; + +static int ext4_errno_to_code(int errno) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(err_translation); i++) + if (err_translation[i].errno == errno) + return err_translation[i].code; + return EXT4_ERR_UNKNOWN; +} + +static void save_error_info(struct super_block *sb, int error, + __u32 ino, __u64 block, + const char *func, unsigned int line) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + + /* We default to EFSCORRUPTED error... */ + if (error == 0) + error = EFSCORRUPTED; + + spin_lock(&sbi->s_error_lock); + sbi->s_add_error_count++; + sbi->s_last_error_code = error; + sbi->s_last_error_line = line; + sbi->s_last_error_ino = ino; + sbi->s_last_error_block = block; + sbi->s_last_error_func = func; + sbi->s_last_error_time = ktime_get_real_seconds(); + if (!sbi->s_first_error_time) { + sbi->s_first_error_code = error; + sbi->s_first_error_line = line; + sbi->s_first_error_ino = ino; + sbi->s_first_error_block = block; + sbi->s_first_error_func = func; + sbi->s_first_error_time = sbi->s_last_error_time; + } + spin_unlock(&sbi->s_error_lock); +} + +/* Deal with the reporting of failure conditions on a filesystem such as + * inconsistencies detected or read IO failures. + * + * On ext2, we can store the error state of the filesystem in the + * superblock. That is not possible on ext4, because we may have other + * write ordering constraints on the superblock which prevent us from + * writing it out straight away; and given that the journal is about to + * be aborted, we can't rely on the current, or future, transactions to + * write out the superblock safely. + * + * We'll just use the jbd2_journal_abort() error code to record an error in + * the journal instead. On recovery, the journal will complain about + * that error until we've noted it down and cleared it. + * + * If force_ro is set, we unconditionally force the filesystem into an + * ABORT|READONLY state, unless the error response on the fs has been set to + * panic in which case we take the easy way out and panic immediately. This is + * used to deal with unrecoverable failures such as journal IO errors or ENOMEM + * at a critical moment in log management. + */ +static void ext4_handle_error(struct super_block *sb, bool force_ro, int error, + __u32 ino, __u64 block, + const char *func, unsigned int line) +{ + journal_t *journal = EXT4_SB(sb)->s_journal; + bool continue_fs = !force_ro && test_opt(sb, ERRORS_CONT); + + EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS; + if (test_opt(sb, WARN_ON_ERROR)) + WARN_ON_ONCE(1); + + if (!continue_fs && !ext4_emergency_ro(sb) && journal) + jbd2_journal_abort(journal, -EIO); + + if (!bdev_read_only(sb->s_bdev)) { + save_error_info(sb, error, ino, block, func, line); + /* + * In case the fs should keep running, we need to writeout + * superblock through the journal. Due to lock ordering + * constraints, it may not be safe to do it right here so we + * defer superblock flushing to a workqueue. We just need to be + * careful when the journal is already shutting down. If we get + * here in that case, just update the sb directly as the last + * transaction won't commit anyway. + */ + if (continue_fs && journal && + !ext4_test_mount_flag(sb, EXT4_MF_JOURNAL_DESTROY)) + schedule_work(&EXT4_SB(sb)->s_sb_upd_work); + else + ext4_commit_super(sb); + } + + /* + * We force ERRORS_RO behavior when system is rebooting. Otherwise we + * could panic during 'reboot -f' as the underlying device got already + * disabled. + */ + if (test_opt(sb, ERRORS_PANIC) && !system_going_down()) { + panic("EXT4-fs (device %s): panic forced after error\n", + sb->s_id); + } + + if (ext4_emergency_ro(sb) || continue_fs) + return; + + ext4_msg(sb, KERN_CRIT, "Remounting filesystem read-only"); + /* + * We don't set SB_RDONLY because that requires sb->s_umount + * semaphore and setting it without proper remount procedure is + * confusing code such as freeze_super() leading to deadlocks + * and other problems. + */ + set_bit(EXT4_FLAGS_EMERGENCY_RO, &EXT4_SB(sb)->s_ext4_flags); +} + +static void update_super_work(struct work_struct *work) +{ + struct ext4_sb_info *sbi = container_of(work, struct ext4_sb_info, + s_sb_upd_work); + journal_t *journal = sbi->s_journal; + handle_t *handle; + + /* + * If the journal is still running, we have to write out superblock + * through the journal to avoid collisions of other journalled sb + * updates. + * + * We use directly jbd2 functions here to avoid recursing back into + * ext4 error handling code during handling of previous errors. + */ + if (!ext4_emergency_state(sbi->s_sb) && + !sb_rdonly(sbi->s_sb) && journal) { + struct buffer_head *sbh = sbi->s_sbh; + bool call_notify_err = false; + + handle = jbd2_journal_start(journal, 1); + if (IS_ERR(handle)) + goto write_directly; + if (jbd2_journal_get_write_access(handle, sbh)) { + jbd2_journal_stop(handle); + goto write_directly; + } + + if (sbi->s_add_error_count > 0) + call_notify_err = true; + + ext4_update_super(sbi->s_sb); + if (buffer_write_io_error(sbh) || !buffer_uptodate(sbh)) { + ext4_msg(sbi->s_sb, KERN_ERR, "previous I/O error to " + "superblock detected"); + clear_buffer_write_io_error(sbh); + set_buffer_uptodate(sbh); + } + + if (jbd2_journal_dirty_metadata(handle, sbh)) { + jbd2_journal_stop(handle); + goto write_directly; + } + jbd2_journal_stop(handle); + + if (call_notify_err) + ext4_notify_error_sysfs(sbi); + + return; + } +write_directly: + /* + * Write through journal failed. Write sb directly to get error info + * out and hope for the best. + */ + ext4_commit_super(sbi->s_sb); + ext4_notify_error_sysfs(sbi); +} + +#define ext4_error_ratelimit(sb) \ + ___ratelimit(&(EXT4_SB(sb)->s_err_ratelimit_state), \ + "EXT4-fs error") + +void __ext4_error(struct super_block *sb, const char *function, + unsigned int line, bool force_ro, int error, __u64 block, + const char *fmt, ...) +{ + struct va_format vaf; + va_list args; + + if (unlikely(ext4_emergency_state(sb))) + return; + + trace_ext4_error(sb, function, line); + if (ext4_error_ratelimit(sb)) { + va_start(args, fmt); + vaf.fmt = fmt; + vaf.va = &args; + printk(KERN_CRIT + "EXT4-fs error (device %s): %s:%d: comm %s: %pV\n", + sb->s_id, function, line, current->comm, &vaf); + va_end(args); + } + fsnotify_sb_error(sb, NULL, error ? error : EFSCORRUPTED); + + ext4_handle_error(sb, force_ro, error, 0, block, function, line); +} + +void __ext4_error_inode(struct inode *inode, const char *function, + unsigned int line, ext4_fsblk_t block, int error, + const char *fmt, ...) +{ + va_list args; + struct va_format vaf; + + if (unlikely(ext4_emergency_state(inode->i_sb))) + return; + + trace_ext4_error(inode->i_sb, function, line); + if (ext4_error_ratelimit(inode->i_sb)) { + va_start(args, fmt); + vaf.fmt = fmt; + vaf.va = &args; + if (block) + printk(KERN_CRIT "EXT4-fs error (device %s): %s:%d: " + "inode #%lu: block %llu: comm %s: %pV\n", + inode->i_sb->s_id, function, line, inode->i_ino, + block, current->comm, &vaf); + else + printk(KERN_CRIT "EXT4-fs error (device %s): %s:%d: " + "inode #%lu: comm %s: %pV\n", + inode->i_sb->s_id, function, line, inode->i_ino, + current->comm, &vaf); + va_end(args); + } + fsnotify_sb_error(inode->i_sb, inode, error ? error : EFSCORRUPTED); + + ext4_handle_error(inode->i_sb, false, error, inode->i_ino, block, + function, line); +} + +void __ext4_error_file(struct file *file, const char *function, + unsigned int line, ext4_fsblk_t block, + const char *fmt, ...) +{ + va_list args; + struct va_format vaf; + struct inode *inode = file_inode(file); + char pathname[80], *path; + + if (unlikely(ext4_emergency_state(inode->i_sb))) + return; + + trace_ext4_error(inode->i_sb, function, line); + if (ext4_error_ratelimit(inode->i_sb)) { + path = file_path(file, pathname, sizeof(pathname)); + if (IS_ERR(path)) + path = "(unknown)"; + va_start(args, fmt); + vaf.fmt = fmt; + vaf.va = &args; + if (block) + printk(KERN_CRIT + "EXT4-fs error (device %s): %s:%d: inode #%lu: " + "block %llu: comm %s: path %s: %pV\n", + inode->i_sb->s_id, function, line, inode->i_ino, + block, current->comm, path, &vaf); + else + printk(KERN_CRIT + "EXT4-fs error (device %s): %s:%d: inode #%lu: " + "comm %s: path %s: %pV\n", + inode->i_sb->s_id, function, line, inode->i_ino, + current->comm, path, &vaf); + va_end(args); + } + fsnotify_sb_error(inode->i_sb, inode, EFSCORRUPTED); + + ext4_handle_error(inode->i_sb, false, EFSCORRUPTED, inode->i_ino, block, + function, line); +} + +const char *ext4_decode_error(struct super_block *sb, int errno, + char nbuf[16]) +{ + char *errstr = NULL; + + switch (errno) { + case -EFSCORRUPTED: + errstr = "Corrupt filesystem"; + break; + case -EFSBADCRC: + errstr = "Filesystem failed CRC"; + break; + case -EIO: + errstr = "IO failure"; + break; + case -ENOMEM: + errstr = "Out of memory"; + break; + case -EROFS: + if (!sb || (EXT4_SB(sb)->s_journal && + EXT4_SB(sb)->s_journal->j_flags & JBD2_ABORT)) + errstr = "Journal has aborted"; + else + errstr = "Readonly filesystem"; + break; + default: + /* If the caller passed in an extra buffer for unknown + * errors, textualise them now. Else we just return + * NULL. */ + if (nbuf) { + /* Check for truncated error codes... */ + if (snprintf(nbuf, 16, "error %d", -errno) >= 0) + errstr = nbuf; + } + break; + } + + return errstr; +} + +/* __ext4_std_error decodes expected errors from journaling functions + * automatically and invokes the appropriate error response. */ + +void __ext4_std_error(struct super_block *sb, const char *function, + unsigned int line, int errno) +{ + char nbuf[16]; + const char *errstr; + + if (unlikely(ext4_emergency_state(sb))) + return; + + /* Special case: if the error is EROFS, and we're not already + * inside a transaction, then there's really no point in logging + * an error. */ + if (errno == -EROFS && journal_current_handle() == NULL && sb_rdonly(sb)) + return; + + if (ext4_error_ratelimit(sb)) { + errstr = ext4_decode_error(sb, errno, nbuf); + printk(KERN_CRIT "EXT4-fs error (device %s) in %s:%d: %s\n", + sb->s_id, function, line, errstr); + } + fsnotify_sb_error(sb, NULL, errno ? errno : EFSCORRUPTED); + + ext4_handle_error(sb, false, -errno, 0, 0, function, line); +} + +void __ext4_msg(struct super_block *sb, + const char *prefix, const char *fmt, ...) +{ + struct va_format vaf; + va_list args; + + if (sb) { + atomic_inc(&EXT4_SB(sb)->s_msg_count); + if (!___ratelimit(&(EXT4_SB(sb)->s_msg_ratelimit_state), + "EXT4-fs")) + return; + } + + va_start(args, fmt); + vaf.fmt = fmt; + vaf.va = &args; + if (sb) + printk("%sEXT4-fs (%s): %pV\n", prefix, sb->s_id, &vaf); + else + printk("%sEXT4-fs: %pV\n", prefix, &vaf); + va_end(args); +} + +static int ext4_warning_ratelimit(struct super_block *sb) +{ + atomic_inc(&EXT4_SB(sb)->s_warning_count); + return ___ratelimit(&(EXT4_SB(sb)->s_warning_ratelimit_state), + "EXT4-fs warning"); +} + +void __ext4_warning(struct super_block *sb, const char *function, + unsigned int line, const char *fmt, ...) +{ + struct va_format vaf; + va_list args; + + if (!ext4_warning_ratelimit(sb)) + return; + + va_start(args, fmt); + vaf.fmt = fmt; + vaf.va = &args; + printk(KERN_WARNING "EXT4-fs warning (device %s): %s:%d: %pV\n", + sb->s_id, function, line, &vaf); + va_end(args); +} + +void __ext4_warning_inode(const struct inode *inode, const char *function, + unsigned int line, const char *fmt, ...) +{ + struct va_format vaf; + va_list args; + + if (!ext4_warning_ratelimit(inode->i_sb)) + return; + + va_start(args, fmt); + vaf.fmt = fmt; + vaf.va = &args; + printk(KERN_WARNING "EXT4-fs warning (device %s): %s:%d: " + "inode #%lu: comm %s: %pV\n", inode->i_sb->s_id, + function, line, inode->i_ino, current->comm, &vaf); + va_end(args); +} + +void __ext4_grp_locked_error(const char *function, unsigned int line, + struct super_block *sb, ext4_group_t grp, + unsigned long ino, ext4_fsblk_t block, + const char *fmt, ...) +__releases(bitlock) +__acquires(bitlock) +{ + struct va_format vaf; + va_list args; + + if (unlikely(ext4_emergency_state(sb))) + return; + + trace_ext4_error(sb, function, line); + if (ext4_error_ratelimit(sb)) { + va_start(args, fmt); + vaf.fmt = fmt; + vaf.va = &args; + printk(KERN_CRIT "EXT4-fs error (device %s): %s:%d: group %u, ", + sb->s_id, function, line, grp); + if (ino) + printk(KERN_CONT "inode %lu: ", ino); + if (block) + printk(KERN_CONT "block %llu:", + (unsigned long long) block); + printk(KERN_CONT "%pV\n", &vaf); + va_end(args); + } + + if (test_opt(sb, ERRORS_CONT)) { + if (test_opt(sb, WARN_ON_ERROR)) + WARN_ON_ONCE(1); + EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS; + if (!bdev_read_only(sb->s_bdev)) { + save_error_info(sb, EFSCORRUPTED, ino, block, function, + line); + schedule_work(&EXT4_SB(sb)->s_sb_upd_work); + } + return; + } + ext4_unlock_group(sb, grp); + ext4_handle_error(sb, false, EFSCORRUPTED, ino, block, function, line); + /* + * We only get here in the ERRORS_RO case; relocking the group + * may be dangerous, but nothing bad will happen since the + * filesystem will have already been marked read/only and the + * journal has been aborted. We return 1 as a hint to callers + * who might what to use the return value from + * ext4_grp_locked_error() to distinguish between the + * ERRORS_CONT and ERRORS_RO case, and perhaps return more + * aggressively from the ext4 function in question, with a + * more appropriate error code. + */ + ext4_lock_group(sb, grp); + return; +} + +void ext4_mark_group_bitmap_corrupted(struct super_block *sb, + ext4_group_t group, + unsigned int flags) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_group_info *grp = ext4_get_group_info(sb, group); + struct ext4_group_desc *gdp = ext4_get_group_desc(sb, group, NULL); + int ret; + + if (!grp || !gdp) + return; + if (flags & EXT4_GROUP_INFO_BBITMAP_CORRUPT) { + ret = ext4_test_and_set_bit(EXT4_GROUP_INFO_BBITMAP_CORRUPT_BIT, + &grp->bb_state); + if (!ret) + percpu_counter_sub(&sbi->s_freeclusters_counter, + grp->bb_free); + } + + if (flags & EXT4_GROUP_INFO_IBITMAP_CORRUPT) { + ret = ext4_test_and_set_bit(EXT4_GROUP_INFO_IBITMAP_CORRUPT_BIT, + &grp->bb_state); + if (!ret && gdp) { + int count; + + count = ext4_free_inodes_count(sb, gdp); + percpu_counter_sub(&sbi->s_freeinodes_counter, + count); + } + } +} + +void ext4_update_dynamic_rev(struct super_block *sb) +{ + struct ext4_super_block *es = EXT4_SB(sb)->s_es; + + if (le32_to_cpu(es->s_rev_level) > EXT4_GOOD_OLD_REV) + return; + + ext4_warning(sb, + "updating to rev %d because of new feature flag, " + "running e2fsck is recommended", + EXT4_DYNAMIC_REV); + + es->s_first_ino = cpu_to_le32(EXT4_GOOD_OLD_FIRST_INO); + es->s_inode_size = cpu_to_le16(EXT4_GOOD_OLD_INODE_SIZE); + es->s_rev_level = cpu_to_le32(EXT4_DYNAMIC_REV); + /* leave es->s_feature_*compat flags alone */ + /* es->s_uuid will be set by e2fsck if empty */ + + /* + * The rest of the superblock fields should be zero, and if not it + * means they are likely already in use, so leave them alone. We + * can leave it up to e2fsck to clean up any inconsistencies there. + */ +} + +static inline struct inode *orphan_list_entry(struct list_head *l) +{ + return &list_entry(l, struct ext4_inode_info, i_orphan)->vfs_inode; +} + +static void dump_orphan_list(struct super_block *sb, struct ext4_sb_info *sbi) +{ + struct list_head *l; + + ext4_msg(sb, KERN_ERR, "sb orphan head is %d", + le32_to_cpu(sbi->s_es->s_last_orphan)); + + printk(KERN_ERR "sb_info orphan list:\n"); + list_for_each(l, &sbi->s_orphan) { + struct inode *inode = orphan_list_entry(l); + printk(KERN_ERR " " + "inode %s:%lu at %p: mode %o, nlink %d, next %d\n", + inode->i_sb->s_id, inode->i_ino, inode, + inode->i_mode, inode->i_nlink, + NEXT_ORPHAN(inode)); + } +} + +#ifdef CONFIG_QUOTA +static int ext4_quota_off(struct super_block *sb, int type); + +static inline void ext4_quotas_off(struct super_block *sb, int type) +{ + BUG_ON(type > EXT4_MAXQUOTAS); + + /* Use our quota_off function to clear inode flags etc. */ + for (type--; type >= 0; type--) + ext4_quota_off(sb, type); +} + +/* + * This is a helper function which is used in the mount/remount + * codepaths (which holds s_umount) to fetch the quota file name. + */ +static inline char *get_qf_name(struct super_block *sb, + struct ext4_sb_info *sbi, + int type) +{ + return rcu_dereference_protected(sbi->s_qf_names[type], + lockdep_is_held(&sb->s_umount)); +} +#else +static inline void ext4_quotas_off(struct super_block *sb, int type) +{ +} +#endif + +static int ext4_percpu_param_init(struct ext4_sb_info *sbi) +{ + ext4_fsblk_t block; + int err; + + block = ext4_count_free_clusters(sbi->s_sb); + ext4_free_blocks_count_set(sbi->s_es, EXT4_C2B(sbi, block)); + err = percpu_counter_init(&sbi->s_freeclusters_counter, block, + GFP_KERNEL); + if (!err) { + unsigned long freei = ext4_count_free_inodes(sbi->s_sb); + sbi->s_es->s_free_inodes_count = cpu_to_le32(freei); + err = percpu_counter_init(&sbi->s_freeinodes_counter, freei, + GFP_KERNEL); + } + if (!err) + err = percpu_counter_init(&sbi->s_dirs_counter, + ext4_count_dirs(sbi->s_sb), GFP_KERNEL); + if (!err) + err = percpu_counter_init(&sbi->s_dirtyclusters_counter, 0, + GFP_KERNEL); + if (!err) + err = percpu_counter_init(&sbi->s_sra_exceeded_retry_limit, 0, + GFP_KERNEL); + if (!err) + err = percpu_init_rwsem(&sbi->s_writepages_rwsem); + + if (err) + ext4_msg(sbi->s_sb, KERN_ERR, "insufficient memory"); + + return err; +} + +static void ext4_percpu_param_destroy(struct ext4_sb_info *sbi) +{ + percpu_counter_destroy(&sbi->s_freeclusters_counter); + percpu_counter_destroy(&sbi->s_freeinodes_counter); + percpu_counter_destroy(&sbi->s_dirs_counter); + percpu_counter_destroy(&sbi->s_dirtyclusters_counter); + percpu_counter_destroy(&sbi->s_sra_exceeded_retry_limit); + percpu_free_rwsem(&sbi->s_writepages_rwsem); +} + +static void ext4_group_desc_free(struct ext4_sb_info *sbi) +{ + struct buffer_head **group_desc; + int i; + + rcu_read_lock(); + group_desc = rcu_dereference(sbi->s_group_desc); + for (i = 0; i < sbi->s_gdb_count; i++) + brelse(group_desc[i]); + kvfree(group_desc); + rcu_read_unlock(); +} + +static void ext4_flex_groups_free(struct ext4_sb_info *sbi) +{ + struct flex_groups **flex_groups; + int i; + + rcu_read_lock(); + flex_groups = rcu_dereference(sbi->s_flex_groups); + if (flex_groups) { + for (i = 0; i < sbi->s_flex_groups_allocated; i++) + kvfree(flex_groups[i]); + kvfree(flex_groups); + } + rcu_read_unlock(); +} + +static void ext4_put_super(struct super_block *sb) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_super_block *es = sbi->s_es; + int aborted = 0; + int err; + + /* + * Unregister sysfs before destroying jbd2 journal. + * Since we could still access attr_journal_task attribute via sysfs + * path which could have sbi->s_journal->j_task as NULL + * Unregister sysfs before flush sbi->s_sb_upd_work. + * Since user may read /proc/fs/ext4/xx/mb_groups during umount, If + * read metadata verify failed then will queue error work. + * update_super_work will call start_this_handle may trigger + * BUG_ON. + */ + ext4_unregister_sysfs(sb); + + if (___ratelimit(&ext4_mount_msg_ratelimit, "EXT4-fs unmount")) + ext4_msg(sb, KERN_INFO, "unmounting filesystem %pU.", + &sb->s_uuid); + + ext4_unregister_li_request(sb); + ext4_quotas_off(sb, EXT4_MAXQUOTAS); + + destroy_workqueue(sbi->rsv_conversion_wq); + ext4_release_orphan_info(sb); + + if (sbi->s_journal) { + aborted = is_journal_aborted(sbi->s_journal); + err = ext4_journal_destroy(sbi, sbi->s_journal); + if ((err < 0) && !aborted) { + ext4_abort(sb, -err, "Couldn't clean up the journal"); + } + } else + flush_work(&sbi->s_sb_upd_work); + + ext4_es_unregister_shrinker(sbi); + timer_shutdown_sync(&sbi->s_err_report); + ext4_release_system_zone(sb); + ext4_mb_release(sb); + ext4_ext_release(sb); + + if (!ext4_emergency_state(sb) && !sb_rdonly(sb)) { + if (!aborted) { + ext4_clear_feature_journal_needs_recovery(sb); + ext4_clear_feature_orphan_present(sb); + es->s_state = cpu_to_le16(sbi->s_mount_state); + } + ext4_commit_super(sb); + } + + ext4_group_desc_free(sbi); + ext4_flex_groups_free(sbi); + + WARN_ON_ONCE(!(sbi->s_mount_state & EXT4_ERROR_FS) && + percpu_counter_sum(&sbi->s_dirtyclusters_counter)); + ext4_percpu_param_destroy(sbi); +#ifdef CONFIG_QUOTA + for (int i = 0; i < EXT4_MAXQUOTAS; i++) + kfree(get_qf_name(sb, sbi, i)); +#endif + + /* Debugging code just in case the in-memory inode orphan list + * isn't empty. The on-disk one can be non-empty if we've + * detected an error and taken the fs readonly, but the + * in-memory list had better be clean by this point. */ + if (!list_empty(&sbi->s_orphan)) + dump_orphan_list(sb, sbi); + ASSERT(list_empty(&sbi->s_orphan)); + + sync_blockdev(sb->s_bdev); + invalidate_bdev(sb->s_bdev); + if (sbi->s_journal_bdev_file) { + /* + * Invalidate the journal device's buffers. We don't want them + * floating about in memory - the physical journal device may + * hotswapped, and it breaks the `ro-after' testing code. + */ + sync_blockdev(file_bdev(sbi->s_journal_bdev_file)); + invalidate_bdev(file_bdev(sbi->s_journal_bdev_file)); + } + + ext4_xattr_destroy_cache(sbi->s_ea_inode_cache); + sbi->s_ea_inode_cache = NULL; + + ext4_xattr_destroy_cache(sbi->s_ea_block_cache); + sbi->s_ea_block_cache = NULL; + + ext4_stop_mmpd(sbi); + + brelse(sbi->s_sbh); + sb->s_fs_info = NULL; + /* + * Now that we are completely done shutting down the + * superblock, we need to actually destroy the kobject. + */ + kobject_put(&sbi->s_kobj); + wait_for_completion(&sbi->s_kobj_unregister); + kfree(sbi->s_blockgroup_lock); + fs_put_dax(sbi->s_daxdev, NULL); + fscrypt_free_dummy_policy(&sbi->s_dummy_enc_policy); +#if IS_ENABLED(CONFIG_UNICODE) + utf8_unload(sb->s_encoding); +#endif + kfree(sbi); +} + +static struct kmem_cache *ext4_inode_cachep; + +/* + * Called inside transaction, so use GFP_NOFS + */ +static struct inode *ext4_alloc_inode(struct super_block *sb) +{ + struct ext4_inode_info *ei; + + ei = alloc_inode_sb(sb, ext4_inode_cachep, GFP_NOFS); + if (!ei) + return NULL; + + inode_set_iversion(&ei->vfs_inode, 1); + ei->i_flags = 0; + spin_lock_init(&ei->i_raw_lock); + ei->i_prealloc_node = RB_ROOT; + atomic_set(&ei->i_prealloc_active, 0); + rwlock_init(&ei->i_prealloc_lock); + ext4_es_init_tree(&ei->i_es_tree); + rwlock_init(&ei->i_es_lock); + INIT_LIST_HEAD(&ei->i_es_list); + ei->i_es_all_nr = 0; + ei->i_es_shk_nr = 0; + ei->i_es_shrink_lblk = 0; + ei->i_reserved_data_blocks = 0; + spin_lock_init(&(ei->i_block_reservation_lock)); + ext4_init_pending_tree(&ei->i_pending_tree); +#ifdef CONFIG_QUOTA + ei->i_reserved_quota = 0; + memset(&ei->i_dquot, 0, sizeof(ei->i_dquot)); +#endif + ei->jinode = NULL; + INIT_LIST_HEAD(&ei->i_rsv_conversion_list); + spin_lock_init(&ei->i_completed_io_lock); + ei->i_sync_tid = 0; + ei->i_datasync_tid = 0; + INIT_WORK(&ei->i_rsv_conversion_work, ext4_end_io_rsv_work); + ext4_fc_init_inode(&ei->vfs_inode); + spin_lock_init(&ei->i_fc_lock); + return &ei->vfs_inode; +} + +static int ext4_drop_inode(struct inode *inode) +{ + int drop = inode_generic_drop(inode); + + if (!drop) + drop = fscrypt_drop_inode(inode); + + trace_ext4_drop_inode(inode, drop); + return drop; +} + +static void ext4_free_in_core_inode(struct inode *inode) +{ + fscrypt_free_inode(inode); + if (!list_empty(&(EXT4_I(inode)->i_fc_list))) { + pr_warn("%s: inode %ld still in fc list", + __func__, inode->i_ino); + } + kmem_cache_free(ext4_inode_cachep, EXT4_I(inode)); +} + +static void ext4_destroy_inode(struct inode *inode) +{ + if (ext4_inode_orphan_tracked(inode)) { + ext4_msg(inode->i_sb, KERN_ERR, + "Inode %lu (%p): inode tracked as orphan!", + inode->i_ino, EXT4_I(inode)); + print_hex_dump(KERN_INFO, "", DUMP_PREFIX_ADDRESS, 16, 4, + EXT4_I(inode), sizeof(struct ext4_inode_info), + true); + dump_stack(); + } + + if (!(EXT4_SB(inode->i_sb)->s_mount_state & EXT4_ERROR_FS) && + WARN_ON_ONCE(EXT4_I(inode)->i_reserved_data_blocks)) + ext4_msg(inode->i_sb, KERN_ERR, + "Inode %lu (%p): i_reserved_data_blocks (%u) not cleared!", + inode->i_ino, EXT4_I(inode), + EXT4_I(inode)->i_reserved_data_blocks); +} + +static void ext4_shutdown(struct super_block *sb) +{ + ext4_force_shutdown(sb, EXT4_GOING_FLAGS_NOLOGFLUSH); +} + +static void init_once(void *foo) +{ + struct ext4_inode_info *ei = foo; + + INIT_LIST_HEAD(&ei->i_orphan); + init_rwsem(&ei->xattr_sem); + init_rwsem(&ei->i_data_sem); + inode_init_once(&ei->vfs_inode); + ext4_fc_init_inode(&ei->vfs_inode); +#ifdef CONFIG_FS_ENCRYPTION + ei->i_crypt_info = NULL; +#endif +#ifdef CONFIG_FS_VERITY + ei->i_verity_info = NULL; +#endif +} + +static int __init init_inodecache(void) +{ + ext4_inode_cachep = kmem_cache_create_usercopy("ext4_inode_cache", + sizeof(struct ext4_inode_info), 0, + SLAB_RECLAIM_ACCOUNT | SLAB_ACCOUNT, + offsetof(struct ext4_inode_info, i_data), + sizeof_field(struct ext4_inode_info, i_data), + init_once); + if (ext4_inode_cachep == NULL) + return -ENOMEM; + return 0; +} + +static void destroy_inodecache(void) +{ + /* + * Make sure all delayed rcu free inodes are flushed before we + * destroy cache. + */ + rcu_barrier(); + kmem_cache_destroy(ext4_inode_cachep); +} + +void ext4_clear_inode(struct inode *inode) +{ + ext4_fc_del(inode); + invalidate_inode_buffers(inode); + clear_inode(inode); + ext4_discard_preallocations(inode); + ext4_es_remove_extent(inode, 0, EXT_MAX_BLOCKS); + dquot_drop(inode); + if (EXT4_I(inode)->jinode) { + jbd2_journal_release_jbd_inode(EXT4_JOURNAL(inode), + EXT4_I(inode)->jinode); + jbd2_free_inode(EXT4_I(inode)->jinode); + EXT4_I(inode)->jinode = NULL; + } + fscrypt_put_encryption_info(inode); + fsverity_cleanup_inode(inode); +} + +static struct inode *ext4_nfs_get_inode(struct super_block *sb, + u64 ino, u32 generation) +{ + struct inode *inode; + + /* + * Currently we don't know the generation for parent directory, so + * a generation of 0 means "accept any" + */ + inode = ext4_iget(sb, ino, EXT4_IGET_HANDLE); + if (IS_ERR(inode)) + return ERR_CAST(inode); + if (generation && inode->i_generation != generation) { + iput(inode); + return ERR_PTR(-ESTALE); + } + + return inode; +} + +static struct dentry *ext4_fh_to_dentry(struct super_block *sb, struct fid *fid, + int fh_len, int fh_type) +{ + return generic_fh_to_dentry(sb, fid, fh_len, fh_type, + ext4_nfs_get_inode); +} + +static struct dentry *ext4_fh_to_parent(struct super_block *sb, struct fid *fid, + int fh_len, int fh_type) +{ + return generic_fh_to_parent(sb, fid, fh_len, fh_type, + ext4_nfs_get_inode); +} + +static int ext4_nfs_commit_metadata(struct inode *inode) +{ + struct writeback_control wbc = { + .sync_mode = WB_SYNC_ALL + }; + + trace_ext4_nfs_commit_metadata(inode); + return ext4_write_inode(inode, &wbc); +} + +#ifdef CONFIG_QUOTA +static const char * const quotatypes[] = INITQFNAMES; +#define QTYPE2NAME(t) (quotatypes[t]) + +static int ext4_write_dquot(struct dquot *dquot); +static int ext4_acquire_dquot(struct dquot *dquot); +static int ext4_release_dquot(struct dquot *dquot); +static int ext4_mark_dquot_dirty(struct dquot *dquot); +static int ext4_write_info(struct super_block *sb, int type); +static int ext4_quota_on(struct super_block *sb, int type, int format_id, + const struct path *path); +static ssize_t ext4_quota_read(struct super_block *sb, int type, char *data, + size_t len, loff_t off); +static ssize_t ext4_quota_write(struct super_block *sb, int type, + const char *data, size_t len, loff_t off); +static int ext4_quota_enable(struct super_block *sb, int type, int format_id, + unsigned int flags); + +static struct dquot __rcu **ext4_get_dquots(struct inode *inode) +{ + return EXT4_I(inode)->i_dquot; +} + +static const struct dquot_operations ext4_quota_operations = { + .get_reserved_space = ext4_get_reserved_space, + .write_dquot = ext4_write_dquot, + .acquire_dquot = ext4_acquire_dquot, + .release_dquot = ext4_release_dquot, + .mark_dirty = ext4_mark_dquot_dirty, + .write_info = ext4_write_info, + .alloc_dquot = dquot_alloc, + .destroy_dquot = dquot_destroy, + .get_projid = ext4_get_projid, + .get_inode_usage = ext4_get_inode_usage, + .get_next_id = dquot_get_next_id, +}; + +static const struct quotactl_ops ext4_qctl_operations = { + .quota_on = ext4_quota_on, + .quota_off = ext4_quota_off, + .quota_sync = dquot_quota_sync, + .get_state = dquot_get_state, + .set_info = dquot_set_dqinfo, + .get_dqblk = dquot_get_dqblk, + .set_dqblk = dquot_set_dqblk, + .get_nextdqblk = dquot_get_next_dqblk, +}; +#endif + +static const struct super_operations ext4_sops = { + .alloc_inode = ext4_alloc_inode, + .free_inode = ext4_free_in_core_inode, + .destroy_inode = ext4_destroy_inode, + .write_inode = ext4_write_inode, + .dirty_inode = ext4_dirty_inode, + .drop_inode = ext4_drop_inode, + .evict_inode = ext4_evict_inode, + .put_super = ext4_put_super, + .sync_fs = ext4_sync_fs, + .freeze_fs = ext4_freeze, + .unfreeze_fs = ext4_unfreeze, + .statfs = ext4_statfs, + .show_options = ext4_show_options, + .shutdown = ext4_shutdown, +#ifdef CONFIG_QUOTA + .quota_read = ext4_quota_read, + .quota_write = ext4_quota_write, + .get_dquots = ext4_get_dquots, +#endif +}; + +static const struct export_operations ext4_export_ops = { + .encode_fh = generic_encode_ino32_fh, + .fh_to_dentry = ext4_fh_to_dentry, + .fh_to_parent = ext4_fh_to_parent, + .get_parent = ext4_get_parent, + .commit_metadata = ext4_nfs_commit_metadata, +}; + +enum { + Opt_bsd_df, Opt_minix_df, Opt_grpid, Opt_nogrpid, + Opt_resgid, Opt_resuid, Opt_sb, + Opt_nouid32, Opt_debug, Opt_removed, + Opt_user_xattr, Opt_acl, + Opt_auto_da_alloc, Opt_noauto_da_alloc, Opt_noload, + Opt_commit, Opt_min_batch_time, Opt_max_batch_time, Opt_journal_dev, + Opt_journal_path, Opt_journal_checksum, Opt_journal_async_commit, + Opt_abort, Opt_data_journal, Opt_data_ordered, Opt_data_writeback, + Opt_data_err_abort, Opt_data_err_ignore, Opt_test_dummy_encryption, + Opt_inlinecrypt, + Opt_usrjquota, Opt_grpjquota, Opt_quota, + Opt_noquota, Opt_barrier, Opt_nobarrier, Opt_err, + Opt_usrquota, Opt_grpquota, Opt_prjquota, + Opt_dax, Opt_dax_always, Opt_dax_inode, Opt_dax_never, + Opt_stripe, Opt_delalloc, Opt_nodelalloc, Opt_warn_on_error, + Opt_nowarn_on_error, Opt_mblk_io_submit, Opt_debug_want_extra_isize, + Opt_nomblk_io_submit, Opt_block_validity, Opt_noblock_validity, + Opt_inode_readahead_blks, Opt_journal_ioprio, + Opt_dioread_nolock, Opt_dioread_lock, + Opt_discard, Opt_nodiscard, Opt_init_itable, Opt_noinit_itable, + Opt_max_dir_size_kb, Opt_nojournal_checksum, Opt_nombcache, + Opt_no_prefetch_block_bitmaps, Opt_mb_optimize_scan, + Opt_errors, Opt_data, Opt_data_err, Opt_jqfmt, Opt_dax_type, +#ifdef CONFIG_EXT4_DEBUG + Opt_fc_debug_max_replay, Opt_fc_debug_force +#endif +}; + +static const struct constant_table ext4_param_errors[] = { + {"continue", EXT4_MOUNT_ERRORS_CONT}, + {"panic", EXT4_MOUNT_ERRORS_PANIC}, + {"remount-ro", EXT4_MOUNT_ERRORS_RO}, + {} +}; + +static const struct constant_table ext4_param_data[] = { + {"journal", EXT4_MOUNT_JOURNAL_DATA}, + {"ordered", EXT4_MOUNT_ORDERED_DATA}, + {"writeback", EXT4_MOUNT_WRITEBACK_DATA}, + {} +}; + +static const struct constant_table ext4_param_data_err[] = { + {"abort", Opt_data_err_abort}, + {"ignore", Opt_data_err_ignore}, + {} +}; + +static const struct constant_table ext4_param_jqfmt[] = { + {"vfsold", QFMT_VFS_OLD}, + {"vfsv0", QFMT_VFS_V0}, + {"vfsv1", QFMT_VFS_V1}, + {} +}; + +static const struct constant_table ext4_param_dax[] = { + {"always", Opt_dax_always}, + {"inode", Opt_dax_inode}, + {"never", Opt_dax_never}, + {} +}; + +/* + * Mount option specification + * We don't use fsparam_flag_no because of the way we set the + * options and the way we show them in _ext4_show_options(). To + * keep the changes to a minimum, let's keep the negative options + * separate for now. + */ +static const struct fs_parameter_spec ext4_param_specs[] = { + fsparam_flag ("bsddf", Opt_bsd_df), + fsparam_flag ("minixdf", Opt_minix_df), + fsparam_flag ("grpid", Opt_grpid), + fsparam_flag ("bsdgroups", Opt_grpid), + fsparam_flag ("nogrpid", Opt_nogrpid), + fsparam_flag ("sysvgroups", Opt_nogrpid), + fsparam_gid ("resgid", Opt_resgid), + fsparam_uid ("resuid", Opt_resuid), + fsparam_u32 ("sb", Opt_sb), + fsparam_enum ("errors", Opt_errors, ext4_param_errors), + fsparam_flag ("nouid32", Opt_nouid32), + fsparam_flag ("debug", Opt_debug), + fsparam_flag ("oldalloc", Opt_removed), + fsparam_flag ("orlov", Opt_removed), + fsparam_flag ("user_xattr", Opt_user_xattr), + fsparam_flag ("acl", Opt_acl), + fsparam_flag ("norecovery", Opt_noload), + fsparam_flag ("noload", Opt_noload), + fsparam_flag ("bh", Opt_removed), + fsparam_flag ("nobh", Opt_removed), + fsparam_u32 ("commit", Opt_commit), + fsparam_u32 ("min_batch_time", Opt_min_batch_time), + fsparam_u32 ("max_batch_time", Opt_max_batch_time), + fsparam_u32 ("journal_dev", Opt_journal_dev), + fsparam_bdev ("journal_path", Opt_journal_path), + fsparam_flag ("journal_checksum", Opt_journal_checksum), + fsparam_flag ("nojournal_checksum", Opt_nojournal_checksum), + fsparam_flag ("journal_async_commit",Opt_journal_async_commit), + fsparam_flag ("abort", Opt_abort), + fsparam_enum ("data", Opt_data, ext4_param_data), + fsparam_enum ("data_err", Opt_data_err, + ext4_param_data_err), + fsparam_string_empty + ("usrjquota", Opt_usrjquota), + fsparam_string_empty + ("grpjquota", Opt_grpjquota), + fsparam_enum ("jqfmt", Opt_jqfmt, ext4_param_jqfmt), + fsparam_flag ("grpquota", Opt_grpquota), + fsparam_flag ("quota", Opt_quota), + fsparam_flag ("noquota", Opt_noquota), + fsparam_flag ("usrquota", Opt_usrquota), + fsparam_flag ("prjquota", Opt_prjquota), + fsparam_flag ("barrier", Opt_barrier), + fsparam_u32 ("barrier", Opt_barrier), + fsparam_flag ("nobarrier", Opt_nobarrier), + fsparam_flag ("i_version", Opt_removed), + fsparam_flag ("dax", Opt_dax), + fsparam_enum ("dax", Opt_dax_type, ext4_param_dax), + fsparam_u32 ("stripe", Opt_stripe), + fsparam_flag ("delalloc", Opt_delalloc), + fsparam_flag ("nodelalloc", Opt_nodelalloc), + fsparam_flag ("warn_on_error", Opt_warn_on_error), + fsparam_flag ("nowarn_on_error", Opt_nowarn_on_error), + fsparam_u32 ("debug_want_extra_isize", + Opt_debug_want_extra_isize), + fsparam_flag ("mblk_io_submit", Opt_removed), + fsparam_flag ("nomblk_io_submit", Opt_removed), + fsparam_flag ("block_validity", Opt_block_validity), + fsparam_flag ("noblock_validity", Opt_noblock_validity), + fsparam_u32 ("inode_readahead_blks", + Opt_inode_readahead_blks), + fsparam_u32 ("journal_ioprio", Opt_journal_ioprio), + fsparam_u32 ("auto_da_alloc", Opt_auto_da_alloc), + fsparam_flag ("auto_da_alloc", Opt_auto_da_alloc), + fsparam_flag ("noauto_da_alloc", Opt_noauto_da_alloc), + fsparam_flag ("dioread_nolock", Opt_dioread_nolock), + fsparam_flag ("nodioread_nolock", Opt_dioread_lock), + fsparam_flag ("dioread_lock", Opt_dioread_lock), + fsparam_flag ("discard", Opt_discard), + fsparam_flag ("nodiscard", Opt_nodiscard), + fsparam_u32 ("init_itable", Opt_init_itable), + fsparam_flag ("init_itable", Opt_init_itable), + fsparam_flag ("noinit_itable", Opt_noinit_itable), +#ifdef CONFIG_EXT4_DEBUG + fsparam_flag ("fc_debug_force", Opt_fc_debug_force), + fsparam_u32 ("fc_debug_max_replay", Opt_fc_debug_max_replay), +#endif + fsparam_u32 ("max_dir_size_kb", Opt_max_dir_size_kb), + fsparam_flag ("test_dummy_encryption", + Opt_test_dummy_encryption), + fsparam_string ("test_dummy_encryption", + Opt_test_dummy_encryption), + fsparam_flag ("inlinecrypt", Opt_inlinecrypt), + fsparam_flag ("nombcache", Opt_nombcache), + fsparam_flag ("no_mbcache", Opt_nombcache), /* for backward compatibility */ + fsparam_flag ("prefetch_block_bitmaps", + Opt_removed), + fsparam_flag ("no_prefetch_block_bitmaps", + Opt_no_prefetch_block_bitmaps), + fsparam_s32 ("mb_optimize_scan", Opt_mb_optimize_scan), + fsparam_string ("check", Opt_removed), /* mount option from ext2/3 */ + fsparam_flag ("nocheck", Opt_removed), /* mount option from ext2/3 */ + fsparam_flag ("reservation", Opt_removed), /* mount option from ext2/3 */ + fsparam_flag ("noreservation", Opt_removed), /* mount option from ext2/3 */ + fsparam_u32 ("journal", Opt_removed), /* mount option from ext2/3 */ + {} +}; + + +#define MOPT_SET 0x0001 +#define MOPT_CLEAR 0x0002 +#define MOPT_NOSUPPORT 0x0004 +#define MOPT_EXPLICIT 0x0008 +#ifdef CONFIG_QUOTA +#define MOPT_Q 0 +#define MOPT_QFMT 0x0010 +#else +#define MOPT_Q MOPT_NOSUPPORT +#define MOPT_QFMT MOPT_NOSUPPORT +#endif +#define MOPT_NO_EXT2 0x0020 +#define MOPT_NO_EXT3 0x0040 +#define MOPT_EXT4_ONLY (MOPT_NO_EXT2 | MOPT_NO_EXT3) +#define MOPT_SKIP 0x0080 +#define MOPT_2 0x0100 + +static const struct mount_opts { + int token; + int mount_opt; + int flags; +} ext4_mount_opts[] = { + {Opt_minix_df, EXT4_MOUNT_MINIX_DF, MOPT_SET}, + {Opt_bsd_df, EXT4_MOUNT_MINIX_DF, MOPT_CLEAR}, + {Opt_grpid, EXT4_MOUNT_GRPID, MOPT_SET}, + {Opt_nogrpid, EXT4_MOUNT_GRPID, MOPT_CLEAR}, + {Opt_block_validity, EXT4_MOUNT_BLOCK_VALIDITY, MOPT_SET}, + {Opt_noblock_validity, EXT4_MOUNT_BLOCK_VALIDITY, MOPT_CLEAR}, + {Opt_dioread_nolock, EXT4_MOUNT_DIOREAD_NOLOCK, + MOPT_EXT4_ONLY | MOPT_SET}, + {Opt_dioread_lock, EXT4_MOUNT_DIOREAD_NOLOCK, + MOPT_EXT4_ONLY | MOPT_CLEAR}, + {Opt_discard, EXT4_MOUNT_DISCARD, MOPT_SET}, + {Opt_nodiscard, EXT4_MOUNT_DISCARD, MOPT_CLEAR}, + {Opt_delalloc, EXT4_MOUNT_DELALLOC, + MOPT_EXT4_ONLY | MOPT_SET | MOPT_EXPLICIT}, + {Opt_nodelalloc, EXT4_MOUNT_DELALLOC, + MOPT_EXT4_ONLY | MOPT_CLEAR}, + {Opt_warn_on_error, EXT4_MOUNT_WARN_ON_ERROR, MOPT_SET}, + {Opt_nowarn_on_error, EXT4_MOUNT_WARN_ON_ERROR, MOPT_CLEAR}, + {Opt_commit, 0, MOPT_NO_EXT2}, + {Opt_nojournal_checksum, EXT4_MOUNT_JOURNAL_CHECKSUM, + MOPT_EXT4_ONLY | MOPT_CLEAR}, + {Opt_journal_checksum, EXT4_MOUNT_JOURNAL_CHECKSUM, + MOPT_EXT4_ONLY | MOPT_SET | MOPT_EXPLICIT}, + {Opt_journal_async_commit, (EXT4_MOUNT_JOURNAL_ASYNC_COMMIT | + EXT4_MOUNT_JOURNAL_CHECKSUM), + MOPT_EXT4_ONLY | MOPT_SET | MOPT_EXPLICIT}, + {Opt_noload, EXT4_MOUNT_NOLOAD, MOPT_NO_EXT2 | MOPT_SET}, + {Opt_data_err, EXT4_MOUNT_DATA_ERR_ABORT, MOPT_NO_EXT2}, + {Opt_barrier, EXT4_MOUNT_BARRIER, MOPT_SET}, + {Opt_nobarrier, EXT4_MOUNT_BARRIER, MOPT_CLEAR}, + {Opt_noauto_da_alloc, EXT4_MOUNT_NO_AUTO_DA_ALLOC, MOPT_SET}, + {Opt_auto_da_alloc, EXT4_MOUNT_NO_AUTO_DA_ALLOC, MOPT_CLEAR}, + {Opt_noinit_itable, EXT4_MOUNT_INIT_INODE_TABLE, MOPT_CLEAR}, + {Opt_dax_type, 0, MOPT_EXT4_ONLY}, + {Opt_journal_dev, 0, MOPT_NO_EXT2}, + {Opt_journal_path, 0, MOPT_NO_EXT2}, + {Opt_journal_ioprio, 0, MOPT_NO_EXT2}, + {Opt_data, 0, MOPT_NO_EXT2}, + {Opt_user_xattr, EXT4_MOUNT_XATTR_USER, MOPT_SET}, +#ifdef CONFIG_EXT4_FS_POSIX_ACL + {Opt_acl, EXT4_MOUNT_POSIX_ACL, MOPT_SET}, +#else + {Opt_acl, 0, MOPT_NOSUPPORT}, +#endif + {Opt_nouid32, EXT4_MOUNT_NO_UID32, MOPT_SET}, + {Opt_debug, EXT4_MOUNT_DEBUG, MOPT_SET}, + {Opt_quota, EXT4_MOUNT_QUOTA | EXT4_MOUNT_USRQUOTA, MOPT_SET | MOPT_Q}, + {Opt_usrquota, EXT4_MOUNT_QUOTA | EXT4_MOUNT_USRQUOTA, + MOPT_SET | MOPT_Q}, + {Opt_grpquota, EXT4_MOUNT_QUOTA | EXT4_MOUNT_GRPQUOTA, + MOPT_SET | MOPT_Q}, + {Opt_prjquota, EXT4_MOUNT_QUOTA | EXT4_MOUNT_PRJQUOTA, + MOPT_SET | MOPT_Q}, + {Opt_noquota, (EXT4_MOUNT_QUOTA | EXT4_MOUNT_USRQUOTA | + EXT4_MOUNT_GRPQUOTA | EXT4_MOUNT_PRJQUOTA), + MOPT_CLEAR | MOPT_Q}, + {Opt_usrjquota, 0, MOPT_Q}, + {Opt_grpjquota, 0, MOPT_Q}, + {Opt_jqfmt, 0, MOPT_QFMT}, + {Opt_nombcache, EXT4_MOUNT_NO_MBCACHE, MOPT_SET}, + {Opt_no_prefetch_block_bitmaps, EXT4_MOUNT_NO_PREFETCH_BLOCK_BITMAPS, + MOPT_SET}, +#ifdef CONFIG_EXT4_DEBUG + {Opt_fc_debug_force, EXT4_MOUNT2_JOURNAL_FAST_COMMIT, + MOPT_SET | MOPT_2 | MOPT_EXT4_ONLY}, +#endif + {Opt_abort, EXT4_MOUNT2_ABORT, MOPT_SET | MOPT_2}, + {Opt_err, 0, 0} +}; + +#if IS_ENABLED(CONFIG_UNICODE) +static const struct ext4_sb_encodings { + __u16 magic; + char *name; + unsigned int version; +} ext4_sb_encoding_map[] = { + {EXT4_ENC_UTF8_12_1, "utf8", UNICODE_AGE(12, 1, 0)}, +}; + +static const struct ext4_sb_encodings * +ext4_sb_read_encoding(const struct ext4_super_block *es) +{ + __u16 magic = le16_to_cpu(es->s_encoding); + int i; + + for (i = 0; i < ARRAY_SIZE(ext4_sb_encoding_map); i++) + if (magic == ext4_sb_encoding_map[i].magic) + return &ext4_sb_encoding_map[i]; + + return NULL; +} +#endif + +#define EXT4_SPEC_JQUOTA (1 << 0) +#define EXT4_SPEC_JQFMT (1 << 1) +#define EXT4_SPEC_DATAJ (1 << 2) +#define EXT4_SPEC_SB_BLOCK (1 << 3) +#define EXT4_SPEC_JOURNAL_DEV (1 << 4) +#define EXT4_SPEC_JOURNAL_IOPRIO (1 << 5) +#define EXT4_SPEC_s_want_extra_isize (1 << 7) +#define EXT4_SPEC_s_max_batch_time (1 << 8) +#define EXT4_SPEC_s_min_batch_time (1 << 9) +#define EXT4_SPEC_s_inode_readahead_blks (1 << 10) +#define EXT4_SPEC_s_li_wait_mult (1 << 11) +#define EXT4_SPEC_s_max_dir_size_kb (1 << 12) +#define EXT4_SPEC_s_stripe (1 << 13) +#define EXT4_SPEC_s_resuid (1 << 14) +#define EXT4_SPEC_s_resgid (1 << 15) +#define EXT4_SPEC_s_commit_interval (1 << 16) +#define EXT4_SPEC_s_fc_debug_max_replay (1 << 17) +#define EXT4_SPEC_s_sb_block (1 << 18) +#define EXT4_SPEC_mb_optimize_scan (1 << 19) + +struct ext4_fs_context { + char *s_qf_names[EXT4_MAXQUOTAS]; + struct fscrypt_dummy_policy dummy_enc_policy; + int s_jquota_fmt; /* Format of quota to use */ +#ifdef CONFIG_EXT4_DEBUG + int s_fc_debug_max_replay; +#endif + unsigned short qname_spec; + unsigned long vals_s_flags; /* Bits to set in s_flags */ + unsigned long mask_s_flags; /* Bits changed in s_flags */ + unsigned long journal_devnum; + unsigned long s_commit_interval; + unsigned long s_stripe; + unsigned int s_inode_readahead_blks; + unsigned int s_want_extra_isize; + unsigned int s_li_wait_mult; + unsigned int s_max_dir_size_kb; + unsigned int journal_ioprio; + unsigned int vals_s_mount_opt; + unsigned int mask_s_mount_opt; + unsigned int vals_s_mount_opt2; + unsigned int mask_s_mount_opt2; + unsigned int opt_flags; /* MOPT flags */ + unsigned int spec; + u32 s_max_batch_time; + u32 s_min_batch_time; + kuid_t s_resuid; + kgid_t s_resgid; + ext4_fsblk_t s_sb_block; +}; + +static void ext4_fc_free(struct fs_context *fc) +{ + struct ext4_fs_context *ctx = fc->fs_private; + int i; + + if (!ctx) + return; + + for (i = 0; i < EXT4_MAXQUOTAS; i++) + kfree(ctx->s_qf_names[i]); + + fscrypt_free_dummy_policy(&ctx->dummy_enc_policy); + kfree(ctx); +} + +int ext4_init_fs_context(struct fs_context *fc) +{ + struct ext4_fs_context *ctx; + + ctx = kzalloc(sizeof(struct ext4_fs_context), GFP_KERNEL); + if (!ctx) + return -ENOMEM; + + fc->fs_private = ctx; + fc->ops = &ext4_context_ops; + + /* i_version is always enabled now */ + fc->sb_flags |= SB_I_VERSION; + + return 0; +} + +#ifdef CONFIG_QUOTA +/* + * Note the name of the specified quota file. + */ +static int note_qf_name(struct fs_context *fc, int qtype, + struct fs_parameter *param) +{ + struct ext4_fs_context *ctx = fc->fs_private; + char *qname; + + if (param->size < 1) { + ext4_msg(NULL, KERN_ERR, "Missing quota name"); + return -EINVAL; + } + if (strchr(param->string, '/')) { + ext4_msg(NULL, KERN_ERR, + "quotafile must be on filesystem root"); + return -EINVAL; + } + if (ctx->s_qf_names[qtype]) { + if (strcmp(ctx->s_qf_names[qtype], param->string) != 0) { + ext4_msg(NULL, KERN_ERR, + "%s quota file already specified", + QTYPE2NAME(qtype)); + return -EINVAL; + } + return 0; + } + + qname = kmemdup_nul(param->string, param->size, GFP_KERNEL); + if (!qname) { + ext4_msg(NULL, KERN_ERR, + "Not enough memory for storing quotafile name"); + return -ENOMEM; + } + ctx->s_qf_names[qtype] = qname; + ctx->qname_spec |= 1 << qtype; + ctx->spec |= EXT4_SPEC_JQUOTA; + return 0; +} + +/* + * Clear the name of the specified quota file. + */ +static int unnote_qf_name(struct fs_context *fc, int qtype) +{ + struct ext4_fs_context *ctx = fc->fs_private; + + kfree(ctx->s_qf_names[qtype]); + + ctx->s_qf_names[qtype] = NULL; + ctx->qname_spec |= 1 << qtype; + ctx->spec |= EXT4_SPEC_JQUOTA; + return 0; +} +#endif + +static int ext4_parse_test_dummy_encryption(const struct fs_parameter *param, + struct ext4_fs_context *ctx) +{ + int err; + + if (!IS_ENABLED(CONFIG_FS_ENCRYPTION)) { + ext4_msg(NULL, KERN_WARNING, + "test_dummy_encryption option not supported"); + return -EINVAL; + } + err = fscrypt_parse_test_dummy_encryption(param, + &ctx->dummy_enc_policy); + if (err == -EINVAL) { + ext4_msg(NULL, KERN_WARNING, + "Value of option \"%s\" is unrecognized", param->key); + } else if (err == -EEXIST) { + ext4_msg(NULL, KERN_WARNING, + "Conflicting test_dummy_encryption options"); + return -EINVAL; + } + return err; +} + +#define EXT4_SET_CTX(name) \ +static inline __maybe_unused \ +void ctx_set_##name(struct ext4_fs_context *ctx, unsigned long flag) \ +{ \ + ctx->mask_s_##name |= flag; \ + ctx->vals_s_##name |= flag; \ +} + +#define EXT4_CLEAR_CTX(name) \ +static inline __maybe_unused \ +void ctx_clear_##name(struct ext4_fs_context *ctx, unsigned long flag) \ +{ \ + ctx->mask_s_##name |= flag; \ + ctx->vals_s_##name &= ~flag; \ +} + +#define EXT4_TEST_CTX(name) \ +static inline unsigned long \ +ctx_test_##name(struct ext4_fs_context *ctx, unsigned long flag) \ +{ \ + return (ctx->vals_s_##name & flag); \ +} + +EXT4_SET_CTX(flags); /* set only */ +EXT4_SET_CTX(mount_opt); +EXT4_CLEAR_CTX(mount_opt); +EXT4_TEST_CTX(mount_opt); +EXT4_SET_CTX(mount_opt2); +EXT4_CLEAR_CTX(mount_opt2); +EXT4_TEST_CTX(mount_opt2); + +static int ext4_parse_param(struct fs_context *fc, struct fs_parameter *param) +{ + struct ext4_fs_context *ctx = fc->fs_private; + struct fs_parse_result result; + const struct mount_opts *m; + int is_remount; + int token; + + token = fs_parse(fc, ext4_param_specs, param, &result); + if (token < 0) + return token; + is_remount = fc->purpose == FS_CONTEXT_FOR_RECONFIGURE; + + for (m = ext4_mount_opts; m->token != Opt_err; m++) + if (token == m->token) + break; + + ctx->opt_flags |= m->flags; + + if (m->flags & MOPT_EXPLICIT) { + if (m->mount_opt & EXT4_MOUNT_DELALLOC) { + ctx_set_mount_opt2(ctx, EXT4_MOUNT2_EXPLICIT_DELALLOC); + } else if (m->mount_opt & EXT4_MOUNT_JOURNAL_CHECKSUM) { + ctx_set_mount_opt2(ctx, + EXT4_MOUNT2_EXPLICIT_JOURNAL_CHECKSUM); + } else + return -EINVAL; + } + + if (m->flags & MOPT_NOSUPPORT) { + ext4_msg(NULL, KERN_ERR, "%s option not supported", + param->key); + return 0; + } + + switch (token) { +#ifdef CONFIG_QUOTA + case Opt_usrjquota: + if (!*param->string) + return unnote_qf_name(fc, USRQUOTA); + else + return note_qf_name(fc, USRQUOTA, param); + case Opt_grpjquota: + if (!*param->string) + return unnote_qf_name(fc, GRPQUOTA); + else + return note_qf_name(fc, GRPQUOTA, param); +#endif + case Opt_sb: + if (fc->purpose == FS_CONTEXT_FOR_RECONFIGURE) { + ext4_msg(NULL, KERN_WARNING, + "Ignoring %s option on remount", param->key); + } else { + ctx->s_sb_block = result.uint_32; + ctx->spec |= EXT4_SPEC_s_sb_block; + } + return 0; + case Opt_removed: + ext4_msg(NULL, KERN_WARNING, "Ignoring removed %s option", + param->key); + return 0; + case Opt_inlinecrypt: +#ifdef CONFIG_FS_ENCRYPTION_INLINE_CRYPT + ctx_set_flags(ctx, SB_INLINECRYPT); +#else + ext4_msg(NULL, KERN_ERR, "inline encryption not supported"); +#endif + return 0; + case Opt_errors: + ctx_clear_mount_opt(ctx, EXT4_MOUNT_ERRORS_MASK); + ctx_set_mount_opt(ctx, result.uint_32); + return 0; +#ifdef CONFIG_QUOTA + case Opt_jqfmt: + ctx->s_jquota_fmt = result.uint_32; + ctx->spec |= EXT4_SPEC_JQFMT; + return 0; +#endif + case Opt_data: + ctx_clear_mount_opt(ctx, EXT4_MOUNT_DATA_FLAGS); + ctx_set_mount_opt(ctx, result.uint_32); + ctx->spec |= EXT4_SPEC_DATAJ; + return 0; + case Opt_commit: + if (result.uint_32 == 0) + result.uint_32 = JBD2_DEFAULT_MAX_COMMIT_AGE; + else if (result.uint_32 > INT_MAX / HZ) { + ext4_msg(NULL, KERN_ERR, + "Invalid commit interval %d, " + "must be smaller than %d", + result.uint_32, INT_MAX / HZ); + return -EINVAL; + } + ctx->s_commit_interval = HZ * result.uint_32; + ctx->spec |= EXT4_SPEC_s_commit_interval; + return 0; + case Opt_debug_want_extra_isize: + if ((result.uint_32 & 1) || (result.uint_32 < 4)) { + ext4_msg(NULL, KERN_ERR, + "Invalid want_extra_isize %d", result.uint_32); + return -EINVAL; + } + ctx->s_want_extra_isize = result.uint_32; + ctx->spec |= EXT4_SPEC_s_want_extra_isize; + return 0; + case Opt_max_batch_time: + ctx->s_max_batch_time = result.uint_32; + ctx->spec |= EXT4_SPEC_s_max_batch_time; + return 0; + case Opt_min_batch_time: + ctx->s_min_batch_time = result.uint_32; + ctx->spec |= EXT4_SPEC_s_min_batch_time; + return 0; + case Opt_inode_readahead_blks: + if (result.uint_32 && + (result.uint_32 > (1 << 30) || + !is_power_of_2(result.uint_32))) { + ext4_msg(NULL, KERN_ERR, + "EXT4-fs: inode_readahead_blks must be " + "0 or a power of 2 smaller than 2^31"); + return -EINVAL; + } + ctx->s_inode_readahead_blks = result.uint_32; + ctx->spec |= EXT4_SPEC_s_inode_readahead_blks; + return 0; + case Opt_init_itable: + ctx_set_mount_opt(ctx, EXT4_MOUNT_INIT_INODE_TABLE); + ctx->s_li_wait_mult = EXT4_DEF_LI_WAIT_MULT; + if (param->type == fs_value_is_string) + ctx->s_li_wait_mult = result.uint_32; + ctx->spec |= EXT4_SPEC_s_li_wait_mult; + return 0; + case Opt_max_dir_size_kb: + ctx->s_max_dir_size_kb = result.uint_32; + ctx->spec |= EXT4_SPEC_s_max_dir_size_kb; + return 0; +#ifdef CONFIG_EXT4_DEBUG + case Opt_fc_debug_max_replay: + ctx->s_fc_debug_max_replay = result.uint_32; + ctx->spec |= EXT4_SPEC_s_fc_debug_max_replay; + return 0; +#endif + case Opt_stripe: + ctx->s_stripe = result.uint_32; + ctx->spec |= EXT4_SPEC_s_stripe; + return 0; + case Opt_resuid: + ctx->s_resuid = result.uid; + ctx->spec |= EXT4_SPEC_s_resuid; + return 0; + case Opt_resgid: + ctx->s_resgid = result.gid; + ctx->spec |= EXT4_SPEC_s_resgid; + return 0; + case Opt_journal_dev: + if (is_remount) { + ext4_msg(NULL, KERN_ERR, + "Cannot specify journal on remount"); + return -EINVAL; + } + ctx->journal_devnum = result.uint_32; + ctx->spec |= EXT4_SPEC_JOURNAL_DEV; + return 0; + case Opt_journal_path: + { + struct inode *journal_inode; + struct path path; + int error; + + if (is_remount) { + ext4_msg(NULL, KERN_ERR, + "Cannot specify journal on remount"); + return -EINVAL; + } + + error = fs_lookup_param(fc, param, 1, LOOKUP_FOLLOW, &path); + if (error) { + ext4_msg(NULL, KERN_ERR, "error: could not find " + "journal device path"); + return -EINVAL; + } + + journal_inode = d_inode(path.dentry); + ctx->journal_devnum = new_encode_dev(journal_inode->i_rdev); + ctx->spec |= EXT4_SPEC_JOURNAL_DEV; + path_put(&path); + return 0; + } + case Opt_journal_ioprio: + if (result.uint_32 > 7) { + ext4_msg(NULL, KERN_ERR, "Invalid journal IO priority" + " (must be 0-7)"); + return -EINVAL; + } + ctx->journal_ioprio = + IOPRIO_PRIO_VALUE(IOPRIO_CLASS_BE, result.uint_32); + ctx->spec |= EXT4_SPEC_JOURNAL_IOPRIO; + return 0; + case Opt_test_dummy_encryption: + return ext4_parse_test_dummy_encryption(param, ctx); + case Opt_dax: + case Opt_dax_type: +#ifdef CONFIG_FS_DAX + { + int type = (token == Opt_dax) ? + Opt_dax : result.uint_32; + + switch (type) { + case Opt_dax: + case Opt_dax_always: + ctx_set_mount_opt(ctx, EXT4_MOUNT_DAX_ALWAYS); + ctx_clear_mount_opt2(ctx, EXT4_MOUNT2_DAX_NEVER); + break; + case Opt_dax_never: + ctx_set_mount_opt2(ctx, EXT4_MOUNT2_DAX_NEVER); + ctx_clear_mount_opt(ctx, EXT4_MOUNT_DAX_ALWAYS); + break; + case Opt_dax_inode: + ctx_clear_mount_opt(ctx, EXT4_MOUNT_DAX_ALWAYS); + ctx_clear_mount_opt2(ctx, EXT4_MOUNT2_DAX_NEVER); + /* Strictly for printing options */ + ctx_set_mount_opt2(ctx, EXT4_MOUNT2_DAX_INODE); + break; + } + return 0; + } +#else + ext4_msg(NULL, KERN_INFO, "dax option not supported"); + return -EINVAL; +#endif + case Opt_data_err: + if (result.uint_32 == Opt_data_err_abort) + ctx_set_mount_opt(ctx, m->mount_opt); + else if (result.uint_32 == Opt_data_err_ignore) + ctx_clear_mount_opt(ctx, m->mount_opt); + return 0; + case Opt_mb_optimize_scan: + if (result.int_32 == 1) { + ctx_set_mount_opt2(ctx, EXT4_MOUNT2_MB_OPTIMIZE_SCAN); + ctx->spec |= EXT4_SPEC_mb_optimize_scan; + } else if (result.int_32 == 0) { + ctx_clear_mount_opt2(ctx, EXT4_MOUNT2_MB_OPTIMIZE_SCAN); + ctx->spec |= EXT4_SPEC_mb_optimize_scan; + } else { + ext4_msg(NULL, KERN_WARNING, + "mb_optimize_scan should be set to 0 or 1."); + return -EINVAL; + } + return 0; + } + + /* + * At this point we should only be getting options requiring MOPT_SET, + * or MOPT_CLEAR. Anything else is a bug + */ + if (m->token == Opt_err) { + ext4_msg(NULL, KERN_WARNING, "buggy handling of option %s", + param->key); + WARN_ON(1); + return -EINVAL; + } + + else { + unsigned int set = 0; + + if ((param->type == fs_value_is_flag) || + result.uint_32 > 0) + set = 1; + + if (m->flags & MOPT_CLEAR) + set = !set; + else if (unlikely(!(m->flags & MOPT_SET))) { + ext4_msg(NULL, KERN_WARNING, + "buggy handling of option %s", + param->key); + WARN_ON(1); + return -EINVAL; + } + if (m->flags & MOPT_2) { + if (set != 0) + ctx_set_mount_opt2(ctx, m->mount_opt); + else + ctx_clear_mount_opt2(ctx, m->mount_opt); + } else { + if (set != 0) + ctx_set_mount_opt(ctx, m->mount_opt); + else + ctx_clear_mount_opt(ctx, m->mount_opt); + } + } + + return 0; +} + +static int parse_options(struct fs_context *fc, char *options) +{ + struct fs_parameter param; + int ret; + char *key; + + if (!options) + return 0; + + while ((key = strsep(&options, ",")) != NULL) { + if (*key) { + size_t v_len = 0; + char *value = strchr(key, '='); + + param.type = fs_value_is_flag; + param.string = NULL; + + if (value) { + if (value == key) + continue; + + *value++ = 0; + v_len = strlen(value); + param.string = kmemdup_nul(value, v_len, + GFP_KERNEL); + if (!param.string) + return -ENOMEM; + param.type = fs_value_is_string; + } + + param.key = key; + param.size = v_len; + + ret = ext4_parse_param(fc, ¶m); + kfree(param.string); + if (ret < 0) + return ret; + } + } + + ret = ext4_validate_options(fc); + if (ret < 0) + return ret; + + return 0; +} + +static int parse_apply_sb_mount_options(struct super_block *sb, + struct ext4_fs_context *m_ctx) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + char s_mount_opts[65]; + struct ext4_fs_context *s_ctx = NULL; + struct fs_context *fc = NULL; + int ret = -ENOMEM; + + if (!sbi->s_es->s_mount_opts[0]) + return 0; + + strscpy_pad(s_mount_opts, sbi->s_es->s_mount_opts); + + fc = kzalloc(sizeof(struct fs_context), GFP_KERNEL); + if (!fc) + return -ENOMEM; + + s_ctx = kzalloc(sizeof(struct ext4_fs_context), GFP_KERNEL); + if (!s_ctx) + goto out_free; + + fc->fs_private = s_ctx; + fc->s_fs_info = sbi; + + ret = parse_options(fc, s_mount_opts); + if (ret < 0) + goto parse_failed; + + ret = ext4_check_opt_consistency(fc, sb); + if (ret < 0) { +parse_failed: + ext4_msg(sb, KERN_WARNING, + "failed to parse options in superblock: %s", + s_mount_opts); + ret = 0; + goto out_free; + } + + if (s_ctx->spec & EXT4_SPEC_JOURNAL_DEV) + m_ctx->journal_devnum = s_ctx->journal_devnum; + if (s_ctx->spec & EXT4_SPEC_JOURNAL_IOPRIO) + m_ctx->journal_ioprio = s_ctx->journal_ioprio; + + ext4_apply_options(fc, sb); + ret = 0; + +out_free: + ext4_fc_free(fc); + kfree(fc); + return ret; +} + +static void ext4_apply_quota_options(struct fs_context *fc, + struct super_block *sb) +{ +#ifdef CONFIG_QUOTA + bool quota_feature = ext4_has_feature_quota(sb); + struct ext4_fs_context *ctx = fc->fs_private; + struct ext4_sb_info *sbi = EXT4_SB(sb); + char *qname; + int i; + + if (quota_feature) + return; + + if (ctx->spec & EXT4_SPEC_JQUOTA) { + for (i = 0; i < EXT4_MAXQUOTAS; i++) { + if (!(ctx->qname_spec & (1 << i))) + continue; + + qname = ctx->s_qf_names[i]; /* May be NULL */ + if (qname) + set_opt(sb, QUOTA); + ctx->s_qf_names[i] = NULL; + qname = rcu_replace_pointer(sbi->s_qf_names[i], qname, + lockdep_is_held(&sb->s_umount)); + if (qname) + kfree_rcu_mightsleep(qname); + } + } + + if (ctx->spec & EXT4_SPEC_JQFMT) + sbi->s_jquota_fmt = ctx->s_jquota_fmt; +#endif +} + +/* + * Check quota settings consistency. + */ +static int ext4_check_quota_consistency(struct fs_context *fc, + struct super_block *sb) +{ +#ifdef CONFIG_QUOTA + struct ext4_fs_context *ctx = fc->fs_private; + struct ext4_sb_info *sbi = EXT4_SB(sb); + bool quota_feature = ext4_has_feature_quota(sb); + bool quota_loaded = sb_any_quota_loaded(sb); + bool usr_qf_name, grp_qf_name, usrquota, grpquota; + int quota_flags, i; + + /* + * We do the test below only for project quotas. 'usrquota' and + * 'grpquota' mount options are allowed even without quota feature + * to support legacy quotas in quota files. + */ + if (ctx_test_mount_opt(ctx, EXT4_MOUNT_PRJQUOTA) && + !ext4_has_feature_project(sb)) { + ext4_msg(NULL, KERN_ERR, "Project quota feature not enabled. " + "Cannot enable project quota enforcement."); + return -EINVAL; + } + + quota_flags = EXT4_MOUNT_QUOTA | EXT4_MOUNT_USRQUOTA | + EXT4_MOUNT_GRPQUOTA | EXT4_MOUNT_PRJQUOTA; + if (quota_loaded && + ctx->mask_s_mount_opt & quota_flags && + !ctx_test_mount_opt(ctx, quota_flags)) + goto err_quota_change; + + if (ctx->spec & EXT4_SPEC_JQUOTA) { + + for (i = 0; i < EXT4_MAXQUOTAS; i++) { + if (!(ctx->qname_spec & (1 << i))) + continue; + + if (quota_loaded && + !!sbi->s_qf_names[i] != !!ctx->s_qf_names[i]) + goto err_jquota_change; + + if (sbi->s_qf_names[i] && ctx->s_qf_names[i] && + strcmp(get_qf_name(sb, sbi, i), + ctx->s_qf_names[i]) != 0) + goto err_jquota_specified; + } + + if (quota_feature) { + ext4_msg(NULL, KERN_INFO, + "Journaled quota options ignored when " + "QUOTA feature is enabled"); + return 0; + } + } + + if (ctx->spec & EXT4_SPEC_JQFMT) { + if (sbi->s_jquota_fmt != ctx->s_jquota_fmt && quota_loaded) + goto err_jquota_change; + if (quota_feature) { + ext4_msg(NULL, KERN_INFO, "Quota format mount options " + "ignored when QUOTA feature is enabled"); + return 0; + } + } + + /* Make sure we don't mix old and new quota format */ + usr_qf_name = (get_qf_name(sb, sbi, USRQUOTA) || + ctx->s_qf_names[USRQUOTA]); + grp_qf_name = (get_qf_name(sb, sbi, GRPQUOTA) || + ctx->s_qf_names[GRPQUOTA]); + + usrquota = (ctx_test_mount_opt(ctx, EXT4_MOUNT_USRQUOTA) || + test_opt(sb, USRQUOTA)); + + grpquota = (ctx_test_mount_opt(ctx, EXT4_MOUNT_GRPQUOTA) || + test_opt(sb, GRPQUOTA)); + + if (usr_qf_name) { + ctx_clear_mount_opt(ctx, EXT4_MOUNT_USRQUOTA); + usrquota = false; + } + if (grp_qf_name) { + ctx_clear_mount_opt(ctx, EXT4_MOUNT_GRPQUOTA); + grpquota = false; + } + + if (usr_qf_name || grp_qf_name) { + if (usrquota || grpquota) { + ext4_msg(NULL, KERN_ERR, "old and new quota " + "format mixing"); + return -EINVAL; + } + + if (!(ctx->spec & EXT4_SPEC_JQFMT || sbi->s_jquota_fmt)) { + ext4_msg(NULL, KERN_ERR, "journaled quota format " + "not specified"); + return -EINVAL; + } + } + + return 0; + +err_quota_change: + ext4_msg(NULL, KERN_ERR, + "Cannot change quota options when quota turned on"); + return -EINVAL; +err_jquota_change: + ext4_msg(NULL, KERN_ERR, "Cannot change journaled quota " + "options when quota turned on"); + return -EINVAL; +err_jquota_specified: + ext4_msg(NULL, KERN_ERR, "%s quota file already specified", + QTYPE2NAME(i)); + return -EINVAL; +#else + return 0; +#endif +} + +static int ext4_check_test_dummy_encryption(const struct fs_context *fc, + struct super_block *sb) +{ + const struct ext4_fs_context *ctx = fc->fs_private; + const struct ext4_sb_info *sbi = EXT4_SB(sb); + + if (!fscrypt_is_dummy_policy_set(&ctx->dummy_enc_policy)) + return 0; + + if (!ext4_has_feature_encrypt(sb)) { + ext4_msg(NULL, KERN_WARNING, + "test_dummy_encryption requires encrypt feature"); + return -EINVAL; + } + /* + * This mount option is just for testing, and it's not worthwhile to + * implement the extra complexity (e.g. RCU protection) that would be + * needed to allow it to be set or changed during remount. We do allow + * it to be specified during remount, but only if there is no change. + */ + if (fc->purpose == FS_CONTEXT_FOR_RECONFIGURE) { + if (fscrypt_dummy_policies_equal(&sbi->s_dummy_enc_policy, + &ctx->dummy_enc_policy)) + return 0; + ext4_msg(NULL, KERN_WARNING, + "Can't set or change test_dummy_encryption on remount"); + return -EINVAL; + } + /* Also make sure s_mount_opts didn't contain a conflicting value. */ + if (fscrypt_is_dummy_policy_set(&sbi->s_dummy_enc_policy)) { + if (fscrypt_dummy_policies_equal(&sbi->s_dummy_enc_policy, + &ctx->dummy_enc_policy)) + return 0; + ext4_msg(NULL, KERN_WARNING, + "Conflicting test_dummy_encryption options"); + return -EINVAL; + } + return 0; +} + +static void ext4_apply_test_dummy_encryption(struct ext4_fs_context *ctx, + struct super_block *sb) +{ + if (!fscrypt_is_dummy_policy_set(&ctx->dummy_enc_policy) || + /* if already set, it was already verified to be the same */ + fscrypt_is_dummy_policy_set(&EXT4_SB(sb)->s_dummy_enc_policy)) + return; + EXT4_SB(sb)->s_dummy_enc_policy = ctx->dummy_enc_policy; + memset(&ctx->dummy_enc_policy, 0, sizeof(ctx->dummy_enc_policy)); + ext4_msg(sb, KERN_WARNING, "Test dummy encryption mode enabled"); +} + +static int ext4_check_opt_consistency(struct fs_context *fc, + struct super_block *sb) +{ + struct ext4_fs_context *ctx = fc->fs_private; + struct ext4_sb_info *sbi = fc->s_fs_info; + int is_remount = fc->purpose == FS_CONTEXT_FOR_RECONFIGURE; + int err; + + if ((ctx->opt_flags & MOPT_NO_EXT2) && IS_EXT2_SB(sb)) { + ext4_msg(NULL, KERN_ERR, + "Mount option(s) incompatible with ext2"); + return -EINVAL; + } + if ((ctx->opt_flags & MOPT_NO_EXT3) && IS_EXT3_SB(sb)) { + ext4_msg(NULL, KERN_ERR, + "Mount option(s) incompatible with ext3"); + return -EINVAL; + } + + if (ctx->s_want_extra_isize > + (sbi->s_inode_size - EXT4_GOOD_OLD_INODE_SIZE)) { + ext4_msg(NULL, KERN_ERR, + "Invalid want_extra_isize %d", + ctx->s_want_extra_isize); + return -EINVAL; + } + + err = ext4_check_test_dummy_encryption(fc, sb); + if (err) + return err; + + if ((ctx->spec & EXT4_SPEC_DATAJ) && is_remount) { + if (!sbi->s_journal) { + ext4_msg(NULL, KERN_WARNING, + "Remounting file system with no journal " + "so ignoring journalled data option"); + ctx_clear_mount_opt(ctx, EXT4_MOUNT_DATA_FLAGS); + } else if (ctx_test_mount_opt(ctx, EXT4_MOUNT_DATA_FLAGS) != + test_opt(sb, DATA_FLAGS)) { + ext4_msg(NULL, KERN_ERR, "Cannot change data mode " + "on remount"); + return -EINVAL; + } + } + + if (is_remount) { + if (!sbi->s_journal && + ctx_test_mount_opt(ctx, EXT4_MOUNT_DATA_ERR_ABORT)) { + ext4_msg(NULL, KERN_WARNING, + "Remounting fs w/o journal so ignoring data_err option"); + ctx_clear_mount_opt(ctx, EXT4_MOUNT_DATA_ERR_ABORT); + } + + if (ctx_test_mount_opt(ctx, EXT4_MOUNT_DAX_ALWAYS) && + (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA)) { + ext4_msg(NULL, KERN_ERR, "can't mount with " + "both data=journal and dax"); + return -EINVAL; + } + + if (ctx_test_mount_opt(ctx, EXT4_MOUNT_DAX_ALWAYS) && + (!(sbi->s_mount_opt & EXT4_MOUNT_DAX_ALWAYS) || + (sbi->s_mount_opt2 & EXT4_MOUNT2_DAX_NEVER))) { +fail_dax_change_remount: + ext4_msg(NULL, KERN_ERR, "can't change " + "dax mount option while remounting"); + return -EINVAL; + } else if (ctx_test_mount_opt2(ctx, EXT4_MOUNT2_DAX_NEVER) && + (!(sbi->s_mount_opt2 & EXT4_MOUNT2_DAX_NEVER) || + (sbi->s_mount_opt & EXT4_MOUNT_DAX_ALWAYS))) { + goto fail_dax_change_remount; + } else if (ctx_test_mount_opt2(ctx, EXT4_MOUNT2_DAX_INODE) && + ((sbi->s_mount_opt & EXT4_MOUNT_DAX_ALWAYS) || + (sbi->s_mount_opt2 & EXT4_MOUNT2_DAX_NEVER) || + !(sbi->s_mount_opt2 & EXT4_MOUNT2_DAX_INODE))) { + goto fail_dax_change_remount; + } + } + + return ext4_check_quota_consistency(fc, sb); +} + +static void ext4_apply_options(struct fs_context *fc, struct super_block *sb) +{ + struct ext4_fs_context *ctx = fc->fs_private; + struct ext4_sb_info *sbi = fc->s_fs_info; + + sbi->s_mount_opt &= ~ctx->mask_s_mount_opt; + sbi->s_mount_opt |= ctx->vals_s_mount_opt; + sbi->s_mount_opt2 &= ~ctx->mask_s_mount_opt2; + sbi->s_mount_opt2 |= ctx->vals_s_mount_opt2; + sb->s_flags &= ~ctx->mask_s_flags; + sb->s_flags |= ctx->vals_s_flags; + +#define APPLY(X) ({ if (ctx->spec & EXT4_SPEC_##X) sbi->X = ctx->X; }) + APPLY(s_commit_interval); + APPLY(s_stripe); + APPLY(s_max_batch_time); + APPLY(s_min_batch_time); + APPLY(s_want_extra_isize); + APPLY(s_inode_readahead_blks); + APPLY(s_max_dir_size_kb); + APPLY(s_li_wait_mult); + APPLY(s_resgid); + APPLY(s_resuid); + +#ifdef CONFIG_EXT4_DEBUG + APPLY(s_fc_debug_max_replay); +#endif + + ext4_apply_quota_options(fc, sb); + ext4_apply_test_dummy_encryption(ctx, sb); +} + + +static int ext4_validate_options(struct fs_context *fc) +{ +#ifdef CONFIG_QUOTA + struct ext4_fs_context *ctx = fc->fs_private; + char *usr_qf_name, *grp_qf_name; + + usr_qf_name = ctx->s_qf_names[USRQUOTA]; + grp_qf_name = ctx->s_qf_names[GRPQUOTA]; + + if (usr_qf_name || grp_qf_name) { + if (ctx_test_mount_opt(ctx, EXT4_MOUNT_USRQUOTA) && usr_qf_name) + ctx_clear_mount_opt(ctx, EXT4_MOUNT_USRQUOTA); + + if (ctx_test_mount_opt(ctx, EXT4_MOUNT_GRPQUOTA) && grp_qf_name) + ctx_clear_mount_opt(ctx, EXT4_MOUNT_GRPQUOTA); + + if (ctx_test_mount_opt(ctx, EXT4_MOUNT_USRQUOTA) || + ctx_test_mount_opt(ctx, EXT4_MOUNT_GRPQUOTA)) { + ext4_msg(NULL, KERN_ERR, "old and new quota " + "format mixing"); + return -EINVAL; + } + } +#endif + return 1; +} + +static inline void ext4_show_quota_options(struct seq_file *seq, + struct super_block *sb) +{ +#if defined(CONFIG_QUOTA) + struct ext4_sb_info *sbi = EXT4_SB(sb); + char *usr_qf_name, *grp_qf_name; + + if (sbi->s_jquota_fmt) { + char *fmtname = ""; + + switch (sbi->s_jquota_fmt) { + case QFMT_VFS_OLD: + fmtname = "vfsold"; + break; + case QFMT_VFS_V0: + fmtname = "vfsv0"; + break; + case QFMT_VFS_V1: + fmtname = "vfsv1"; + break; + } + seq_printf(seq, ",jqfmt=%s", fmtname); + } + + rcu_read_lock(); + usr_qf_name = rcu_dereference(sbi->s_qf_names[USRQUOTA]); + grp_qf_name = rcu_dereference(sbi->s_qf_names[GRPQUOTA]); + if (usr_qf_name) + seq_show_option(seq, "usrjquota", usr_qf_name); + if (grp_qf_name) + seq_show_option(seq, "grpjquota", grp_qf_name); + rcu_read_unlock(); +#endif +} + +static const char *token2str(int token) +{ + const struct fs_parameter_spec *spec; + + for (spec = ext4_param_specs; spec->name != NULL; spec++) + if (spec->opt == token && !spec->type) + break; + return spec->name; +} + +/* + * Show an option if + * - it's set to a non-default value OR + * - if the per-sb default is different from the global default + */ +static int _ext4_show_options(struct seq_file *seq, struct super_block *sb, + int nodefs) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_super_block *es = sbi->s_es; + int def_errors; + const struct mount_opts *m; + char sep = nodefs ? '\n' : ','; + +#define SEQ_OPTS_PUTS(str) seq_printf(seq, "%c" str, sep) +#define SEQ_OPTS_PRINT(str, arg) seq_printf(seq, "%c" str, sep, arg) + + if (sbi->s_sb_block != 1) + SEQ_OPTS_PRINT("sb=%llu", sbi->s_sb_block); + + for (m = ext4_mount_opts; m->token != Opt_err; m++) { + int want_set = m->flags & MOPT_SET; + int opt_2 = m->flags & MOPT_2; + unsigned int mount_opt, def_mount_opt; + + if (((m->flags & (MOPT_SET|MOPT_CLEAR)) == 0) || + m->flags & MOPT_SKIP) + continue; + + if (opt_2) { + mount_opt = sbi->s_mount_opt2; + def_mount_opt = sbi->s_def_mount_opt2; + } else { + mount_opt = sbi->s_mount_opt; + def_mount_opt = sbi->s_def_mount_opt; + } + /* skip if same as the default */ + if (!nodefs && !(m->mount_opt & (mount_opt ^ def_mount_opt))) + continue; + /* select Opt_noFoo vs Opt_Foo */ + if ((want_set && + (mount_opt & m->mount_opt) != m->mount_opt) || + (!want_set && (mount_opt & m->mount_opt))) + continue; + SEQ_OPTS_PRINT("%s", token2str(m->token)); + } + + if (nodefs || !uid_eq(sbi->s_resuid, make_kuid(&init_user_ns, EXT4_DEF_RESUID)) || + ext4_get_resuid(es) != EXT4_DEF_RESUID) + SEQ_OPTS_PRINT("resuid=%u", + from_kuid_munged(&init_user_ns, sbi->s_resuid)); + if (nodefs || !gid_eq(sbi->s_resgid, make_kgid(&init_user_ns, EXT4_DEF_RESGID)) || + ext4_get_resgid(es) != EXT4_DEF_RESGID) + SEQ_OPTS_PRINT("resgid=%u", + from_kgid_munged(&init_user_ns, sbi->s_resgid)); + def_errors = nodefs ? -1 : le16_to_cpu(es->s_errors); + if (test_opt(sb, ERRORS_RO) && def_errors != EXT4_ERRORS_RO) + SEQ_OPTS_PUTS("errors=remount-ro"); + if (test_opt(sb, ERRORS_CONT) && def_errors != EXT4_ERRORS_CONTINUE) + SEQ_OPTS_PUTS("errors=continue"); + if (test_opt(sb, ERRORS_PANIC) && def_errors != EXT4_ERRORS_PANIC) + SEQ_OPTS_PUTS("errors=panic"); + if (nodefs || sbi->s_commit_interval != JBD2_DEFAULT_MAX_COMMIT_AGE*HZ) + SEQ_OPTS_PRINT("commit=%lu", sbi->s_commit_interval / HZ); + if (nodefs || sbi->s_min_batch_time != EXT4_DEF_MIN_BATCH_TIME) + SEQ_OPTS_PRINT("min_batch_time=%u", sbi->s_min_batch_time); + if (nodefs || sbi->s_max_batch_time != EXT4_DEF_MAX_BATCH_TIME) + SEQ_OPTS_PRINT("max_batch_time=%u", sbi->s_max_batch_time); + if (nodefs && sb->s_flags & SB_I_VERSION) + SEQ_OPTS_PUTS("i_version"); + if (nodefs || sbi->s_stripe) + SEQ_OPTS_PRINT("stripe=%lu", sbi->s_stripe); + if (nodefs || EXT4_MOUNT_DATA_FLAGS & + (sbi->s_mount_opt ^ sbi->s_def_mount_opt)) { + if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA) + SEQ_OPTS_PUTS("data=journal"); + else if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_ORDERED_DATA) + SEQ_OPTS_PUTS("data=ordered"); + else if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_WRITEBACK_DATA) + SEQ_OPTS_PUTS("data=writeback"); + } + if (nodefs || + sbi->s_inode_readahead_blks != EXT4_DEF_INODE_READAHEAD_BLKS) + SEQ_OPTS_PRINT("inode_readahead_blks=%u", + sbi->s_inode_readahead_blks); + + if (test_opt(sb, INIT_INODE_TABLE) && (nodefs || + (sbi->s_li_wait_mult != EXT4_DEF_LI_WAIT_MULT))) + SEQ_OPTS_PRINT("init_itable=%u", sbi->s_li_wait_mult); + if (nodefs || sbi->s_max_dir_size_kb) + SEQ_OPTS_PRINT("max_dir_size_kb=%u", sbi->s_max_dir_size_kb); + if (test_opt(sb, DATA_ERR_ABORT)) + SEQ_OPTS_PUTS("data_err=abort"); + + fscrypt_show_test_dummy_encryption(seq, sep, sb); + + if (sb->s_flags & SB_INLINECRYPT) + SEQ_OPTS_PUTS("inlinecrypt"); + + if (test_opt(sb, DAX_ALWAYS)) { + if (IS_EXT2_SB(sb)) + SEQ_OPTS_PUTS("dax"); + else + SEQ_OPTS_PUTS("dax=always"); + } else if (test_opt2(sb, DAX_NEVER)) { + SEQ_OPTS_PUTS("dax=never"); + } else if (test_opt2(sb, DAX_INODE)) { + SEQ_OPTS_PUTS("dax=inode"); + } + + if (sbi->s_groups_count >= MB_DEFAULT_LINEAR_SCAN_THRESHOLD && + !test_opt2(sb, MB_OPTIMIZE_SCAN)) { + SEQ_OPTS_PUTS("mb_optimize_scan=0"); + } else if (sbi->s_groups_count < MB_DEFAULT_LINEAR_SCAN_THRESHOLD && + test_opt2(sb, MB_OPTIMIZE_SCAN)) { + SEQ_OPTS_PUTS("mb_optimize_scan=1"); + } + + if (nodefs && !test_opt(sb, NO_PREFETCH_BLOCK_BITMAPS)) + SEQ_OPTS_PUTS("prefetch_block_bitmaps"); + + if (ext4_emergency_ro(sb)) + SEQ_OPTS_PUTS("emergency_ro"); + + if (ext4_forced_shutdown(sb)) + SEQ_OPTS_PUTS("shutdown"); + + ext4_show_quota_options(seq, sb); + return 0; +} + +static int ext4_show_options(struct seq_file *seq, struct dentry *root) +{ + return _ext4_show_options(seq, root->d_sb, 0); +} + +int ext4_seq_options_show(struct seq_file *seq, void *offset) +{ + struct super_block *sb = seq->private; + int rc; + + seq_puts(seq, sb_rdonly(sb) ? "ro" : "rw"); + rc = _ext4_show_options(seq, sb, 1); + seq_putc(seq, '\n'); + return rc; +} + +static int ext4_setup_super(struct super_block *sb, struct ext4_super_block *es, + int read_only) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + int err = 0; + + if (le32_to_cpu(es->s_rev_level) > EXT4_MAX_SUPP_REV) { + ext4_msg(sb, KERN_ERR, "revision level too high, " + "forcing read-only mode"); + err = -EROFS; + goto done; + } + if (read_only) + goto done; + if (!(sbi->s_mount_state & EXT4_VALID_FS)) + ext4_msg(sb, KERN_WARNING, "warning: mounting unchecked fs, " + "running e2fsck is recommended"); + else if (sbi->s_mount_state & EXT4_ERROR_FS) + ext4_msg(sb, KERN_WARNING, + "warning: mounting fs with errors, " + "running e2fsck is recommended"); + else if ((__s16) le16_to_cpu(es->s_max_mnt_count) > 0 && + le16_to_cpu(es->s_mnt_count) >= + (unsigned short) (__s16) le16_to_cpu(es->s_max_mnt_count)) + ext4_msg(sb, KERN_WARNING, + "warning: maximal mount count reached, " + "running e2fsck is recommended"); + else if (le32_to_cpu(es->s_checkinterval) && + (ext4_get_tstamp(es, s_lastcheck) + + le32_to_cpu(es->s_checkinterval) <= ktime_get_real_seconds())) + ext4_msg(sb, KERN_WARNING, + "warning: checktime reached, " + "running e2fsck is recommended"); + if (!sbi->s_journal) + es->s_state &= cpu_to_le16(~EXT4_VALID_FS); + if (!(__s16) le16_to_cpu(es->s_max_mnt_count)) + es->s_max_mnt_count = cpu_to_le16(EXT4_DFL_MAX_MNT_COUNT); + le16_add_cpu(&es->s_mnt_count, 1); + ext4_update_tstamp(es, s_mtime); + if (sbi->s_journal) { + ext4_set_feature_journal_needs_recovery(sb); + if (ext4_has_feature_orphan_file(sb)) + ext4_set_feature_orphan_present(sb); + } + + err = ext4_commit_super(sb); +done: + if (test_opt(sb, DEBUG)) + printk(KERN_INFO "[EXT4 FS bs=%lu, gc=%u, " + "bpg=%lu, ipg=%lu, mo=%04x, mo2=%04x]\n", + sb->s_blocksize, + sbi->s_groups_count, + EXT4_BLOCKS_PER_GROUP(sb), + EXT4_INODES_PER_GROUP(sb), + sbi->s_mount_opt, sbi->s_mount_opt2); + return err; +} + +int ext4_alloc_flex_bg_array(struct super_block *sb, ext4_group_t ngroup) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct flex_groups **old_groups, **new_groups; + int size, i, j; + + if (!sbi->s_log_groups_per_flex) + return 0; + + size = ext4_flex_group(sbi, ngroup - 1) + 1; + if (size <= sbi->s_flex_groups_allocated) + return 0; + + new_groups = kvzalloc(roundup_pow_of_two(size * + sizeof(*sbi->s_flex_groups)), GFP_KERNEL); + if (!new_groups) { + ext4_msg(sb, KERN_ERR, + "not enough memory for %d flex group pointers", size); + return -ENOMEM; + } + for (i = sbi->s_flex_groups_allocated; i < size; i++) { + new_groups[i] = kvzalloc(roundup_pow_of_two( + sizeof(struct flex_groups)), + GFP_KERNEL); + if (!new_groups[i]) { + for (j = sbi->s_flex_groups_allocated; j < i; j++) + kvfree(new_groups[j]); + kvfree(new_groups); + ext4_msg(sb, KERN_ERR, + "not enough memory for %d flex groups", size); + return -ENOMEM; + } + } + rcu_read_lock(); + old_groups = rcu_dereference(sbi->s_flex_groups); + if (old_groups) + memcpy(new_groups, old_groups, + (sbi->s_flex_groups_allocated * + sizeof(struct flex_groups *))); + rcu_read_unlock(); + rcu_assign_pointer(sbi->s_flex_groups, new_groups); + sbi->s_flex_groups_allocated = size; + if (old_groups) + ext4_kvfree_array_rcu(old_groups); + return 0; +} + +static int ext4_fill_flex_info(struct super_block *sb) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_group_desc *gdp = NULL; + struct flex_groups *fg; + ext4_group_t flex_group; + int i, err; + + sbi->s_log_groups_per_flex = sbi->s_es->s_log_groups_per_flex; + if (sbi->s_log_groups_per_flex < 1 || sbi->s_log_groups_per_flex > 31) { + sbi->s_log_groups_per_flex = 0; + return 1; + } + + err = ext4_alloc_flex_bg_array(sb, sbi->s_groups_count); + if (err) + goto failed; + + for (i = 0; i < sbi->s_groups_count; i++) { + gdp = ext4_get_group_desc(sb, i, NULL); + + flex_group = ext4_flex_group(sbi, i); + fg = sbi_array_rcu_deref(sbi, s_flex_groups, flex_group); + atomic_add(ext4_free_inodes_count(sb, gdp), &fg->free_inodes); + atomic64_add(ext4_free_group_clusters(sb, gdp), + &fg->free_clusters); + atomic_add(ext4_used_dirs_count(sb, gdp), &fg->used_dirs); + } + + return 1; +failed: + return 0; +} + +static __le16 ext4_group_desc_csum(struct super_block *sb, __u32 block_group, + struct ext4_group_desc *gdp) +{ + int offset = offsetof(struct ext4_group_desc, bg_checksum); + __u16 crc = 0; + __le32 le_group = cpu_to_le32(block_group); + struct ext4_sb_info *sbi = EXT4_SB(sb); + + if (ext4_has_feature_metadata_csum(sbi->s_sb)) { + /* Use new metadata_csum algorithm */ + __u32 csum32; + __u16 dummy_csum = 0; + + csum32 = ext4_chksum(sbi->s_csum_seed, (__u8 *)&le_group, + sizeof(le_group)); + csum32 = ext4_chksum(csum32, (__u8 *)gdp, offset); + csum32 = ext4_chksum(csum32, (__u8 *)&dummy_csum, + sizeof(dummy_csum)); + offset += sizeof(dummy_csum); + if (offset < sbi->s_desc_size) + csum32 = ext4_chksum(csum32, (__u8 *)gdp + offset, + sbi->s_desc_size - offset); + + crc = csum32 & 0xFFFF; + goto out; + } + + /* old crc16 code */ + if (!ext4_has_feature_gdt_csum(sb)) + return 0; + + crc = crc16(~0, sbi->s_es->s_uuid, sizeof(sbi->s_es->s_uuid)); + crc = crc16(crc, (__u8 *)&le_group, sizeof(le_group)); + crc = crc16(crc, (__u8 *)gdp, offset); + offset += sizeof(gdp->bg_checksum); /* skip checksum */ + /* for checksum of struct ext4_group_desc do the rest...*/ + if (ext4_has_feature_64bit(sb) && offset < sbi->s_desc_size) + crc = crc16(crc, (__u8 *)gdp + offset, + sbi->s_desc_size - offset); + +out: + return cpu_to_le16(crc); +} + +int ext4_group_desc_csum_verify(struct super_block *sb, __u32 block_group, + struct ext4_group_desc *gdp) +{ + if (ext4_has_group_desc_csum(sb) && + (gdp->bg_checksum != ext4_group_desc_csum(sb, block_group, gdp))) + return 0; + + return 1; +} + +void ext4_group_desc_csum_set(struct super_block *sb, __u32 block_group, + struct ext4_group_desc *gdp) +{ + if (!ext4_has_group_desc_csum(sb)) + return; + gdp->bg_checksum = ext4_group_desc_csum(sb, block_group, gdp); +} + +/* Called at mount-time, super-block is locked */ +static int ext4_check_descriptors(struct super_block *sb, + ext4_fsblk_t sb_block, + ext4_group_t *first_not_zeroed) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + ext4_fsblk_t first_block = le32_to_cpu(sbi->s_es->s_first_data_block); + ext4_fsblk_t last_block; + ext4_fsblk_t last_bg_block = sb_block + ext4_bg_num_gdb(sb, 0); + ext4_fsblk_t block_bitmap; + ext4_fsblk_t inode_bitmap; + ext4_fsblk_t inode_table; + int flexbg_flag = 0; + ext4_group_t i, grp = sbi->s_groups_count; + + if (ext4_has_feature_flex_bg(sb)) + flexbg_flag = 1; + + ext4_debug("Checking group descriptors"); + + for (i = 0; i < sbi->s_groups_count; i++) { + struct ext4_group_desc *gdp = ext4_get_group_desc(sb, i, NULL); + + if (i == sbi->s_groups_count - 1 || flexbg_flag) + last_block = ext4_blocks_count(sbi->s_es) - 1; + else + last_block = first_block + + (EXT4_BLOCKS_PER_GROUP(sb) - 1); + + if ((grp == sbi->s_groups_count) && + !(gdp->bg_flags & cpu_to_le16(EXT4_BG_INODE_ZEROED))) + grp = i; + + block_bitmap = ext4_block_bitmap(sb, gdp); + if (block_bitmap == sb_block) { + ext4_msg(sb, KERN_ERR, "ext4_check_descriptors: " + "Block bitmap for group %u overlaps " + "superblock", i); + if (!sb_rdonly(sb)) + return 0; + } + if (block_bitmap >= sb_block + 1 && + block_bitmap <= last_bg_block) { + ext4_msg(sb, KERN_ERR, "ext4_check_descriptors: " + "Block bitmap for group %u overlaps " + "block group descriptors", i); + if (!sb_rdonly(sb)) + return 0; + } + if (block_bitmap < first_block || block_bitmap > last_block) { + ext4_msg(sb, KERN_ERR, "ext4_check_descriptors: " + "Block bitmap for group %u not in group " + "(block %llu)!", i, block_bitmap); + return 0; + } + inode_bitmap = ext4_inode_bitmap(sb, gdp); + if (inode_bitmap == sb_block) { + ext4_msg(sb, KERN_ERR, "ext4_check_descriptors: " + "Inode bitmap for group %u overlaps " + "superblock", i); + if (!sb_rdonly(sb)) + return 0; + } + if (inode_bitmap >= sb_block + 1 && + inode_bitmap <= last_bg_block) { + ext4_msg(sb, KERN_ERR, "ext4_check_descriptors: " + "Inode bitmap for group %u overlaps " + "block group descriptors", i); + if (!sb_rdonly(sb)) + return 0; + } + if (inode_bitmap < first_block || inode_bitmap > last_block) { + ext4_msg(sb, KERN_ERR, "ext4_check_descriptors: " + "Inode bitmap for group %u not in group " + "(block %llu)!", i, inode_bitmap); + return 0; + } + inode_table = ext4_inode_table(sb, gdp); + if (inode_table == sb_block) { + ext4_msg(sb, KERN_ERR, "ext4_check_descriptors: " + "Inode table for group %u overlaps " + "superblock", i); + if (!sb_rdonly(sb)) + return 0; + } + if (inode_table >= sb_block + 1 && + inode_table <= last_bg_block) { + ext4_msg(sb, KERN_ERR, "ext4_check_descriptors: " + "Inode table for group %u overlaps " + "block group descriptors", i); + if (!sb_rdonly(sb)) + return 0; + } + if (inode_table < first_block || + inode_table + sbi->s_itb_per_group - 1 > last_block) { + ext4_msg(sb, KERN_ERR, "ext4_check_descriptors: " + "Inode table for group %u not in group " + "(block %llu)!", i, inode_table); + return 0; + } + ext4_lock_group(sb, i); + if (!ext4_group_desc_csum_verify(sb, i, gdp)) { + ext4_msg(sb, KERN_ERR, "ext4_check_descriptors: " + "Checksum for group %u failed (%u!=%u)", + i, le16_to_cpu(ext4_group_desc_csum(sb, i, + gdp)), le16_to_cpu(gdp->bg_checksum)); + if (!sb_rdonly(sb)) { + ext4_unlock_group(sb, i); + return 0; + } + } + ext4_unlock_group(sb, i); + if (!flexbg_flag) + first_block += EXT4_BLOCKS_PER_GROUP(sb); + } + if (NULL != first_not_zeroed) + *first_not_zeroed = grp; + return 1; +} + +/* + * Maximal extent format file size. + * Resulting logical blkno at s_maxbytes must fit in our on-disk + * extent format containers, within a sector_t, and within i_blocks + * in the vfs. ext4 inode has 48 bits of i_block in fsblock units, + * so that won't be a limiting factor. + * + * However there is other limiting factor. We do store extents in the form + * of starting block and length, hence the resulting length of the extent + * covering maximum file size must fit into on-disk format containers as + * well. Given that length is always by 1 unit bigger than max unit (because + * we count 0 as well) we have to lower the s_maxbytes by one fs block. + * + * Note, this does *not* consider any metadata overhead for vfs i_blocks. + */ +static loff_t ext4_max_size(int blkbits, int has_huge_files) +{ + loff_t res; + loff_t upper_limit = MAX_LFS_FILESIZE; + + BUILD_BUG_ON(sizeof(blkcnt_t) < sizeof(u64)); + + if (!has_huge_files) { + upper_limit = (1LL << 32) - 1; + + /* total blocks in file system block size */ + upper_limit >>= (blkbits - 9); + upper_limit <<= blkbits; + } + + /* + * 32-bit extent-start container, ee_block. We lower the maxbytes + * by one fs block, so ee_len can cover the extent of maximum file + * size + */ + res = (1LL << 32) - 1; + res <<= blkbits; + + /* Sanity check against vm- & vfs- imposed limits */ + if (res > upper_limit) + res = upper_limit; + + return res; +} + +/* + * Maximal bitmap file size. There is a direct, and {,double-,triple-}indirect + * block limit, and also a limit of (2^48 - 1) 512-byte sectors in i_blocks. + * We need to be 1 filesystem block less than the 2^48 sector limit. + */ +static loff_t ext4_max_bitmap_size(int bits, int has_huge_files) +{ + loff_t upper_limit, res = EXT4_NDIR_BLOCKS; + int meta_blocks; + unsigned int ppb = 1 << (bits - 2); + + /* + * This is calculated to be the largest file size for a dense, block + * mapped file such that the file's total number of 512-byte sectors, + * including data and all indirect blocks, does not exceed (2^48 - 1). + * + * __u32 i_blocks_lo and _u16 i_blocks_high represent the total + * number of 512-byte sectors of the file. + */ + if (!has_huge_files) { + /* + * !has_huge_files or implies that the inode i_block field + * represents total file blocks in 2^32 512-byte sectors == + * size of vfs inode i_blocks * 8 + */ + upper_limit = (1LL << 32) - 1; + + /* total blocks in file system block size */ + upper_limit >>= (bits - 9); + + } else { + /* + * We use 48 bit ext4_inode i_blocks + * With EXT4_HUGE_FILE_FL set the i_blocks + * represent total number of blocks in + * file system block size + */ + upper_limit = (1LL << 48) - 1; + + } + + /* Compute how many blocks we can address by block tree */ + res += ppb; + res += ppb * ppb; + res += ((loff_t)ppb) * ppb * ppb; + /* Compute how many metadata blocks are needed */ + meta_blocks = 1; + meta_blocks += 1 + ppb; + meta_blocks += 1 + ppb + ppb * ppb; + /* Does block tree limit file size? */ + if (res + meta_blocks <= upper_limit) + goto check_lfs; + + res = upper_limit; + /* How many metadata blocks are needed for addressing upper_limit? */ + upper_limit -= EXT4_NDIR_BLOCKS; + /* indirect blocks */ + meta_blocks = 1; + upper_limit -= ppb; + /* double indirect blocks */ + if (upper_limit < ppb * ppb) { + meta_blocks += 1 + DIV_ROUND_UP_ULL(upper_limit, ppb); + res -= meta_blocks; + goto check_lfs; + } + meta_blocks += 1 + ppb; + upper_limit -= ppb * ppb; + /* tripple indirect blocks for the rest */ + meta_blocks += 1 + DIV_ROUND_UP_ULL(upper_limit, ppb) + + DIV_ROUND_UP_ULL(upper_limit, ppb*ppb); + res -= meta_blocks; +check_lfs: + res <<= bits; + if (res > MAX_LFS_FILESIZE) + res = MAX_LFS_FILESIZE; + + return res; +} + +static ext4_fsblk_t descriptor_loc(struct super_block *sb, + ext4_fsblk_t logical_sb_block, int nr) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + ext4_group_t bg, first_meta_bg; + int has_super = 0; + + first_meta_bg = le32_to_cpu(sbi->s_es->s_first_meta_bg); + + if (!ext4_has_feature_meta_bg(sb) || nr < first_meta_bg) + return logical_sb_block + nr + 1; + bg = sbi->s_desc_per_block * nr; + if (ext4_bg_has_super(sb, bg)) + has_super = 1; + + /* + * If we have a meta_bg fs with 1k blocks, group 0's GDT is at + * block 2, not 1. If s_first_data_block == 0 (bigalloc is enabled + * on modern mke2fs or blksize > 1k on older mke2fs) then we must + * compensate. + */ + if (sb->s_blocksize == 1024 && nr == 0 && + le32_to_cpu(sbi->s_es->s_first_data_block) == 0) + has_super++; + + return (has_super + ext4_group_first_block_no(sb, bg)); +} + +/** + * ext4_get_stripe_size: Get the stripe size. + * @sbi: In memory super block info + * + * If we have specified it via mount option, then + * use the mount option value. If the value specified at mount time is + * greater than the blocks per group use the super block value. + * If the super block value is greater than blocks per group return 0. + * Allocator needs it be less than blocks per group. + * + */ +static unsigned long ext4_get_stripe_size(struct ext4_sb_info *sbi) +{ + unsigned long stride = le16_to_cpu(sbi->s_es->s_raid_stride); + unsigned long stripe_width = + le32_to_cpu(sbi->s_es->s_raid_stripe_width); + int ret; + + if (sbi->s_stripe && sbi->s_stripe <= sbi->s_blocks_per_group) + ret = sbi->s_stripe; + else if (stripe_width && stripe_width <= sbi->s_blocks_per_group) + ret = stripe_width; + else if (stride && stride <= sbi->s_blocks_per_group) + ret = stride; + else + ret = 0; + + /* + * If the stripe width is 1, this makes no sense and + * we set it to 0 to turn off stripe handling code. + */ + if (ret <= 1) + ret = 0; + + return ret; +} + +/* + * Check whether this filesystem can be mounted based on + * the features present and the RDONLY/RDWR mount requested. + * Returns 1 if this filesystem can be mounted as requested, + * 0 if it cannot be. + */ +int ext4_feature_set_ok(struct super_block *sb, int readonly) +{ + if (ext4_has_unknown_ext4_incompat_features(sb)) { + ext4_msg(sb, KERN_ERR, + "Couldn't mount because of " + "unsupported optional features (%x)", + (le32_to_cpu(EXT4_SB(sb)->s_es->s_feature_incompat) & + ~EXT4_FEATURE_INCOMPAT_SUPP)); + return 0; + } + + if (!IS_ENABLED(CONFIG_UNICODE) && ext4_has_feature_casefold(sb)) { + ext4_msg(sb, KERN_ERR, + "Filesystem with casefold feature cannot be " + "mounted without CONFIG_UNICODE"); + return 0; + } + + if (readonly) + return 1; + + if (ext4_has_feature_readonly(sb)) { + ext4_msg(sb, KERN_INFO, "filesystem is read-only"); + sb->s_flags |= SB_RDONLY; + return 1; + } + + /* Check that feature set is OK for a read-write mount */ + if (ext4_has_unknown_ext4_ro_compat_features(sb)) { + ext4_msg(sb, KERN_ERR, "couldn't mount RDWR because of " + "unsupported optional features (%x)", + (le32_to_cpu(EXT4_SB(sb)->s_es->s_feature_ro_compat) & + ~EXT4_FEATURE_RO_COMPAT_SUPP)); + return 0; + } + if (ext4_has_feature_bigalloc(sb) && !ext4_has_feature_extents(sb)) { + ext4_msg(sb, KERN_ERR, + "Can't support bigalloc feature without " + "extents feature\n"); + return 0; + } + +#if !IS_ENABLED(CONFIG_QUOTA) || !IS_ENABLED(CONFIG_QFMT_V2) + if (!readonly && (ext4_has_feature_quota(sb) || + ext4_has_feature_project(sb))) { + ext4_msg(sb, KERN_ERR, + "The kernel was not built with CONFIG_QUOTA and CONFIG_QFMT_V2"); + return 0; + } +#endif /* CONFIG_QUOTA */ + return 1; +} + +/* + * This function is called once a day if we have errors logged + * on the file system + */ +static void print_daily_error_info(struct timer_list *t) +{ + struct ext4_sb_info *sbi = timer_container_of(sbi, t, s_err_report); + struct super_block *sb = sbi->s_sb; + struct ext4_super_block *es = sbi->s_es; + + if (es->s_error_count) + /* fsck newer than v1.41.13 is needed to clean this condition. */ + ext4_msg(sb, KERN_NOTICE, "error count since last fsck: %u", + le32_to_cpu(es->s_error_count)); + if (es->s_first_error_time) { + printk(KERN_NOTICE "EXT4-fs (%s): initial error at time %llu: %.*s:%d", + sb->s_id, + ext4_get_tstamp(es, s_first_error_time), + (int) sizeof(es->s_first_error_func), + es->s_first_error_func, + le32_to_cpu(es->s_first_error_line)); + if (es->s_first_error_ino) + printk(KERN_CONT ": inode %u", + le32_to_cpu(es->s_first_error_ino)); + if (es->s_first_error_block) + printk(KERN_CONT ": block %llu", (unsigned long long) + le64_to_cpu(es->s_first_error_block)); + printk(KERN_CONT "\n"); + } + if (es->s_last_error_time) { + printk(KERN_NOTICE "EXT4-fs (%s): last error at time %llu: %.*s:%d", + sb->s_id, + ext4_get_tstamp(es, s_last_error_time), + (int) sizeof(es->s_last_error_func), + es->s_last_error_func, + le32_to_cpu(es->s_last_error_line)); + if (es->s_last_error_ino) + printk(KERN_CONT ": inode %u", + le32_to_cpu(es->s_last_error_ino)); + if (es->s_last_error_block) + printk(KERN_CONT ": block %llu", (unsigned long long) + le64_to_cpu(es->s_last_error_block)); + printk(KERN_CONT "\n"); + } + mod_timer(&sbi->s_err_report, jiffies + 24*60*60*HZ); /* Once a day */ +} + +/* Find next suitable group and run ext4_init_inode_table */ +static int ext4_run_li_request(struct ext4_li_request *elr) +{ + struct ext4_group_desc *gdp = NULL; + struct super_block *sb = elr->lr_super; + ext4_group_t ngroups = EXT4_SB(sb)->s_groups_count; + ext4_group_t group = elr->lr_next_group; + unsigned int prefetch_ios = 0; + int ret = 0; + int nr = EXT4_SB(sb)->s_mb_prefetch; + u64 start_time; + + if (elr->lr_mode == EXT4_LI_MODE_PREFETCH_BBITMAP) { + elr->lr_next_group = ext4_mb_prefetch(sb, group, nr, &prefetch_ios); + ext4_mb_prefetch_fini(sb, elr->lr_next_group, nr); + trace_ext4_prefetch_bitmaps(sb, group, elr->lr_next_group, nr); + if (group >= elr->lr_next_group) { + ret = 1; + if (elr->lr_first_not_zeroed != ngroups && + !ext4_emergency_state(sb) && !sb_rdonly(sb) && + test_opt(sb, INIT_INODE_TABLE)) { + elr->lr_next_group = elr->lr_first_not_zeroed; + elr->lr_mode = EXT4_LI_MODE_ITABLE; + ret = 0; + } + } + return ret; + } + + for (; group < ngroups; group++) { + gdp = ext4_get_group_desc(sb, group, NULL); + if (!gdp) { + ret = 1; + break; + } + + if (!(gdp->bg_flags & cpu_to_le16(EXT4_BG_INODE_ZEROED))) + break; + } + + if (group >= ngroups) + ret = 1; + + if (!ret) { + start_time = ktime_get_ns(); + ret = ext4_init_inode_table(sb, group, + elr->lr_timeout ? 0 : 1); + trace_ext4_lazy_itable_init(sb, group); + if (elr->lr_timeout == 0) { + elr->lr_timeout = nsecs_to_jiffies((ktime_get_ns() - start_time) * + EXT4_SB(elr->lr_super)->s_li_wait_mult); + } + elr->lr_next_sched = jiffies + elr->lr_timeout; + elr->lr_next_group = group + 1; + } + return ret; +} + +/* + * Remove lr_request from the list_request and free the + * request structure. Should be called with li_list_mtx held + */ +static void ext4_remove_li_request(struct ext4_li_request *elr) +{ + if (!elr) + return; + + list_del(&elr->lr_request); + EXT4_SB(elr->lr_super)->s_li_request = NULL; + kfree(elr); +} + +static void ext4_unregister_li_request(struct super_block *sb) +{ + mutex_lock(&ext4_li_mtx); + if (!ext4_li_info) { + mutex_unlock(&ext4_li_mtx); + return; + } + + mutex_lock(&ext4_li_info->li_list_mtx); + ext4_remove_li_request(EXT4_SB(sb)->s_li_request); + mutex_unlock(&ext4_li_info->li_list_mtx); + mutex_unlock(&ext4_li_mtx); +} + +static struct task_struct *ext4_lazyinit_task; + +/* + * This is the function where ext4lazyinit thread lives. It walks + * through the request list searching for next scheduled filesystem. + * When such a fs is found, run the lazy initialization request + * (ext4_rn_li_request) and keep track of the time spend in this + * function. Based on that time we compute next schedule time of + * the request. When walking through the list is complete, compute + * next waking time and put itself into sleep. + */ +static int ext4_lazyinit_thread(void *arg) +{ + struct ext4_lazy_init *eli = arg; + struct list_head *pos, *n; + struct ext4_li_request *elr; + unsigned long next_wakeup, cur; + + BUG_ON(NULL == eli); + set_freezable(); + +cont_thread: + while (true) { + bool next_wakeup_initialized = false; + + next_wakeup = 0; + mutex_lock(&eli->li_list_mtx); + if (list_empty(&eli->li_request_list)) { + mutex_unlock(&eli->li_list_mtx); + goto exit_thread; + } + list_for_each_safe(pos, n, &eli->li_request_list) { + int err = 0; + int progress = 0; + elr = list_entry(pos, struct ext4_li_request, + lr_request); + + if (time_before(jiffies, elr->lr_next_sched)) { + if (!next_wakeup_initialized || + time_before(elr->lr_next_sched, next_wakeup)) { + next_wakeup = elr->lr_next_sched; + next_wakeup_initialized = true; + } + continue; + } + if (down_read_trylock(&elr->lr_super->s_umount)) { + if (sb_start_write_trylock(elr->lr_super)) { + progress = 1; + /* + * We hold sb->s_umount, sb can not + * be removed from the list, it is + * now safe to drop li_list_mtx + */ + mutex_unlock(&eli->li_list_mtx); + err = ext4_run_li_request(elr); + sb_end_write(elr->lr_super); + mutex_lock(&eli->li_list_mtx); + n = pos->next; + } + up_read((&elr->lr_super->s_umount)); + } + /* error, remove the lazy_init job */ + if (err) { + ext4_remove_li_request(elr); + continue; + } + if (!progress) { + elr->lr_next_sched = jiffies + + get_random_u32_below(EXT4_DEF_LI_MAX_START_DELAY * HZ); + } + if (!next_wakeup_initialized || + time_before(elr->lr_next_sched, next_wakeup)) { + next_wakeup = elr->lr_next_sched; + next_wakeup_initialized = true; + } + } + mutex_unlock(&eli->li_list_mtx); + + try_to_freeze(); + + cur = jiffies; + if (!next_wakeup_initialized || time_after_eq(cur, next_wakeup)) { + cond_resched(); + continue; + } + + schedule_timeout_interruptible(next_wakeup - cur); + + if (kthread_should_stop()) { + ext4_clear_request_list(); + goto exit_thread; + } + } + +exit_thread: + /* + * It looks like the request list is empty, but we need + * to check it under the li_list_mtx lock, to prevent any + * additions into it, and of course we should lock ext4_li_mtx + * to atomically free the list and ext4_li_info, because at + * this point another ext4 filesystem could be registering + * new one. + */ + mutex_lock(&ext4_li_mtx); + mutex_lock(&eli->li_list_mtx); + if (!list_empty(&eli->li_request_list)) { + mutex_unlock(&eli->li_list_mtx); + mutex_unlock(&ext4_li_mtx); + goto cont_thread; + } + mutex_unlock(&eli->li_list_mtx); + kfree(ext4_li_info); + ext4_li_info = NULL; + mutex_unlock(&ext4_li_mtx); + + return 0; +} + +static void ext4_clear_request_list(void) +{ + struct list_head *pos, *n; + struct ext4_li_request *elr; + + mutex_lock(&ext4_li_info->li_list_mtx); + list_for_each_safe(pos, n, &ext4_li_info->li_request_list) { + elr = list_entry(pos, struct ext4_li_request, + lr_request); + ext4_remove_li_request(elr); + } + mutex_unlock(&ext4_li_info->li_list_mtx); +} + +static int ext4_run_lazyinit_thread(void) +{ + ext4_lazyinit_task = kthread_run(ext4_lazyinit_thread, + ext4_li_info, "ext4lazyinit"); + if (IS_ERR(ext4_lazyinit_task)) { + int err = PTR_ERR(ext4_lazyinit_task); + ext4_clear_request_list(); + kfree(ext4_li_info); + ext4_li_info = NULL; + printk(KERN_CRIT "EXT4-fs: error %d creating inode table " + "initialization thread\n", + err); + return err; + } + ext4_li_info->li_state |= EXT4_LAZYINIT_RUNNING; + return 0; +} + +/* + * Check whether it make sense to run itable init. thread or not. + * If there is at least one uninitialized inode table, return + * corresponding group number, else the loop goes through all + * groups and return total number of groups. + */ +static ext4_group_t ext4_has_uninit_itable(struct super_block *sb) +{ + ext4_group_t group, ngroups = EXT4_SB(sb)->s_groups_count; + struct ext4_group_desc *gdp = NULL; + + if (!ext4_has_group_desc_csum(sb)) + return ngroups; + + for (group = 0; group < ngroups; group++) { + gdp = ext4_get_group_desc(sb, group, NULL); + if (!gdp) + continue; + + if (!(gdp->bg_flags & cpu_to_le16(EXT4_BG_INODE_ZEROED))) + break; + } + + return group; +} + +static int ext4_li_info_new(void) +{ + struct ext4_lazy_init *eli = NULL; + + eli = kzalloc(sizeof(*eli), GFP_KERNEL); + if (!eli) + return -ENOMEM; + + INIT_LIST_HEAD(&eli->li_request_list); + mutex_init(&eli->li_list_mtx); + + eli->li_state |= EXT4_LAZYINIT_QUIT; + + ext4_li_info = eli; + + return 0; +} + +static struct ext4_li_request *ext4_li_request_new(struct super_block *sb, + ext4_group_t start) +{ + struct ext4_li_request *elr; + + elr = kzalloc(sizeof(*elr), GFP_KERNEL); + if (!elr) + return NULL; + + elr->lr_super = sb; + elr->lr_first_not_zeroed = start; + if (test_opt(sb, NO_PREFETCH_BLOCK_BITMAPS)) { + elr->lr_mode = EXT4_LI_MODE_ITABLE; + elr->lr_next_group = start; + } else { + elr->lr_mode = EXT4_LI_MODE_PREFETCH_BBITMAP; + } + + /* + * Randomize first schedule time of the request to + * spread the inode table initialization requests + * better. + */ + elr->lr_next_sched = jiffies + get_random_u32_below(EXT4_DEF_LI_MAX_START_DELAY * HZ); + return elr; +} + +int ext4_register_li_request(struct super_block *sb, + ext4_group_t first_not_zeroed) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_li_request *elr = NULL; + ext4_group_t ngroups = sbi->s_groups_count; + int ret = 0; + + mutex_lock(&ext4_li_mtx); + if (sbi->s_li_request != NULL) { + /* + * Reset timeout so it can be computed again, because + * s_li_wait_mult might have changed. + */ + sbi->s_li_request->lr_timeout = 0; + goto out; + } + + if (ext4_emergency_state(sb) || sb_rdonly(sb) || + (test_opt(sb, NO_PREFETCH_BLOCK_BITMAPS) && + (first_not_zeroed == ngroups || !test_opt(sb, INIT_INODE_TABLE)))) + goto out; + + elr = ext4_li_request_new(sb, first_not_zeroed); + if (!elr) { + ret = -ENOMEM; + goto out; + } + + if (NULL == ext4_li_info) { + ret = ext4_li_info_new(); + if (ret) + goto out; + } + + mutex_lock(&ext4_li_info->li_list_mtx); + list_add(&elr->lr_request, &ext4_li_info->li_request_list); + mutex_unlock(&ext4_li_info->li_list_mtx); + + sbi->s_li_request = elr; + /* + * set elr to NULL here since it has been inserted to + * the request_list and the removal and free of it is + * handled by ext4_clear_request_list from now on. + */ + elr = NULL; + + if (!(ext4_li_info->li_state & EXT4_LAZYINIT_RUNNING)) { + ret = ext4_run_lazyinit_thread(); + if (ret) + goto out; + } +out: + mutex_unlock(&ext4_li_mtx); + if (ret) + kfree(elr); + return ret; +} + +/* + * We do not need to lock anything since this is called on + * module unload. + */ +static void ext4_destroy_lazyinit_thread(void) +{ + /* + * If thread exited earlier + * there's nothing to be done. + */ + if (!ext4_li_info || !ext4_lazyinit_task) + return; + + kthread_stop(ext4_lazyinit_task); +} + +static int set_journal_csum_feature_set(struct super_block *sb) +{ + int ret = 1; + int compat, incompat; + struct ext4_sb_info *sbi = EXT4_SB(sb); + + if (ext4_has_feature_metadata_csum(sb)) { + /* journal checksum v3 */ + compat = 0; + incompat = JBD2_FEATURE_INCOMPAT_CSUM_V3; + } else { + /* journal checksum v1 */ + compat = JBD2_FEATURE_COMPAT_CHECKSUM; + incompat = 0; + } + + jbd2_journal_clear_features(sbi->s_journal, + JBD2_FEATURE_COMPAT_CHECKSUM, 0, + JBD2_FEATURE_INCOMPAT_CSUM_V3 | + JBD2_FEATURE_INCOMPAT_CSUM_V2); + if (test_opt(sb, JOURNAL_ASYNC_COMMIT)) { + ret = jbd2_journal_set_features(sbi->s_journal, + compat, 0, + JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT | + incompat); + } else if (test_opt(sb, JOURNAL_CHECKSUM)) { + ret = jbd2_journal_set_features(sbi->s_journal, + compat, 0, + incompat); + jbd2_journal_clear_features(sbi->s_journal, 0, 0, + JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT); + } else { + jbd2_journal_clear_features(sbi->s_journal, 0, 0, + JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT); + } + + return ret; +} + +/* + * Note: calculating the overhead so we can be compatible with + * historical BSD practice is quite difficult in the face of + * clusters/bigalloc. This is because multiple metadata blocks from + * different block group can end up in the same allocation cluster. + * Calculating the exact overhead in the face of clustered allocation + * requires either O(all block bitmaps) in memory or O(number of block + * groups**2) in time. We will still calculate the superblock for + * older file systems --- and if we come across with a bigalloc file + * system with zero in s_overhead_clusters the estimate will be close to + * correct especially for very large cluster sizes --- but for newer + * file systems, it's better to calculate this figure once at mkfs + * time, and store it in the superblock. If the superblock value is + * present (even for non-bigalloc file systems), we will use it. + */ +static int count_overhead(struct super_block *sb, ext4_group_t grp, + char *buf) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_group_desc *gdp; + ext4_fsblk_t first_block, last_block, b; + ext4_group_t i, ngroups = ext4_get_groups_count(sb); + int s, j, count = 0; + int has_super = ext4_bg_has_super(sb, grp); + + if (!ext4_has_feature_bigalloc(sb)) + return (has_super + ext4_bg_num_gdb(sb, grp) + + (has_super ? le16_to_cpu(sbi->s_es->s_reserved_gdt_blocks) : 0) + + sbi->s_itb_per_group + 2); + + first_block = le32_to_cpu(sbi->s_es->s_first_data_block) + + (grp * EXT4_BLOCKS_PER_GROUP(sb)); + last_block = first_block + EXT4_BLOCKS_PER_GROUP(sb) - 1; + for (i = 0; i < ngroups; i++) { + gdp = ext4_get_group_desc(sb, i, NULL); + b = ext4_block_bitmap(sb, gdp); + if (b >= first_block && b <= last_block) { + ext4_set_bit(EXT4_B2C(sbi, b - first_block), buf); + count++; + } + b = ext4_inode_bitmap(sb, gdp); + if (b >= first_block && b <= last_block) { + ext4_set_bit(EXT4_B2C(sbi, b - first_block), buf); + count++; + } + b = ext4_inode_table(sb, gdp); + if (b >= first_block && b + sbi->s_itb_per_group <= last_block) + for (j = 0; j < sbi->s_itb_per_group; j++, b++) { + int c = EXT4_B2C(sbi, b - first_block); + ext4_set_bit(c, buf); + count++; + } + if (i != grp) + continue; + s = 0; + if (ext4_bg_has_super(sb, grp)) { + ext4_set_bit(s++, buf); + count++; + } + j = ext4_bg_num_gdb(sb, grp); + if (s + j > EXT4_BLOCKS_PER_GROUP(sb)) { + ext4_error(sb, "Invalid number of block group " + "descriptor blocks: %d", j); + j = EXT4_BLOCKS_PER_GROUP(sb) - s; + } + count += j; + for (; j > 0; j--) + ext4_set_bit(EXT4_B2C(sbi, s++), buf); + } + if (!count) + return 0; + return EXT4_CLUSTERS_PER_GROUP(sb) - + ext4_count_free(buf, EXT4_CLUSTERS_PER_GROUP(sb) / 8); +} + +/* + * Compute the overhead and stash it in sbi->s_overhead + */ +int ext4_calculate_overhead(struct super_block *sb) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_super_block *es = sbi->s_es; + struct inode *j_inode; + unsigned int j_blocks, j_inum = le32_to_cpu(es->s_journal_inum); + ext4_group_t i, ngroups = ext4_get_groups_count(sb); + ext4_fsblk_t overhead = 0; + char *buf = (char *) get_zeroed_page(GFP_NOFS); + + if (!buf) + return -ENOMEM; + + /* + * Compute the overhead (FS structures). This is constant + * for a given filesystem unless the number of block groups + * changes so we cache the previous value until it does. + */ + + /* + * All of the blocks before first_data_block are overhead + */ + overhead = EXT4_B2C(sbi, le32_to_cpu(es->s_first_data_block)); + + /* + * Add the overhead found in each block group + */ + for (i = 0; i < ngroups; i++) { + int blks; + + blks = count_overhead(sb, i, buf); + overhead += blks; + if (blks) + memset(buf, 0, PAGE_SIZE); + cond_resched(); + } + + /* + * Add the internal journal blocks whether the journal has been + * loaded or not + */ + if (sbi->s_journal && !sbi->s_journal_bdev_file) + overhead += EXT4_NUM_B2C(sbi, sbi->s_journal->j_total_len); + else if (ext4_has_feature_journal(sb) && !sbi->s_journal && j_inum) { + /* j_inum for internal journal is non-zero */ + j_inode = ext4_get_journal_inode(sb, j_inum); + if (!IS_ERR(j_inode)) { + j_blocks = j_inode->i_size >> sb->s_blocksize_bits; + overhead += EXT4_NUM_B2C(sbi, j_blocks); + iput(j_inode); + } else { + ext4_msg(sb, KERN_ERR, "can't get journal size"); + } + } + sbi->s_overhead = overhead; + smp_wmb(); + free_page((unsigned long) buf); + return 0; +} + +static void ext4_set_resv_clusters(struct super_block *sb) +{ + ext4_fsblk_t resv_clusters; + struct ext4_sb_info *sbi = EXT4_SB(sb); + + /* + * There's no need to reserve anything when we aren't using extents. + * The space estimates are exact, there are no unwritten extents, + * hole punching doesn't need new metadata... This is needed especially + * to keep ext2/3 backward compatibility. + */ + if (!ext4_has_feature_extents(sb)) + return; + /* + * By default we reserve 2% or 4096 clusters, whichever is smaller. + * This should cover the situations where we can not afford to run + * out of space like for example punch hole, or converting + * unwritten extents in delalloc path. In most cases such + * allocation would require 1, or 2 blocks, higher numbers are + * very rare. + */ + resv_clusters = (ext4_blocks_count(sbi->s_es) >> + sbi->s_cluster_bits); + + do_div(resv_clusters, 50); + resv_clusters = min_t(ext4_fsblk_t, resv_clusters, 4096); + + atomic64_set(&sbi->s_resv_clusters, resv_clusters); +} + +static const char *ext4_quota_mode(struct super_block *sb) +{ +#ifdef CONFIG_QUOTA + if (!ext4_quota_capable(sb)) + return "none"; + + if (EXT4_SB(sb)->s_journal && ext4_is_quota_journalled(sb)) + return "journalled"; + else + return "writeback"; +#else + return "disabled"; +#endif +} + +static void ext4_setup_csum_trigger(struct super_block *sb, + enum ext4_journal_trigger_type type, + void (*trigger)( + struct jbd2_buffer_trigger_type *type, + struct buffer_head *bh, + void *mapped_data, + size_t size)) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + + sbi->s_journal_triggers[type].sb = sb; + sbi->s_journal_triggers[type].tr_triggers.t_frozen = trigger; +} + +static void ext4_free_sbi(struct ext4_sb_info *sbi) +{ + if (!sbi) + return; + + kfree(sbi->s_blockgroup_lock); + fs_put_dax(sbi->s_daxdev, NULL); + kfree(sbi); +} + +static struct ext4_sb_info *ext4_alloc_sbi(struct super_block *sb) +{ + struct ext4_sb_info *sbi; + + sbi = kzalloc(sizeof(*sbi), GFP_KERNEL); + if (!sbi) + return NULL; + + sbi->s_daxdev = fs_dax_get_by_bdev(sb->s_bdev, &sbi->s_dax_part_off, + NULL, NULL); + + sbi->s_blockgroup_lock = + kzalloc(sizeof(struct blockgroup_lock), GFP_KERNEL); + + if (!sbi->s_blockgroup_lock) + goto err_out; + + sb->s_fs_info = sbi; + sbi->s_sb = sb; + return sbi; +err_out: + fs_put_dax(sbi->s_daxdev, NULL); + kfree(sbi); + return NULL; +} + +static void ext4_set_def_opts(struct super_block *sb, + struct ext4_super_block *es) +{ + unsigned long def_mount_opts; + + /* Set defaults before we parse the mount options */ + def_mount_opts = le32_to_cpu(es->s_default_mount_opts); + set_opt(sb, INIT_INODE_TABLE); + if (def_mount_opts & EXT4_DEFM_DEBUG) + set_opt(sb, DEBUG); + if (def_mount_opts & EXT4_DEFM_BSDGROUPS) + set_opt(sb, GRPID); + if (def_mount_opts & EXT4_DEFM_UID16) + set_opt(sb, NO_UID32); + /* xattr user namespace & acls are now defaulted on */ + set_opt(sb, XATTR_USER); +#ifdef CONFIG_EXT4_FS_POSIX_ACL + set_opt(sb, POSIX_ACL); +#endif + if (ext4_has_feature_fast_commit(sb)) + set_opt2(sb, JOURNAL_FAST_COMMIT); + /* don't forget to enable journal_csum when metadata_csum is enabled. */ + if (ext4_has_feature_metadata_csum(sb)) + set_opt(sb, JOURNAL_CHECKSUM); + + if ((def_mount_opts & EXT4_DEFM_JMODE) == EXT4_DEFM_JMODE_DATA) + set_opt(sb, JOURNAL_DATA); + else if ((def_mount_opts & EXT4_DEFM_JMODE) == EXT4_DEFM_JMODE_ORDERED) + set_opt(sb, ORDERED_DATA); + else if ((def_mount_opts & EXT4_DEFM_JMODE) == EXT4_DEFM_JMODE_WBACK) + set_opt(sb, WRITEBACK_DATA); + + if (le16_to_cpu(es->s_errors) == EXT4_ERRORS_PANIC) + set_opt(sb, ERRORS_PANIC); + else if (le16_to_cpu(es->s_errors) == EXT4_ERRORS_CONTINUE) + set_opt(sb, ERRORS_CONT); + else + set_opt(sb, ERRORS_RO); + /* block_validity enabled by default; disable with noblock_validity */ + set_opt(sb, BLOCK_VALIDITY); + if (def_mount_opts & EXT4_DEFM_DISCARD) + set_opt(sb, DISCARD); + + if ((def_mount_opts & EXT4_DEFM_NOBARRIER) == 0) + set_opt(sb, BARRIER); + + /* + * enable delayed allocation by default + * Use -o nodelalloc to turn it off + */ + if (!IS_EXT3_SB(sb) && !IS_EXT2_SB(sb) && + ((def_mount_opts & EXT4_DEFM_NODELALLOC) == 0)) + set_opt(sb, DELALLOC); + + if (sb->s_blocksize <= PAGE_SIZE) + set_opt(sb, DIOREAD_NOLOCK); +} + +static int ext4_handle_clustersize(struct super_block *sb) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_super_block *es = sbi->s_es; + int clustersize; + + /* Handle clustersize */ + clustersize = BLOCK_SIZE << le32_to_cpu(es->s_log_cluster_size); + if (ext4_has_feature_bigalloc(sb)) { + if (clustersize < sb->s_blocksize) { + ext4_msg(sb, KERN_ERR, + "cluster size (%d) smaller than " + "block size (%lu)", clustersize, sb->s_blocksize); + return -EINVAL; + } + sbi->s_cluster_bits = le32_to_cpu(es->s_log_cluster_size) - + le32_to_cpu(es->s_log_block_size); + } else { + if (clustersize != sb->s_blocksize) { + ext4_msg(sb, KERN_ERR, + "fragment/cluster size (%d) != " + "block size (%lu)", clustersize, sb->s_blocksize); + return -EINVAL; + } + if (sbi->s_blocks_per_group > sb->s_blocksize * 8) { + ext4_msg(sb, KERN_ERR, + "#blocks per group too big: %lu", + sbi->s_blocks_per_group); + return -EINVAL; + } + sbi->s_cluster_bits = 0; + } + sbi->s_clusters_per_group = le32_to_cpu(es->s_clusters_per_group); + if (sbi->s_clusters_per_group > sb->s_blocksize * 8) { + ext4_msg(sb, KERN_ERR, "#clusters per group too big: %lu", + sbi->s_clusters_per_group); + return -EINVAL; + } + if (sbi->s_blocks_per_group != + (sbi->s_clusters_per_group * (clustersize / sb->s_blocksize))) { + ext4_msg(sb, KERN_ERR, + "blocks per group (%lu) and clusters per group (%lu) inconsistent", + sbi->s_blocks_per_group, sbi->s_clusters_per_group); + return -EINVAL; + } + sbi->s_cluster_ratio = clustersize / sb->s_blocksize; + + /* Do we have standard group size of clustersize * 8 blocks ? */ + if (sbi->s_blocks_per_group == clustersize << 3) + set_opt2(sb, STD_GROUP_SIZE); + + return 0; +} + +/* + * ext4_atomic_write_init: Initializes filesystem min & max atomic write units. + * With non-bigalloc filesystem awu will be based upon filesystem blocksize + * & bdev awu units. + * With bigalloc it will be based upon bigalloc cluster size & bdev awu units. + * @sb: super block + */ +static void ext4_atomic_write_init(struct super_block *sb) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct block_device *bdev = sb->s_bdev; + unsigned int clustersize = EXT4_CLUSTER_SIZE(sb); + + if (!bdev_can_atomic_write(bdev)) + return; + + if (!ext4_has_feature_extents(sb)) + return; + + sbi->s_awu_min = max(sb->s_blocksize, + bdev_atomic_write_unit_min_bytes(bdev)); + sbi->s_awu_max = min(clustersize, + bdev_atomic_write_unit_max_bytes(bdev)); + if (sbi->s_awu_min && sbi->s_awu_max && + sbi->s_awu_min <= sbi->s_awu_max) { + ext4_msg(sb, KERN_NOTICE, "Supports (experimental) DIO atomic writes awu_min: %u, awu_max: %u", + sbi->s_awu_min, sbi->s_awu_max); + } else { + sbi->s_awu_min = 0; + sbi->s_awu_max = 0; + } +} + +static void ext4_fast_commit_init(struct super_block *sb) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + + /* Initialize fast commit stuff */ + atomic_set(&sbi->s_fc_subtid, 0); + INIT_LIST_HEAD(&sbi->s_fc_q[FC_Q_MAIN]); + INIT_LIST_HEAD(&sbi->s_fc_q[FC_Q_STAGING]); + INIT_LIST_HEAD(&sbi->s_fc_dentry_q[FC_Q_MAIN]); + INIT_LIST_HEAD(&sbi->s_fc_dentry_q[FC_Q_STAGING]); + sbi->s_fc_bytes = 0; + ext4_clear_mount_flag(sb, EXT4_MF_FC_INELIGIBLE); + sbi->s_fc_ineligible_tid = 0; + mutex_init(&sbi->s_fc_lock); + memset(&sbi->s_fc_stats, 0, sizeof(sbi->s_fc_stats)); + sbi->s_fc_replay_state.fc_regions = NULL; + sbi->s_fc_replay_state.fc_regions_size = 0; + sbi->s_fc_replay_state.fc_regions_used = 0; + sbi->s_fc_replay_state.fc_regions_valid = 0; + sbi->s_fc_replay_state.fc_modified_inodes = NULL; + sbi->s_fc_replay_state.fc_modified_inodes_size = 0; + sbi->s_fc_replay_state.fc_modified_inodes_used = 0; +} + +static int ext4_inode_info_init(struct super_block *sb, + struct ext4_super_block *es) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + + if (le32_to_cpu(es->s_rev_level) == EXT4_GOOD_OLD_REV) { + sbi->s_inode_size = EXT4_GOOD_OLD_INODE_SIZE; + sbi->s_first_ino = EXT4_GOOD_OLD_FIRST_INO; + } else { + sbi->s_inode_size = le16_to_cpu(es->s_inode_size); + sbi->s_first_ino = le32_to_cpu(es->s_first_ino); + if (sbi->s_first_ino < EXT4_GOOD_OLD_FIRST_INO) { + ext4_msg(sb, KERN_ERR, "invalid first ino: %u", + sbi->s_first_ino); + return -EINVAL; + } + if ((sbi->s_inode_size < EXT4_GOOD_OLD_INODE_SIZE) || + (!is_power_of_2(sbi->s_inode_size)) || + (sbi->s_inode_size > sb->s_blocksize)) { + ext4_msg(sb, KERN_ERR, + "unsupported inode size: %d", + sbi->s_inode_size); + ext4_msg(sb, KERN_ERR, "blocksize: %lu", sb->s_blocksize); + return -EINVAL; + } + /* + * i_atime_extra is the last extra field available for + * [acm]times in struct ext4_inode. Checking for that + * field should suffice to ensure we have extra space + * for all three. + */ + if (sbi->s_inode_size >= offsetof(struct ext4_inode, i_atime_extra) + + sizeof(((struct ext4_inode *)0)->i_atime_extra)) { + sb->s_time_gran = 1; + sb->s_time_max = EXT4_EXTRA_TIMESTAMP_MAX; + } else { + sb->s_time_gran = NSEC_PER_SEC; + sb->s_time_max = EXT4_NON_EXTRA_TIMESTAMP_MAX; + } + sb->s_time_min = EXT4_TIMESTAMP_MIN; + } + + if (sbi->s_inode_size > EXT4_GOOD_OLD_INODE_SIZE) { + sbi->s_want_extra_isize = sizeof(struct ext4_inode) - + EXT4_GOOD_OLD_INODE_SIZE; + if (ext4_has_feature_extra_isize(sb)) { + unsigned v, max = (sbi->s_inode_size - + EXT4_GOOD_OLD_INODE_SIZE); + + v = le16_to_cpu(es->s_want_extra_isize); + if (v > max) { + ext4_msg(sb, KERN_ERR, + "bad s_want_extra_isize: %d", v); + return -EINVAL; + } + if (sbi->s_want_extra_isize < v) + sbi->s_want_extra_isize = v; + + v = le16_to_cpu(es->s_min_extra_isize); + if (v > max) { + ext4_msg(sb, KERN_ERR, + "bad s_min_extra_isize: %d", v); + return -EINVAL; + } + if (sbi->s_want_extra_isize < v) + sbi->s_want_extra_isize = v; + } + } + + return 0; +} + +#if IS_ENABLED(CONFIG_UNICODE) +static int ext4_encoding_init(struct super_block *sb, struct ext4_super_block *es) +{ + const struct ext4_sb_encodings *encoding_info; + struct unicode_map *encoding; + __u16 encoding_flags = le16_to_cpu(es->s_encoding_flags); + + if (!ext4_has_feature_casefold(sb) || sb->s_encoding) + return 0; + + encoding_info = ext4_sb_read_encoding(es); + if (!encoding_info) { + ext4_msg(sb, KERN_ERR, + "Encoding requested by superblock is unknown"); + return -EINVAL; + } + + encoding = utf8_load(encoding_info->version); + if (IS_ERR(encoding)) { + ext4_msg(sb, KERN_ERR, + "can't mount with superblock charset: %s-%u.%u.%u " + "not supported by the kernel. flags: 0x%x.", + encoding_info->name, + unicode_major(encoding_info->version), + unicode_minor(encoding_info->version), + unicode_rev(encoding_info->version), + encoding_flags); + return -EINVAL; + } + ext4_msg(sb, KERN_INFO,"Using encoding defined by superblock: " + "%s-%u.%u.%u with flags 0x%hx", encoding_info->name, + unicode_major(encoding_info->version), + unicode_minor(encoding_info->version), + unicode_rev(encoding_info->version), + encoding_flags); + + sb->s_encoding = encoding; + sb->s_encoding_flags = encoding_flags; + + return 0; +} +#else +static inline int ext4_encoding_init(struct super_block *sb, struct ext4_super_block *es) +{ + return 0; +} +#endif + +static int ext4_init_metadata_csum(struct super_block *sb, struct ext4_super_block *es) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + + /* Warn if metadata_csum and gdt_csum are both set. */ + if (ext4_has_feature_metadata_csum(sb) && + ext4_has_feature_gdt_csum(sb)) + ext4_warning(sb, "metadata_csum and uninit_bg are " + "redundant flags; please run fsck."); + + /* Check for a known checksum algorithm */ + if (!ext4_verify_csum_type(sb, es)) { + ext4_msg(sb, KERN_ERR, "VFS: Found ext4 filesystem with " + "unknown checksum algorithm."); + return -EINVAL; + } + ext4_setup_csum_trigger(sb, EXT4_JTR_ORPHAN_FILE, + ext4_orphan_file_block_trigger); + + /* Check superblock checksum */ + if (!ext4_superblock_csum_verify(sb, es)) { + ext4_msg(sb, KERN_ERR, "VFS: Found ext4 filesystem with " + "invalid superblock checksum. Run e2fsck?"); + return -EFSBADCRC; + } + + /* Precompute checksum seed for all metadata */ + if (ext4_has_feature_csum_seed(sb)) + sbi->s_csum_seed = le32_to_cpu(es->s_checksum_seed); + else if (ext4_has_feature_metadata_csum(sb) || + ext4_has_feature_ea_inode(sb)) + sbi->s_csum_seed = ext4_chksum(~0, es->s_uuid, + sizeof(es->s_uuid)); + return 0; +} + +static int ext4_check_feature_compatibility(struct super_block *sb, + struct ext4_super_block *es, + int silent) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + + if (le32_to_cpu(es->s_rev_level) == EXT4_GOOD_OLD_REV && + (ext4_has_compat_features(sb) || + ext4_has_ro_compat_features(sb) || + ext4_has_incompat_features(sb))) + ext4_msg(sb, KERN_WARNING, + "feature flags set on rev 0 fs, " + "running e2fsck is recommended"); + + if (es->s_creator_os == cpu_to_le32(EXT4_OS_HURD)) { + set_opt2(sb, HURD_COMPAT); + if (ext4_has_feature_64bit(sb)) { + ext4_msg(sb, KERN_ERR, + "The Hurd can't support 64-bit file systems"); + return -EINVAL; + } + + /* + * ea_inode feature uses l_i_version field which is not + * available in HURD_COMPAT mode. + */ + if (ext4_has_feature_ea_inode(sb)) { + ext4_msg(sb, KERN_ERR, + "ea_inode feature is not supported for Hurd"); + return -EINVAL; + } + } + + if (IS_EXT2_SB(sb)) { + if (ext2_feature_set_ok(sb)) + ext4_msg(sb, KERN_INFO, "mounting ext2 file system " + "using the ext4 subsystem"); + else { + /* + * If we're probing be silent, if this looks like + * it's actually an ext[34] filesystem. + */ + if (silent && ext4_feature_set_ok(sb, sb_rdonly(sb))) + return -EINVAL; + ext4_msg(sb, KERN_ERR, "couldn't mount as ext2 due " + "to feature incompatibilities"); + return -EINVAL; + } + } + + if (IS_EXT3_SB(sb)) { + if (ext3_feature_set_ok(sb)) + ext4_msg(sb, KERN_INFO, "mounting ext3 file system " + "using the ext4 subsystem"); + else { + /* + * If we're probing be silent, if this looks like + * it's actually an ext4 filesystem. + */ + if (silent && ext4_feature_set_ok(sb, sb_rdonly(sb))) + return -EINVAL; + ext4_msg(sb, KERN_ERR, "couldn't mount as ext3 due " + "to feature incompatibilities"); + return -EINVAL; + } + } + + /* + * Check feature flags regardless of the revision level, since we + * previously didn't change the revision level when setting the flags, + * so there is a chance incompat flags are set on a rev 0 filesystem. + */ + if (!ext4_feature_set_ok(sb, (sb_rdonly(sb)))) + return -EINVAL; + + if (sbi->s_daxdev) { + if (sb->s_blocksize == PAGE_SIZE) + set_bit(EXT4_FLAGS_BDEV_IS_DAX, &sbi->s_ext4_flags); + else + ext4_msg(sb, KERN_ERR, "unsupported blocksize for DAX\n"); + } + + if (sbi->s_mount_opt & EXT4_MOUNT_DAX_ALWAYS) { + if (ext4_has_feature_inline_data(sb)) { + ext4_msg(sb, KERN_ERR, "Cannot use DAX on a filesystem" + " that may contain inline data"); + return -EINVAL; + } + if (!test_bit(EXT4_FLAGS_BDEV_IS_DAX, &sbi->s_ext4_flags)) { + ext4_msg(sb, KERN_ERR, + "DAX unsupported by block device."); + return -EINVAL; + } + } + + if (ext4_has_feature_encrypt(sb) && es->s_encryption_level) { + ext4_msg(sb, KERN_ERR, "Unsupported encryption level %d", + es->s_encryption_level); + return -EINVAL; + } + + return 0; +} + +static int ext4_check_geometry(struct super_block *sb, + struct ext4_super_block *es) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + __u64 blocks_count; + int err; + + if (le16_to_cpu(sbi->s_es->s_reserved_gdt_blocks) > (sb->s_blocksize / 4)) { + ext4_msg(sb, KERN_ERR, + "Number of reserved GDT blocks insanely large: %d", + le16_to_cpu(sbi->s_es->s_reserved_gdt_blocks)); + return -EINVAL; + } + /* + * Test whether we have more sectors than will fit in sector_t, + * and whether the max offset is addressable by the page cache. + */ + err = generic_check_addressable(sb->s_blocksize_bits, + ext4_blocks_count(es)); + if (err) { + ext4_msg(sb, KERN_ERR, "filesystem" + " too large to mount safely on this system"); + return err; + } + + /* check blocks count against device size */ + blocks_count = sb_bdev_nr_blocks(sb); + if (blocks_count && ext4_blocks_count(es) > blocks_count) { + ext4_msg(sb, KERN_WARNING, "bad geometry: block count %llu " + "exceeds size of device (%llu blocks)", + ext4_blocks_count(es), blocks_count); + return -EINVAL; + } + + /* + * It makes no sense for the first data block to be beyond the end + * of the filesystem. + */ + if (le32_to_cpu(es->s_first_data_block) >= ext4_blocks_count(es)) { + ext4_msg(sb, KERN_WARNING, "bad geometry: first data " + "block %u is beyond end of filesystem (%llu)", + le32_to_cpu(es->s_first_data_block), + ext4_blocks_count(es)); + return -EINVAL; + } + if ((es->s_first_data_block == 0) && (es->s_log_block_size == 0) && + (sbi->s_cluster_ratio == 1)) { + ext4_msg(sb, KERN_WARNING, "bad geometry: first data " + "block is 0 with a 1k block and cluster size"); + return -EINVAL; + } + + blocks_count = (ext4_blocks_count(es) - + le32_to_cpu(es->s_first_data_block) + + EXT4_BLOCKS_PER_GROUP(sb) - 1); + do_div(blocks_count, EXT4_BLOCKS_PER_GROUP(sb)); + if (blocks_count > ((uint64_t)1<<32) - EXT4_DESC_PER_BLOCK(sb)) { + ext4_msg(sb, KERN_WARNING, "groups count too large: %llu " + "(block count %llu, first data block %u, " + "blocks per group %lu)", blocks_count, + ext4_blocks_count(es), + le32_to_cpu(es->s_first_data_block), + EXT4_BLOCKS_PER_GROUP(sb)); + return -EINVAL; + } + sbi->s_groups_count = blocks_count; + sbi->s_blockfile_groups = min_t(ext4_group_t, sbi->s_groups_count, + (EXT4_MAX_BLOCK_FILE_PHYS / EXT4_BLOCKS_PER_GROUP(sb))); + if (((u64)sbi->s_groups_count * sbi->s_inodes_per_group) != + le32_to_cpu(es->s_inodes_count)) { + ext4_msg(sb, KERN_ERR, "inodes count not valid: %u vs %llu", + le32_to_cpu(es->s_inodes_count), + ((u64)sbi->s_groups_count * sbi->s_inodes_per_group)); + return -EINVAL; + } + + return 0; +} + +static int ext4_group_desc_init(struct super_block *sb, + struct ext4_super_block *es, + ext4_fsblk_t logical_sb_block, + ext4_group_t *first_not_zeroed) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + unsigned int db_count; + ext4_fsblk_t block; + int i; + + db_count = (sbi->s_groups_count + EXT4_DESC_PER_BLOCK(sb) - 1) / + EXT4_DESC_PER_BLOCK(sb); + if (ext4_has_feature_meta_bg(sb)) { + if (le32_to_cpu(es->s_first_meta_bg) > db_count) { + ext4_msg(sb, KERN_WARNING, + "first meta block group too large: %u " + "(group descriptor block count %u)", + le32_to_cpu(es->s_first_meta_bg), db_count); + return -EINVAL; + } + } + rcu_assign_pointer(sbi->s_group_desc, + kvmalloc_array(db_count, + sizeof(struct buffer_head *), + GFP_KERNEL)); + if (sbi->s_group_desc == NULL) { + ext4_msg(sb, KERN_ERR, "not enough memory"); + return -ENOMEM; + } + + bgl_lock_init(sbi->s_blockgroup_lock); + + /* Pre-read the descriptors into the buffer cache */ + for (i = 0; i < db_count; i++) { + block = descriptor_loc(sb, logical_sb_block, i); + ext4_sb_breadahead_unmovable(sb, block); + } + + for (i = 0; i < db_count; i++) { + struct buffer_head *bh; + + block = descriptor_loc(sb, logical_sb_block, i); + bh = ext4_sb_bread_unmovable(sb, block); + if (IS_ERR(bh)) { + ext4_msg(sb, KERN_ERR, + "can't read group descriptor %d", i); + sbi->s_gdb_count = i; + return PTR_ERR(bh); + } + rcu_read_lock(); + rcu_dereference(sbi->s_group_desc)[i] = bh; + rcu_read_unlock(); + } + sbi->s_gdb_count = db_count; + if (!ext4_check_descriptors(sb, logical_sb_block, first_not_zeroed)) { + ext4_msg(sb, KERN_ERR, "group descriptors corrupted!"); + return -EFSCORRUPTED; + } + + return 0; +} + +static int ext4_load_and_init_journal(struct super_block *sb, + struct ext4_super_block *es, + struct ext4_fs_context *ctx) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + int err; + + err = ext4_load_journal(sb, es, ctx->journal_devnum); + if (err) + return err; + + if (ext4_has_feature_64bit(sb) && + !jbd2_journal_set_features(EXT4_SB(sb)->s_journal, 0, 0, + JBD2_FEATURE_INCOMPAT_64BIT)) { + ext4_msg(sb, KERN_ERR, "Failed to set 64-bit journal feature"); + goto out; + } + + if (!set_journal_csum_feature_set(sb)) { + ext4_msg(sb, KERN_ERR, "Failed to set journal checksum " + "feature set"); + goto out; + } + + if (test_opt2(sb, JOURNAL_FAST_COMMIT) && + !jbd2_journal_set_features(EXT4_SB(sb)->s_journal, 0, 0, + JBD2_FEATURE_INCOMPAT_FAST_COMMIT)) { + ext4_msg(sb, KERN_ERR, + "Failed to set fast commit journal feature"); + goto out; + } + + /* We have now updated the journal if required, so we can + * validate the data journaling mode. */ + switch (test_opt(sb, DATA_FLAGS)) { + case 0: + /* No mode set, assume a default based on the journal + * capabilities: ORDERED_DATA if the journal can + * cope, else JOURNAL_DATA + */ + if (jbd2_journal_check_available_features + (sbi->s_journal, 0, 0, JBD2_FEATURE_INCOMPAT_REVOKE)) { + set_opt(sb, ORDERED_DATA); + sbi->s_def_mount_opt |= EXT4_MOUNT_ORDERED_DATA; + } else { + set_opt(sb, JOURNAL_DATA); + sbi->s_def_mount_opt |= EXT4_MOUNT_JOURNAL_DATA; + } + break; + + case EXT4_MOUNT_ORDERED_DATA: + case EXT4_MOUNT_WRITEBACK_DATA: + if (!jbd2_journal_check_available_features + (sbi->s_journal, 0, 0, JBD2_FEATURE_INCOMPAT_REVOKE)) { + ext4_msg(sb, KERN_ERR, "Journal does not support " + "requested data journaling mode"); + goto out; + } + break; + default: + break; + } + + if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_ORDERED_DATA && + test_opt(sb, JOURNAL_ASYNC_COMMIT)) { + ext4_msg(sb, KERN_ERR, "can't mount with " + "journal_async_commit in data=ordered mode"); + goto out; + } + + set_task_ioprio(sbi->s_journal->j_task, ctx->journal_ioprio); + + sbi->s_journal->j_submit_inode_data_buffers = + ext4_journal_submit_inode_data_buffers; + sbi->s_journal->j_finish_inode_data_buffers = + ext4_journal_finish_inode_data_buffers; + + return 0; + +out: + ext4_journal_destroy(sbi, sbi->s_journal); + return -EINVAL; +} + +static int ext4_check_journal_data_mode(struct super_block *sb) +{ + if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA) { + printk_once(KERN_WARNING "EXT4-fs: Warning: mounting with " + "data=journal disables delayed allocation, " + "dioread_nolock, O_DIRECT and fast_commit support!\n"); + /* can't mount with both data=journal and dioread_nolock. */ + clear_opt(sb, DIOREAD_NOLOCK); + clear_opt2(sb, JOURNAL_FAST_COMMIT); + if (test_opt2(sb, EXPLICIT_DELALLOC)) { + ext4_msg(sb, KERN_ERR, "can't mount with " + "both data=journal and delalloc"); + return -EINVAL; + } + if (test_opt(sb, DAX_ALWAYS)) { + ext4_msg(sb, KERN_ERR, "can't mount with " + "both data=journal and dax"); + return -EINVAL; + } + if (ext4_has_feature_encrypt(sb)) { + ext4_msg(sb, KERN_WARNING, + "encrypted files will use data=ordered " + "instead of data journaling mode"); + } + if (test_opt(sb, DELALLOC)) + clear_opt(sb, DELALLOC); + } else { + sb->s_iflags |= SB_I_CGROUPWB; + } + + return 0; +} + +static const char *ext4_has_journal_option(struct super_block *sb) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + + if (test_opt(sb, JOURNAL_ASYNC_COMMIT)) + return "journal_async_commit"; + if (test_opt2(sb, EXPLICIT_JOURNAL_CHECKSUM)) + return "journal_checksum"; + if (sbi->s_commit_interval != JBD2_DEFAULT_MAX_COMMIT_AGE*HZ) + return "commit="; + if (EXT4_MOUNT_DATA_FLAGS & + (sbi->s_mount_opt ^ sbi->s_def_mount_opt)) + return "data="; + if (test_opt(sb, DATA_ERR_ABORT)) + return "data_err=abort"; + return NULL; +} + +static int ext4_load_super(struct super_block *sb, ext4_fsblk_t *lsb, + int silent) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_super_block *es; + ext4_fsblk_t logical_sb_block; + unsigned long offset = 0; + struct buffer_head *bh; + int ret = -EINVAL; + int blocksize; + + blocksize = sb_min_blocksize(sb, EXT4_MIN_BLOCK_SIZE); + if (!blocksize) { + ext4_msg(sb, KERN_ERR, "unable to set blocksize"); + return -EINVAL; + } + + /* + * The ext4 superblock will not be buffer aligned for other than 1kB + * block sizes. We need to calculate the offset from buffer start. + */ + if (blocksize != EXT4_MIN_BLOCK_SIZE) { + logical_sb_block = sbi->s_sb_block * EXT4_MIN_BLOCK_SIZE; + offset = do_div(logical_sb_block, blocksize); + } else { + logical_sb_block = sbi->s_sb_block; + } + + bh = ext4_sb_bread_unmovable(sb, logical_sb_block); + if (IS_ERR(bh)) { + ext4_msg(sb, KERN_ERR, "unable to read superblock"); + return PTR_ERR(bh); + } + /* + * Note: s_es must be initialized as soon as possible because + * some ext4 macro-instructions depend on its value + */ + es = (struct ext4_super_block *) (bh->b_data + offset); + sbi->s_es = es; + sb->s_magic = le16_to_cpu(es->s_magic); + if (sb->s_magic != EXT4_SUPER_MAGIC) { + if (!silent) + ext4_msg(sb, KERN_ERR, "VFS: Can't find ext4 filesystem"); + goto out; + } + + if (le32_to_cpu(es->s_log_block_size) > + (EXT4_MAX_BLOCK_LOG_SIZE - EXT4_MIN_BLOCK_LOG_SIZE)) { + ext4_msg(sb, KERN_ERR, + "Invalid log block size: %u", + le32_to_cpu(es->s_log_block_size)); + goto out; + } + if (le32_to_cpu(es->s_log_cluster_size) > + (EXT4_MAX_CLUSTER_LOG_SIZE - EXT4_MIN_BLOCK_LOG_SIZE)) { + ext4_msg(sb, KERN_ERR, + "Invalid log cluster size: %u", + le32_to_cpu(es->s_log_cluster_size)); + goto out; + } + + blocksize = EXT4_MIN_BLOCK_SIZE << le32_to_cpu(es->s_log_block_size); + + /* + * If the default block size is not the same as the real block size, + * we need to reload it. + */ + if (sb->s_blocksize == blocksize) { + *lsb = logical_sb_block; + sbi->s_sbh = bh; + return 0; + } + + /* + * bh must be released before kill_bdev(), otherwise + * it won't be freed and its page also. kill_bdev() + * is called by sb_set_blocksize(). + */ + brelse(bh); + /* Validate the filesystem blocksize */ + if (!sb_set_blocksize(sb, blocksize)) { + ext4_msg(sb, KERN_ERR, "bad block size %d", + blocksize); + bh = NULL; + goto out; + } + + logical_sb_block = sbi->s_sb_block * EXT4_MIN_BLOCK_SIZE; + offset = do_div(logical_sb_block, blocksize); + bh = ext4_sb_bread_unmovable(sb, logical_sb_block); + if (IS_ERR(bh)) { + ext4_msg(sb, KERN_ERR, "Can't read superblock on 2nd try"); + ret = PTR_ERR(bh); + bh = NULL; + goto out; + } + es = (struct ext4_super_block *)(bh->b_data + offset); + sbi->s_es = es; + if (es->s_magic != cpu_to_le16(EXT4_SUPER_MAGIC)) { + ext4_msg(sb, KERN_ERR, "Magic mismatch, very weird!"); + goto out; + } + *lsb = logical_sb_block; + sbi->s_sbh = bh; + return 0; +out: + brelse(bh); + return ret; +} + +static int ext4_hash_info_init(struct super_block *sb) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_super_block *es = sbi->s_es; + unsigned int i; + + sbi->s_def_hash_version = es->s_def_hash_version; + + if (sbi->s_def_hash_version > DX_HASH_LAST) { + ext4_msg(sb, KERN_ERR, + "Invalid default hash set in the superblock"); + return -EINVAL; + } else if (sbi->s_def_hash_version == DX_HASH_SIPHASH) { + ext4_msg(sb, KERN_ERR, + "SIPHASH is not a valid default hash value"); + return -EINVAL; + } + + for (i = 0; i < 4; i++) + sbi->s_hash_seed[i] = le32_to_cpu(es->s_hash_seed[i]); + + if (ext4_has_feature_dir_index(sb)) { + i = le32_to_cpu(es->s_flags); + if (i & EXT2_FLAGS_UNSIGNED_HASH) + sbi->s_hash_unsigned = 3; + else if ((i & EXT2_FLAGS_SIGNED_HASH) == 0) { +#ifdef __CHAR_UNSIGNED__ + if (!sb_rdonly(sb)) + es->s_flags |= + cpu_to_le32(EXT2_FLAGS_UNSIGNED_HASH); + sbi->s_hash_unsigned = 3; +#else + if (!sb_rdonly(sb)) + es->s_flags |= + cpu_to_le32(EXT2_FLAGS_SIGNED_HASH); +#endif + } + } + return 0; +} + +static int ext4_block_group_meta_init(struct super_block *sb, int silent) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_super_block *es = sbi->s_es; + int has_huge_files; + + has_huge_files = ext4_has_feature_huge_file(sb); + sbi->s_bitmap_maxbytes = ext4_max_bitmap_size(sb->s_blocksize_bits, + has_huge_files); + sb->s_maxbytes = ext4_max_size(sb->s_blocksize_bits, has_huge_files); + + sbi->s_desc_size = le16_to_cpu(es->s_desc_size); + if (ext4_has_feature_64bit(sb)) { + if (sbi->s_desc_size < EXT4_MIN_DESC_SIZE_64BIT || + sbi->s_desc_size > EXT4_MAX_DESC_SIZE || + !is_power_of_2(sbi->s_desc_size)) { + ext4_msg(sb, KERN_ERR, + "unsupported descriptor size %lu", + sbi->s_desc_size); + return -EINVAL; + } + } else + sbi->s_desc_size = EXT4_MIN_DESC_SIZE; + + sbi->s_blocks_per_group = le32_to_cpu(es->s_blocks_per_group); + sbi->s_inodes_per_group = le32_to_cpu(es->s_inodes_per_group); + + sbi->s_inodes_per_block = sb->s_blocksize / EXT4_INODE_SIZE(sb); + if (sbi->s_inodes_per_block == 0 || sbi->s_blocks_per_group == 0) { + if (!silent) + ext4_msg(sb, KERN_ERR, "VFS: Can't find ext4 filesystem"); + return -EINVAL; + } + if (sbi->s_inodes_per_group < sbi->s_inodes_per_block || + sbi->s_inodes_per_group > sb->s_blocksize * 8) { + ext4_msg(sb, KERN_ERR, "invalid inodes per group: %lu\n", + sbi->s_inodes_per_group); + return -EINVAL; + } + sbi->s_itb_per_group = sbi->s_inodes_per_group / + sbi->s_inodes_per_block; + sbi->s_desc_per_block = sb->s_blocksize / EXT4_DESC_SIZE(sb); + sbi->s_mount_state = le16_to_cpu(es->s_state) & ~EXT4_FC_REPLAY; + sbi->s_addr_per_block_bits = ilog2(EXT4_ADDR_PER_BLOCK(sb)); + sbi->s_desc_per_block_bits = ilog2(EXT4_DESC_PER_BLOCK(sb)); + + return 0; +} + +/* + * It's hard to get stripe aligned blocks if stripe is not aligned with + * cluster, just disable stripe and alert user to simplify code and avoid + * stripe aligned allocation which will rarely succeed. + */ +static bool ext4_is_stripe_incompatible(struct super_block *sb, unsigned long stripe) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + return (stripe > 0 && sbi->s_cluster_ratio > 1 && + stripe % sbi->s_cluster_ratio != 0); +} + +static int __ext4_fill_super(struct fs_context *fc, struct super_block *sb) +{ + struct ext4_super_block *es = NULL; + struct ext4_sb_info *sbi = EXT4_SB(sb); + ext4_fsblk_t logical_sb_block; + struct inode *root; + int needs_recovery; + int err; + ext4_group_t first_not_zeroed; + struct ext4_fs_context *ctx = fc->fs_private; + int silent = fc->sb_flags & SB_SILENT; + + /* Set defaults for the variables that will be set during parsing */ + if (!(ctx->spec & EXT4_SPEC_JOURNAL_IOPRIO)) + ctx->journal_ioprio = EXT4_DEF_JOURNAL_IOPRIO; + + sbi->s_inode_readahead_blks = EXT4_DEF_INODE_READAHEAD_BLKS; + sbi->s_sectors_written_start = + part_stat_read(sb->s_bdev, sectors[STAT_WRITE]); + + err = ext4_load_super(sb, &logical_sb_block, silent); + if (err) + goto out_fail; + + es = sbi->s_es; + sbi->s_kbytes_written = le64_to_cpu(es->s_kbytes_written); + + err = ext4_init_metadata_csum(sb, es); + if (err) + goto failed_mount; + + ext4_set_def_opts(sb, es); + + sbi->s_resuid = make_kuid(&init_user_ns, ext4_get_resuid(es)); + sbi->s_resgid = make_kgid(&init_user_ns, ext4_get_resuid(es)); + sbi->s_commit_interval = JBD2_DEFAULT_MAX_COMMIT_AGE * HZ; + sbi->s_min_batch_time = EXT4_DEF_MIN_BATCH_TIME; + sbi->s_max_batch_time = EXT4_DEF_MAX_BATCH_TIME; + sbi->s_sb_update_kb = EXT4_DEF_SB_UPDATE_INTERVAL_KB; + sbi->s_sb_update_sec = EXT4_DEF_SB_UPDATE_INTERVAL_SEC; + + /* + * set default s_li_wait_mult for lazyinit, for the case there is + * no mount option specified. + */ + sbi->s_li_wait_mult = EXT4_DEF_LI_WAIT_MULT; + + err = ext4_inode_info_init(sb, es); + if (err) + goto failed_mount; + + err = parse_apply_sb_mount_options(sb, ctx); + if (err < 0) + goto failed_mount; + + sbi->s_def_mount_opt = sbi->s_mount_opt; + sbi->s_def_mount_opt2 = sbi->s_mount_opt2; + + err = ext4_check_opt_consistency(fc, sb); + if (err < 0) + goto failed_mount; + + ext4_apply_options(fc, sb); + + err = ext4_encoding_init(sb, es); + if (err) + goto failed_mount; + + err = ext4_check_journal_data_mode(sb); + if (err) + goto failed_mount; + + sb->s_flags = (sb->s_flags & ~SB_POSIXACL) | + (test_opt(sb, POSIX_ACL) ? SB_POSIXACL : 0); + + /* HSM events are allowed by default. */ + sb->s_iflags |= SB_I_ALLOW_HSM; + + err = ext4_check_feature_compatibility(sb, es, silent); + if (err) + goto failed_mount; + + err = ext4_block_group_meta_init(sb, silent); + if (err) + goto failed_mount; + + err = ext4_hash_info_init(sb); + if (err) + goto failed_mount; + + err = ext4_handle_clustersize(sb); + if (err) + goto failed_mount; + + err = ext4_check_geometry(sb, es); + if (err) + goto failed_mount; + + timer_setup(&sbi->s_err_report, print_daily_error_info, 0); + spin_lock_init(&sbi->s_error_lock); + INIT_WORK(&sbi->s_sb_upd_work, update_super_work); + + err = ext4_group_desc_init(sb, es, logical_sb_block, &first_not_zeroed); + if (err) + goto failed_mount3; + + err = ext4_es_register_shrinker(sbi); + if (err) + goto failed_mount3; + + sbi->s_stripe = ext4_get_stripe_size(sbi); + if (ext4_is_stripe_incompatible(sb, sbi->s_stripe)) { + ext4_msg(sb, KERN_WARNING, + "stripe (%lu) is not aligned with cluster size (%u), " + "stripe is disabled", + sbi->s_stripe, sbi->s_cluster_ratio); + sbi->s_stripe = 0; + } + sbi->s_extent_max_zeroout_kb = 32; + + /* + * set up enough so that it can read an inode + */ + sb->s_op = &ext4_sops; + sb->s_export_op = &ext4_export_ops; + sb->s_xattr = ext4_xattr_handlers; +#ifdef CONFIG_FS_ENCRYPTION + sb->s_cop = &ext4_cryptops; +#endif +#ifdef CONFIG_FS_VERITY + sb->s_vop = &ext4_verityops; +#endif +#ifdef CONFIG_QUOTA + sb->dq_op = &ext4_quota_operations; + if (ext4_has_feature_quota(sb)) + sb->s_qcop = &dquot_quotactl_sysfile_ops; + else + sb->s_qcop = &ext4_qctl_operations; + sb->s_quota_types = QTYPE_MASK_USR | QTYPE_MASK_GRP | QTYPE_MASK_PRJ; +#endif + super_set_uuid(sb, es->s_uuid, sizeof(es->s_uuid)); + super_set_sysfs_name_bdev(sb); + + INIT_LIST_HEAD(&sbi->s_orphan); /* unlinked but open files */ + mutex_init(&sbi->s_orphan_lock); + + spin_lock_init(&sbi->s_bdev_wb_lock); + + ext4_atomic_write_init(sb); + ext4_fast_commit_init(sb); + + sb->s_root = NULL; + + needs_recovery = (es->s_last_orphan != 0 || + ext4_has_feature_orphan_present(sb) || + ext4_has_feature_journal_needs_recovery(sb)); + + if (ext4_has_feature_mmp(sb) && !sb_rdonly(sb)) { + err = ext4_multi_mount_protect(sb, le64_to_cpu(es->s_mmp_block)); + if (err) + goto failed_mount3a; + } + + err = -EINVAL; + /* + * The first inode we look at is the journal inode. Don't try + * root first: it may be modified in the journal! + */ + if (!test_opt(sb, NOLOAD) && ext4_has_feature_journal(sb)) { + err = ext4_load_and_init_journal(sb, es, ctx); + if (err) + goto failed_mount3a; + if (bdev_read_only(sb->s_bdev)) + needs_recovery = 0; + } else if (test_opt(sb, NOLOAD) && !sb_rdonly(sb) && + ext4_has_feature_journal_needs_recovery(sb)) { + ext4_msg(sb, KERN_ERR, "required journal recovery " + "suppressed and not mounted read-only"); + goto failed_mount3a; + } else { + const char *journal_option; + + /* Nojournal mode, all journal mount options are illegal */ + journal_option = ext4_has_journal_option(sb); + if (journal_option != NULL) { + ext4_msg(sb, KERN_ERR, + "can't mount with %s, fs mounted w/o journal", + journal_option); + goto failed_mount3a; + } + + sbi->s_def_mount_opt &= ~EXT4_MOUNT_JOURNAL_CHECKSUM; + clear_opt(sb, JOURNAL_CHECKSUM); + clear_opt(sb, DATA_FLAGS); + clear_opt2(sb, JOURNAL_FAST_COMMIT); + sbi->s_journal = NULL; + needs_recovery = 0; + } + + if (!test_opt(sb, NO_MBCACHE)) { + sbi->s_ea_block_cache = ext4_xattr_create_cache(); + if (!sbi->s_ea_block_cache) { + ext4_msg(sb, KERN_ERR, + "Failed to create ea_block_cache"); + err = -EINVAL; + goto failed_mount_wq; + } + + if (ext4_has_feature_ea_inode(sb)) { + sbi->s_ea_inode_cache = ext4_xattr_create_cache(); + if (!sbi->s_ea_inode_cache) { + ext4_msg(sb, KERN_ERR, + "Failed to create ea_inode_cache"); + err = -EINVAL; + goto failed_mount_wq; + } + } + } + + /* + * Get the # of file system overhead blocks from the + * superblock if present. + */ + sbi->s_overhead = le32_to_cpu(es->s_overhead_clusters); + /* ignore the precalculated value if it is ridiculous */ + if (sbi->s_overhead > ext4_blocks_count(es)) + sbi->s_overhead = 0; + /* + * If the bigalloc feature is not enabled recalculating the + * overhead doesn't take long, so we might as well just redo + * it to make sure we are using the correct value. + */ + if (!ext4_has_feature_bigalloc(sb)) + sbi->s_overhead = 0; + if (sbi->s_overhead == 0) { + err = ext4_calculate_overhead(sb); + if (err) + goto failed_mount_wq; + } + + /* + * The maximum number of concurrent works can be high and + * concurrency isn't really necessary. Limit it to 1. + */ + EXT4_SB(sb)->rsv_conversion_wq = + alloc_workqueue("ext4-rsv-conversion", WQ_MEM_RECLAIM | WQ_UNBOUND, 1); + if (!EXT4_SB(sb)->rsv_conversion_wq) { + printk(KERN_ERR "EXT4-fs: failed to create workqueue\n"); + err = -ENOMEM; + goto failed_mount4; + } + + /* + * The jbd2_journal_load will have done any necessary log recovery, + * so we can safely mount the rest of the filesystem now. + */ + + root = ext4_iget(sb, EXT4_ROOT_INO, EXT4_IGET_SPECIAL); + if (IS_ERR(root)) { + ext4_msg(sb, KERN_ERR, "get root inode failed"); + err = PTR_ERR(root); + root = NULL; + goto failed_mount4; + } + if (!S_ISDIR(root->i_mode) || !root->i_blocks || !root->i_size) { + ext4_msg(sb, KERN_ERR, "corrupt root inode, run e2fsck"); + iput(root); + err = -EFSCORRUPTED; + goto failed_mount4; + } + + generic_set_sb_d_ops(sb); + sb->s_root = d_make_root(root); + if (!sb->s_root) { + ext4_msg(sb, KERN_ERR, "get root dentry failed"); + err = -ENOMEM; + goto failed_mount4; + } + + err = ext4_setup_super(sb, es, sb_rdonly(sb)); + if (err == -EROFS) { + sb->s_flags |= SB_RDONLY; + } else if (err) + goto failed_mount4a; + + ext4_set_resv_clusters(sb); + + if (test_opt(sb, BLOCK_VALIDITY)) { + err = ext4_setup_system_zone(sb); + if (err) { + ext4_msg(sb, KERN_ERR, "failed to initialize system " + "zone (%d)", err); + goto failed_mount4a; + } + } + ext4_fc_replay_cleanup(sb); + + ext4_ext_init(sb); + + /* + * Enable optimize_scan if number of groups is > threshold. This can be + * turned off by passing "mb_optimize_scan=0". This can also be + * turned on forcefully by passing "mb_optimize_scan=1". + */ + if (!(ctx->spec & EXT4_SPEC_mb_optimize_scan)) { + if (sbi->s_groups_count >= MB_DEFAULT_LINEAR_SCAN_THRESHOLD) + set_opt2(sb, MB_OPTIMIZE_SCAN); + else + clear_opt2(sb, MB_OPTIMIZE_SCAN); + } + + err = ext4_mb_init(sb); + if (err) { + ext4_msg(sb, KERN_ERR, "failed to initialize mballoc (%d)", + err); + goto failed_mount5; + } + + /* + * We can only set up the journal commit callback once + * mballoc is initialized + */ + if (sbi->s_journal) + sbi->s_journal->j_commit_callback = + ext4_journal_commit_callback; + + err = ext4_percpu_param_init(sbi); + if (err) + goto failed_mount6; + + if (ext4_has_feature_flex_bg(sb)) + if (!ext4_fill_flex_info(sb)) { + ext4_msg(sb, KERN_ERR, + "unable to initialize " + "flex_bg meta info!"); + err = -ENOMEM; + goto failed_mount6; + } + + err = ext4_register_li_request(sb, first_not_zeroed); + if (err) + goto failed_mount6; + + err = ext4_init_orphan_info(sb); + if (err) + goto failed_mount7; +#ifdef CONFIG_QUOTA + /* Enable quota usage during mount. */ + if (ext4_has_feature_quota(sb) && !sb_rdonly(sb)) { + err = ext4_enable_quotas(sb); + if (err) + goto failed_mount8; + } +#endif /* CONFIG_QUOTA */ + + /* + * Save the original bdev mapping's wb_err value which could be + * used to detect the metadata async write error. + */ + errseq_check_and_advance(&sb->s_bdev->bd_mapping->wb_err, + &sbi->s_bdev_wb_err); + EXT4_SB(sb)->s_mount_state |= EXT4_ORPHAN_FS; + ext4_orphan_cleanup(sb, es); + EXT4_SB(sb)->s_mount_state &= ~EXT4_ORPHAN_FS; + /* + * Update the checksum after updating free space/inode counters and + * ext4_orphan_cleanup. Otherwise the superblock can have an incorrect + * checksum in the buffer cache until it is written out and + * e2fsprogs programs trying to open a file system immediately + * after it is mounted can fail. + */ + ext4_superblock_csum_set(sb); + if (needs_recovery) { + ext4_msg(sb, KERN_INFO, "recovery complete"); + err = ext4_mark_recovery_complete(sb, es); + if (err) + goto failed_mount9; + } + + if (test_opt(sb, DISCARD) && !bdev_max_discard_sectors(sb->s_bdev)) { + ext4_msg(sb, KERN_WARNING, + "mounting with \"discard\" option, but the device does not support discard"); + clear_opt(sb, DISCARD); + } + + if (es->s_error_count) + mod_timer(&sbi->s_err_report, jiffies + 300*HZ); /* 5 minutes */ + + /* Enable message ratelimiting. Default is 10 messages per 5 secs. */ + ratelimit_state_init(&sbi->s_err_ratelimit_state, 5 * HZ, 10); + ratelimit_state_init(&sbi->s_warning_ratelimit_state, 5 * HZ, 10); + ratelimit_state_init(&sbi->s_msg_ratelimit_state, 5 * HZ, 10); + atomic_set(&sbi->s_warning_count, 0); + atomic_set(&sbi->s_msg_count, 0); + + /* Register sysfs after all initializations are complete. */ + err = ext4_register_sysfs(sb); + if (err) + goto failed_mount9; + + return 0; + +failed_mount9: + ext4_quotas_off(sb, EXT4_MAXQUOTAS); +failed_mount8: __maybe_unused + ext4_release_orphan_info(sb); +failed_mount7: + ext4_unregister_li_request(sb); +failed_mount6: + ext4_mb_release(sb); + ext4_flex_groups_free(sbi); + ext4_percpu_param_destroy(sbi); +failed_mount5: + ext4_ext_release(sb); + ext4_release_system_zone(sb); +failed_mount4a: + dput(sb->s_root); + sb->s_root = NULL; +failed_mount4: + ext4_msg(sb, KERN_ERR, "mount failed"); + if (EXT4_SB(sb)->rsv_conversion_wq) + destroy_workqueue(EXT4_SB(sb)->rsv_conversion_wq); +failed_mount_wq: + ext4_xattr_destroy_cache(sbi->s_ea_inode_cache); + sbi->s_ea_inode_cache = NULL; + + ext4_xattr_destroy_cache(sbi->s_ea_block_cache); + sbi->s_ea_block_cache = NULL; + + if (sbi->s_journal) { + ext4_journal_destroy(sbi, sbi->s_journal); + } +failed_mount3a: + ext4_es_unregister_shrinker(sbi); +failed_mount3: + /* flush s_sb_upd_work before sbi destroy */ + flush_work(&sbi->s_sb_upd_work); + ext4_stop_mmpd(sbi); + timer_delete_sync(&sbi->s_err_report); + ext4_group_desc_free(sbi); +failed_mount: +#if IS_ENABLED(CONFIG_UNICODE) + utf8_unload(sb->s_encoding); +#endif + +#ifdef CONFIG_QUOTA + for (unsigned int i = 0; i < EXT4_MAXQUOTAS; i++) + kfree(get_qf_name(sb, sbi, i)); +#endif + fscrypt_free_dummy_policy(&sbi->s_dummy_enc_policy); + brelse(sbi->s_sbh); + if (sbi->s_journal_bdev_file) { + invalidate_bdev(file_bdev(sbi->s_journal_bdev_file)); + bdev_fput(sbi->s_journal_bdev_file); + } +out_fail: + invalidate_bdev(sb->s_bdev); + sb->s_fs_info = NULL; + return err; +} + +static int ext4_fill_super(struct super_block *sb, struct fs_context *fc) +{ + struct ext4_fs_context *ctx = fc->fs_private; + struct ext4_sb_info *sbi; + const char *descr; + int ret; + + sbi = ext4_alloc_sbi(sb); + if (!sbi) + return -ENOMEM; + + fc->s_fs_info = sbi; + + /* Cleanup superblock name */ + strreplace(sb->s_id, '/', '!'); + + sbi->s_sb_block = 1; /* Default super block location */ + if (ctx->spec & EXT4_SPEC_s_sb_block) + sbi->s_sb_block = ctx->s_sb_block; + + ret = __ext4_fill_super(fc, sb); + if (ret < 0) + goto free_sbi; + + if (sbi->s_journal) { + if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA) + descr = " journalled data mode"; + else if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_ORDERED_DATA) + descr = " ordered data mode"; + else + descr = " writeback data mode"; + } else + descr = "out journal"; + + if (___ratelimit(&ext4_mount_msg_ratelimit, "EXT4-fs mount")) + ext4_msg(sb, KERN_INFO, "mounted filesystem %pU %s with%s. " + "Quota mode: %s.", &sb->s_uuid, + sb_rdonly(sb) ? "ro" : "r/w", descr, + ext4_quota_mode(sb)); + + /* Update the s_overhead_clusters if necessary */ + ext4_update_overhead(sb, false); + return 0; + +free_sbi: + ext4_free_sbi(sbi); + fc->s_fs_info = NULL; + return ret; +} + +static int ext4_get_tree(struct fs_context *fc) +{ + return get_tree_bdev(fc, ext4_fill_super); +} + +/* + * Setup any per-fs journal parameters now. We'll do this both on + * initial mount, once the journal has been initialised but before we've + * done any recovery; and again on any subsequent remount. + */ +static void ext4_init_journal_params(struct super_block *sb, journal_t *journal) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + + journal->j_commit_interval = sbi->s_commit_interval; + journal->j_min_batch_time = sbi->s_min_batch_time; + journal->j_max_batch_time = sbi->s_max_batch_time; + ext4_fc_init(sb, journal); + + write_lock(&journal->j_state_lock); + if (test_opt(sb, BARRIER)) + journal->j_flags |= JBD2_BARRIER; + else + journal->j_flags &= ~JBD2_BARRIER; + /* + * Always enable journal cycle record option, letting the journal + * records log transactions continuously between each mount. + */ + journal->j_flags |= JBD2_CYCLE_RECORD; + write_unlock(&journal->j_state_lock); +} + +static struct inode *ext4_get_journal_inode(struct super_block *sb, + unsigned int journal_inum) +{ + struct inode *journal_inode; + + /* + * Test for the existence of a valid inode on disk. Bad things + * happen if we iget() an unused inode, as the subsequent iput() + * will try to delete it. + */ + journal_inode = ext4_iget(sb, journal_inum, EXT4_IGET_SPECIAL); + if (IS_ERR(journal_inode)) { + ext4_msg(sb, KERN_ERR, "no journal found"); + return ERR_CAST(journal_inode); + } + if (!journal_inode->i_nlink) { + make_bad_inode(journal_inode); + iput(journal_inode); + ext4_msg(sb, KERN_ERR, "journal inode is deleted"); + return ERR_PTR(-EFSCORRUPTED); + } + if (!S_ISREG(journal_inode->i_mode) || IS_ENCRYPTED(journal_inode)) { + ext4_msg(sb, KERN_ERR, "invalid journal inode"); + iput(journal_inode); + return ERR_PTR(-EFSCORRUPTED); + } + + ext4_debug("Journal inode found at %p: %lld bytes\n", + journal_inode, journal_inode->i_size); + return journal_inode; +} + +static int ext4_journal_bmap(journal_t *journal, sector_t *block) +{ + struct ext4_map_blocks map; + int ret; + + if (journal->j_inode == NULL) + return 0; + + map.m_lblk = *block; + map.m_len = 1; + ret = ext4_map_blocks(NULL, journal->j_inode, &map, 0); + if (ret <= 0) { + ext4_msg(journal->j_inode->i_sb, KERN_CRIT, + "journal bmap failed: block %llu ret %d\n", + *block, ret); + jbd2_journal_abort(journal, ret ? ret : -EIO); + return ret; + } + *block = map.m_pblk; + return 0; +} + +static journal_t *ext4_open_inode_journal(struct super_block *sb, + unsigned int journal_inum) +{ + struct inode *journal_inode; + journal_t *journal; + + journal_inode = ext4_get_journal_inode(sb, journal_inum); + if (IS_ERR(journal_inode)) + return ERR_CAST(journal_inode); + + journal = jbd2_journal_init_inode(journal_inode); + if (IS_ERR(journal)) { + ext4_msg(sb, KERN_ERR, "Could not load journal inode"); + iput(journal_inode); + return ERR_CAST(journal); + } + journal->j_private = sb; + journal->j_bmap = ext4_journal_bmap; + ext4_init_journal_params(sb, journal); + return journal; +} + +static struct file *ext4_get_journal_blkdev(struct super_block *sb, + dev_t j_dev, ext4_fsblk_t *j_start, + ext4_fsblk_t *j_len) +{ + struct buffer_head *bh; + struct block_device *bdev; + struct file *bdev_file; + int hblock, blocksize; + ext4_fsblk_t sb_block; + unsigned long offset; + struct ext4_super_block *es; + int errno; + + bdev_file = bdev_file_open_by_dev(j_dev, + BLK_OPEN_READ | BLK_OPEN_WRITE | BLK_OPEN_RESTRICT_WRITES, + sb, &fs_holder_ops); + if (IS_ERR(bdev_file)) { + ext4_msg(sb, KERN_ERR, + "failed to open journal device unknown-block(%u,%u) %ld", + MAJOR(j_dev), MINOR(j_dev), PTR_ERR(bdev_file)); + return bdev_file; + } + + bdev = file_bdev(bdev_file); + blocksize = sb->s_blocksize; + hblock = bdev_logical_block_size(bdev); + if (blocksize < hblock) { + ext4_msg(sb, KERN_ERR, + "blocksize too small for journal device"); + errno = -EINVAL; + goto out_bdev; + } + + sb_block = EXT4_MIN_BLOCK_SIZE / blocksize; + offset = EXT4_MIN_BLOCK_SIZE % blocksize; + set_blocksize(bdev_file, blocksize); + bh = __bread(bdev, sb_block, blocksize); + if (!bh) { + ext4_msg(sb, KERN_ERR, "couldn't read superblock of " + "external journal"); + errno = -EINVAL; + goto out_bdev; + } + + es = (struct ext4_super_block *) (bh->b_data + offset); + if ((le16_to_cpu(es->s_magic) != EXT4_SUPER_MAGIC) || + !(le32_to_cpu(es->s_feature_incompat) & + EXT4_FEATURE_INCOMPAT_JOURNAL_DEV)) { + ext4_msg(sb, KERN_ERR, "external journal has bad superblock"); + errno = -EFSCORRUPTED; + goto out_bh; + } + + if ((le32_to_cpu(es->s_feature_ro_compat) & + EXT4_FEATURE_RO_COMPAT_METADATA_CSUM) && + es->s_checksum != ext4_superblock_csum(es)) { + ext4_msg(sb, KERN_ERR, "external journal has corrupt superblock"); + errno = -EFSCORRUPTED; + goto out_bh; + } + + if (memcmp(EXT4_SB(sb)->s_es->s_journal_uuid, es->s_uuid, 16)) { + ext4_msg(sb, KERN_ERR, "journal UUID does not match"); + errno = -EFSCORRUPTED; + goto out_bh; + } + + *j_start = sb_block + 1; + *j_len = ext4_blocks_count(es); + brelse(bh); + return bdev_file; + +out_bh: + brelse(bh); +out_bdev: + bdev_fput(bdev_file); + return ERR_PTR(errno); +} + +static journal_t *ext4_open_dev_journal(struct super_block *sb, + dev_t j_dev) +{ + journal_t *journal; + ext4_fsblk_t j_start; + ext4_fsblk_t j_len; + struct file *bdev_file; + int errno = 0; + + bdev_file = ext4_get_journal_blkdev(sb, j_dev, &j_start, &j_len); + if (IS_ERR(bdev_file)) + return ERR_CAST(bdev_file); + + journal = jbd2_journal_init_dev(file_bdev(bdev_file), sb->s_bdev, j_start, + j_len, sb->s_blocksize); + if (IS_ERR(journal)) { + ext4_msg(sb, KERN_ERR, "failed to create device journal"); + errno = PTR_ERR(journal); + goto out_bdev; + } + if (be32_to_cpu(journal->j_superblock->s_nr_users) != 1) { + ext4_msg(sb, KERN_ERR, "External journal has more than one " + "user (unsupported) - %d", + be32_to_cpu(journal->j_superblock->s_nr_users)); + errno = -EINVAL; + goto out_journal; + } + journal->j_private = sb; + EXT4_SB(sb)->s_journal_bdev_file = bdev_file; + ext4_init_journal_params(sb, journal); + return journal; + +out_journal: + ext4_journal_destroy(EXT4_SB(sb), journal); +out_bdev: + bdev_fput(bdev_file); + return ERR_PTR(errno); +} + +static int ext4_load_journal(struct super_block *sb, + struct ext4_super_block *es, + unsigned long journal_devnum) +{ + journal_t *journal; + unsigned int journal_inum = le32_to_cpu(es->s_journal_inum); + dev_t journal_dev; + int err = 0; + int really_read_only; + int journal_dev_ro; + + if (WARN_ON_ONCE(!ext4_has_feature_journal(sb))) + return -EFSCORRUPTED; + + if (journal_devnum && + journal_devnum != le32_to_cpu(es->s_journal_dev)) { + ext4_msg(sb, KERN_INFO, "external journal device major/minor " + "numbers have changed"); + journal_dev = new_decode_dev(journal_devnum); + } else + journal_dev = new_decode_dev(le32_to_cpu(es->s_journal_dev)); + + if (journal_inum && journal_dev) { + ext4_msg(sb, KERN_ERR, + "filesystem has both journal inode and journal device!"); + return -EINVAL; + } + + if (journal_inum) { + journal = ext4_open_inode_journal(sb, journal_inum); + if (IS_ERR(journal)) + return PTR_ERR(journal); + } else { + journal = ext4_open_dev_journal(sb, journal_dev); + if (IS_ERR(journal)) + return PTR_ERR(journal); + } + + journal_dev_ro = bdev_read_only(journal->j_dev); + really_read_only = bdev_read_only(sb->s_bdev) | journal_dev_ro; + + if (journal_dev_ro && !sb_rdonly(sb)) { + ext4_msg(sb, KERN_ERR, + "journal device read-only, try mounting with '-o ro'"); + err = -EROFS; + goto err_out; + } + + /* + * Are we loading a blank journal or performing recovery after a + * crash? For recovery, we need to check in advance whether we + * can get read-write access to the device. + */ + if (ext4_has_feature_journal_needs_recovery(sb)) { + if (sb_rdonly(sb)) { + ext4_msg(sb, KERN_INFO, "INFO: recovery " + "required on readonly filesystem"); + if (really_read_only) { + ext4_msg(sb, KERN_ERR, "write access " + "unavailable, cannot proceed " + "(try mounting with noload)"); + err = -EROFS; + goto err_out; + } + ext4_msg(sb, KERN_INFO, "write access will " + "be enabled during recovery"); + } + } + + if (!(journal->j_flags & JBD2_BARRIER)) + ext4_msg(sb, KERN_INFO, "barriers disabled"); + + if (!ext4_has_feature_journal_needs_recovery(sb)) + err = jbd2_journal_wipe(journal, !really_read_only); + if (!err) { + char *save = kmalloc(EXT4_S_ERR_LEN, GFP_KERNEL); + __le16 orig_state; + bool changed = false; + + if (save) + memcpy(save, ((char *) es) + + EXT4_S_ERR_START, EXT4_S_ERR_LEN); + err = jbd2_journal_load(journal); + if (save && memcmp(((char *) es) + EXT4_S_ERR_START, + save, EXT4_S_ERR_LEN)) { + memcpy(((char *) es) + EXT4_S_ERR_START, + save, EXT4_S_ERR_LEN); + changed = true; + } + kfree(save); + orig_state = es->s_state; + es->s_state |= cpu_to_le16(EXT4_SB(sb)->s_mount_state & + EXT4_ERROR_FS); + if (orig_state != es->s_state) + changed = true; + /* Write out restored error information to the superblock */ + if (changed && !really_read_only) { + int err2; + err2 = ext4_commit_super(sb); + err = err ? : err2; + } + } + + if (err) { + ext4_msg(sb, KERN_ERR, "error loading journal"); + goto err_out; + } + + EXT4_SB(sb)->s_journal = journal; + err = ext4_clear_journal_err(sb, es); + if (err) { + ext4_journal_destroy(EXT4_SB(sb), journal); + return err; + } + + if (!really_read_only && journal_devnum && + journal_devnum != le32_to_cpu(es->s_journal_dev)) { + es->s_journal_dev = cpu_to_le32(journal_devnum); + ext4_commit_super(sb); + } + if (!really_read_only && journal_inum && + journal_inum != le32_to_cpu(es->s_journal_inum)) { + es->s_journal_inum = cpu_to_le32(journal_inum); + ext4_commit_super(sb); + } + + return 0; + +err_out: + ext4_journal_destroy(EXT4_SB(sb), journal); + return err; +} + +/* Copy state of EXT4_SB(sb) into buffer for on-disk superblock */ +static void ext4_update_super(struct super_block *sb) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_super_block *es = sbi->s_es; + struct buffer_head *sbh = sbi->s_sbh; + + lock_buffer(sbh); + /* + * If the file system is mounted read-only, don't update the + * superblock write time. This avoids updating the superblock + * write time when we are mounting the root file system + * read/only but we need to replay the journal; at that point, + * for people who are east of GMT and who make their clock + * tick in localtime for Windows bug-for-bug compatibility, + * the clock is set in the future, and this will cause e2fsck + * to complain and force a full file system check. + */ + if (!sb_rdonly(sb)) + ext4_update_tstamp(es, s_wtime); + es->s_kbytes_written = + cpu_to_le64(sbi->s_kbytes_written + + ((part_stat_read(sb->s_bdev, sectors[STAT_WRITE]) - + sbi->s_sectors_written_start) >> 1)); + if (percpu_counter_initialized(&sbi->s_freeclusters_counter)) + ext4_free_blocks_count_set(es, + EXT4_C2B(sbi, percpu_counter_sum_positive( + &sbi->s_freeclusters_counter))); + if (percpu_counter_initialized(&sbi->s_freeinodes_counter)) + es->s_free_inodes_count = + cpu_to_le32(percpu_counter_sum_positive( + &sbi->s_freeinodes_counter)); + /* Copy error information to the on-disk superblock */ + spin_lock(&sbi->s_error_lock); + if (sbi->s_add_error_count > 0) { + es->s_state |= cpu_to_le16(EXT4_ERROR_FS); + if (!es->s_first_error_time && !es->s_first_error_time_hi) { + __ext4_update_tstamp(&es->s_first_error_time, + &es->s_first_error_time_hi, + sbi->s_first_error_time); + strtomem_pad(es->s_first_error_func, + sbi->s_first_error_func, 0); + es->s_first_error_line = + cpu_to_le32(sbi->s_first_error_line); + es->s_first_error_ino = + cpu_to_le32(sbi->s_first_error_ino); + es->s_first_error_block = + cpu_to_le64(sbi->s_first_error_block); + es->s_first_error_errcode = + ext4_errno_to_code(sbi->s_first_error_code); + } + __ext4_update_tstamp(&es->s_last_error_time, + &es->s_last_error_time_hi, + sbi->s_last_error_time); + strtomem_pad(es->s_last_error_func, sbi->s_last_error_func, 0); + es->s_last_error_line = cpu_to_le32(sbi->s_last_error_line); + es->s_last_error_ino = cpu_to_le32(sbi->s_last_error_ino); + es->s_last_error_block = cpu_to_le64(sbi->s_last_error_block); + es->s_last_error_errcode = + ext4_errno_to_code(sbi->s_last_error_code); + /* + * Start the daily error reporting function if it hasn't been + * started already + */ + if (!es->s_error_count) + mod_timer(&sbi->s_err_report, jiffies + 24*60*60*HZ); + le32_add_cpu(&es->s_error_count, sbi->s_add_error_count); + sbi->s_add_error_count = 0; + } + spin_unlock(&sbi->s_error_lock); + + ext4_superblock_csum_set(sb); + unlock_buffer(sbh); +} + +static int ext4_commit_super(struct super_block *sb) +{ + struct buffer_head *sbh = EXT4_SB(sb)->s_sbh; + + if (!sbh) + return -EINVAL; + + ext4_update_super(sb); + + lock_buffer(sbh); + /* Buffer got discarded which means block device got invalidated */ + if (!buffer_mapped(sbh)) { + unlock_buffer(sbh); + return -EIO; + } + + if (buffer_write_io_error(sbh) || !buffer_uptodate(sbh)) { + /* + * Oh, dear. A previous attempt to write the + * superblock failed. This could happen because the + * USB device was yanked out. Or it could happen to + * be a transient write error and maybe the block will + * be remapped. Nothing we can do but to retry the + * write and hope for the best. + */ + ext4_msg(sb, KERN_ERR, "previous I/O error to " + "superblock detected"); + clear_buffer_write_io_error(sbh); + set_buffer_uptodate(sbh); + } + get_bh(sbh); + /* Clear potential dirty bit if it was journalled update */ + clear_buffer_dirty(sbh); + sbh->b_end_io = end_buffer_write_sync; + submit_bh(REQ_OP_WRITE | REQ_SYNC | + (test_opt(sb, BARRIER) ? REQ_FUA : 0), sbh); + wait_on_buffer(sbh); + if (buffer_write_io_error(sbh)) { + ext4_msg(sb, KERN_ERR, "I/O error while writing " + "superblock"); + clear_buffer_write_io_error(sbh); + set_buffer_uptodate(sbh); + return -EIO; + } + return 0; +} + +/* + * Have we just finished recovery? If so, and if we are mounting (or + * remounting) the filesystem readonly, then we will end up with a + * consistent fs on disk. Record that fact. + */ +static int ext4_mark_recovery_complete(struct super_block *sb, + struct ext4_super_block *es) +{ + int err; + journal_t *journal = EXT4_SB(sb)->s_journal; + + if (!ext4_has_feature_journal(sb)) { + if (journal != NULL) { + ext4_error(sb, "Journal got removed while the fs was " + "mounted!"); + return -EFSCORRUPTED; + } + return 0; + } + jbd2_journal_lock_updates(journal); + err = jbd2_journal_flush(journal, 0); + if (err < 0) + goto out; + + if (sb_rdonly(sb) && (ext4_has_feature_journal_needs_recovery(sb) || + ext4_has_feature_orphan_present(sb))) { + if (!ext4_orphan_file_empty(sb)) { + ext4_error(sb, "Orphan file not empty on read-only fs."); + err = -EFSCORRUPTED; + goto out; + } + ext4_clear_feature_journal_needs_recovery(sb); + ext4_clear_feature_orphan_present(sb); + ext4_commit_super(sb); + } +out: + jbd2_journal_unlock_updates(journal); + return err; +} + +/* + * If we are mounting (or read-write remounting) a filesystem whose journal + * has recorded an error from a previous lifetime, move that error to the + * main filesystem now. + */ +static int ext4_clear_journal_err(struct super_block *sb, + struct ext4_super_block *es) +{ + journal_t *journal; + int j_errno; + const char *errstr; + + if (!ext4_has_feature_journal(sb)) { + ext4_error(sb, "Journal got removed while the fs was mounted!"); + return -EFSCORRUPTED; + } + + journal = EXT4_SB(sb)->s_journal; + + /* + * Now check for any error status which may have been recorded in the + * journal by a prior ext4_error() or ext4_abort() + */ + + j_errno = jbd2_journal_errno(journal); + if (j_errno) { + char nbuf[16]; + + errstr = ext4_decode_error(sb, j_errno, nbuf); + ext4_warning(sb, "Filesystem error recorded " + "from previous mount: %s", errstr); + + EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS; + es->s_state |= cpu_to_le16(EXT4_ERROR_FS); + j_errno = ext4_commit_super(sb); + if (j_errno) + return j_errno; + ext4_warning(sb, "Marked fs in need of filesystem check."); + + jbd2_journal_clear_err(journal); + jbd2_journal_update_sb_errno(journal); + } + return 0; +} + +/* + * Force the running and committing transactions to commit, + * and wait on the commit. + */ +int ext4_force_commit(struct super_block *sb) +{ + return ext4_journal_force_commit(EXT4_SB(sb)->s_journal); +} + +static int ext4_sync_fs(struct super_block *sb, int wait) +{ + int ret = 0; + tid_t target; + bool needs_barrier = false; + struct ext4_sb_info *sbi = EXT4_SB(sb); + + ret = ext4_emergency_state(sb); + if (unlikely(ret)) + return ret; + + trace_ext4_sync_fs(sb, wait); + flush_workqueue(sbi->rsv_conversion_wq); + /* + * Writeback quota in non-journalled quota case - journalled quota has + * no dirty dquots + */ + dquot_writeback_dquots(sb, -1); + /* + * Data writeback is possible w/o journal transaction, so barrier must + * being sent at the end of the function. But we can skip it if + * transaction_commit will do it for us. + */ + if (sbi->s_journal) { + target = jbd2_get_latest_transaction(sbi->s_journal); + if (wait && sbi->s_journal->j_flags & JBD2_BARRIER && + !jbd2_trans_will_send_data_barrier(sbi->s_journal, target)) + needs_barrier = true; + + if (jbd2_journal_start_commit(sbi->s_journal, &target)) { + if (wait) + ret = jbd2_log_wait_commit(sbi->s_journal, + target); + } + } else if (wait && test_opt(sb, BARRIER)) + needs_barrier = true; + if (needs_barrier) { + int err; + err = blkdev_issue_flush(sb->s_bdev); + if (!ret) + ret = err; + } + + return ret; +} + +/* + * LVM calls this function before a (read-only) snapshot is created. This + * gives us a chance to flush the journal completely and mark the fs clean. + * + * Note that only this function cannot bring a filesystem to be in a clean + * state independently. It relies on upper layer to stop all data & metadata + * modifications. + */ +static int ext4_freeze(struct super_block *sb) +{ + int error = 0; + journal_t *journal = EXT4_SB(sb)->s_journal; + + if (journal) { + /* Now we set up the journal barrier. */ + jbd2_journal_lock_updates(journal); + + /* + * Don't clear the needs_recovery flag if we failed to + * flush the journal. + */ + error = jbd2_journal_flush(journal, 0); + if (error < 0) + goto out; + + /* Journal blocked and flushed, clear needs_recovery flag. */ + ext4_clear_feature_journal_needs_recovery(sb); + if (ext4_orphan_file_empty(sb)) + ext4_clear_feature_orphan_present(sb); + } + + error = ext4_commit_super(sb); +out: + if (journal) + /* we rely on upper layer to stop further updates */ + jbd2_journal_unlock_updates(journal); + return error; +} + +/* + * Called by LVM after the snapshot is done. We need to reset the RECOVER + * flag here, even though the filesystem is not technically dirty yet. + */ +static int ext4_unfreeze(struct super_block *sb) +{ + if (ext4_emergency_state(sb)) + return 0; + + if (EXT4_SB(sb)->s_journal) { + /* Reset the needs_recovery flag before the fs is unlocked. */ + ext4_set_feature_journal_needs_recovery(sb); + if (ext4_has_feature_orphan_file(sb)) + ext4_set_feature_orphan_present(sb); + } + + ext4_commit_super(sb); + return 0; +} + +/* + * Structure to save mount options for ext4_remount's benefit + */ +struct ext4_mount_options { + unsigned long s_mount_opt; + unsigned long s_mount_opt2; + kuid_t s_resuid; + kgid_t s_resgid; + unsigned long s_commit_interval; + u32 s_min_batch_time, s_max_batch_time; +#ifdef CONFIG_QUOTA + int s_jquota_fmt; + char *s_qf_names[EXT4_MAXQUOTAS]; +#endif +}; + +static int __ext4_remount(struct fs_context *fc, struct super_block *sb) +{ + struct ext4_fs_context *ctx = fc->fs_private; + struct ext4_super_block *es; + struct ext4_sb_info *sbi = EXT4_SB(sb); + unsigned long old_sb_flags; + struct ext4_mount_options old_opts; + ext4_group_t g; + int err = 0; + int alloc_ctx; +#ifdef CONFIG_QUOTA + int enable_quota = 0; + int i, j; + char *to_free[EXT4_MAXQUOTAS]; +#endif + + + /* Store the original options */ + old_sb_flags = sb->s_flags; + old_opts.s_mount_opt = sbi->s_mount_opt; + old_opts.s_mount_opt2 = sbi->s_mount_opt2; + old_opts.s_resuid = sbi->s_resuid; + old_opts.s_resgid = sbi->s_resgid; + old_opts.s_commit_interval = sbi->s_commit_interval; + old_opts.s_min_batch_time = sbi->s_min_batch_time; + old_opts.s_max_batch_time = sbi->s_max_batch_time; +#ifdef CONFIG_QUOTA + old_opts.s_jquota_fmt = sbi->s_jquota_fmt; + for (i = 0; i < EXT4_MAXQUOTAS; i++) + if (sbi->s_qf_names[i]) { + char *qf_name = get_qf_name(sb, sbi, i); + + old_opts.s_qf_names[i] = kstrdup(qf_name, GFP_KERNEL); + if (!old_opts.s_qf_names[i]) { + for (j = 0; j < i; j++) + kfree(old_opts.s_qf_names[j]); + return -ENOMEM; + } + } else + old_opts.s_qf_names[i] = NULL; +#endif + if (!(ctx->spec & EXT4_SPEC_JOURNAL_IOPRIO)) { + if (sbi->s_journal && sbi->s_journal->j_task->io_context) + ctx->journal_ioprio = + sbi->s_journal->j_task->io_context->ioprio; + else + ctx->journal_ioprio = EXT4_DEF_JOURNAL_IOPRIO; + + } + + if ((ctx->spec & EXT4_SPEC_s_stripe) && + ext4_is_stripe_incompatible(sb, ctx->s_stripe)) { + ext4_msg(sb, KERN_WARNING, + "stripe (%lu) is not aligned with cluster size (%u), " + "stripe is disabled", + ctx->s_stripe, sbi->s_cluster_ratio); + ctx->s_stripe = 0; + } + + /* + * Changing the DIOREAD_NOLOCK or DELALLOC mount options may cause + * two calls to ext4_should_dioread_nolock() to return inconsistent + * values, triggering WARN_ON in ext4_add_complete_io(). we grab + * here s_writepages_rwsem to avoid race between writepages ops and + * remount. + */ + alloc_ctx = ext4_writepages_down_write(sb); + ext4_apply_options(fc, sb); + ext4_writepages_up_write(sb, alloc_ctx); + + if ((old_opts.s_mount_opt & EXT4_MOUNT_JOURNAL_CHECKSUM) ^ + test_opt(sb, JOURNAL_CHECKSUM)) { + ext4_msg(sb, KERN_ERR, "changing journal_checksum " + "during remount not supported; ignoring"); + sbi->s_mount_opt ^= EXT4_MOUNT_JOURNAL_CHECKSUM; + } + + if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA) { + if (test_opt2(sb, EXPLICIT_DELALLOC)) { + ext4_msg(sb, KERN_ERR, "can't mount with " + "both data=journal and delalloc"); + err = -EINVAL; + goto restore_opts; + } + if (test_opt(sb, DIOREAD_NOLOCK)) { + ext4_msg(sb, KERN_ERR, "can't mount with " + "both data=journal and dioread_nolock"); + err = -EINVAL; + goto restore_opts; + } + } else if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_ORDERED_DATA) { + if (test_opt(sb, JOURNAL_ASYNC_COMMIT)) { + ext4_msg(sb, KERN_ERR, "can't mount with " + "journal_async_commit in data=ordered mode"); + err = -EINVAL; + goto restore_opts; + } + } + + if ((sbi->s_mount_opt ^ old_opts.s_mount_opt) & EXT4_MOUNT_NO_MBCACHE) { + ext4_msg(sb, KERN_ERR, "can't enable nombcache during remount"); + err = -EINVAL; + goto restore_opts; + } + + if ((old_opts.s_mount_opt & EXT4_MOUNT_DELALLOC) && + !test_opt(sb, DELALLOC)) { + ext4_msg(sb, KERN_ERR, "can't disable delalloc during remount"); + err = -EINVAL; + goto restore_opts; + } + + sb->s_flags = (sb->s_flags & ~SB_POSIXACL) | + (test_opt(sb, POSIX_ACL) ? SB_POSIXACL : 0); + + es = sbi->s_es; + + if (sbi->s_journal) { + ext4_init_journal_params(sb, sbi->s_journal); + set_task_ioprio(sbi->s_journal->j_task, ctx->journal_ioprio); + } + + /* Flush outstanding errors before changing fs state */ + flush_work(&sbi->s_sb_upd_work); + + if ((bool)(fc->sb_flags & SB_RDONLY) != sb_rdonly(sb)) { + if (ext4_emergency_state(sb)) { + err = -EROFS; + goto restore_opts; + } + + if (fc->sb_flags & SB_RDONLY) { + err = sync_filesystem(sb); + if (err < 0) + goto restore_opts; + err = dquot_suspend(sb, -1); + if (err < 0) + goto restore_opts; + + /* + * First of all, the unconditional stuff we have to do + * to disable replay of the journal when we next remount + */ + sb->s_flags |= SB_RDONLY; + + /* + * OK, test if we are remounting a valid rw partition + * readonly, and if so set the rdonly flag and then + * mark the partition as valid again. + */ + if (!(es->s_state & cpu_to_le16(EXT4_VALID_FS)) && + (sbi->s_mount_state & EXT4_VALID_FS)) + es->s_state = cpu_to_le16(sbi->s_mount_state); + + if (sbi->s_journal) { + /* + * We let remount-ro finish even if marking fs + * as clean failed... + */ + ext4_mark_recovery_complete(sb, es); + } + } else { + /* Make sure we can mount this feature set readwrite */ + if (ext4_has_feature_readonly(sb) || + !ext4_feature_set_ok(sb, 0)) { + err = -EROFS; + goto restore_opts; + } + /* + * Make sure the group descriptor checksums + * are sane. If they aren't, refuse to remount r/w. + */ + for (g = 0; g < sbi->s_groups_count; g++) { + struct ext4_group_desc *gdp = + ext4_get_group_desc(sb, g, NULL); + + if (!ext4_group_desc_csum_verify(sb, g, gdp)) { + ext4_msg(sb, KERN_ERR, + "ext4_remount: Checksum for group %u failed (%u!=%u)", + g, le16_to_cpu(ext4_group_desc_csum(sb, g, gdp)), + le16_to_cpu(gdp->bg_checksum)); + err = -EFSBADCRC; + goto restore_opts; + } + } + + /* + * If we have an unprocessed orphan list hanging + * around from a previously readonly bdev mount, + * require a full umount/remount for now. + */ + if (es->s_last_orphan || !ext4_orphan_file_empty(sb)) { + ext4_msg(sb, KERN_WARNING, "Couldn't " + "remount RDWR because of unprocessed " + "orphan inode list. Please " + "umount/remount instead"); + err = -EINVAL; + goto restore_opts; + } + + /* + * Mounting a RDONLY partition read-write, so reread + * and store the current valid flag. (It may have + * been changed by e2fsck since we originally mounted + * the partition.) + */ + if (sbi->s_journal) { + err = ext4_clear_journal_err(sb, es); + if (err) + goto restore_opts; + } + sbi->s_mount_state = (le16_to_cpu(es->s_state) & + ~EXT4_FC_REPLAY); + + err = ext4_setup_super(sb, es, 0); + if (err) + goto restore_opts; + + sb->s_flags &= ~SB_RDONLY; + if (ext4_has_feature_mmp(sb)) { + err = ext4_multi_mount_protect(sb, + le64_to_cpu(es->s_mmp_block)); + if (err) + goto restore_opts; + } +#ifdef CONFIG_QUOTA + enable_quota = 1; +#endif + } + } + + /* + * Handle creation of system zone data early because it can fail. + * Releasing of existing data is done when we are sure remount will + * succeed. + */ + if (test_opt(sb, BLOCK_VALIDITY) && !sbi->s_system_blks) { + err = ext4_setup_system_zone(sb); + if (err) + goto restore_opts; + } + + if (sbi->s_journal == NULL && !(old_sb_flags & SB_RDONLY)) { + err = ext4_commit_super(sb); + if (err) + goto restore_opts; + } + +#ifdef CONFIG_QUOTA + if (enable_quota) { + if (sb_any_quota_suspended(sb)) + dquot_resume(sb, -1); + else if (ext4_has_feature_quota(sb)) { + err = ext4_enable_quotas(sb); + if (err) + goto restore_opts; + } + } + /* Release old quota file names */ + for (i = 0; i < EXT4_MAXQUOTAS; i++) + kfree(old_opts.s_qf_names[i]); +#endif + if (!test_opt(sb, BLOCK_VALIDITY) && sbi->s_system_blks) + ext4_release_system_zone(sb); + + /* + * Reinitialize lazy itable initialization thread based on + * current settings + */ + if (sb_rdonly(sb) || !test_opt(sb, INIT_INODE_TABLE)) + ext4_unregister_li_request(sb); + else { + ext4_group_t first_not_zeroed; + first_not_zeroed = ext4_has_uninit_itable(sb); + ext4_register_li_request(sb, first_not_zeroed); + } + + if (!ext4_has_feature_mmp(sb) || sb_rdonly(sb)) + ext4_stop_mmpd(sbi); + + /* + * Handle aborting the filesystem as the last thing during remount to + * avoid obsure errors during remount when some option changes fail to + * apply due to shutdown filesystem. + */ + if (test_opt2(sb, ABORT)) + ext4_abort(sb, ESHUTDOWN, "Abort forced by user"); + + return 0; + +restore_opts: + /* + * If there was a failing r/w to ro transition, we may need to + * re-enable quota + */ + if (sb_rdonly(sb) && !(old_sb_flags & SB_RDONLY) && + sb_any_quota_suspended(sb)) + dquot_resume(sb, -1); + + alloc_ctx = ext4_writepages_down_write(sb); + sb->s_flags = old_sb_flags; + sbi->s_mount_opt = old_opts.s_mount_opt; + sbi->s_mount_opt2 = old_opts.s_mount_opt2; + sbi->s_resuid = old_opts.s_resuid; + sbi->s_resgid = old_opts.s_resgid; + sbi->s_commit_interval = old_opts.s_commit_interval; + sbi->s_min_batch_time = old_opts.s_min_batch_time; + sbi->s_max_batch_time = old_opts.s_max_batch_time; + ext4_writepages_up_write(sb, alloc_ctx); + + if (!test_opt(sb, BLOCK_VALIDITY) && sbi->s_system_blks) + ext4_release_system_zone(sb); +#ifdef CONFIG_QUOTA + sbi->s_jquota_fmt = old_opts.s_jquota_fmt; + for (i = 0; i < EXT4_MAXQUOTAS; i++) { + to_free[i] = get_qf_name(sb, sbi, i); + rcu_assign_pointer(sbi->s_qf_names[i], old_opts.s_qf_names[i]); + } + synchronize_rcu(); + for (i = 0; i < EXT4_MAXQUOTAS; i++) + kfree(to_free[i]); +#endif + if (!ext4_has_feature_mmp(sb) || sb_rdonly(sb)) + ext4_stop_mmpd(sbi); + return err; +} + +static int ext4_reconfigure(struct fs_context *fc) +{ + struct super_block *sb = fc->root->d_sb; + int ret; + bool old_ro = sb_rdonly(sb); + + fc->s_fs_info = EXT4_SB(sb); + + ret = ext4_check_opt_consistency(fc, sb); + if (ret < 0) + return ret; + + ret = __ext4_remount(fc, sb); + if (ret < 0) + return ret; + + ext4_msg(sb, KERN_INFO, "re-mounted %pU%s.", + &sb->s_uuid, + (old_ro != sb_rdonly(sb)) ? (sb_rdonly(sb) ? " ro" : " r/w") : ""); + + return 0; +} + +#ifdef CONFIG_QUOTA +static int ext4_statfs_project(struct super_block *sb, + kprojid_t projid, struct kstatfs *buf) +{ + struct kqid qid; + struct dquot *dquot; + u64 limit; + u64 curblock; + + qid = make_kqid_projid(projid); + dquot = dqget(sb, qid); + if (IS_ERR(dquot)) + return PTR_ERR(dquot); + spin_lock(&dquot->dq_dqb_lock); + + limit = min_not_zero(dquot->dq_dqb.dqb_bsoftlimit, + dquot->dq_dqb.dqb_bhardlimit); + limit >>= sb->s_blocksize_bits; + + if (limit) { + uint64_t remaining = 0; + + curblock = (dquot->dq_dqb.dqb_curspace + + dquot->dq_dqb.dqb_rsvspace) >> sb->s_blocksize_bits; + if (limit > curblock) + remaining = limit - curblock; + + buf->f_blocks = min(buf->f_blocks, limit); + buf->f_bfree = min(buf->f_bfree, remaining); + buf->f_bavail = min(buf->f_bavail, remaining); + } + + limit = min_not_zero(dquot->dq_dqb.dqb_isoftlimit, + dquot->dq_dqb.dqb_ihardlimit); + if (limit) { + uint64_t remaining = 0; + + if (limit > dquot->dq_dqb.dqb_curinodes) + remaining = limit - dquot->dq_dqb.dqb_curinodes; + + buf->f_files = min(buf->f_files, limit); + buf->f_ffree = min(buf->f_ffree, remaining); + } + + spin_unlock(&dquot->dq_dqb_lock); + dqput(dquot); + return 0; +} +#endif + +static int ext4_statfs(struct dentry *dentry, struct kstatfs *buf) +{ + struct super_block *sb = dentry->d_sb; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_super_block *es = sbi->s_es; + ext4_fsblk_t overhead = 0, resv_blocks; + s64 bfree; + resv_blocks = EXT4_C2B(sbi, atomic64_read(&sbi->s_resv_clusters)); + + if (!test_opt(sb, MINIX_DF)) + overhead = sbi->s_overhead; + + buf->f_type = EXT4_SUPER_MAGIC; + buf->f_bsize = sb->s_blocksize; + buf->f_blocks = ext4_blocks_count(es) - EXT4_C2B(sbi, overhead); + bfree = percpu_counter_sum_positive(&sbi->s_freeclusters_counter) - + percpu_counter_sum_positive(&sbi->s_dirtyclusters_counter); + /* prevent underflow in case that few free space is available */ + buf->f_bfree = EXT4_C2B(sbi, max_t(s64, bfree, 0)); + buf->f_bavail = buf->f_bfree - + (ext4_r_blocks_count(es) + resv_blocks); + if (buf->f_bfree < (ext4_r_blocks_count(es) + resv_blocks)) + buf->f_bavail = 0; + buf->f_files = le32_to_cpu(es->s_inodes_count); + buf->f_ffree = percpu_counter_sum_positive(&sbi->s_freeinodes_counter); + buf->f_namelen = EXT4_NAME_LEN; + buf->f_fsid = uuid_to_fsid(es->s_uuid); + +#ifdef CONFIG_QUOTA + if (ext4_test_inode_flag(dentry->d_inode, EXT4_INODE_PROJINHERIT) && + sb_has_quota_limits_enabled(sb, PRJQUOTA)) + ext4_statfs_project(sb, EXT4_I(dentry->d_inode)->i_projid, buf); +#endif + return 0; +} + + +#ifdef CONFIG_QUOTA + +/* + * Helper functions so that transaction is started before we acquire dqio_sem + * to keep correct lock ordering of transaction > dqio_sem + */ +static inline struct inode *dquot_to_inode(struct dquot *dquot) +{ + return sb_dqopt(dquot->dq_sb)->files[dquot->dq_id.type]; +} + +static int ext4_write_dquot(struct dquot *dquot) +{ + int ret, err; + handle_t *handle; + struct inode *inode; + + inode = dquot_to_inode(dquot); + handle = ext4_journal_start(inode, EXT4_HT_QUOTA, + EXT4_QUOTA_TRANS_BLOCKS(dquot->dq_sb)); + if (IS_ERR(handle)) + return PTR_ERR(handle); + ret = dquot_commit(dquot); + if (ret < 0) + ext4_error_err(dquot->dq_sb, -ret, + "Failed to commit dquot type %d", + dquot->dq_id.type); + err = ext4_journal_stop(handle); + if (!ret) + ret = err; + return ret; +} + +static int ext4_acquire_dquot(struct dquot *dquot) +{ + int ret, err; + handle_t *handle; + + handle = ext4_journal_start(dquot_to_inode(dquot), EXT4_HT_QUOTA, + EXT4_QUOTA_INIT_BLOCKS(dquot->dq_sb)); + if (IS_ERR(handle)) + return PTR_ERR(handle); + ret = dquot_acquire(dquot); + if (ret < 0) + ext4_error_err(dquot->dq_sb, -ret, + "Failed to acquire dquot type %d", + dquot->dq_id.type); + err = ext4_journal_stop(handle); + if (!ret) + ret = err; + return ret; +} + +static int ext4_release_dquot(struct dquot *dquot) +{ + int ret, err; + handle_t *handle; + bool freeze_protected = false; + + /* + * Trying to sb_start_intwrite() in a running transaction + * can result in a deadlock. Further, running transactions + * are already protected from freezing. + */ + if (!ext4_journal_current_handle()) { + sb_start_intwrite(dquot->dq_sb); + freeze_protected = true; + } + + handle = ext4_journal_start(dquot_to_inode(dquot), EXT4_HT_QUOTA, + EXT4_QUOTA_DEL_BLOCKS(dquot->dq_sb)); + if (IS_ERR(handle)) { + /* Release dquot anyway to avoid endless cycle in dqput() */ + dquot_release(dquot); + if (freeze_protected) + sb_end_intwrite(dquot->dq_sb); + return PTR_ERR(handle); + } + ret = dquot_release(dquot); + if (ret < 0) + ext4_error_err(dquot->dq_sb, -ret, + "Failed to release dquot type %d", + dquot->dq_id.type); + err = ext4_journal_stop(handle); + if (!ret) + ret = err; + + if (freeze_protected) + sb_end_intwrite(dquot->dq_sb); + + return ret; +} + +static int ext4_mark_dquot_dirty(struct dquot *dquot) +{ + struct super_block *sb = dquot->dq_sb; + + if (ext4_is_quota_journalled(sb)) { + dquot_mark_dquot_dirty(dquot); + return ext4_write_dquot(dquot); + } else { + return dquot_mark_dquot_dirty(dquot); + } +} + +static int ext4_write_info(struct super_block *sb, int type) +{ + int ret, err; + handle_t *handle; + + /* Data block + inode block */ + handle = ext4_journal_start_sb(sb, EXT4_HT_QUOTA, 2); + if (IS_ERR(handle)) + return PTR_ERR(handle); + ret = dquot_commit_info(sb, type); + err = ext4_journal_stop(handle); + if (!ret) + ret = err; + return ret; +} + +static void lockdep_set_quota_inode(struct inode *inode, int subclass) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + + /* The first argument of lockdep_set_subclass has to be + * *exactly* the same as the argument to init_rwsem() --- in + * this case, in init_once() --- or lockdep gets unhappy + * because the name of the lock is set using the + * stringification of the argument to init_rwsem(). + */ + (void) ei; /* shut up clang warning if !CONFIG_LOCKDEP */ + lockdep_set_subclass(&ei->i_data_sem, subclass); +} + +/* + * Standard function to be called on quota_on + */ +static int ext4_quota_on(struct super_block *sb, int type, int format_id, + const struct path *path) +{ + int err; + + if (!test_opt(sb, QUOTA)) + return -EINVAL; + + /* Quotafile not on the same filesystem? */ + if (path->dentry->d_sb != sb) + return -EXDEV; + + /* Quota already enabled for this file? */ + if (IS_NOQUOTA(d_inode(path->dentry))) + return -EBUSY; + + /* Journaling quota? */ + if (EXT4_SB(sb)->s_qf_names[type]) { + /* Quotafile not in fs root? */ + if (path->dentry->d_parent != sb->s_root) + ext4_msg(sb, KERN_WARNING, + "Quota file not on filesystem root. " + "Journaled quota will not work"); + sb_dqopt(sb)->flags |= DQUOT_NOLIST_DIRTY; + } else { + /* + * Clear the flag just in case mount options changed since + * last time. + */ + sb_dqopt(sb)->flags &= ~DQUOT_NOLIST_DIRTY; + } + + lockdep_set_quota_inode(path->dentry->d_inode, I_DATA_SEM_QUOTA); + err = dquot_quota_on(sb, type, format_id, path); + if (!err) { + struct inode *inode = d_inode(path->dentry); + handle_t *handle; + + /* + * Set inode flags to prevent userspace from messing with quota + * files. If this fails, we return success anyway since quotas + * are already enabled and this is not a hard failure. + */ + inode_lock(inode); + handle = ext4_journal_start(inode, EXT4_HT_QUOTA, 1); + if (IS_ERR(handle)) + goto unlock_inode; + EXT4_I(inode)->i_flags |= EXT4_NOATIME_FL | EXT4_IMMUTABLE_FL; + inode_set_flags(inode, S_NOATIME | S_IMMUTABLE, + S_NOATIME | S_IMMUTABLE); + err = ext4_mark_inode_dirty(handle, inode); + ext4_journal_stop(handle); + unlock_inode: + inode_unlock(inode); + if (err) + dquot_quota_off(sb, type); + } + if (err) + lockdep_set_quota_inode(path->dentry->d_inode, + I_DATA_SEM_NORMAL); + return err; +} + +static inline bool ext4_check_quota_inum(int type, unsigned long qf_inum) +{ + switch (type) { + case USRQUOTA: + return qf_inum == EXT4_USR_QUOTA_INO; + case GRPQUOTA: + return qf_inum == EXT4_GRP_QUOTA_INO; + case PRJQUOTA: + return qf_inum >= EXT4_GOOD_OLD_FIRST_INO; + default: + BUG(); + } +} + +static int ext4_quota_enable(struct super_block *sb, int type, int format_id, + unsigned int flags) +{ + int err; + struct inode *qf_inode; + unsigned long qf_inums[EXT4_MAXQUOTAS] = { + le32_to_cpu(EXT4_SB(sb)->s_es->s_usr_quota_inum), + le32_to_cpu(EXT4_SB(sb)->s_es->s_grp_quota_inum), + le32_to_cpu(EXT4_SB(sb)->s_es->s_prj_quota_inum) + }; + + BUG_ON(!ext4_has_feature_quota(sb)); + + if (!qf_inums[type]) + return -EPERM; + + if (!ext4_check_quota_inum(type, qf_inums[type])) { + ext4_error(sb, "Bad quota inum: %lu, type: %d", + qf_inums[type], type); + return -EUCLEAN; + } + + qf_inode = ext4_iget(sb, qf_inums[type], EXT4_IGET_SPECIAL); + if (IS_ERR(qf_inode)) { + ext4_error(sb, "Bad quota inode: %lu, type: %d", + qf_inums[type], type); + return PTR_ERR(qf_inode); + } + + /* Don't account quota for quota files to avoid recursion */ + qf_inode->i_flags |= S_NOQUOTA; + lockdep_set_quota_inode(qf_inode, I_DATA_SEM_QUOTA); + err = dquot_load_quota_inode(qf_inode, type, format_id, flags); + if (err) + lockdep_set_quota_inode(qf_inode, I_DATA_SEM_NORMAL); + iput(qf_inode); + + return err; +} + +/* Enable usage tracking for all quota types. */ +int ext4_enable_quotas(struct super_block *sb) +{ + int type, err = 0; + unsigned long qf_inums[EXT4_MAXQUOTAS] = { + le32_to_cpu(EXT4_SB(sb)->s_es->s_usr_quota_inum), + le32_to_cpu(EXT4_SB(sb)->s_es->s_grp_quota_inum), + le32_to_cpu(EXT4_SB(sb)->s_es->s_prj_quota_inum) + }; + bool quota_mopt[EXT4_MAXQUOTAS] = { + test_opt(sb, USRQUOTA), + test_opt(sb, GRPQUOTA), + test_opt(sb, PRJQUOTA), + }; + + sb_dqopt(sb)->flags |= DQUOT_QUOTA_SYS_FILE | DQUOT_NOLIST_DIRTY; + for (type = 0; type < EXT4_MAXQUOTAS; type++) { + if (qf_inums[type]) { + err = ext4_quota_enable(sb, type, QFMT_VFS_V1, + DQUOT_USAGE_ENABLED | + (quota_mopt[type] ? DQUOT_LIMITS_ENABLED : 0)); + if (err) { + ext4_warning(sb, + "Failed to enable quota tracking " + "(type=%d, err=%d, ino=%lu). " + "Please run e2fsck to fix.", type, + err, qf_inums[type]); + + ext4_quotas_off(sb, type); + return err; + } + } + } + return 0; +} + +static int ext4_quota_off(struct super_block *sb, int type) +{ + struct inode *inode = sb_dqopt(sb)->files[type]; + handle_t *handle; + int err; + + /* Force all delayed allocation blocks to be allocated. + * Caller already holds s_umount sem */ + if (test_opt(sb, DELALLOC)) + sync_filesystem(sb); + + if (!inode || !igrab(inode)) + goto out; + + err = dquot_quota_off(sb, type); + if (err || ext4_has_feature_quota(sb)) + goto out_put; + /* + * When the filesystem was remounted read-only first, we cannot cleanup + * inode flags here. Bad luck but people should be using QUOTA feature + * these days anyway. + */ + if (sb_rdonly(sb)) + goto out_put; + + inode_lock(inode); + /* + * Update modification times of quota files when userspace can + * start looking at them. If we fail, we return success anyway since + * this is not a hard failure and quotas are already disabled. + */ + handle = ext4_journal_start(inode, EXT4_HT_QUOTA, 1); + if (IS_ERR(handle)) { + err = PTR_ERR(handle); + goto out_unlock; + } + EXT4_I(inode)->i_flags &= ~(EXT4_NOATIME_FL | EXT4_IMMUTABLE_FL); + inode_set_flags(inode, 0, S_NOATIME | S_IMMUTABLE); + inode_set_mtime_to_ts(inode, inode_set_ctime_current(inode)); + err = ext4_mark_inode_dirty(handle, inode); + ext4_journal_stop(handle); +out_unlock: + inode_unlock(inode); +out_put: + lockdep_set_quota_inode(inode, I_DATA_SEM_NORMAL); + iput(inode); + return err; +out: + return dquot_quota_off(sb, type); +} + +/* Read data from quotafile - avoid pagecache and such because we cannot afford + * acquiring the locks... As quota files are never truncated and quota code + * itself serializes the operations (and no one else should touch the files) + * we don't have to be afraid of races */ +static ssize_t ext4_quota_read(struct super_block *sb, int type, char *data, + size_t len, loff_t off) +{ + struct inode *inode = sb_dqopt(sb)->files[type]; + ext4_lblk_t blk = off >> EXT4_BLOCK_SIZE_BITS(sb); + int offset = off & (sb->s_blocksize - 1); + int tocopy; + size_t toread; + struct buffer_head *bh; + loff_t i_size = i_size_read(inode); + + if (off > i_size) + return 0; + if (off+len > i_size) + len = i_size-off; + toread = len; + while (toread > 0) { + tocopy = min_t(unsigned long, sb->s_blocksize - offset, toread); + bh = ext4_bread(NULL, inode, blk, 0); + if (IS_ERR(bh)) + return PTR_ERR(bh); + if (!bh) /* A hole? */ + memset(data, 0, tocopy); + else + memcpy(data, bh->b_data+offset, tocopy); + brelse(bh); + offset = 0; + toread -= tocopy; + data += tocopy; + blk++; + } + return len; +} + +/* Write to quotafile (we know the transaction is already started and has + * enough credits) */ +static ssize_t ext4_quota_write(struct super_block *sb, int type, + const char *data, size_t len, loff_t off) +{ + struct inode *inode = sb_dqopt(sb)->files[type]; + ext4_lblk_t blk = off >> EXT4_BLOCK_SIZE_BITS(sb); + int err = 0, err2 = 0, offset = off & (sb->s_blocksize - 1); + int retries = 0; + struct buffer_head *bh; + handle_t *handle = journal_current_handle(); + + if (!handle) { + ext4_msg(sb, KERN_WARNING, "Quota write (off=%llu, len=%llu)" + " cancelled because transaction is not started", + (unsigned long long)off, (unsigned long long)len); + return -EIO; + } + /* + * Since we account only one data block in transaction credits, + * then it is impossible to cross a block boundary. + */ + if (sb->s_blocksize - offset < len) { + ext4_msg(sb, KERN_WARNING, "Quota write (off=%llu, len=%llu)" + " cancelled because not block aligned", + (unsigned long long)off, (unsigned long long)len); + return -EIO; + } + + do { + bh = ext4_bread(handle, inode, blk, + EXT4_GET_BLOCKS_CREATE | + EXT4_GET_BLOCKS_METADATA_NOFAIL); + } while (PTR_ERR(bh) == -ENOSPC && + ext4_should_retry_alloc(inode->i_sb, &retries)); + if (IS_ERR(bh)) + return PTR_ERR(bh); + if (!bh) + goto out; + BUFFER_TRACE(bh, "get write access"); + err = ext4_journal_get_write_access(handle, sb, bh, EXT4_JTR_NONE); + if (err) { + brelse(bh); + return err; + } + lock_buffer(bh); + memcpy(bh->b_data+offset, data, len); + flush_dcache_folio(bh->b_folio); + unlock_buffer(bh); + err = ext4_handle_dirty_metadata(handle, NULL, bh); + brelse(bh); +out: + if (inode->i_size < off + len) { + i_size_write(inode, off + len); + EXT4_I(inode)->i_disksize = inode->i_size; + err2 = ext4_mark_inode_dirty(handle, inode); + if (unlikely(err2 && !err)) + err = err2; + } + return err ? err : len; +} +#endif + +#if !defined(CONFIG_EXT2_FS) && !defined(CONFIG_EXT2_FS_MODULE) && defined(CONFIG_EXT4_USE_FOR_EXT2) +static inline void register_as_ext2(void) +{ + int err = register_filesystem(&ext2_fs_type); + if (err) + printk(KERN_WARNING + "EXT4-fs: Unable to register as ext2 (%d)\n", err); +} + +static inline void unregister_as_ext2(void) +{ + unregister_filesystem(&ext2_fs_type); +} + +static inline int ext2_feature_set_ok(struct super_block *sb) +{ + if (ext4_has_unknown_ext2_incompat_features(sb)) + return 0; + if (sb_rdonly(sb)) + return 1; + if (ext4_has_unknown_ext2_ro_compat_features(sb)) + return 0; + return 1; +} +#else +static inline void register_as_ext2(void) { } +static inline void unregister_as_ext2(void) { } +static inline int ext2_feature_set_ok(struct super_block *sb) { return 0; } +#endif + +static inline void register_as_ext3(void) +{ + int err = register_filesystem(&ext3_fs_type); + if (err) + printk(KERN_WARNING + "EXT4-fs: Unable to register as ext3 (%d)\n", err); +} + +static inline void unregister_as_ext3(void) +{ + unregister_filesystem(&ext3_fs_type); +} + +static inline int ext3_feature_set_ok(struct super_block *sb) +{ + if (ext4_has_unknown_ext3_incompat_features(sb)) + return 0; + if (!ext4_has_feature_journal(sb)) + return 0; + if (sb_rdonly(sb)) + return 1; + if (ext4_has_unknown_ext3_ro_compat_features(sb)) + return 0; + return 1; +} + +static void ext4_kill_sb(struct super_block *sb) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct file *bdev_file = sbi ? sbi->s_journal_bdev_file : NULL; + + kill_block_super(sb); + + if (bdev_file) + bdev_fput(bdev_file); +} + +static struct file_system_type ext4_fs_type = { + .owner = THIS_MODULE, + .name = "ext4", + .init_fs_context = ext4_init_fs_context, + .parameters = ext4_param_specs, + .kill_sb = ext4_kill_sb, + .fs_flags = FS_REQUIRES_DEV | FS_ALLOW_IDMAP | FS_MGTIME, +}; +MODULE_ALIAS_FS("ext4"); + +static int __init ext4_init_fs(void) +{ + int err; + + ratelimit_state_init(&ext4_mount_msg_ratelimit, 30 * HZ, 64); + ext4_li_info = NULL; + + /* Build-time check for flags consistency */ + ext4_check_flag_values(); + + err = ext4_init_es(); + if (err) + return err; + + err = ext4_init_pending(); + if (err) + goto out7; + + err = ext4_init_post_read_processing(); + if (err) + goto out6; + + err = ext4_init_pageio(); + if (err) + goto out5; + + err = ext4_init_system_zone(); + if (err) + goto out4; + + err = ext4_init_sysfs(); + if (err) + goto out3; + + err = ext4_init_mballoc(); + if (err) + goto out2; + err = init_inodecache(); + if (err) + goto out1; + + err = ext4_fc_init_dentry_cache(); + if (err) + goto out05; + + register_as_ext3(); + register_as_ext2(); + err = register_filesystem(&ext4_fs_type); + if (err) + goto out; + + return 0; +out: + unregister_as_ext2(); + unregister_as_ext3(); + ext4_fc_destroy_dentry_cache(); +out05: + destroy_inodecache(); +out1: + ext4_exit_mballoc(); +out2: + ext4_exit_sysfs(); +out3: + ext4_exit_system_zone(); +out4: + ext4_exit_pageio(); +out5: + ext4_exit_post_read_processing(); +out6: + ext4_exit_pending(); +out7: + ext4_exit_es(); + + return err; +} + +static void __exit ext4_exit_fs(void) +{ + ext4_destroy_lazyinit_thread(); + unregister_as_ext2(); + unregister_as_ext3(); + unregister_filesystem(&ext4_fs_type); + ext4_fc_destroy_dentry_cache(); + destroy_inodecache(); + ext4_exit_mballoc(); + ext4_exit_sysfs(); + ext4_exit_system_zone(); + ext4_exit_pageio(); + ext4_exit_post_read_processing(); + ext4_exit_es(); + ext4_exit_pending(); +} + +MODULE_AUTHOR("Remy Card, Stephen Tweedie, Andrew Morton, Andreas Dilger, Theodore Ts'o and others"); +MODULE_DESCRIPTION("Fourth Extended Filesystem"); +MODULE_LICENSE("GPL"); +module_init(ext4_init_fs) +module_exit(ext4_exit_fs) -- 2.43.0

From: Simon Glass <simon.glass@canonical.com> Copy acl.c, acl.h, balloc.c, bitmap.c, block_validity.c, crypto.c, and dir.c from Linux v6.18 fs/ext4 directory. - acl: POSIX ACL support - balloc: block allocation helpers - bitmap: bitmap manipulation utilities - block_validity: block validity checking - crypto: filesystem encryption support - dir: directory entry handling Co-developed-by: Claude Opus 4.5 <noreply@anthropic.com> --- fs/ext4l/acl.c | 304 +++++++++++ fs/ext4l/acl.h | 74 +++ fs/ext4l/balloc.c | 1003 +++++++++++++++++++++++++++++++++++++ fs/ext4l/bitmap.c | 99 ++++ fs/ext4l/block_validity.c | 370 ++++++++++++++ fs/ext4l/crypto.c | 241 +++++++++ fs/ext4l/dir.c | 693 +++++++++++++++++++++++++ 7 files changed, 2784 insertions(+) create mode 100644 fs/ext4l/acl.c create mode 100644 fs/ext4l/acl.h create mode 100644 fs/ext4l/balloc.c create mode 100644 fs/ext4l/bitmap.c create mode 100644 fs/ext4l/block_validity.c create mode 100644 fs/ext4l/crypto.c create mode 100644 fs/ext4l/dir.c diff --git a/fs/ext4l/acl.c b/fs/ext4l/acl.c new file mode 100644 index 00000000000..3bffe862f95 --- /dev/null +++ b/fs/ext4l/acl.c @@ -0,0 +1,304 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * linux/fs/ext4/acl.c + * + * Copyright (C) 2001-2003 Andreas Gruenbacher, <agruen@suse.de> + */ + +#include <linux/quotaops.h> +#include "ext4_jbd2.h" +#include "ext4.h" +#include "xattr.h" +#include "acl.h" + +/* + * Convert from filesystem to in-memory representation. + */ +static struct posix_acl * +ext4_acl_from_disk(const void *value, size_t size) +{ + const char *end = (char *)value + size; + int n, count; + struct posix_acl *acl; + + if (!value) + return NULL; + if (size < sizeof(ext4_acl_header)) + return ERR_PTR(-EINVAL); + if (((ext4_acl_header *)value)->a_version != + cpu_to_le32(EXT4_ACL_VERSION)) + return ERR_PTR(-EINVAL); + value = (char *)value + sizeof(ext4_acl_header); + count = ext4_acl_count(size); + if (count < 0) + return ERR_PTR(-EINVAL); + if (count == 0) + return NULL; + acl = posix_acl_alloc(count, GFP_NOFS); + if (!acl) + return ERR_PTR(-ENOMEM); + for (n = 0; n < count; n++) { + ext4_acl_entry *entry = + (ext4_acl_entry *)value; + if ((char *)value + sizeof(ext4_acl_entry_short) > end) + goto fail; + acl->a_entries[n].e_tag = le16_to_cpu(entry->e_tag); + acl->a_entries[n].e_perm = le16_to_cpu(entry->e_perm); + + switch (acl->a_entries[n].e_tag) { + case ACL_USER_OBJ: + case ACL_GROUP_OBJ: + case ACL_MASK: + case ACL_OTHER: + value = (char *)value + + sizeof(ext4_acl_entry_short); + break; + + case ACL_USER: + value = (char *)value + sizeof(ext4_acl_entry); + if ((char *)value > end) + goto fail; + acl->a_entries[n].e_uid = + make_kuid(&init_user_ns, + le32_to_cpu(entry->e_id)); + break; + case ACL_GROUP: + value = (char *)value + sizeof(ext4_acl_entry); + if ((char *)value > end) + goto fail; + acl->a_entries[n].e_gid = + make_kgid(&init_user_ns, + le32_to_cpu(entry->e_id)); + break; + + default: + goto fail; + } + } + if (value != end) + goto fail; + return acl; + +fail: + posix_acl_release(acl); + return ERR_PTR(-EINVAL); +} + +/* + * Convert from in-memory to filesystem representation. + */ +static void * +ext4_acl_to_disk(const struct posix_acl *acl, size_t *size) +{ + ext4_acl_header *ext_acl; + char *e; + size_t n; + + *size = ext4_acl_size(acl->a_count); + ext_acl = kmalloc(sizeof(ext4_acl_header) + acl->a_count * + sizeof(ext4_acl_entry), GFP_NOFS); + if (!ext_acl) + return ERR_PTR(-ENOMEM); + ext_acl->a_version = cpu_to_le32(EXT4_ACL_VERSION); + e = (char *)ext_acl + sizeof(ext4_acl_header); + for (n = 0; n < acl->a_count; n++) { + const struct posix_acl_entry *acl_e = &acl->a_entries[n]; + ext4_acl_entry *entry = (ext4_acl_entry *)e; + entry->e_tag = cpu_to_le16(acl_e->e_tag); + entry->e_perm = cpu_to_le16(acl_e->e_perm); + switch (acl_e->e_tag) { + case ACL_USER: + entry->e_id = cpu_to_le32( + from_kuid(&init_user_ns, acl_e->e_uid)); + e += sizeof(ext4_acl_entry); + break; + case ACL_GROUP: + entry->e_id = cpu_to_le32( + from_kgid(&init_user_ns, acl_e->e_gid)); + e += sizeof(ext4_acl_entry); + break; + + case ACL_USER_OBJ: + case ACL_GROUP_OBJ: + case ACL_MASK: + case ACL_OTHER: + e += sizeof(ext4_acl_entry_short); + break; + + default: + goto fail; + } + } + return (char *)ext_acl; + +fail: + kfree(ext_acl); + return ERR_PTR(-EINVAL); +} + +/* + * Inode operation get_posix_acl(). + * + * inode->i_rwsem: don't care + */ +struct posix_acl * +ext4_get_acl(struct inode *inode, int type, bool rcu) +{ + int name_index; + char *value = NULL; + struct posix_acl *acl; + int retval; + + if (rcu) + return ERR_PTR(-ECHILD); + + switch (type) { + case ACL_TYPE_ACCESS: + name_index = EXT4_XATTR_INDEX_POSIX_ACL_ACCESS; + break; + case ACL_TYPE_DEFAULT: + name_index = EXT4_XATTR_INDEX_POSIX_ACL_DEFAULT; + break; + default: + BUG(); + } + retval = ext4_xattr_get(inode, name_index, "", NULL, 0); + if (retval > 0) { + value = kmalloc(retval, GFP_NOFS); + if (!value) + return ERR_PTR(-ENOMEM); + retval = ext4_xattr_get(inode, name_index, "", value, retval); + } + if (retval > 0) + acl = ext4_acl_from_disk(value, retval); + else if (retval == -ENODATA || retval == -ENOSYS) + acl = NULL; + else + acl = ERR_PTR(retval); + kfree(value); + + return acl; +} + +/* + * Set the access or default ACL of an inode. + * + * inode->i_rwsem: down unless called from ext4_new_inode + */ +static int +__ext4_set_acl(handle_t *handle, struct inode *inode, int type, + struct posix_acl *acl, int xattr_flags) +{ + int name_index; + void *value = NULL; + size_t size = 0; + int error; + + switch (type) { + case ACL_TYPE_ACCESS: + name_index = EXT4_XATTR_INDEX_POSIX_ACL_ACCESS; + break; + + case ACL_TYPE_DEFAULT: + name_index = EXT4_XATTR_INDEX_POSIX_ACL_DEFAULT; + if (!S_ISDIR(inode->i_mode)) + return acl ? -EACCES : 0; + break; + + default: + return -EINVAL; + } + if (acl) { + value = ext4_acl_to_disk(acl, &size); + if (IS_ERR(value)) + return (int)PTR_ERR(value); + } + + error = ext4_xattr_set_handle(handle, inode, name_index, "", + value, size, xattr_flags); + + kfree(value); + if (!error) + set_cached_acl(inode, type, acl); + + return error; +} + +int +ext4_set_acl(struct mnt_idmap *idmap, struct dentry *dentry, + struct posix_acl *acl, int type) +{ + handle_t *handle; + int error, credits, retries = 0; + size_t acl_size = acl ? ext4_acl_size(acl->a_count) : 0; + struct inode *inode = d_inode(dentry); + umode_t mode = inode->i_mode; + int update_mode = 0; + + error = dquot_initialize(inode); + if (error) + return error; +retry: + error = ext4_xattr_set_credits(inode, acl_size, false /* is_create */, + &credits); + if (error) + return error; + + handle = ext4_journal_start(inode, EXT4_HT_XATTR, credits); + if (IS_ERR(handle)) + return PTR_ERR(handle); + + if ((type == ACL_TYPE_ACCESS) && acl) { + error = posix_acl_update_mode(idmap, inode, &mode, &acl); + if (error) + goto out_stop; + if (mode != inode->i_mode) + update_mode = 1; + } + + error = __ext4_set_acl(handle, inode, type, acl, 0 /* xattr_flags */); + if (!error && update_mode) { + inode->i_mode = mode; + inode_set_ctime_current(inode); + error = ext4_mark_inode_dirty(handle, inode); + } +out_stop: + ext4_journal_stop(handle); + if (error == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries)) + goto retry; + return error; +} + +/* + * Initialize the ACLs of a new inode. Called from ext4_new_inode. + * + * dir->i_rwsem: down + * inode->i_rwsem: up (access to inode is still exclusive) + */ +int +ext4_init_acl(handle_t *handle, struct inode *inode, struct inode *dir) +{ + struct posix_acl *default_acl, *acl; + int error; + + error = posix_acl_create(dir, &inode->i_mode, &default_acl, &acl); + if (error) + return error; + + if (default_acl) { + error = __ext4_set_acl(handle, inode, ACL_TYPE_DEFAULT, + default_acl, XATTR_CREATE); + posix_acl_release(default_acl); + } else { + inode->i_default_acl = NULL; + } + if (acl) { + if (!error) + error = __ext4_set_acl(handle, inode, ACL_TYPE_ACCESS, + acl, XATTR_CREATE); + posix_acl_release(acl); + } else { + inode->i_acl = NULL; + } + return error; +} diff --git a/fs/ext4l/acl.h b/fs/ext4l/acl.h new file mode 100644 index 00000000000..0c5a79c3b5d --- /dev/null +++ b/fs/ext4l/acl.h @@ -0,0 +1,74 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + File: fs/ext4/acl.h + + (C) 2001 Andreas Gruenbacher, <a.gruenbacher@computer.org> +*/ + +#include <linux/posix_acl_xattr.h> + +#define EXT4_ACL_VERSION 0x0001 + +typedef struct { + __le16 e_tag; + __le16 e_perm; + __le32 e_id; +} ext4_acl_entry; + +typedef struct { + __le16 e_tag; + __le16 e_perm; +} ext4_acl_entry_short; + +typedef struct { + __le32 a_version; +} ext4_acl_header; + +static inline size_t ext4_acl_size(int count) +{ + if (count <= 4) { + return sizeof(ext4_acl_header) + + count * sizeof(ext4_acl_entry_short); + } else { + return sizeof(ext4_acl_header) + + 4 * sizeof(ext4_acl_entry_short) + + (count - 4) * sizeof(ext4_acl_entry); + } +} + +static inline int ext4_acl_count(size_t size) +{ + ssize_t s; + size -= sizeof(ext4_acl_header); + s = size - 4 * sizeof(ext4_acl_entry_short); + if (s < 0) { + if (size % sizeof(ext4_acl_entry_short)) + return -1; + return size / sizeof(ext4_acl_entry_short); + } else { + if (s % sizeof(ext4_acl_entry)) + return -1; + return s / sizeof(ext4_acl_entry) + 4; + } +} + +#ifdef CONFIG_EXT4_FS_POSIX_ACL + +/* acl.c */ +struct posix_acl *ext4_get_acl(struct inode *inode, int type, bool rcu); +int ext4_set_acl(struct mnt_idmap *idmap, struct dentry *dentry, + struct posix_acl *acl, int type); +extern int ext4_init_acl(handle_t *, struct inode *, struct inode *); + +#else /* CONFIG_EXT4_FS_POSIX_ACL */ +#include <linux/sched.h> +#define ext4_get_acl NULL +#define ext4_set_acl NULL + +static inline int +ext4_init_acl(handle_t *handle, struct inode *inode, struct inode *dir) +{ + return 0; +} +#endif /* CONFIG_EXT4_FS_POSIX_ACL */ + diff --git a/fs/ext4l/balloc.c b/fs/ext4l/balloc.c new file mode 100644 index 00000000000..c9329ed5c09 --- /dev/null +++ b/fs/ext4l/balloc.c @@ -0,0 +1,1003 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * linux/fs/ext4/balloc.c + * + * Copyright (C) 1992, 1993, 1994, 1995 + * Remy Card (card@masi.ibp.fr) + * Laboratoire MASI - Institut Blaise Pascal + * Universite Pierre et Marie Curie (Paris VI) + * + * Enhanced block allocation by Stephen Tweedie (sct@redhat.com), 1993 + * Big-endian to little-endian byte-swapping/bitmaps by + * David S. Miller (davem@caip.rutgers.edu), 1995 + */ + +#include <linux/time.h> +#include <linux/capability.h> +#include <linux/fs.h> +#include <linux/quotaops.h> +#include <linux/buffer_head.h> +#include "ext4.h" +#include "ext4_jbd2.h" +#include "mballoc.h" + +#include <trace/events/ext4.h> +#include <kunit/static_stub.h> + +static unsigned ext4_num_base_meta_clusters(struct super_block *sb, + ext4_group_t block_group); +/* + * balloc.c contains the blocks allocation and deallocation routines + */ + +/* + * Calculate block group number for a given block number + */ +ext4_group_t ext4_get_group_number(struct super_block *sb, + ext4_fsblk_t block) +{ + ext4_group_t group; + + if (test_opt2(sb, STD_GROUP_SIZE)) + group = (block - + le32_to_cpu(EXT4_SB(sb)->s_es->s_first_data_block)) >> + (EXT4_BLOCK_SIZE_BITS(sb) + EXT4_CLUSTER_BITS(sb) + 3); + else + ext4_get_group_no_and_offset(sb, block, &group, NULL); + return group; +} + +/* + * Calculate the block group number and offset into the block/cluster + * allocation bitmap, given a block number + */ +void ext4_get_group_no_and_offset(struct super_block *sb, ext4_fsblk_t blocknr, + ext4_group_t *blockgrpp, ext4_grpblk_t *offsetp) +{ + struct ext4_super_block *es = EXT4_SB(sb)->s_es; + ext4_grpblk_t offset; + + blocknr = blocknr - le32_to_cpu(es->s_first_data_block); + offset = do_div(blocknr, EXT4_BLOCKS_PER_GROUP(sb)) >> + EXT4_SB(sb)->s_cluster_bits; + if (offsetp) + *offsetp = offset; + if (blockgrpp) + *blockgrpp = blocknr; + +} + +/* + * Check whether the 'block' lives within the 'block_group'. Returns 1 if so + * and 0 otherwise. + */ +static inline int ext4_block_in_group(struct super_block *sb, + ext4_fsblk_t block, + ext4_group_t block_group) +{ + ext4_group_t actual_group; + + actual_group = ext4_get_group_number(sb, block); + return (actual_group == block_group) ? 1 : 0; +} + +/* + * Return the number of clusters used for file system metadata; this + * represents the overhead needed by the file system. + */ +static unsigned ext4_num_overhead_clusters(struct super_block *sb, + ext4_group_t block_group, + struct ext4_group_desc *gdp) +{ + unsigned base_clusters, num_clusters; + int block_cluster = -1, inode_cluster; + int itbl_cluster_start = -1, itbl_cluster_end = -1; + ext4_fsblk_t start = ext4_group_first_block_no(sb, block_group); + ext4_fsblk_t end = start + EXT4_BLOCKS_PER_GROUP(sb) - 1; + ext4_fsblk_t itbl_blk_start, itbl_blk_end; + struct ext4_sb_info *sbi = EXT4_SB(sb); + + /* This is the number of clusters used by the superblock, + * block group descriptors, and reserved block group + * descriptor blocks */ + base_clusters = ext4_num_base_meta_clusters(sb, block_group); + num_clusters = base_clusters; + + /* + * Account and record inode table clusters if any cluster + * is in the block group, or inode table cluster range is + * [-1, -1] and won't overlap with block/inode bitmap cluster + * accounted below. + */ + itbl_blk_start = ext4_inode_table(sb, gdp); + itbl_blk_end = itbl_blk_start + sbi->s_itb_per_group - 1; + if (itbl_blk_start <= end && itbl_blk_end >= start) { + itbl_blk_start = max(itbl_blk_start, start); + itbl_blk_end = min(itbl_blk_end, end); + + itbl_cluster_start = EXT4_B2C(sbi, itbl_blk_start - start); + itbl_cluster_end = EXT4_B2C(sbi, itbl_blk_end - start); + + num_clusters += itbl_cluster_end - itbl_cluster_start + 1; + /* check if border cluster is overlapped */ + if (itbl_cluster_start == base_clusters - 1) + num_clusters--; + } + + /* + * For the allocation bitmaps, we first need to check to see + * if the block is in the block group. If it is, then check + * to see if the cluster is already accounted for in the clusters + * used for the base metadata cluster and inode tables cluster. + * Normally all of these blocks are contiguous, so the special + * case handling shouldn't be necessary except for *very* + * unusual file system layouts. + */ + if (ext4_block_in_group(sb, ext4_block_bitmap(sb, gdp), block_group)) { + block_cluster = EXT4_B2C(sbi, + ext4_block_bitmap(sb, gdp) - start); + if (block_cluster >= base_clusters && + (block_cluster < itbl_cluster_start || + block_cluster > itbl_cluster_end)) + num_clusters++; + } + + if (ext4_block_in_group(sb, ext4_inode_bitmap(sb, gdp), block_group)) { + inode_cluster = EXT4_B2C(sbi, + ext4_inode_bitmap(sb, gdp) - start); + /* + * Additional check if inode bitmap is in just accounted + * block_cluster + */ + if (inode_cluster != block_cluster && + inode_cluster >= base_clusters && + (inode_cluster < itbl_cluster_start || + inode_cluster > itbl_cluster_end)) + num_clusters++; + } + + return num_clusters; +} + +static unsigned int num_clusters_in_group(struct super_block *sb, + ext4_group_t block_group) +{ + unsigned int blocks; + + if (block_group == ext4_get_groups_count(sb) - 1) { + /* + * Even though mke2fs always initializes the first and + * last group, just in case some other tool was used, + * we need to make sure we calculate the right free + * blocks. + */ + blocks = ext4_blocks_count(EXT4_SB(sb)->s_es) - + ext4_group_first_block_no(sb, block_group); + } else + blocks = EXT4_BLOCKS_PER_GROUP(sb); + return EXT4_NUM_B2C(EXT4_SB(sb), blocks); +} + +/* Initializes an uninitialized block bitmap */ +static int ext4_init_block_bitmap(struct super_block *sb, + struct buffer_head *bh, + ext4_group_t block_group, + struct ext4_group_desc *gdp) +{ + unsigned int bit, bit_max; + struct ext4_sb_info *sbi = EXT4_SB(sb); + ext4_fsblk_t start, tmp; + + ASSERT(buffer_locked(bh)); + + if (!ext4_group_desc_csum_verify(sb, block_group, gdp)) { + ext4_mark_group_bitmap_corrupted(sb, block_group, + EXT4_GROUP_INFO_BBITMAP_CORRUPT | + EXT4_GROUP_INFO_IBITMAP_CORRUPT); + return -EFSBADCRC; + } + memset(bh->b_data, 0, sb->s_blocksize); + + bit_max = ext4_num_base_meta_clusters(sb, block_group); + if ((bit_max >> 3) >= bh->b_size) + return -EFSCORRUPTED; + + for (bit = 0; bit < bit_max; bit++) + ext4_set_bit(bit, bh->b_data); + + start = ext4_group_first_block_no(sb, block_group); + + /* Set bits for block and inode bitmaps, and inode table */ + tmp = ext4_block_bitmap(sb, gdp); + if (ext4_block_in_group(sb, tmp, block_group)) + ext4_set_bit(EXT4_B2C(sbi, tmp - start), bh->b_data); + + tmp = ext4_inode_bitmap(sb, gdp); + if (ext4_block_in_group(sb, tmp, block_group)) + ext4_set_bit(EXT4_B2C(sbi, tmp - start), bh->b_data); + + tmp = ext4_inode_table(sb, gdp); + for (; tmp < ext4_inode_table(sb, gdp) + + sbi->s_itb_per_group; tmp++) { + if (ext4_block_in_group(sb, tmp, block_group)) + ext4_set_bit(EXT4_B2C(sbi, tmp - start), bh->b_data); + } + + /* + * Also if the number of blocks within the group is less than + * the blocksize * 8 ( which is the size of bitmap ), set rest + * of the block bitmap to 1 + */ + ext4_mark_bitmap_end(num_clusters_in_group(sb, block_group), + sb->s_blocksize * 8, bh->b_data); + return 0; +} + +/* Return the number of free blocks in a block group. It is used when + * the block bitmap is uninitialized, so we can't just count the bits + * in the bitmap. */ +unsigned ext4_free_clusters_after_init(struct super_block *sb, + ext4_group_t block_group, + struct ext4_group_desc *gdp) +{ + return num_clusters_in_group(sb, block_group) - + ext4_num_overhead_clusters(sb, block_group, gdp); +} + +/* + * The free blocks are managed by bitmaps. A file system contains several + * blocks groups. Each group contains 1 bitmap block for blocks, 1 bitmap + * block for inodes, N blocks for the inode table and data blocks. + * + * The file system contains group descriptors which are located after the + * super block. Each descriptor contains the number of the bitmap block and + * the free blocks count in the block. The descriptors are loaded in memory + * when a file system is mounted (see ext4_fill_super). + */ + +/** + * ext4_get_group_desc() -- load group descriptor from disk + * @sb: super block + * @block_group: given block group + * @bh: pointer to the buffer head to store the block + * group descriptor + */ +struct ext4_group_desc * ext4_get_group_desc(struct super_block *sb, + ext4_group_t block_group, + struct buffer_head **bh) +{ + unsigned int group_desc; + unsigned int offset; + ext4_group_t ngroups = ext4_get_groups_count(sb); + struct ext4_group_desc *desc; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct buffer_head *bh_p; + + KUNIT_STATIC_STUB_REDIRECT(ext4_get_group_desc, + sb, block_group, bh); + + if (block_group >= ngroups) { + ext4_error(sb, "block_group >= groups_count - block_group = %u," + " groups_count = %u", block_group, ngroups); + + return NULL; + } + + group_desc = block_group >> EXT4_DESC_PER_BLOCK_BITS(sb); + offset = block_group & (EXT4_DESC_PER_BLOCK(sb) - 1); + bh_p = sbi_array_rcu_deref(sbi, s_group_desc, group_desc); + /* + * sbi_array_rcu_deref returns with rcu unlocked, this is ok since + * the pointer being dereferenced won't be dereferenced again. By + * looking at the usage in add_new_gdb() the value isn't modified, + * just the pointer, and so it remains valid. + */ + if (!bh_p) { + ext4_error(sb, "Group descriptor not loaded - " + "block_group = %u, group_desc = %u, desc = %u", + block_group, group_desc, offset); + return NULL; + } + + desc = (struct ext4_group_desc *)( + (__u8 *)bh_p->b_data + + offset * EXT4_DESC_SIZE(sb)); + if (bh) + *bh = bh_p; + return desc; +} + +static ext4_fsblk_t ext4_valid_block_bitmap_padding(struct super_block *sb, + ext4_group_t block_group, + struct buffer_head *bh) +{ + ext4_grpblk_t next_zero_bit; + unsigned long bitmap_size = sb->s_blocksize * 8; + unsigned int offset = num_clusters_in_group(sb, block_group); + + if (bitmap_size <= offset) + return 0; + + next_zero_bit = ext4_find_next_zero_bit(bh->b_data, bitmap_size, offset); + + return (next_zero_bit < bitmap_size ? next_zero_bit : 0); +} + +struct ext4_group_info *ext4_get_group_info(struct super_block *sb, + ext4_group_t group) +{ + struct ext4_group_info **grp_info; + long indexv, indexh; + + if (unlikely(group >= EXT4_SB(sb)->s_groups_count)) + return NULL; + indexv = group >> (EXT4_DESC_PER_BLOCK_BITS(sb)); + indexh = group & ((EXT4_DESC_PER_BLOCK(sb)) - 1); + grp_info = sbi_array_rcu_deref(EXT4_SB(sb), s_group_info, indexv); + return grp_info[indexh]; +} + +/* + * Return the block number which was discovered to be invalid, or 0 if + * the block bitmap is valid. + */ +static ext4_fsblk_t ext4_valid_block_bitmap(struct super_block *sb, + struct ext4_group_desc *desc, + ext4_group_t block_group, + struct buffer_head *bh) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + ext4_grpblk_t offset; + ext4_grpblk_t next_zero_bit; + ext4_grpblk_t max_bit = EXT4_CLUSTERS_PER_GROUP(sb); + ext4_fsblk_t blk; + ext4_fsblk_t group_first_block; + + if (ext4_has_feature_flex_bg(sb)) { + /* with FLEX_BG, the inode/block bitmaps and itable + * blocks may not be in the group at all + * so the bitmap validation will be skipped for those groups + * or it has to also read the block group where the bitmaps + * are located to verify they are set. + */ + return 0; + } + group_first_block = ext4_group_first_block_no(sb, block_group); + + /* check whether block bitmap block number is set */ + blk = ext4_block_bitmap(sb, desc); + offset = blk - group_first_block; + if (offset < 0 || EXT4_B2C(sbi, offset) >= max_bit || + !ext4_test_bit(EXT4_B2C(sbi, offset), bh->b_data)) + /* bad block bitmap */ + return blk; + + /* check whether the inode bitmap block number is set */ + blk = ext4_inode_bitmap(sb, desc); + offset = blk - group_first_block; + if (offset < 0 || EXT4_B2C(sbi, offset) >= max_bit || + !ext4_test_bit(EXT4_B2C(sbi, offset), bh->b_data)) + /* bad block bitmap */ + return blk; + + /* check whether the inode table block number is set */ + blk = ext4_inode_table(sb, desc); + offset = blk - group_first_block; + if (offset < 0 || EXT4_B2C(sbi, offset) >= max_bit || + EXT4_B2C(sbi, offset + sbi->s_itb_per_group - 1) >= max_bit) + return blk; + next_zero_bit = ext4_find_next_zero_bit(bh->b_data, + EXT4_B2C(sbi, offset + sbi->s_itb_per_group - 1) + 1, + EXT4_B2C(sbi, offset)); + if (next_zero_bit < + EXT4_B2C(sbi, offset + sbi->s_itb_per_group - 1) + 1) + /* bad bitmap for inode tables */ + return blk; + return 0; +} + +static int ext4_validate_block_bitmap(struct super_block *sb, + struct ext4_group_desc *desc, + ext4_group_t block_group, + struct buffer_head *bh) +{ + ext4_fsblk_t blk; + struct ext4_group_info *grp; + + if (EXT4_SB(sb)->s_mount_state & EXT4_FC_REPLAY) + return 0; + + grp = ext4_get_group_info(sb, block_group); + + if (buffer_verified(bh)) + return 0; + if (!grp || EXT4_MB_GRP_BBITMAP_CORRUPT(grp)) + return -EFSCORRUPTED; + + ext4_lock_group(sb, block_group); + if (buffer_verified(bh)) + goto verified; + if (unlikely(!ext4_block_bitmap_csum_verify(sb, desc, bh) || + ext4_simulate_fail(sb, EXT4_SIM_BBITMAP_CRC))) { + ext4_unlock_group(sb, block_group); + ext4_error(sb, "bg %u: bad block bitmap checksum", block_group); + ext4_mark_group_bitmap_corrupted(sb, block_group, + EXT4_GROUP_INFO_BBITMAP_CORRUPT); + return -EFSBADCRC; + } + blk = ext4_valid_block_bitmap(sb, desc, block_group, bh); + if (unlikely(blk != 0)) { + ext4_unlock_group(sb, block_group); + ext4_error(sb, "bg %u: block %llu: invalid block bitmap", + block_group, blk); + ext4_mark_group_bitmap_corrupted(sb, block_group, + EXT4_GROUP_INFO_BBITMAP_CORRUPT); + return -EFSCORRUPTED; + } + blk = ext4_valid_block_bitmap_padding(sb, block_group, bh); + if (unlikely(blk != 0)) { + ext4_unlock_group(sb, block_group); + ext4_error(sb, "bg %u: block %llu: padding at end of block bitmap is not set", + block_group, blk); + ext4_mark_group_bitmap_corrupted(sb, block_group, + EXT4_GROUP_INFO_BBITMAP_CORRUPT); + return -EFSCORRUPTED; + } + set_buffer_verified(bh); +verified: + ext4_unlock_group(sb, block_group); + return 0; +} + +/** + * ext4_read_block_bitmap_nowait() + * @sb: super block + * @block_group: given block group + * @ignore_locked: ignore locked buffers + * + * Read the bitmap for a given block_group,and validate the + * bits for block/inode/inode tables are set in the bitmaps + * + * Return buffer_head on success or an ERR_PTR in case of failure. + */ +struct buffer_head * +ext4_read_block_bitmap_nowait(struct super_block *sb, ext4_group_t block_group, + bool ignore_locked) +{ + struct ext4_group_desc *desc; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct buffer_head *bh; + ext4_fsblk_t bitmap_blk; + int err; + + KUNIT_STATIC_STUB_REDIRECT(ext4_read_block_bitmap_nowait, + sb, block_group, ignore_locked); + + desc = ext4_get_group_desc(sb, block_group, NULL); + if (!desc) + return ERR_PTR(-EFSCORRUPTED); + bitmap_blk = ext4_block_bitmap(sb, desc); + if ((bitmap_blk <= le32_to_cpu(sbi->s_es->s_first_data_block)) || + (bitmap_blk >= ext4_blocks_count(sbi->s_es))) { + ext4_error(sb, "Invalid block bitmap block %llu in " + "block_group %u", bitmap_blk, block_group); + ext4_mark_group_bitmap_corrupted(sb, block_group, + EXT4_GROUP_INFO_BBITMAP_CORRUPT); + return ERR_PTR(-EFSCORRUPTED); + } + bh = sb_getblk(sb, bitmap_blk); + if (unlikely(!bh)) { + ext4_warning(sb, "Cannot get buffer for block bitmap - " + "block_group = %u, block_bitmap = %llu", + block_group, bitmap_blk); + return ERR_PTR(-ENOMEM); + } + + if (ignore_locked && buffer_locked(bh)) { + /* buffer under IO already, return if called for prefetching */ + put_bh(bh); + return NULL; + } + + if (bitmap_uptodate(bh)) + goto verify; + + lock_buffer(bh); + if (bitmap_uptodate(bh)) { + unlock_buffer(bh); + goto verify; + } + ext4_lock_group(sb, block_group); + if (ext4_has_group_desc_csum(sb) && + (desc->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT))) { + if (block_group == 0) { + ext4_unlock_group(sb, block_group); + unlock_buffer(bh); + ext4_error(sb, "Block bitmap for bg 0 marked " + "uninitialized"); + err = -EFSCORRUPTED; + goto out; + } + err = ext4_init_block_bitmap(sb, bh, block_group, desc); + if (err) { + ext4_unlock_group(sb, block_group); + unlock_buffer(bh); + ext4_error(sb, "Failed to init block bitmap for group " + "%u: %d", block_group, err); + goto out; + } + set_bitmap_uptodate(bh); + set_buffer_uptodate(bh); + set_buffer_verified(bh); + ext4_unlock_group(sb, block_group); + unlock_buffer(bh); + return bh; + } + ext4_unlock_group(sb, block_group); + if (buffer_uptodate(bh)) { + /* + * if not uninit if bh is uptodate, + * bitmap is also uptodate + */ + set_bitmap_uptodate(bh); + unlock_buffer(bh); + goto verify; + } + /* + * submit the buffer_head for reading + */ + set_buffer_new(bh); + trace_ext4_read_block_bitmap_load(sb, block_group, ignore_locked); + ext4_read_bh_nowait(bh, REQ_META | REQ_PRIO | + (ignore_locked ? REQ_RAHEAD : 0), + ext4_end_bitmap_read, + ext4_simulate_fail(sb, EXT4_SIM_BBITMAP_EIO)); + return bh; +verify: + err = ext4_validate_block_bitmap(sb, desc, block_group, bh); + if (err) + goto out; + return bh; +out: + put_bh(bh); + return ERR_PTR(err); +} + +/* Returns 0 on success, -errno on error */ +int ext4_wait_block_bitmap(struct super_block *sb, ext4_group_t block_group, + struct buffer_head *bh) +{ + struct ext4_group_desc *desc; + + KUNIT_STATIC_STUB_REDIRECT(ext4_wait_block_bitmap, + sb, block_group, bh); + + if (!buffer_new(bh)) + return 0; + desc = ext4_get_group_desc(sb, block_group, NULL); + if (!desc) + return -EFSCORRUPTED; + wait_on_buffer(bh); + if (!buffer_uptodate(bh)) { + ext4_error_err(sb, EIO, "Cannot read block bitmap - " + "block_group = %u, block_bitmap = %llu", + block_group, (unsigned long long) bh->b_blocknr); + ext4_mark_group_bitmap_corrupted(sb, block_group, + EXT4_GROUP_INFO_BBITMAP_CORRUPT); + return -EIO; + } + clear_buffer_new(bh); + /* Panic or remount fs read-only if block bitmap is invalid */ + return ext4_validate_block_bitmap(sb, desc, block_group, bh); +} + +struct buffer_head * +ext4_read_block_bitmap(struct super_block *sb, ext4_group_t block_group) +{ + struct buffer_head *bh; + int err; + + bh = ext4_read_block_bitmap_nowait(sb, block_group, false); + if (IS_ERR(bh)) + return bh; + err = ext4_wait_block_bitmap(sb, block_group, bh); + if (err) { + put_bh(bh); + return ERR_PTR(err); + } + return bh; +} + +/** + * ext4_has_free_clusters() + * @sbi: in-core super block structure. + * @nclusters: number of needed blocks + * @flags: flags from ext4_mb_new_blocks() + * + * Check if filesystem has nclusters free & available for allocation. + * On success return 1, return 0 on failure. + */ +static int ext4_has_free_clusters(struct ext4_sb_info *sbi, + s64 nclusters, unsigned int flags) +{ + s64 free_clusters, dirty_clusters, rsv, resv_clusters; + struct percpu_counter *fcc = &sbi->s_freeclusters_counter; + struct percpu_counter *dcc = &sbi->s_dirtyclusters_counter; + + free_clusters = percpu_counter_read_positive(fcc); + dirty_clusters = percpu_counter_read_positive(dcc); + resv_clusters = atomic64_read(&sbi->s_resv_clusters); + + /* + * r_blocks_count should always be multiple of the cluster ratio so + * we are safe to do a plane bit shift only. + */ + rsv = (ext4_r_blocks_count(sbi->s_es) >> sbi->s_cluster_bits) + + resv_clusters; + + if (free_clusters - (nclusters + rsv + dirty_clusters) < + EXT4_FREECLUSTERS_WATERMARK) { + free_clusters = percpu_counter_sum_positive(fcc); + dirty_clusters = percpu_counter_sum_positive(dcc); + } + /* Check whether we have space after accounting for current + * dirty clusters & root reserved clusters. + */ + if (free_clusters >= (rsv + nclusters + dirty_clusters)) + return 1; + + /* Hm, nope. Are (enough) root reserved clusters available? */ + if (uid_eq(sbi->s_resuid, current_fsuid()) || + (!gid_eq(sbi->s_resgid, GLOBAL_ROOT_GID) && in_group_p(sbi->s_resgid)) || + (flags & EXT4_MB_USE_ROOT_BLOCKS) || + capable(CAP_SYS_RESOURCE)) { + + if (free_clusters >= (nclusters + dirty_clusters + + resv_clusters)) + return 1; + } + /* No free blocks. Let's see if we can dip into reserved pool */ + if (flags & EXT4_MB_USE_RESERVED) { + if (free_clusters >= (nclusters + dirty_clusters)) + return 1; + } + + return 0; +} + +int ext4_claim_free_clusters(struct ext4_sb_info *sbi, + s64 nclusters, unsigned int flags) +{ + if (ext4_has_free_clusters(sbi, nclusters, flags)) { + percpu_counter_add(&sbi->s_dirtyclusters_counter, nclusters); + return 0; + } else + return -ENOSPC; +} + +/** + * ext4_should_retry_alloc() - check if a block allocation should be retried + * @sb: superblock + * @retries: number of retry attempts made so far + * + * ext4_should_retry_alloc() is called when ENOSPC is returned while + * attempting to allocate blocks. If there's an indication that a pending + * journal transaction might free some space and allow another attempt to + * succeed, this function will wait for the current or committing transaction + * to complete and then return TRUE. + */ +int ext4_should_retry_alloc(struct super_block *sb, int *retries) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + + if (!sbi->s_journal) + return 0; + + if (++(*retries) > 3) { + percpu_counter_inc(&sbi->s_sra_exceeded_retry_limit); + return 0; + } + + /* + * if there's no indication that blocks are about to be freed it's + * possible we just missed a transaction commit that did so + */ + smp_mb(); + if (atomic_read(&sbi->s_mb_free_pending) == 0) { + if (test_opt(sb, DISCARD)) { + atomic_inc(&sbi->s_retry_alloc_pending); + flush_work(&sbi->s_discard_work); + atomic_dec(&sbi->s_retry_alloc_pending); + } + return ext4_has_free_clusters(sbi, 1, 0); + } + + /* + * it's possible we've just missed a transaction commit here, + * so ignore the returned status + */ + ext4_debug("%s: retrying operation after ENOSPC\n", sb->s_id); + (void) jbd2_journal_force_commit_nested(sbi->s_journal); + return 1; +} + +/* + * ext4_new_meta_blocks() -- allocate block for meta data (indexing) blocks + * + * @handle: handle to this transaction + * @inode: file inode + * @goal: given target block(filesystem wide) + * @count: pointer to total number of clusters needed + * @errp: error code + * + * Return 1st allocated block number on success, *count stores total account + * error stores in errp pointer + */ +ext4_fsblk_t ext4_new_meta_blocks(handle_t *handle, struct inode *inode, + ext4_fsblk_t goal, unsigned int flags, + unsigned long *count, int *errp) +{ + struct ext4_allocation_request ar; + ext4_fsblk_t ret; + + memset(&ar, 0, sizeof(ar)); + /* Fill with neighbour allocated blocks */ + ar.inode = inode; + ar.goal = goal; + ar.len = count ? *count : 1; + ar.flags = flags; + + ret = ext4_mb_new_blocks(handle, &ar, errp); + if (count) + *count = ar.len; + /* + * Account for the allocated meta blocks. We will never + * fail EDQUOT for metdata, but we do account for it. + */ + if (!(*errp) && (flags & EXT4_MB_DELALLOC_RESERVED)) { + dquot_alloc_block_nofail(inode, + EXT4_C2B(EXT4_SB(inode->i_sb), ar.len)); + } + return ret; +} + +/** + * ext4_count_free_clusters() -- count filesystem free clusters + * @sb: superblock + * + * Adds up the number of free clusters from each block group. + */ +ext4_fsblk_t ext4_count_free_clusters(struct super_block *sb) +{ + ext4_fsblk_t desc_count; + struct ext4_group_desc *gdp; + ext4_group_t i; + ext4_group_t ngroups = ext4_get_groups_count(sb); + struct ext4_group_info *grp; +#ifdef EXT4FS_DEBUG + struct ext4_super_block *es; + ext4_fsblk_t bitmap_count; + unsigned int x; + struct buffer_head *bitmap_bh = NULL; + + es = EXT4_SB(sb)->s_es; + desc_count = 0; + bitmap_count = 0; + gdp = NULL; + + for (i = 0; i < ngroups; i++) { + gdp = ext4_get_group_desc(sb, i, NULL); + if (!gdp) + continue; + grp = NULL; + if (EXT4_SB(sb)->s_group_info) + grp = ext4_get_group_info(sb, i); + if (!grp || !EXT4_MB_GRP_BBITMAP_CORRUPT(grp)) + desc_count += ext4_free_group_clusters(sb, gdp); + brelse(bitmap_bh); + bitmap_bh = ext4_read_block_bitmap(sb, i); + if (IS_ERR(bitmap_bh)) { + bitmap_bh = NULL; + continue; + } + + x = ext4_count_free(bitmap_bh->b_data, + EXT4_CLUSTERS_PER_GROUP(sb) / 8); + printk(KERN_DEBUG "group %u: stored = %d, counted = %u\n", + i, ext4_free_group_clusters(sb, gdp), x); + bitmap_count += x; + } + brelse(bitmap_bh); + printk(KERN_DEBUG "ext4_count_free_clusters: stored = %llu" + ", computed = %llu, %llu\n", + EXT4_NUM_B2C(EXT4_SB(sb), ext4_free_blocks_count(es)), + desc_count, bitmap_count); + return bitmap_count; +#else + desc_count = 0; + for (i = 0; i < ngroups; i++) { + gdp = ext4_get_group_desc(sb, i, NULL); + if (!gdp) + continue; + grp = NULL; + if (EXT4_SB(sb)->s_group_info) + grp = ext4_get_group_info(sb, i); + if (!grp || !EXT4_MB_GRP_BBITMAP_CORRUPT(grp)) + desc_count += ext4_free_group_clusters(sb, gdp); + } + + return desc_count; +#endif +} + +static inline int test_root(ext4_group_t a, int b) +{ + while (1) { + if (a < b) + return 0; + if (a == b) + return 1; + if ((a % b) != 0) + return 0; + a = a / b; + } +} + +/** + * ext4_bg_has_super - number of blocks used by the superblock in group + * @sb: superblock for filesystem + * @group: group number to check + * + * Return the number of blocks used by the superblock (primary or backup) + * in this group. Currently this will be only 0 or 1. + */ +int ext4_bg_has_super(struct super_block *sb, ext4_group_t group) +{ + struct ext4_super_block *es = EXT4_SB(sb)->s_es; + + if (group == 0) + return 1; + if (ext4_has_feature_sparse_super2(sb)) { + if (group == le32_to_cpu(es->s_backup_bgs[0]) || + group == le32_to_cpu(es->s_backup_bgs[1])) + return 1; + return 0; + } + if ((group <= 1) || !ext4_has_feature_sparse_super(sb)) + return 1; + if (!(group & 1)) + return 0; + if (test_root(group, 3) || (test_root(group, 5)) || + test_root(group, 7)) + return 1; + + return 0; +} + +static unsigned long ext4_bg_num_gdb_meta(struct super_block *sb, + ext4_group_t group) +{ + unsigned long metagroup = group / EXT4_DESC_PER_BLOCK(sb); + ext4_group_t first = metagroup * EXT4_DESC_PER_BLOCK(sb); + ext4_group_t last = first + EXT4_DESC_PER_BLOCK(sb) - 1; + + if (group == first || group == first + 1 || group == last) + return 1; + return 0; +} + +static unsigned long ext4_bg_num_gdb_nometa(struct super_block *sb, + ext4_group_t group) +{ + if (!ext4_bg_has_super(sb, group)) + return 0; + + if (ext4_has_feature_meta_bg(sb)) + return le32_to_cpu(EXT4_SB(sb)->s_es->s_first_meta_bg); + else + return EXT4_SB(sb)->s_gdb_count; +} + +/** + * ext4_bg_num_gdb - number of blocks used by the group table in group + * @sb: superblock for filesystem + * @group: group number to check + * + * Return the number of blocks used by the group descriptor table + * (primary or backup) in this group. In the future there may be a + * different number of descriptor blocks in each group. + */ +unsigned long ext4_bg_num_gdb(struct super_block *sb, ext4_group_t group) +{ + unsigned long first_meta_bg = + le32_to_cpu(EXT4_SB(sb)->s_es->s_first_meta_bg); + unsigned long metagroup = group / EXT4_DESC_PER_BLOCK(sb); + + if (!ext4_has_feature_meta_bg(sb) || metagroup < first_meta_bg) + return ext4_bg_num_gdb_nometa(sb, group); + + return ext4_bg_num_gdb_meta(sb,group); + +} + +/* + * This function returns the number of file system metadata blocks at + * the beginning of a block group, including the reserved gdt blocks. + */ +unsigned int ext4_num_base_meta_blocks(struct super_block *sb, + ext4_group_t block_group) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + unsigned num; + + /* Check for superblock and gdt backups in this group */ + num = ext4_bg_has_super(sb, block_group); + + if (!ext4_has_feature_meta_bg(sb) || + block_group < le32_to_cpu(sbi->s_es->s_first_meta_bg) * + sbi->s_desc_per_block) { + if (num) { + num += ext4_bg_num_gdb_nometa(sb, block_group); + num += le16_to_cpu(sbi->s_es->s_reserved_gdt_blocks); + } + } else { /* For META_BG_BLOCK_GROUPS */ + num += ext4_bg_num_gdb_meta(sb, block_group); + } + return num; +} + +static unsigned int ext4_num_base_meta_clusters(struct super_block *sb, + ext4_group_t block_group) +{ + return EXT4_NUM_B2C(EXT4_SB(sb), ext4_num_base_meta_blocks(sb, block_group)); +} + +/** + * ext4_inode_to_goal_block - return a hint for block allocation + * @inode: inode for block allocation + * + * Return the ideal location to start allocating blocks for a + * newly created inode. + */ +ext4_fsblk_t ext4_inode_to_goal_block(struct inode *inode) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + ext4_group_t block_group; + ext4_grpblk_t colour; + int flex_size = ext4_flex_bg_size(EXT4_SB(inode->i_sb)); + ext4_fsblk_t bg_start; + ext4_fsblk_t last_block; + + block_group = ei->i_block_group; + if (flex_size >= EXT4_FLEX_SIZE_DIR_ALLOC_SCHEME) { + /* + * If there are at least EXT4_FLEX_SIZE_DIR_ALLOC_SCHEME + * block groups per flexgroup, reserve the first block + * group for directories and special files. Regular + * files will start at the second block group. This + * tends to speed up directory access and improves + * fsck times. + */ + block_group &= ~(flex_size-1); + if (S_ISREG(inode->i_mode)) + block_group++; + } + bg_start = ext4_group_first_block_no(inode->i_sb, block_group); + last_block = ext4_blocks_count(EXT4_SB(inode->i_sb)->s_es) - 1; + + /* + * If we are doing delayed allocation, we don't need take + * colour into account. + */ + if (test_opt(inode->i_sb, DELALLOC)) + return bg_start; + + if (bg_start + EXT4_BLOCKS_PER_GROUP(inode->i_sb) <= last_block) + colour = (task_pid_nr(current) % 16) * + (EXT4_BLOCKS_PER_GROUP(inode->i_sb) / 16); + else + colour = (task_pid_nr(current) % 16) * + ((last_block - bg_start) / 16); + return bg_start + colour; +} + diff --git a/fs/ext4l/bitmap.c b/fs/ext4l/bitmap.c new file mode 100644 index 00000000000..87760fabdd2 --- /dev/null +++ b/fs/ext4l/bitmap.c @@ -0,0 +1,99 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * linux/fs/ext4/bitmap.c + * + * Copyright (C) 1992, 1993, 1994, 1995 + * Remy Card (card@masi.ibp.fr) + * Laboratoire MASI - Institut Blaise Pascal + * Universite Pierre et Marie Curie (Paris VI) + */ + +#include <linux/buffer_head.h> +#include "ext4.h" + +unsigned int ext4_count_free(char *bitmap, unsigned int numchars) +{ + return numchars * BITS_PER_BYTE - memweight(bitmap, numchars); +} + +int ext4_inode_bitmap_csum_verify(struct super_block *sb, + struct ext4_group_desc *gdp, + struct buffer_head *bh) +{ + __u32 hi; + __u32 provided, calculated; + struct ext4_sb_info *sbi = EXT4_SB(sb); + int sz; + + if (!ext4_has_feature_metadata_csum(sb)) + return 1; + + sz = EXT4_INODES_PER_GROUP(sb) >> 3; + provided = le16_to_cpu(gdp->bg_inode_bitmap_csum_lo); + calculated = ext4_chksum(sbi->s_csum_seed, (__u8 *)bh->b_data, sz); + if (sbi->s_desc_size >= EXT4_BG_INODE_BITMAP_CSUM_HI_END) { + hi = le16_to_cpu(gdp->bg_inode_bitmap_csum_hi); + provided |= (hi << 16); + } else + calculated &= 0xFFFF; + + return provided == calculated; +} + +void ext4_inode_bitmap_csum_set(struct super_block *sb, + struct ext4_group_desc *gdp, + struct buffer_head *bh) +{ + __u32 csum; + struct ext4_sb_info *sbi = EXT4_SB(sb); + int sz; + + if (!ext4_has_feature_metadata_csum(sb)) + return; + + sz = EXT4_INODES_PER_GROUP(sb) >> 3; + csum = ext4_chksum(sbi->s_csum_seed, (__u8 *)bh->b_data, sz); + gdp->bg_inode_bitmap_csum_lo = cpu_to_le16(csum & 0xFFFF); + if (sbi->s_desc_size >= EXT4_BG_INODE_BITMAP_CSUM_HI_END) + gdp->bg_inode_bitmap_csum_hi = cpu_to_le16(csum >> 16); +} + +int ext4_block_bitmap_csum_verify(struct super_block *sb, + struct ext4_group_desc *gdp, + struct buffer_head *bh) +{ + __u32 hi; + __u32 provided, calculated; + struct ext4_sb_info *sbi = EXT4_SB(sb); + int sz = EXT4_CLUSTERS_PER_GROUP(sb) / 8; + + if (!ext4_has_feature_metadata_csum(sb)) + return 1; + + provided = le16_to_cpu(gdp->bg_block_bitmap_csum_lo); + calculated = ext4_chksum(sbi->s_csum_seed, (__u8 *)bh->b_data, sz); + if (sbi->s_desc_size >= EXT4_BG_BLOCK_BITMAP_CSUM_HI_END) { + hi = le16_to_cpu(gdp->bg_block_bitmap_csum_hi); + provided |= (hi << 16); + } else + calculated &= 0xFFFF; + + return provided == calculated; +} + +void ext4_block_bitmap_csum_set(struct super_block *sb, + struct ext4_group_desc *gdp, + struct buffer_head *bh) +{ + int sz = EXT4_CLUSTERS_PER_GROUP(sb) / 8; + __u32 csum; + struct ext4_sb_info *sbi = EXT4_SB(sb); + + if (!ext4_has_feature_metadata_csum(sb)) + return; + + csum = ext4_chksum(sbi->s_csum_seed, (__u8 *)bh->b_data, sz); + gdp->bg_block_bitmap_csum_lo = cpu_to_le16(csum & 0xFFFF); + if (sbi->s_desc_size >= EXT4_BG_BLOCK_BITMAP_CSUM_HI_END) + gdp->bg_block_bitmap_csum_hi = cpu_to_le16(csum >> 16); +} diff --git a/fs/ext4l/block_validity.c b/fs/ext4l/block_validity.c new file mode 100644 index 00000000000..e8c5525afc6 --- /dev/null +++ b/fs/ext4l/block_validity.c @@ -0,0 +1,370 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * linux/fs/ext4/block_validity.c + * + * Copyright (C) 2009 + * Theodore Ts'o (tytso@mit.edu) + * + * Track which blocks in the filesystem are metadata blocks that + * should never be used as data blocks by files or directories. + */ + +#include <linux/time.h> +#include <linux/fs.h> +#include <linux/namei.h> +#include <linux/quotaops.h> +#include <linux/buffer_head.h> +#include <linux/swap.h> +#include <linux/pagemap.h> +#include <linux/blkdev.h> +#include <linux/slab.h> +#include "ext4.h" + +struct ext4_system_zone { + struct rb_node node; + ext4_fsblk_t start_blk; + unsigned int count; + u32 ino; +}; + +static struct kmem_cache *ext4_system_zone_cachep; + +int __init ext4_init_system_zone(void) +{ + ext4_system_zone_cachep = KMEM_CACHE(ext4_system_zone, 0); + if (ext4_system_zone_cachep == NULL) + return -ENOMEM; + return 0; +} + +void ext4_exit_system_zone(void) +{ + rcu_barrier(); + kmem_cache_destroy(ext4_system_zone_cachep); +} + +static inline int can_merge(struct ext4_system_zone *entry1, + struct ext4_system_zone *entry2) +{ + if ((entry1->start_blk + entry1->count) == entry2->start_blk && + entry1->ino == entry2->ino) + return 1; + return 0; +} + +static void release_system_zone(struct ext4_system_blocks *system_blks) +{ + struct ext4_system_zone *entry, *n; + + rbtree_postorder_for_each_entry_safe(entry, n, + &system_blks->root, node) + kmem_cache_free(ext4_system_zone_cachep, entry); +} + +/* + * Mark a range of blocks as belonging to the "system zone" --- that + * is, filesystem metadata blocks which should never be used by + * inodes. + */ +static int add_system_zone(struct ext4_system_blocks *system_blks, + ext4_fsblk_t start_blk, + unsigned int count, u32 ino) +{ + struct ext4_system_zone *new_entry, *entry; + struct rb_node **n = &system_blks->root.rb_node, *node; + struct rb_node *parent = NULL, *new_node; + + while (*n) { + parent = *n; + entry = rb_entry(parent, struct ext4_system_zone, node); + if (start_blk < entry->start_blk) + n = &(*n)->rb_left; + else if (start_blk >= (entry->start_blk + entry->count)) + n = &(*n)->rb_right; + else /* Unexpected overlap of system zones. */ + return -EFSCORRUPTED; + } + + new_entry = kmem_cache_alloc(ext4_system_zone_cachep, + GFP_KERNEL); + if (!new_entry) + return -ENOMEM; + new_entry->start_blk = start_blk; + new_entry->count = count; + new_entry->ino = ino; + new_node = &new_entry->node; + + rb_link_node(new_node, parent, n); + rb_insert_color(new_node, &system_blks->root); + + /* Can we merge to the left? */ + node = rb_prev(new_node); + if (node) { + entry = rb_entry(node, struct ext4_system_zone, node); + if (can_merge(entry, new_entry)) { + new_entry->start_blk = entry->start_blk; + new_entry->count += entry->count; + rb_erase(node, &system_blks->root); + kmem_cache_free(ext4_system_zone_cachep, entry); + } + } + + /* Can we merge to the right? */ + node = rb_next(new_node); + if (node) { + entry = rb_entry(node, struct ext4_system_zone, node); + if (can_merge(new_entry, entry)) { + new_entry->count += entry->count; + rb_erase(node, &system_blks->root); + kmem_cache_free(ext4_system_zone_cachep, entry); + } + } + return 0; +} + +static void debug_print_tree(struct ext4_sb_info *sbi) +{ + struct rb_node *node; + struct ext4_system_zone *entry; + struct ext4_system_blocks *system_blks; + int first = 1; + + printk(KERN_INFO "System zones: "); + rcu_read_lock(); + system_blks = rcu_dereference(sbi->s_system_blks); + node = rb_first(&system_blks->root); + while (node) { + entry = rb_entry(node, struct ext4_system_zone, node); + printk(KERN_CONT "%s%llu-%llu", first ? "" : ", ", + entry->start_blk, entry->start_blk + entry->count - 1); + first = 0; + node = rb_next(node); + } + rcu_read_unlock(); + printk(KERN_CONT "\n"); +} + +static int ext4_protect_reserved_inode(struct super_block *sb, + struct ext4_system_blocks *system_blks, + u32 ino) +{ + struct inode *inode; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_map_blocks map; + u32 i = 0, num; + int err = 0, n; + + if ((ino < EXT4_ROOT_INO) || + (ino > le32_to_cpu(sbi->s_es->s_inodes_count))) + return -EINVAL; + inode = ext4_iget(sb, ino, EXT4_IGET_SPECIAL); + if (IS_ERR(inode)) + return PTR_ERR(inode); + num = (inode->i_size + sb->s_blocksize - 1) >> sb->s_blocksize_bits; + while (i < num) { + cond_resched(); + map.m_lblk = i; + map.m_len = num - i; + n = ext4_map_blocks(NULL, inode, &map, 0); + if (n < 0) { + err = n; + break; + } + if (n == 0) { + i++; + } else { + err = add_system_zone(system_blks, map.m_pblk, n, ino); + if (err < 0) { + if (err == -EFSCORRUPTED) { + EXT4_ERROR_INODE_ERR(inode, -err, + "blocks %llu-%llu from inode overlap system zone", + map.m_pblk, + map.m_pblk + map.m_len - 1); + } + break; + } + i += n; + } + } + iput(inode); + return err; +} + +static void ext4_destroy_system_zone(struct rcu_head *rcu) +{ + struct ext4_system_blocks *system_blks; + + system_blks = container_of(rcu, struct ext4_system_blocks, rcu); + release_system_zone(system_blks); + kfree(system_blks); +} + +/* + * Build system zone rbtree which is used for block validity checking. + * + * The update of system_blks pointer in this function is protected by + * sb->s_umount semaphore. However we have to be careful as we can be + * racing with ext4_inode_block_valid() calls reading system_blks rbtree + * protected only by RCU. That's why we first build the rbtree and then + * swap it in place. + */ +int ext4_setup_system_zone(struct super_block *sb) +{ + ext4_group_t ngroups = ext4_get_groups_count(sb); + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_system_blocks *system_blks; + struct ext4_group_desc *gdp; + ext4_group_t i; + int ret; + + system_blks = kzalloc(sizeof(*system_blks), GFP_KERNEL); + if (!system_blks) + return -ENOMEM; + + for (i=0; i < ngroups; i++) { + unsigned int meta_blks = ext4_num_base_meta_blocks(sb, i); + + cond_resched(); + if (meta_blks != 0) { + ret = add_system_zone(system_blks, + ext4_group_first_block_no(sb, i), + meta_blks, 0); + if (ret) + goto err; + } + gdp = ext4_get_group_desc(sb, i, NULL); + ret = add_system_zone(system_blks, + ext4_block_bitmap(sb, gdp), 1, 0); + if (ret) + goto err; + ret = add_system_zone(system_blks, + ext4_inode_bitmap(sb, gdp), 1, 0); + if (ret) + goto err; + ret = add_system_zone(system_blks, + ext4_inode_table(sb, gdp), + sbi->s_itb_per_group, 0); + if (ret) + goto err; + } + if (ext4_has_feature_journal(sb) && sbi->s_es->s_journal_inum) { + ret = ext4_protect_reserved_inode(sb, system_blks, + le32_to_cpu(sbi->s_es->s_journal_inum)); + if (ret) + goto err; + } + + /* + * System blks rbtree complete, announce it once to prevent racing + * with ext4_inode_block_valid() accessing the rbtree at the same + * time. + */ + rcu_assign_pointer(sbi->s_system_blks, system_blks); + + if (test_opt(sb, DEBUG)) + debug_print_tree(sbi); + return 0; +err: + release_system_zone(system_blks); + kfree(system_blks); + return ret; +} + +/* + * Called when the filesystem is unmounted or when remounting it with + * noblock_validity specified. + * + * The update of system_blks pointer in this function is protected by + * sb->s_umount semaphore. However we have to be careful as we can be + * racing with ext4_inode_block_valid() calls reading system_blks rbtree + * protected only by RCU. So we first clear the system_blks pointer and + * then free the rbtree only after RCU grace period expires. + */ +void ext4_release_system_zone(struct super_block *sb) +{ + struct ext4_system_blocks *system_blks; + + system_blks = rcu_dereference_protected(EXT4_SB(sb)->s_system_blks, + lockdep_is_held(&sb->s_umount)); + rcu_assign_pointer(EXT4_SB(sb)->s_system_blks, NULL); + + if (system_blks) + call_rcu(&system_blks->rcu, ext4_destroy_system_zone); +} + +int ext4_sb_block_valid(struct super_block *sb, struct inode *inode, + ext4_fsblk_t start_blk, unsigned int count) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_system_blocks *system_blks; + struct ext4_system_zone *entry; + struct rb_node *n; + int ret = 1; + + if ((start_blk <= le32_to_cpu(sbi->s_es->s_first_data_block)) || + (start_blk + count < start_blk) || + (start_blk + count > ext4_blocks_count(sbi->s_es))) + return 0; + + /* + * Lock the system zone to prevent it being released concurrently + * when doing a remount which inverse current "[no]block_validity" + * mount option. + */ + rcu_read_lock(); + system_blks = rcu_dereference(sbi->s_system_blks); + if (system_blks == NULL) + goto out_rcu; + + n = system_blks->root.rb_node; + while (n) { + entry = rb_entry(n, struct ext4_system_zone, node); + if (start_blk + count - 1 < entry->start_blk) + n = n->rb_left; + else if (start_blk >= (entry->start_blk + entry->count)) + n = n->rb_right; + else { + ret = 0; + if (inode) + ret = (entry->ino == inode->i_ino); + break; + } + } +out_rcu: + rcu_read_unlock(); + return ret; +} + +/* + * Returns 1 if the passed-in block region (start_blk, + * start_blk+count) is valid; 0 if some part of the block region + * overlaps with some other filesystem metadata blocks. + */ +int ext4_inode_block_valid(struct inode *inode, ext4_fsblk_t start_blk, + unsigned int count) +{ + return ext4_sb_block_valid(inode->i_sb, inode, start_blk, count); +} + +int ext4_check_blockref(const char *function, unsigned int line, + struct inode *inode, __le32 *p, unsigned int max) +{ + __le32 *bref = p; + unsigned int blk; + journal_t *journal = EXT4_SB(inode->i_sb)->s_journal; + + if (journal && inode == journal->j_inode) + return 0; + + while (bref < p+max) { + blk = le32_to_cpu(*bref++); + if (blk && + unlikely(!ext4_inode_block_valid(inode, blk, 1))) { + ext4_error_inode(inode, function, line, blk, + "invalid block"); + return -EFSCORRUPTED; + } + } + return 0; +} + diff --git a/fs/ext4l/crypto.c b/fs/ext4l/crypto.c new file mode 100644 index 00000000000..cf0a0970c09 --- /dev/null +++ b/fs/ext4l/crypto.c @@ -0,0 +1,241 @@ +// SPDX-License-Identifier: GPL-2.0 + +#include <linux/quotaops.h> +#include <linux/uuid.h> + +#include "ext4.h" +#include "xattr.h" +#include "ext4_jbd2.h" + +static void ext4_fname_from_fscrypt_name(struct ext4_filename *dst, + const struct fscrypt_name *src) +{ + memset(dst, 0, sizeof(*dst)); + + dst->usr_fname = src->usr_fname; + dst->disk_name = src->disk_name; + dst->hinfo.hash = src->hash; + dst->hinfo.minor_hash = src->minor_hash; + dst->crypto_buf = src->crypto_buf; +} + +int ext4_fname_setup_filename(struct inode *dir, const struct qstr *iname, + int lookup, struct ext4_filename *fname) +{ + struct fscrypt_name name; + int err; + + err = fscrypt_setup_filename(dir, iname, lookup, &name); + if (err) + return err; + + ext4_fname_from_fscrypt_name(fname, &name); + + err = ext4_fname_setup_ci_filename(dir, iname, fname); + if (err) + ext4_fname_free_filename(fname); + + return err; +} + +int ext4_fname_prepare_lookup(struct inode *dir, struct dentry *dentry, + struct ext4_filename *fname) +{ + struct fscrypt_name name; + int err; + + err = fscrypt_prepare_lookup(dir, dentry, &name); + if (err) + return err; + + ext4_fname_from_fscrypt_name(fname, &name); + + err = ext4_fname_setup_ci_filename(dir, &dentry->d_name, fname); + if (err) + ext4_fname_free_filename(fname); + return err; +} + +void ext4_fname_free_filename(struct ext4_filename *fname) +{ + struct fscrypt_name name; + + name.crypto_buf = fname->crypto_buf; + fscrypt_free_filename(&name); + + fname->crypto_buf.name = NULL; + fname->usr_fname = NULL; + fname->disk_name.name = NULL; + + ext4_fname_free_ci_filename(fname); +} + +static bool uuid_is_zero(__u8 u[16]) +{ + int i; + + for (i = 0; i < 16; i++) + if (u[i]) + return false; + return true; +} + +int ext4_ioctl_get_encryption_pwsalt(struct file *filp, void __user *arg) +{ + struct super_block *sb = file_inode(filp)->i_sb; + struct ext4_sb_info *sbi = EXT4_SB(sb); + int err, err2; + handle_t *handle; + + if (!ext4_has_feature_encrypt(sb)) + return -EOPNOTSUPP; + + if (uuid_is_zero(sbi->s_es->s_encrypt_pw_salt)) { + err = mnt_want_write_file(filp); + if (err) + return err; + handle = ext4_journal_start_sb(sb, EXT4_HT_MISC, 1); + if (IS_ERR(handle)) { + err = PTR_ERR(handle); + goto pwsalt_err_exit; + } + err = ext4_journal_get_write_access(handle, sb, sbi->s_sbh, + EXT4_JTR_NONE); + if (err) + goto pwsalt_err_journal; + lock_buffer(sbi->s_sbh); + generate_random_uuid(sbi->s_es->s_encrypt_pw_salt); + ext4_superblock_csum_set(sb); + unlock_buffer(sbi->s_sbh); + err = ext4_handle_dirty_metadata(handle, NULL, sbi->s_sbh); +pwsalt_err_journal: + err2 = ext4_journal_stop(handle); + if (err2 && !err) + err = err2; +pwsalt_err_exit: + mnt_drop_write_file(filp); + if (err) + return err; + } + + if (copy_to_user(arg, sbi->s_es->s_encrypt_pw_salt, 16)) + return -EFAULT; + return 0; +} + +static int ext4_get_context(struct inode *inode, void *ctx, size_t len) +{ + return ext4_xattr_get(inode, EXT4_XATTR_INDEX_ENCRYPTION, + EXT4_XATTR_NAME_ENCRYPTION_CONTEXT, ctx, len); +} + +static int ext4_set_context(struct inode *inode, const void *ctx, size_t len, + void *fs_data) +{ + handle_t *handle = fs_data; + int res, res2, credits, retries = 0; + + /* + * Encrypting the root directory is not allowed because e2fsck expects + * lost+found to exist and be unencrypted, and encrypting the root + * directory would imply encrypting the lost+found directory as well as + * the filename "lost+found" itself. + */ + if (inode->i_ino == EXT4_ROOT_INO) + return -EPERM; + + if (WARN_ON_ONCE(IS_DAX(inode) && i_size_read(inode))) + return -EINVAL; + + if (ext4_test_inode_flag(inode, EXT4_INODE_DAX)) + return -EOPNOTSUPP; + + res = ext4_convert_inline_data(inode); + if (res) + return res; + + /* + * If a journal handle was specified, then the encryption context is + * being set on a new inode via inheritance and is part of a larger + * transaction to create the inode. Otherwise the encryption context is + * being set on an existing inode in its own transaction. Only in the + * latter case should the "retry on ENOSPC" logic be used. + */ + + if (handle) { + res = ext4_xattr_set_handle(handle, inode, + EXT4_XATTR_INDEX_ENCRYPTION, + EXT4_XATTR_NAME_ENCRYPTION_CONTEXT, + ctx, len, 0); + if (!res) { + ext4_set_inode_flag(inode, EXT4_INODE_ENCRYPT); + ext4_clear_inode_state(inode, + EXT4_STATE_MAY_INLINE_DATA); + /* + * Update inode->i_flags - S_ENCRYPTED will be enabled, + * S_DAX may be disabled + */ + ext4_set_inode_flags(inode, false); + } + return res; + } + + res = dquot_initialize(inode); + if (res) + return res; +retry: + res = ext4_xattr_set_credits(inode, len, false /* is_create */, + &credits); + if (res) + return res; + + handle = ext4_journal_start(inode, EXT4_HT_MISC, credits); + if (IS_ERR(handle)) + return PTR_ERR(handle); + + res = ext4_xattr_set_handle(handle, inode, EXT4_XATTR_INDEX_ENCRYPTION, + EXT4_XATTR_NAME_ENCRYPTION_CONTEXT, + ctx, len, 0); + if (!res) { + ext4_set_inode_flag(inode, EXT4_INODE_ENCRYPT); + /* + * Update inode->i_flags - S_ENCRYPTED will be enabled, + * S_DAX may be disabled + */ + ext4_set_inode_flags(inode, false); + res = ext4_mark_inode_dirty(handle, inode); + if (res) + EXT4_ERROR_INODE(inode, "Failed to mark inode dirty"); + } + res2 = ext4_journal_stop(handle); + + if (res == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries)) + goto retry; + if (!res) + res = res2; + return res; +} + +static const union fscrypt_policy *ext4_get_dummy_policy(struct super_block *sb) +{ + return EXT4_SB(sb)->s_dummy_enc_policy.policy; +} + +static bool ext4_has_stable_inodes(struct super_block *sb) +{ + return ext4_has_feature_stable_inodes(sb); +} + +const struct fscrypt_operations ext4_cryptops = { + .inode_info_offs = (int)offsetof(struct ext4_inode_info, i_crypt_info) - + (int)offsetof(struct ext4_inode_info, vfs_inode), + .needs_bounce_pages = 1, + .has_32bit_inodes = 1, + .supports_subblock_data_units = 1, + .legacy_key_prefix = "ext4:", + .get_context = ext4_get_context, + .set_context = ext4_set_context, + .get_dummy_policy = ext4_get_dummy_policy, + .empty_dir = ext4_empty_dir, + .has_stable_inodes = ext4_has_stable_inodes, +}; diff --git a/fs/ext4l/dir.c b/fs/ext4l/dir.c new file mode 100644 index 00000000000..d4164c507a9 --- /dev/null +++ b/fs/ext4l/dir.c @@ -0,0 +1,693 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * linux/fs/ext4/dir.c + * + * Copyright (C) 1992, 1993, 1994, 1995 + * Remy Card (card@masi.ibp.fr) + * Laboratoire MASI - Institut Blaise Pascal + * Universite Pierre et Marie Curie (Paris VI) + * + * from + * + * linux/fs/minix/dir.c + * + * Copyright (C) 1991, 1992 Linus Torvalds + * + * ext4 directory handling functions + * + * Big-endian to little-endian byte-swapping/bitmaps by + * David S. Miller (davem@caip.rutgers.edu), 1995 + * + * Hash Tree Directory indexing (c) 2001 Daniel Phillips + * + */ + +#include <linux/fs.h> +#include <linux/buffer_head.h> +#include <linux/slab.h> +#include <linux/iversion.h> +#include <linux/unicode.h> +#include "ext4.h" +#include "xattr.h" + +static int ext4_dx_readdir(struct file *, struct dir_context *); + +/** + * is_dx_dir() - check if a directory is using htree indexing + * @inode: directory inode + * + * Check if the given dir-inode refers to an htree-indexed directory + * (or a directory which could potentially get converted to use htree + * indexing). + * + * Return 1 if it is a dx dir, 0 if not + */ +static int is_dx_dir(struct inode *inode) +{ + struct super_block *sb = inode->i_sb; + + if (ext4_has_feature_dir_index(inode->i_sb) && + ((ext4_test_inode_flag(inode, EXT4_INODE_INDEX)) || + ((inode->i_size >> sb->s_blocksize_bits) == 1) || + ext4_has_inline_data(inode))) + return 1; + + return 0; +} + +static bool is_fake_dir_entry(struct ext4_dir_entry_2 *de) +{ + /* Check if . or .. , or skip if namelen is 0 */ + if ((de->name_len > 0) && (de->name_len <= 2) && (de->name[0] == '.') && + (de->name[1] == '.' || de->name[1] == '\0')) + return true; + /* Check if this is a csum entry */ + if (de->file_type == EXT4_FT_DIR_CSUM) + return true; + return false; +} + +/* + * Return 0 if the directory entry is OK, and 1 if there is a problem + * + * Note: this is the opposite of what ext2 and ext3 historically returned... + * + * bh passed here can be an inode block or a dir data block, depending + * on the inode inline data flag. + */ +int __ext4_check_dir_entry(const char *function, unsigned int line, + struct inode *dir, struct file *filp, + struct ext4_dir_entry_2 *de, + struct buffer_head *bh, char *buf, int size, + unsigned int offset) +{ + const char *error_msg = NULL; + const int rlen = ext4_rec_len_from_disk(de->rec_len, + dir->i_sb->s_blocksize); + const int next_offset = ((char *) de - buf) + rlen; + bool fake = is_fake_dir_entry(de); + bool has_csum = ext4_has_feature_metadata_csum(dir->i_sb); + + if (unlikely(rlen < ext4_dir_rec_len(1, fake ? NULL : dir))) + error_msg = "rec_len is smaller than minimal"; + else if (unlikely(rlen % 4 != 0)) + error_msg = "rec_len % 4 != 0"; + else if (unlikely(rlen < ext4_dir_rec_len(de->name_len, + fake ? NULL : dir))) + error_msg = "rec_len is too small for name_len"; + else if (unlikely(next_offset > size)) + error_msg = "directory entry overrun"; + else if (unlikely(next_offset > size - ext4_dir_rec_len(1, + has_csum ? NULL : dir) && + next_offset != size)) + error_msg = "directory entry too close to block end"; + else if (unlikely(le32_to_cpu(de->inode) > + le32_to_cpu(EXT4_SB(dir->i_sb)->s_es->s_inodes_count))) + error_msg = "inode out of bounds"; + else if (unlikely(next_offset == size && de->name_len == 1 && + de->name[0] == '.')) + error_msg = "'.' directory cannot be the last in data block"; + else + return 0; + + if (filp) + ext4_error_file(filp, function, line, bh->b_blocknr, + "bad entry in directory: %s - offset=%u, " + "inode=%u, rec_len=%d, size=%d fake=%d", + error_msg, offset, le32_to_cpu(de->inode), + rlen, size, fake); + else + ext4_error_inode(dir, function, line, bh->b_blocknr, + "bad entry in directory: %s - offset=%u, " + "inode=%u, rec_len=%d, size=%d fake=%d", + error_msg, offset, le32_to_cpu(de->inode), + rlen, size, fake); + + return 1; +} + +static int ext4_readdir(struct file *file, struct dir_context *ctx) +{ + unsigned int offset; + int i; + struct ext4_dir_entry_2 *de; + int err; + struct inode *inode = file_inode(file); + struct super_block *sb = inode->i_sb; + struct buffer_head *bh = NULL; + struct fscrypt_str fstr = FSTR_INIT(NULL, 0); + struct dir_private_info *info = file->private_data; + + err = fscrypt_prepare_readdir(inode); + if (err) + return err; + + if (is_dx_dir(inode)) { + err = ext4_dx_readdir(file, ctx); + if (err != ERR_BAD_DX_DIR) + return err; + + /* Can we just clear INDEX flag to ignore htree information? */ + if (!ext4_has_feature_metadata_csum(sb)) { + /* + * We don't set the inode dirty flag since it's not + * critical that it gets flushed back to the disk. + */ + ext4_clear_inode_flag(inode, EXT4_INODE_INDEX); + } + } + + if (ext4_has_inline_data(inode)) { + int has_inline_data = 1; + err = ext4_read_inline_dir(file, ctx, + &has_inline_data); + if (has_inline_data) + return err; + } + + if (IS_ENCRYPTED(inode)) { + err = fscrypt_fname_alloc_buffer(EXT4_NAME_LEN, &fstr); + if (err < 0) + return err; + } + + while (ctx->pos < inode->i_size) { + struct ext4_map_blocks map; + + if (fatal_signal_pending(current)) { + err = -ERESTARTSYS; + goto errout; + } + cond_resched(); + offset = ctx->pos & (sb->s_blocksize - 1); + map.m_lblk = ctx->pos >> EXT4_BLOCK_SIZE_BITS(sb); + map.m_len = 1; + err = ext4_map_blocks(NULL, inode, &map, 0); + if (err == 0) { + /* m_len should never be zero but let's avoid + * an infinite loop if it somehow is */ + if (map.m_len == 0) + map.m_len = 1; + ctx->pos += map.m_len * sb->s_blocksize; + continue; + } + if (err > 0) { + pgoff_t index = map.m_pblk >> + (PAGE_SHIFT - inode->i_blkbits); + if (!ra_has_index(&file->f_ra, index)) + page_cache_sync_readahead( + sb->s_bdev->bd_mapping, + &file->f_ra, file, + index, 1); + file->f_ra.prev_pos = (loff_t)index << PAGE_SHIFT; + bh = ext4_bread(NULL, inode, map.m_lblk, 0); + if (IS_ERR(bh)) { + err = PTR_ERR(bh); + bh = NULL; + goto errout; + } + } + + if (!bh) { + /* corrupt size? Maybe no more blocks to read */ + if (ctx->pos > inode->i_blocks << 9) + break; + ctx->pos += sb->s_blocksize - offset; + continue; + } + + /* Check the checksum */ + if (!buffer_verified(bh) && + !ext4_dirblock_csum_verify(inode, bh)) { + EXT4_ERROR_FILE(file, 0, "directory fails checksum " + "at offset %llu", + (unsigned long long)ctx->pos); + ctx->pos += sb->s_blocksize - offset; + brelse(bh); + bh = NULL; + continue; + } + set_buffer_verified(bh); + + /* If the dir block has changed since the last call to + * readdir(2), then we might be pointing to an invalid + * dirent right now. Scan from the start of the block + * to make sure. */ + if (!inode_eq_iversion(inode, info->cookie)) { + for (i = 0; i < sb->s_blocksize && i < offset; ) { + de = (struct ext4_dir_entry_2 *) + (bh->b_data + i); + /* It's too expensive to do a full + * dirent test each time round this + * loop, but we do have to test at + * least that it is non-zero. A + * failure will be detected in the + * dirent test below. */ + if (ext4_rec_len_from_disk(de->rec_len, + sb->s_blocksize) < ext4_dir_rec_len(1, + inode)) + break; + i += ext4_rec_len_from_disk(de->rec_len, + sb->s_blocksize); + } + offset = i; + ctx->pos = (ctx->pos & ~(sb->s_blocksize - 1)) + | offset; + info->cookie = inode_query_iversion(inode); + } + + while (ctx->pos < inode->i_size + && offset < sb->s_blocksize) { + de = (struct ext4_dir_entry_2 *) (bh->b_data + offset); + if (ext4_check_dir_entry(inode, file, de, bh, + bh->b_data, bh->b_size, + offset)) { + /* + * On error, skip to the next block + */ + ctx->pos = (ctx->pos | + (sb->s_blocksize - 1)) + 1; + break; + } + offset += ext4_rec_len_from_disk(de->rec_len, + sb->s_blocksize); + if (le32_to_cpu(de->inode)) { + if (!IS_ENCRYPTED(inode)) { + if (!dir_emit(ctx, de->name, + de->name_len, + le32_to_cpu(de->inode), + get_dtype(sb, de->file_type))) + goto done; + } else { + int save_len = fstr.len; + struct fscrypt_str de_name = + FSTR_INIT(de->name, + de->name_len); + u32 hash; + u32 minor_hash; + + if (IS_CASEFOLDED(inode)) { + hash = EXT4_DIRENT_HASH(de); + minor_hash = EXT4_DIRENT_MINOR_HASH(de); + } else { + hash = 0; + minor_hash = 0; + } + + /* Directory is encrypted */ + err = fscrypt_fname_disk_to_usr(inode, + hash, minor_hash, &de_name, &fstr); + de_name = fstr; + fstr.len = save_len; + if (err) + goto errout; + if (!dir_emit(ctx, + de_name.name, de_name.len, + le32_to_cpu(de->inode), + get_dtype(sb, de->file_type))) + goto done; + } + } + ctx->pos += ext4_rec_len_from_disk(de->rec_len, + sb->s_blocksize); + } + if ((ctx->pos < inode->i_size) && !dir_relax_shared(inode)) + goto done; + brelse(bh); + bh = NULL; + } +done: + err = 0; +errout: + fscrypt_fname_free_buffer(&fstr); + brelse(bh); + return err; +} + +static inline int is_32bit_api(void) +{ +#ifdef CONFIG_COMPAT + return in_compat_syscall(); +#else + return (BITS_PER_LONG == 32); +#endif +} + +/* + * These functions convert from the major/minor hash to an f_pos + * value for dx directories + * + * Upper layer (for example NFS) should specify FMODE_32BITHASH or + * FMODE_64BITHASH explicitly. On the other hand, we allow ext4 to be mounted + * directly on both 32-bit and 64-bit nodes, under such case, neither + * FMODE_32BITHASH nor FMODE_64BITHASH is specified. + */ +static inline loff_t hash2pos(struct file *filp, __u32 major, __u32 minor) +{ + if ((filp->f_mode & FMODE_32BITHASH) || + (!(filp->f_mode & FMODE_64BITHASH) && is_32bit_api())) + return major >> 1; + else + return ((__u64)(major >> 1) << 32) | (__u64)minor; +} + +static inline __u32 pos2maj_hash(struct file *filp, loff_t pos) +{ + if ((filp->f_mode & FMODE_32BITHASH) || + (!(filp->f_mode & FMODE_64BITHASH) && is_32bit_api())) + return (pos << 1) & 0xffffffff; + else + return ((pos >> 32) << 1) & 0xffffffff; +} + +static inline __u32 pos2min_hash(struct file *filp, loff_t pos) +{ + if ((filp->f_mode & FMODE_32BITHASH) || + (!(filp->f_mode & FMODE_64BITHASH) && is_32bit_api())) + return 0; + else + return pos & 0xffffffff; +} + +/* + * Return 32- or 64-bit end-of-file for dx directories + */ +static inline loff_t ext4_get_htree_eof(struct file *filp) +{ + if ((filp->f_mode & FMODE_32BITHASH) || + (!(filp->f_mode & FMODE_64BITHASH) && is_32bit_api())) + return EXT4_HTREE_EOF_32BIT; + else + return EXT4_HTREE_EOF_64BIT; +} + + +/* + * ext4_dir_llseek() calls generic_file_llseek_size to handle htree + * directories, where the "offset" is in terms of the filename hash + * value instead of the byte offset. + * + * Because we may return a 64-bit hash that is well beyond offset limits, + * we need to pass the max hash as the maximum allowable offset in + * the htree directory case. + * + * For non-htree, ext4_llseek already chooses the proper max offset. + */ +static loff_t ext4_dir_llseek(struct file *file, loff_t offset, int whence) +{ + struct inode *inode = file->f_mapping->host; + struct dir_private_info *info = file->private_data; + int dx_dir = is_dx_dir(inode); + loff_t ret, htree_max = ext4_get_htree_eof(file); + + if (likely(dx_dir)) + ret = generic_file_llseek_size(file, offset, whence, + htree_max, htree_max); + else + ret = ext4_llseek(file, offset, whence); + info->cookie = inode_peek_iversion(inode) - 1; + return ret; +} + +/* + * This structure holds the nodes of the red-black tree used to store + * the directory entry in hash order. + */ +struct fname { + __u32 hash; + __u32 minor_hash; + struct rb_node rb_hash; + struct fname *next; + __u32 inode; + __u8 name_len; + __u8 file_type; + char name[] __counted_by(name_len); +}; + +/* + * This function implements a non-recursive way of freeing all of the + * nodes in the red-black tree. + */ +static void free_rb_tree_fname(struct rb_root *root) +{ + struct fname *fname, *next; + + rbtree_postorder_for_each_entry_safe(fname, next, root, rb_hash) + while (fname) { + struct fname *old = fname; + fname = fname->next; + kfree(old); + } + + *root = RB_ROOT; +} + +static void ext4_htree_init_dir_info(struct file *filp, loff_t pos) +{ + struct dir_private_info *p = filp->private_data; + + if (is_dx_dir(file_inode(filp)) && !p->initialized) { + p->curr_hash = pos2maj_hash(filp, pos); + p->curr_minor_hash = pos2min_hash(filp, pos); + p->initialized = true; + } +} + +void ext4_htree_free_dir_info(struct dir_private_info *p) +{ + free_rb_tree_fname(&p->root); + kfree(p); +} + +/* + * Given a directory entry, enter it into the fname rb tree. + * + * When filename encryption is enabled, the dirent will hold the + * encrypted filename, while the htree will hold decrypted filename. + * The decrypted filename is passed in via ent_name. parameter. + */ +int ext4_htree_store_dirent(struct file *dir_file, __u32 hash, + __u32 minor_hash, + struct ext4_dir_entry_2 *dirent, + struct fscrypt_str *ent_name) +{ + struct rb_node **p, *parent = NULL; + struct fname *fname, *new_fn; + struct dir_private_info *info; + + info = dir_file->private_data; + p = &info->root.rb_node; + + /* Create and allocate the fname structure */ + new_fn = kzalloc(struct_size(new_fn, name, ent_name->len + 1), + GFP_KERNEL); + if (!new_fn) + return -ENOMEM; + new_fn->hash = hash; + new_fn->minor_hash = minor_hash; + new_fn->inode = le32_to_cpu(dirent->inode); + new_fn->name_len = ent_name->len; + new_fn->file_type = dirent->file_type; + memcpy(new_fn->name, ent_name->name, ent_name->len); + + while (*p) { + parent = *p; + fname = rb_entry(parent, struct fname, rb_hash); + + /* + * If the hash and minor hash match up, then we put + * them on a linked list. This rarely happens... + */ + if ((new_fn->hash == fname->hash) && + (new_fn->minor_hash == fname->minor_hash)) { + new_fn->next = fname->next; + fname->next = new_fn; + return 0; + } + + if (new_fn->hash < fname->hash) + p = &(*p)->rb_left; + else if (new_fn->hash > fname->hash) + p = &(*p)->rb_right; + else if (new_fn->minor_hash < fname->minor_hash) + p = &(*p)->rb_left; + else /* if (new_fn->minor_hash > fname->minor_hash) */ + p = &(*p)->rb_right; + } + + rb_link_node(&new_fn->rb_hash, parent, p); + rb_insert_color(&new_fn->rb_hash, &info->root); + return 0; +} + + + +/* + * This is a helper function for ext4_dx_readdir. It calls filldir + * for all entries on the fname linked list. (Normally there is only + * one entry on the linked list, unless there are 62 bit hash collisions.) + */ +static int call_filldir(struct file *file, struct dir_context *ctx, + struct fname *fname) +{ + struct dir_private_info *info = file->private_data; + struct inode *inode = file_inode(file); + struct super_block *sb = inode->i_sb; + + if (!fname) { + ext4_msg(sb, KERN_ERR, "%s:%d: inode #%lu: comm %s: " + "called with null fname?!?", __func__, __LINE__, + inode->i_ino, current->comm); + return 0; + } + ctx->pos = hash2pos(file, fname->hash, fname->minor_hash); + while (fname) { + if (!dir_emit(ctx, fname->name, + fname->name_len, + fname->inode, + get_dtype(sb, fname->file_type))) { + info->extra_fname = fname; + return 1; + } + fname = fname->next; + } + return 0; +} + +static int ext4_dx_readdir(struct file *file, struct dir_context *ctx) +{ + struct dir_private_info *info = file->private_data; + struct inode *inode = file_inode(file); + struct fname *fname; + int ret = 0; + + ext4_htree_init_dir_info(file, ctx->pos); + + if (ctx->pos == ext4_get_htree_eof(file)) + return 0; /* EOF */ + + /* Some one has messed with f_pos; reset the world */ + if (info->last_pos != ctx->pos) { + free_rb_tree_fname(&info->root); + info->curr_node = NULL; + info->extra_fname = NULL; + info->curr_hash = pos2maj_hash(file, ctx->pos); + info->curr_minor_hash = pos2min_hash(file, ctx->pos); + } + + /* + * If there are any leftover names on the hash collision + * chain, return them first. + */ + if (info->extra_fname) { + if (call_filldir(file, ctx, info->extra_fname)) + goto finished; + info->extra_fname = NULL; + goto next_node; + } else if (!info->curr_node) + info->curr_node = rb_first(&info->root); + + while (1) { + /* + * Fill the rbtree if we have no more entries, + * or the inode has changed since we last read in the + * cached entries. + */ + if ((!info->curr_node) || + !inode_eq_iversion(inode, info->cookie)) { + info->curr_node = NULL; + free_rb_tree_fname(&info->root); + info->cookie = inode_query_iversion(inode); + ret = ext4_htree_fill_tree(file, info->curr_hash, + info->curr_minor_hash, + &info->next_hash); + if (ret < 0) + goto finished; + if (ret == 0) { + ctx->pos = ext4_get_htree_eof(file); + break; + } + info->curr_node = rb_first(&info->root); + } + + fname = rb_entry(info->curr_node, struct fname, rb_hash); + info->curr_hash = fname->hash; + info->curr_minor_hash = fname->minor_hash; + if (call_filldir(file, ctx, fname)) + break; + next_node: + info->curr_node = rb_next(info->curr_node); + if (info->curr_node) { + fname = rb_entry(info->curr_node, struct fname, + rb_hash); + info->curr_hash = fname->hash; + info->curr_minor_hash = fname->minor_hash; + } else { + if (info->next_hash == ~0) { + ctx->pos = ext4_get_htree_eof(file); + break; + } + info->curr_hash = info->next_hash; + info->curr_minor_hash = 0; + } + } +finished: + info->last_pos = ctx->pos; + return ret < 0 ? ret : 0; +} + +static int ext4_release_dir(struct inode *inode, struct file *filp) +{ + if (filp->private_data) + ext4_htree_free_dir_info(filp->private_data); + + return 0; +} + +int ext4_check_all_de(struct inode *dir, struct buffer_head *bh, void *buf, + int buf_size) +{ + struct ext4_dir_entry_2 *de; + int rlen; + unsigned int offset = 0; + char *top; + + de = buf; + top = buf + buf_size; + while ((char *) de < top) { + if (ext4_check_dir_entry(dir, NULL, de, bh, + buf, buf_size, offset)) + return -EFSCORRUPTED; + rlen = ext4_rec_len_from_disk(de->rec_len, buf_size); + de = (struct ext4_dir_entry_2 *)((char *)de + rlen); + offset += rlen; + } + if ((char *) de > top) + return -EFSCORRUPTED; + + return 0; +} + +static int ext4_dir_open(struct inode *inode, struct file *file) +{ + struct dir_private_info *info; + + info = kzalloc(sizeof(*info), GFP_KERNEL); + if (!info) + return -ENOMEM; + file->private_data = info; + return 0; +} + +const struct file_operations ext4_dir_operations = { + .open = ext4_dir_open, + .llseek = ext4_dir_llseek, + .read = generic_read_dir, + .iterate_shared = ext4_readdir, + .unlocked_ioctl = ext4_ioctl, +#ifdef CONFIG_COMPAT + .compat_ioctl = ext4_compat_ioctl, +#endif + .fsync = ext4_sync_file, + .release = ext4_release_dir, +}; -- 2.43.0

From: Simon Glass <simon.glass@canonical.com> Copy ext4_jbd2.c, ext4_jbd2.h, fast_commit.c, and fast_commit.h from Linux v6.18 fs/ext4 directory. - ext4_jbd2: ext4-specific JBD2 journal interface wrappers - fast_commit: fast commit journal optimization Co-developed-by: Claude Opus 4.5 <noreply@anthropic.com> --- fs/ext4l/ext4_jbd2.c | 408 +++++++ fs/ext4l/ext4_jbd2.h | 461 ++++++++ fs/ext4l/fast_commit.c | 2343 ++++++++++++++++++++++++++++++++++++++++ fs/ext4l/fast_commit.h | 186 ++++ 4 files changed, 3398 insertions(+) create mode 100644 fs/ext4l/ext4_jbd2.c create mode 100644 fs/ext4l/ext4_jbd2.h create mode 100644 fs/ext4l/fast_commit.c create mode 100644 fs/ext4l/fast_commit.h diff --git a/fs/ext4l/ext4_jbd2.c b/fs/ext4l/ext4_jbd2.c new file mode 100644 index 00000000000..a0e66bc1009 --- /dev/null +++ b/fs/ext4l/ext4_jbd2.c @@ -0,0 +1,408 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Interface between ext4 and JBD + */ + +#include "ext4_jbd2.h" + +#include <trace/events/ext4.h> + +int ext4_inode_journal_mode(struct inode *inode) +{ + if (EXT4_JOURNAL(inode) == NULL) + return EXT4_INODE_WRITEBACK_DATA_MODE; /* writeback */ + /* We do not support data journalling with delayed allocation */ + if (!S_ISREG(inode->i_mode) || + ext4_test_inode_flag(inode, EXT4_INODE_EA_INODE) || + test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA || + (ext4_test_inode_flag(inode, EXT4_INODE_JOURNAL_DATA) && + !test_opt(inode->i_sb, DELALLOC) && + !mapping_large_folio_support(inode->i_mapping))) { + /* We do not support data journalling for encrypted data */ + if (S_ISREG(inode->i_mode) && IS_ENCRYPTED(inode)) + return EXT4_INODE_ORDERED_DATA_MODE; /* ordered */ + return EXT4_INODE_JOURNAL_DATA_MODE; /* journal data */ + } + if (test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_ORDERED_DATA) + return EXT4_INODE_ORDERED_DATA_MODE; /* ordered */ + if (test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_WRITEBACK_DATA) + return EXT4_INODE_WRITEBACK_DATA_MODE; /* writeback */ + BUG(); +} + +/* Just increment the non-pointer handle value */ +static handle_t *ext4_get_nojournal(void) +{ + handle_t *handle = current->journal_info; + unsigned long ref_cnt = (unsigned long)handle; + + BUG_ON(ref_cnt >= EXT4_NOJOURNAL_MAX_REF_COUNT); + + ref_cnt++; + handle = (handle_t *)ref_cnt; + + current->journal_info = handle; + return handle; +} + + +/* Decrement the non-pointer handle value */ +static void ext4_put_nojournal(handle_t *handle) +{ + unsigned long ref_cnt = (unsigned long)handle; + + BUG_ON(ref_cnt == 0); + + ref_cnt--; + handle = (handle_t *)ref_cnt; + + current->journal_info = handle; +} + +/* + * Wrappers for jbd2_journal_start/end. + */ +static int ext4_journal_check_start(struct super_block *sb) +{ + int ret; + journal_t *journal; + + might_sleep(); + + ret = ext4_emergency_state(sb); + if (unlikely(ret)) + return ret; + + if (WARN_ON_ONCE(sb_rdonly(sb))) + return -EROFS; + + WARN_ON(sb->s_writers.frozen == SB_FREEZE_COMPLETE); + journal = EXT4_SB(sb)->s_journal; + /* + * Special case here: if the journal has aborted behind our + * backs (eg. EIO in the commit thread), then we still need to + * take the FS itself readonly cleanly. + */ + if (journal && is_journal_aborted(journal)) { + ext4_abort(sb, -journal->j_errno, "Detected aborted journal"); + return -EROFS; + } + return 0; +} + +handle_t *__ext4_journal_start_sb(struct inode *inode, + struct super_block *sb, unsigned int line, + int type, int blocks, int rsv_blocks, + int revoke_creds) +{ + journal_t *journal; + int err; + if (inode) + trace_ext4_journal_start_inode(inode, blocks, rsv_blocks, + revoke_creds, type, + _RET_IP_); + else + trace_ext4_journal_start_sb(sb, blocks, rsv_blocks, + revoke_creds, type, + _RET_IP_); + err = ext4_journal_check_start(sb); + if (err < 0) + return ERR_PTR(err); + + journal = EXT4_SB(sb)->s_journal; + if (!journal || (EXT4_SB(sb)->s_mount_state & EXT4_FC_REPLAY)) + return ext4_get_nojournal(); + return jbd2__journal_start(journal, blocks, rsv_blocks, revoke_creds, + GFP_NOFS, type, line); +} + +int __ext4_journal_stop(const char *where, unsigned int line, handle_t *handle) +{ + struct super_block *sb; + int err; + int rc; + + if (!ext4_handle_valid(handle)) { + ext4_put_nojournal(handle); + return 0; + } + + err = handle->h_err; + if (!handle->h_transaction) { + rc = jbd2_journal_stop(handle); + return err ? err : rc; + } + + sb = handle->h_transaction->t_journal->j_private; + rc = jbd2_journal_stop(handle); + + if (!err) + err = rc; + if (err) + __ext4_std_error(sb, where, line, err); + return err; +} + +handle_t *__ext4_journal_start_reserved(handle_t *handle, unsigned int line, + int type) +{ + struct super_block *sb; + int err; + + if (!ext4_handle_valid(handle)) + return ext4_get_nojournal(); + + sb = handle->h_journal->j_private; + trace_ext4_journal_start_reserved(sb, + jbd2_handle_buffer_credits(handle), _RET_IP_); + err = ext4_journal_check_start(sb); + if (err < 0) { + jbd2_journal_free_reserved(handle); + return ERR_PTR(err); + } + + err = jbd2_journal_start_reserved(handle, type, line); + if (err < 0) + return ERR_PTR(err); + return handle; +} + +int __ext4_journal_ensure_credits(handle_t *handle, int check_cred, + int extend_cred, int revoke_cred) +{ + if (!ext4_handle_valid(handle)) + return 0; + if (is_handle_aborted(handle)) + return -EROFS; + if (jbd2_handle_buffer_credits(handle) >= check_cred && + handle->h_revoke_credits >= revoke_cred) + return 0; + extend_cred = max(0, extend_cred - jbd2_handle_buffer_credits(handle)); + revoke_cred = max(0, revoke_cred - handle->h_revoke_credits); + return ext4_journal_extend(handle, extend_cred, revoke_cred); +} + +static void ext4_journal_abort_handle(const char *caller, unsigned int line, + const char *err_fn, + struct buffer_head *bh, + handle_t *handle, int err) +{ + char nbuf[16]; + const char *errstr = ext4_decode_error(NULL, err, nbuf); + + BUG_ON(!ext4_handle_valid(handle)); + + if (bh) + BUFFER_TRACE(bh, "abort"); + + if (!handle->h_err) + handle->h_err = err; + + if (is_handle_aborted(handle)) + return; + + printk(KERN_ERR "EXT4-fs: %s:%d: aborting transaction: %s in %s\n", + caller, line, errstr, err_fn); + + jbd2_journal_abort_handle(handle); +} + +static void ext4_check_bdev_write_error(struct super_block *sb) +{ + struct address_space *mapping = sb->s_bdev->bd_mapping; + struct ext4_sb_info *sbi = EXT4_SB(sb); + int err; + + /* + * If the block device has write error flag, it may have failed to + * async write out metadata buffers in the background. In this case, + * we could read old data from disk and write it out again, which + * may lead to on-disk filesystem inconsistency. + */ + if (errseq_check(&mapping->wb_err, READ_ONCE(sbi->s_bdev_wb_err))) { + spin_lock(&sbi->s_bdev_wb_lock); + err = errseq_check_and_advance(&mapping->wb_err, &sbi->s_bdev_wb_err); + spin_unlock(&sbi->s_bdev_wb_lock); + if (err) + ext4_error_err(sb, -err, + "Error while async write back metadata"); + } +} + +int __ext4_journal_get_write_access(const char *where, unsigned int line, + handle_t *handle, struct super_block *sb, + struct buffer_head *bh, + enum ext4_journal_trigger_type trigger_type) +{ + int err; + + might_sleep(); + + if (ext4_handle_valid(handle)) { + err = jbd2_journal_get_write_access(handle, bh); + if (err) { + ext4_journal_abort_handle(where, line, __func__, bh, + handle, err); + return err; + } + } else + ext4_check_bdev_write_error(sb); + if (trigger_type == EXT4_JTR_NONE || + !ext4_has_feature_metadata_csum(sb)) + return 0; + BUG_ON(trigger_type >= EXT4_JOURNAL_TRIGGER_COUNT); + jbd2_journal_set_triggers(bh, + &EXT4_SB(sb)->s_journal_triggers[trigger_type].tr_triggers); + return 0; +} + +/* + * The ext4 forget function must perform a revoke if we are freeing data + * which has been journaled. Metadata (eg. indirect blocks) must be + * revoked in all cases. + * + * "bh" may be NULL: a metadata block may have been freed from memory + * but there may still be a record of it in the journal, and that record + * still needs to be revoked. + */ +int __ext4_forget(const char *where, unsigned int line, handle_t *handle, + int is_metadata, struct inode *inode, + struct buffer_head *bh, ext4_fsblk_t blocknr) +{ + int err; + + might_sleep(); + + trace_ext4_forget(inode, is_metadata, blocknr); + BUFFER_TRACE(bh, "enter"); + + ext4_debug("forgetting bh %p: is_metadata=%d, mode %o, data mode %x\n", + bh, is_metadata, inode->i_mode, + test_opt(inode->i_sb, DATA_FLAGS)); + + /* + * In the no journal case, we should wait for the ongoing buffer + * to complete and do a forget. + */ + if (!ext4_handle_valid(handle)) { + if (bh) { + clear_buffer_dirty(bh); + wait_on_buffer(bh); + __bforget(bh); + } + return 0; + } + + /* Never use the revoke function if we are doing full data + * journaling: there is no need to, and a V1 superblock won't + * support it. Otherwise, only skip the revoke on un-journaled + * data blocks. */ + + if (test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA || + (!is_metadata && !ext4_should_journal_data(inode))) { + if (bh) { + BUFFER_TRACE(bh, "call jbd2_journal_forget"); + err = jbd2_journal_forget(handle, bh); + if (err) + ext4_journal_abort_handle(where, line, __func__, + bh, handle, err); + return err; + } + return 0; + } + + /* + * data!=journal && (is_metadata || should_journal_data(inode)) + */ + BUFFER_TRACE(bh, "call jbd2_journal_revoke"); + err = jbd2_journal_revoke(handle, blocknr, bh); + if (err) { + ext4_journal_abort_handle(where, line, __func__, + bh, handle, err); + __ext4_error(inode->i_sb, where, line, true, -err, 0, + "error %d when attempting revoke", err); + } + BUFFER_TRACE(bh, "exit"); + return err; +} + +int __ext4_journal_get_create_access(const char *where, unsigned int line, + handle_t *handle, struct super_block *sb, + struct buffer_head *bh, + enum ext4_journal_trigger_type trigger_type) +{ + int err; + + if (!ext4_handle_valid(handle)) + return 0; + + err = jbd2_journal_get_create_access(handle, bh); + if (err) { + ext4_journal_abort_handle(where, line, __func__, bh, handle, + err); + return err; + } + if (trigger_type == EXT4_JTR_NONE || + !ext4_has_feature_metadata_csum(sb)) + return 0; + BUG_ON(trigger_type >= EXT4_JOURNAL_TRIGGER_COUNT); + jbd2_journal_set_triggers(bh, + &EXT4_SB(sb)->s_journal_triggers[trigger_type].tr_triggers); + return 0; +} + +int __ext4_handle_dirty_metadata(const char *where, unsigned int line, + handle_t *handle, struct inode *inode, + struct buffer_head *bh) +{ + int err = 0; + + might_sleep(); + + set_buffer_meta(bh); + set_buffer_prio(bh); + set_buffer_uptodate(bh); + if (ext4_handle_valid(handle)) { + err = jbd2_journal_dirty_metadata(handle, bh); + /* Errors can only happen due to aborted journal or a nasty bug */ + if (!is_handle_aborted(handle) && WARN_ON_ONCE(err)) { + ext4_journal_abort_handle(where, line, __func__, bh, + handle, err); + if (inode == NULL) { + pr_err("EXT4: jbd2_journal_dirty_metadata " + "failed: handle type %u started at " + "line %u, credits %u/%u, errcode %d", + handle->h_type, + handle->h_line_no, + handle->h_requested_credits, + jbd2_handle_buffer_credits(handle), err); + return err; + } + ext4_error_inode(inode, where, line, + bh->b_blocknr, + "journal_dirty_metadata failed: " + "handle type %u started at line %u, " + "credits %u/%u, errcode %d", + handle->h_type, + handle->h_line_no, + handle->h_requested_credits, + jbd2_handle_buffer_credits(handle), + err); + } + } else { + if (inode) + mark_buffer_dirty_inode(bh, inode); + else + mark_buffer_dirty(bh); + if (inode && inode_needs_sync(inode)) { + sync_dirty_buffer(bh); + if (buffer_req(bh) && !buffer_uptodate(bh)) { + ext4_error_inode_err(inode, where, line, + bh->b_blocknr, EIO, + "IO error syncing itable block"); + err = -EIO; + } + } + } + return err; +} diff --git a/fs/ext4l/ext4_jbd2.h b/fs/ext4l/ext4_jbd2.h new file mode 100644 index 00000000000..63d17c5201b --- /dev/null +++ b/fs/ext4l/ext4_jbd2.h @@ -0,0 +1,461 @@ +// SPDX-License-Identifier: GPL-2.0+ +/* + * ext4_jbd2.h + * + * Written by Stephen C. Tweedie <sct@redhat.com>, 1999 + * + * Copyright 1998--1999 Red Hat corp --- All Rights Reserved + * + * Ext4-specific journaling extensions. + */ + +#ifndef _EXT4_JBD2_H +#define _EXT4_JBD2_H + +#include <linux/fs.h> +#include <linux/jbd2.h> +#include "ext4.h" + +#define EXT4_JOURNAL(inode) (EXT4_SB((inode)->i_sb)->s_journal) + +/* Define the number of blocks we need to account to a transaction to + * modify one block of data. + * + * We may have to touch one inode, one bitmap buffer, up to three + * indirection blocks, the group and superblock summaries, and the data + * block to complete the transaction. + * + * For extents-enabled fs we may have to allocate and modify up to + * 5 levels of tree, data block (for each of these we need bitmap + group + * summaries), root which is stored in the inode, sb + */ + +#define EXT4_SINGLEDATA_TRANS_BLOCKS(sb) \ + (ext4_has_feature_extents(sb) ? 20U : 8U) + +/* Extended attribute operations touch at most two data buffers, + * two bitmap buffers, and two group summaries, in addition to the inode + * and the superblock, which are already accounted for. */ + +#define EXT4_XATTR_TRANS_BLOCKS 6U + +/* Define the minimum size for a transaction which modifies data. This + * needs to take into account the fact that we may end up modifying two + * quota files too (one for the group, one for the user quota). The + * superblock only gets updated once, of course, so don't bother + * counting that again for the quota updates. */ + +#define EXT4_DATA_TRANS_BLOCKS(sb) (EXT4_SINGLEDATA_TRANS_BLOCKS(sb) + \ + EXT4_XATTR_TRANS_BLOCKS - 2 + \ + EXT4_MAXQUOTAS_TRANS_BLOCKS(sb)) + +/* + * Define the number of metadata blocks we need to account to modify data. + * + * This include super block, inode block, quota blocks and xattr blocks + */ +#define EXT4_META_TRANS_BLOCKS(sb) (EXT4_XATTR_TRANS_BLOCKS + \ + EXT4_MAXQUOTAS_TRANS_BLOCKS(sb)) + +/* Define an arbitrary limit for the amount of data we will anticipate + * writing to any given transaction. For unbounded transactions such as + * write(2) and truncate(2) we can write more than this, but we always + * start off at the maximum transaction size and grow the transaction + * optimistically as we go. */ + +#define EXT4_MAX_TRANS_DATA 64U + +/* We break up a large truncate or write transaction once the handle's + * buffer credits gets this low, we need either to extend the + * transaction or to start a new one. Reserve enough space here for + * inode, bitmap, superblock, group and indirection updates for at least + * one block, plus two quota updates. Quota allocations are not + * needed. */ + +#define EXT4_RESERVE_TRANS_BLOCKS 12U + +/* + * Number of credits needed if we need to insert an entry into a + * directory. For each new index block, we need 4 blocks (old index + * block, new index block, bitmap block, bg summary). For normal + * htree directories there are 2 levels; if the largedir feature + * enabled it's 3 levels. + */ +#define EXT4_INDEX_EXTRA_TRANS_BLOCKS 12U + +#ifdef CONFIG_QUOTA +/* Amount of blocks needed for quota update - we know that the structure was + * allocated so we need to update only data block */ +#define EXT4_QUOTA_TRANS_BLOCKS(sb) ((ext4_quota_capable(sb)) ? 1 : 0) +/* Amount of blocks needed for quota insert/delete - we do some block writes + * but inode, sb and group updates are done only once */ +#define EXT4_QUOTA_INIT_BLOCKS(sb) ((ext4_quota_capable(sb)) ?\ + (DQUOT_INIT_ALLOC*(EXT4_SINGLEDATA_TRANS_BLOCKS(sb)-3)\ + +3+DQUOT_INIT_REWRITE) : 0) + +#define EXT4_QUOTA_DEL_BLOCKS(sb) ((ext4_quota_capable(sb)) ?\ + (DQUOT_DEL_ALLOC*(EXT4_SINGLEDATA_TRANS_BLOCKS(sb)-3)\ + +3+DQUOT_DEL_REWRITE) : 0) +#else +#define EXT4_QUOTA_TRANS_BLOCKS(sb) 0 +#define EXT4_QUOTA_INIT_BLOCKS(sb) 0 +#define EXT4_QUOTA_DEL_BLOCKS(sb) 0 +#endif +#define EXT4_MAXQUOTAS_TRANS_BLOCKS(sb) (EXT4_MAXQUOTAS*EXT4_QUOTA_TRANS_BLOCKS(sb)) +#define EXT4_MAXQUOTAS_INIT_BLOCKS(sb) (EXT4_MAXQUOTAS*EXT4_QUOTA_INIT_BLOCKS(sb)) +#define EXT4_MAXQUOTAS_DEL_BLOCKS(sb) (EXT4_MAXQUOTAS*EXT4_QUOTA_DEL_BLOCKS(sb)) + +/* + * Ext4 handle operation types -- for logging purposes + */ +#define EXT4_HT_MISC 0 +#define EXT4_HT_INODE 1 +#define EXT4_HT_WRITE_PAGE 2 +#define EXT4_HT_MAP_BLOCKS 3 +#define EXT4_HT_DIR 4 +#define EXT4_HT_TRUNCATE 5 +#define EXT4_HT_QUOTA 6 +#define EXT4_HT_RESIZE 7 +#define EXT4_HT_MIGRATE 8 +#define EXT4_HT_MOVE_EXTENTS 9 +#define EXT4_HT_XATTR 10 +#define EXT4_HT_EXT_CONVERT 11 +#define EXT4_HT_MAX 12 + +int +ext4_mark_iloc_dirty(handle_t *handle, + struct inode *inode, + struct ext4_iloc *iloc); + +/* + * On success, We end up with an outstanding reference count against + * iloc->bh. This _must_ be cleaned up later. + */ + +int ext4_reserve_inode_write(handle_t *handle, struct inode *inode, + struct ext4_iloc *iloc); + +#define ext4_mark_inode_dirty(__h, __i) \ + __ext4_mark_inode_dirty((__h), (__i), __func__, __LINE__) +int __ext4_mark_inode_dirty(handle_t *handle, struct inode *inode, + const char *func, unsigned int line); + +int ext4_expand_extra_isize(struct inode *inode, + unsigned int new_extra_isize, + struct ext4_iloc *iloc); +/* + * Wrapper functions with which ext4 calls into JBD. + */ +int __ext4_journal_get_write_access(const char *where, unsigned int line, + handle_t *handle, struct super_block *sb, + struct buffer_head *bh, + enum ext4_journal_trigger_type trigger_type); + +int __ext4_forget(const char *where, unsigned int line, handle_t *handle, + int is_metadata, struct inode *inode, + struct buffer_head *bh, ext4_fsblk_t blocknr); + +int __ext4_journal_get_create_access(const char *where, unsigned int line, + handle_t *handle, struct super_block *sb, + struct buffer_head *bh, + enum ext4_journal_trigger_type trigger_type); + +int __ext4_handle_dirty_metadata(const char *where, unsigned int line, + handle_t *handle, struct inode *inode, + struct buffer_head *bh); + +#define ext4_journal_get_write_access(handle, sb, bh, trigger_type) \ + __ext4_journal_get_write_access(__func__, __LINE__, (handle), (sb), \ + (bh), (trigger_type)) +#define ext4_forget(handle, is_metadata, inode, bh, block_nr) \ + __ext4_forget(__func__, __LINE__, (handle), (is_metadata), (inode), \ + (bh), (block_nr)) +#define ext4_journal_get_create_access(handle, sb, bh, trigger_type) \ + __ext4_journal_get_create_access(__func__, __LINE__, (handle), (sb), \ + (bh), (trigger_type)) +#define ext4_handle_dirty_metadata(handle, inode, bh) \ + __ext4_handle_dirty_metadata(__func__, __LINE__, (handle), (inode), \ + (bh)) + +handle_t *__ext4_journal_start_sb(struct inode *inode, struct super_block *sb, + unsigned int line, int type, int blocks, + int rsv_blocks, int revoke_creds); +int __ext4_journal_stop(const char *where, unsigned int line, handle_t *handle); + +#define EXT4_NOJOURNAL_MAX_REF_COUNT ((unsigned long) 4096) + +/* Note: Do not use this for NULL handles. This is only to determine if + * a properly allocated handle is using a journal or not. */ +static inline int ext4_handle_valid(handle_t *handle) +{ + if ((unsigned long)handle < EXT4_NOJOURNAL_MAX_REF_COUNT) + return 0; + return 1; +} + +static inline void ext4_handle_sync(handle_t *handle) +{ + if (ext4_handle_valid(handle)) + handle->h_sync = 1; +} + +static inline int ext4_handle_is_aborted(handle_t *handle) +{ + if (ext4_handle_valid(handle)) + return is_handle_aborted(handle); + return 0; +} + +static inline int ext4_free_metadata_revoke_credits(struct super_block *sb, + int blocks) +{ + /* Freeing each metadata block can result in freeing one cluster */ + return blocks * EXT4_SB(sb)->s_cluster_ratio; +} + +static inline int ext4_trans_default_revoke_credits(struct super_block *sb) +{ + return ext4_free_metadata_revoke_credits(sb, 8); +} + +#define ext4_journal_start_sb(sb, type, nblocks) \ + __ext4_journal_start_sb(NULL, (sb), __LINE__, (type), (nblocks), 0,\ + ext4_trans_default_revoke_credits(sb)) + +#define ext4_journal_start(inode, type, nblocks) \ + __ext4_journal_start((inode), __LINE__, (type), (nblocks), 0, \ + ext4_trans_default_revoke_credits((inode)->i_sb)) + +#define ext4_journal_start_with_reserve(inode, type, blocks, rsv_blocks)\ + __ext4_journal_start((inode), __LINE__, (type), (blocks), (rsv_blocks),\ + ext4_trans_default_revoke_credits((inode)->i_sb)) + +#define ext4_journal_start_with_revoke(inode, type, blocks, revoke_creds) \ + __ext4_journal_start((inode), __LINE__, (type), (blocks), 0, \ + (revoke_creds)) + +static inline handle_t *__ext4_journal_start(struct inode *inode, + unsigned int line, int type, + int blocks, int rsv_blocks, + int revoke_creds) +{ + return __ext4_journal_start_sb(inode, inode->i_sb, line, type, blocks, + rsv_blocks, revoke_creds); +} + +#define ext4_journal_stop(handle) \ + __ext4_journal_stop(__func__, __LINE__, (handle)) + +#define ext4_journal_start_reserved(handle, type) \ + __ext4_journal_start_reserved((handle), __LINE__, (type)) + +handle_t *__ext4_journal_start_reserved(handle_t *handle, unsigned int line, + int type); + +static inline handle_t *ext4_journal_current_handle(void) +{ + return journal_current_handle(); +} + +static inline int ext4_journal_extend(handle_t *handle, int nblocks, int revoke) +{ + if (ext4_handle_valid(handle)) + return jbd2_journal_extend(handle, nblocks, revoke); + return 0; +} + +static inline int ext4_journal_restart(handle_t *handle, int nblocks, + int revoke) +{ + if (ext4_handle_valid(handle)) + return jbd2__journal_restart(handle, nblocks, revoke, GFP_NOFS); + return 0; +} + +int __ext4_journal_ensure_credits(handle_t *handle, int check_cred, + int extend_cred, int revoke_cred); + + +/* + * Ensure @handle has at least @check_creds credits available. If not, + * transaction will be extended or restarted to contain at least @extend_cred + * credits. Before restarting transaction @fn is executed to allow for cleanup + * before the transaction is restarted. + * + * The return value is < 0 in case of error, 0 in case the handle has enough + * credits or transaction extension succeeded, 1 in case transaction had to be + * restarted. + */ +#define ext4_journal_ensure_credits_fn(handle, check_cred, extend_cred, \ + revoke_cred, fn) \ +({ \ + __label__ __ensure_end; \ + int err = __ext4_journal_ensure_credits((handle), (check_cred), \ + (extend_cred), (revoke_cred)); \ + \ + if (err <= 0) \ + goto __ensure_end; \ + err = (fn); \ + if (err < 0) \ + goto __ensure_end; \ + err = ext4_journal_restart((handle), (extend_cred), (revoke_cred)); \ + if (err == 0) \ + err = 1; \ +__ensure_end: \ + err; \ +}) + +/* + * Ensure given handle has at least requested amount of credits available, + * possibly restarting transaction if needed. We also make sure the transaction + * has space for at least ext4_trans_default_revoke_credits(sb) revoke records + * as freeing one or two blocks is very common pattern and requesting this is + * very cheap. + */ +static inline int ext4_journal_ensure_credits(handle_t *handle, int credits, + int revoke_creds) +{ + return ext4_journal_ensure_credits_fn(handle, credits, credits, + revoke_creds, 0); +} + +static inline int ext4_journal_blocks_per_folio(struct inode *inode) +{ + if (EXT4_JOURNAL(inode) != NULL) + return jbd2_journal_blocks_per_folio(inode); + return 0; +} + +static inline int ext4_journal_force_commit(journal_t *journal) +{ + if (journal) + return jbd2_journal_force_commit(journal); + return 0; +} + +static inline int ext4_jbd2_inode_add_write(handle_t *handle, + struct inode *inode, loff_t start_byte, loff_t length) +{ + if (ext4_handle_valid(handle)) + return jbd2_journal_inode_ranged_write(handle, + EXT4_I(inode)->jinode, start_byte, length); + return 0; +} + +static inline int ext4_jbd2_inode_add_wait(handle_t *handle, + struct inode *inode, loff_t start_byte, loff_t length) +{ + if (ext4_handle_valid(handle)) + return jbd2_journal_inode_ranged_wait(handle, + EXT4_I(inode)->jinode, start_byte, length); + return 0; +} + +static inline void ext4_update_inode_fsync_trans(handle_t *handle, + struct inode *inode, + int datasync) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + + if (ext4_handle_valid(handle) && !is_handle_aborted(handle)) { + ei->i_sync_tid = handle->h_transaction->t_tid; + if (datasync) + ei->i_datasync_tid = handle->h_transaction->t_tid; + } +} + +/* super.c */ +int ext4_force_commit(struct super_block *sb); + +/* + * Ext4 inode journal modes + */ +#define EXT4_INODE_JOURNAL_DATA_MODE 0x01 /* journal data mode */ +#define EXT4_INODE_ORDERED_DATA_MODE 0x02 /* ordered data mode */ +#define EXT4_INODE_WRITEBACK_DATA_MODE 0x04 /* writeback data mode */ + +int ext4_inode_journal_mode(struct inode *inode); + +static inline int ext4_should_journal_data(struct inode *inode) +{ + return ext4_inode_journal_mode(inode) & EXT4_INODE_JOURNAL_DATA_MODE; +} + +static inline int ext4_should_order_data(struct inode *inode) +{ + return ext4_inode_journal_mode(inode) & EXT4_INODE_ORDERED_DATA_MODE; +} + +static inline int ext4_should_writeback_data(struct inode *inode) +{ + return ext4_inode_journal_mode(inode) & EXT4_INODE_WRITEBACK_DATA_MODE; +} + +static inline int ext4_free_data_revoke_credits(struct inode *inode, int blocks) +{ + if (test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA) + return 0; + if (!ext4_should_journal_data(inode)) + return 0; + /* + * Data blocks in one extent are contiguous, just account for partial + * clusters at extent boundaries + */ + return blocks + 2*(EXT4_SB(inode->i_sb)->s_cluster_ratio - 1); +} + +/* + * This function controls whether or not we should try to go down the + * dioread_nolock code paths, which makes it safe to avoid taking + * i_rwsem for direct I/O reads. This only works for extent-based + * files, and it doesn't work if data journaling is enabled, since the + * dioread_nolock code uses b_private to pass information back to the + * I/O completion handler, and this conflicts with the jbd's use of + * b_private. + */ +static inline int ext4_should_dioread_nolock(struct inode *inode) +{ + if (!test_opt(inode->i_sb, DIOREAD_NOLOCK)) + return 0; + if (!S_ISREG(inode->i_mode)) + return 0; + if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) + return 0; + if (ext4_should_journal_data(inode)) + return 0; + /* temporary fix to prevent generic/422 test failures */ + if (!test_opt(inode->i_sb, DELALLOC)) + return 0; + return 1; +} + +/* + * Pass journal explicitly as it may not be cached in the sbi->s_journal in some + * cases + */ +static inline int ext4_journal_destroy(struct ext4_sb_info *sbi, journal_t *journal) +{ + int err = 0; + + /* + * At this point only two things can be operating on the journal. + * JBD2 thread performing transaction commit and s_sb_upd_work + * issuing sb update through the journal. Once we set + * EXT4_JOURNAL_DESTROY, new ext4_handle_error() calls will not + * queue s_sb_upd_work and ext4_force_commit() makes sure any + * ext4_handle_error() calls from the running transaction commit are + * finished. Hence no new s_sb_upd_work can be queued after we + * flush it here. + */ + ext4_set_mount_flag(sbi->s_sb, EXT4_MF_JOURNAL_DESTROY); + + ext4_force_commit(sbi->s_sb); + flush_work(&sbi->s_sb_upd_work); + + err = jbd2_journal_destroy(journal); + sbi->s_journal = NULL; + + return err; +} + +#endif /* _EXT4_JBD2_H */ diff --git a/fs/ext4l/fast_commit.c b/fs/ext4l/fast_commit.c new file mode 100644 index 00000000000..fa66b08de99 --- /dev/null +++ b/fs/ext4l/fast_commit.c @@ -0,0 +1,2343 @@ +// SPDX-License-Identifier: GPL-2.0 + +/* + * fs/ext4/fast_commit.c + * + * Written by Harshad Shirwadkar <harshadshirwadkar@gmail.com> + * + * Ext4 fast commits routines. + */ +#include "ext4.h" +#include "ext4_jbd2.h" +#include "ext4_extents.h" +#include "mballoc.h" + +#include <linux/lockdep.h> +/* + * Ext4 Fast Commits + * ----------------- + * + * Ext4 fast commits implement fine grained journalling for Ext4. + * + * Fast commits are organized as a log of tag-length-value (TLV) structs. (See + * struct ext4_fc_tl). Each TLV contains some delta that is replayed TLV by + * TLV during the recovery phase. For the scenarios for which we currently + * don't have replay code, fast commit falls back to full commits. + * Fast commits record delta in one of the following three categories. + * + * (A) Directory entry updates: + * + * - EXT4_FC_TAG_UNLINK - records directory entry unlink + * - EXT4_FC_TAG_LINK - records directory entry link + * - EXT4_FC_TAG_CREAT - records inode and directory entry creation + * + * (B) File specific data range updates: + * + * - EXT4_FC_TAG_ADD_RANGE - records addition of new blocks to an inode + * - EXT4_FC_TAG_DEL_RANGE - records deletion of blocks from an inode + * + * (C) Inode metadata (mtime / ctime etc): + * + * - EXT4_FC_TAG_INODE - record the inode that should be replayed + * during recovery. Note that iblocks field is + * not replayed and instead derived during + * replay. + * Commit Operation + * ---------------- + * With fast commits, we maintain all the directory entry operations in the + * order in which they are issued in an in-memory queue. This queue is flushed + * to disk during the commit operation. We also maintain a list of inodes + * that need to be committed during a fast commit in another in memory queue of + * inodes. During the commit operation, we commit in the following order: + * + * [1] Prepare all the inodes to write out their data by setting + * "EXT4_STATE_FC_FLUSHING_DATA". This ensures that inode cannot be + * deleted while it is being flushed. + * [2] Flush data buffers to disk and clear "EXT4_STATE_FC_FLUSHING_DATA" + * state. + * [3] Lock the journal by calling jbd2_journal_lock_updates. This ensures that + * all the exsiting handles finish and no new handles can start. + * [4] Mark all the fast commit eligible inodes as undergoing fast commit + * by setting "EXT4_STATE_FC_COMMITTING" state. + * [5] Unlock the journal by calling jbd2_journal_unlock_updates. This allows + * starting of new handles. If new handles try to start an update on + * any of the inodes that are being committed, ext4_fc_track_inode() + * will block until those inodes have finished the fast commit. + * [6] Commit all the directory entry updates in the fast commit space. + * [7] Commit all the changed inodes in the fast commit space and clear + * "EXT4_STATE_FC_COMMITTING" for these inodes. + * [8] Write tail tag (this tag ensures the atomicity, please read the following + * section for more details). + * + * All the inode updates must be enclosed within jbd2_jounrnal_start() + * and jbd2_journal_stop() similar to JBD2 journaling. + * + * Fast Commit Ineligibility + * ------------------------- + * + * Not all operations are supported by fast commits today (e.g extended + * attributes). Fast commit ineligibility is marked by calling + * ext4_fc_mark_ineligible(): This makes next fast commit operation to fall back + * to full commit. + * + * Atomicity of commits + * -------------------- + * In order to guarantee atomicity during the commit operation, fast commit + * uses "EXT4_FC_TAG_TAIL" tag that marks a fast commit as complete. Tail + * tag contains CRC of the contents and TID of the transaction after which + * this fast commit should be applied. Recovery code replays fast commit + * logs only if there's at least 1 valid tail present. For every fast commit + * operation, there is 1 tail. This means, we may end up with multiple tails + * in the fast commit space. Here's an example: + * + * - Create a new file A and remove existing file B + * - fsync() + * - Append contents to file A + * - Truncate file A + * - fsync() + * + * The fast commit space at the end of above operations would look like this: + * [HEAD] [CREAT A] [UNLINK B] [TAIL] [ADD_RANGE A] [DEL_RANGE A] [TAIL] + * |<--- Fast Commit 1 --->|<--- Fast Commit 2 ---->| + * + * Replay code should thus check for all the valid tails in the FC area. + * + * Fast Commit Replay Idempotence + * ------------------------------ + * + * Fast commits tags are idempotent in nature provided the recovery code follows + * certain rules. The guiding principle that the commit path follows while + * committing is that it stores the result of a particular operation instead of + * storing the procedure. + * + * Let's consider this rename operation: 'mv /a /b'. Let's assume dirent '/a' + * was associated with inode 10. During fast commit, instead of storing this + * operation as a procedure "rename a to b", we store the resulting file system + * state as a "series" of outcomes: + * + * - Link dirent b to inode 10 + * - Unlink dirent a + * - Inode <10> with valid refcount + * + * Now when recovery code runs, it needs "enforce" this state on the file + * system. This is what guarantees idempotence of fast commit replay. + * + * Let's take an example of a procedure that is not idempotent and see how fast + * commits make it idempotent. Consider following sequence of operations: + * + * rm A; mv B A; read A + * (x) (y) (z) + * + * (x), (y) and (z) are the points at which we can crash. If we store this + * sequence of operations as is then the replay is not idempotent. Let's say + * while in replay, we crash at (z). During the second replay, file A (which was + * actually created as a result of "mv B A" operation) would get deleted. Thus, + * file named A would be absent when we try to read A. So, this sequence of + * operations is not idempotent. However, as mentioned above, instead of storing + * the procedure fast commits store the outcome of each procedure. Thus the fast + * commit log for above procedure would be as follows: + * + * (Let's assume dirent A was linked to inode 10 and dirent B was linked to + * inode 11 before the replay) + * + * [Unlink A] [Link A to inode 11] [Unlink B] [Inode 11] + * (w) (x) (y) (z) + * + * If we crash at (z), we will have file A linked to inode 11. During the second + * replay, we will remove file A (inode 11). But we will create it back and make + * it point to inode 11. We won't find B, so we'll just skip that step. At this + * point, the refcount for inode 11 is not reliable, but that gets fixed by the + * replay of last inode 11 tag. Crashes at points (w), (x) and (y) get handled + * similarly. Thus, by converting a non-idempotent procedure into a series of + * idempotent outcomes, fast commits ensured idempotence during the replay. + * + * Locking + * ------- + * sbi->s_fc_lock protects the fast commit inodes queue and the fast commit + * dentry queue. ei->i_fc_lock protects the fast commit related info in a given + * inode. Most of the code avoids acquiring both the locks, but if one must do + * that then sbi->s_fc_lock must be acquired before ei->i_fc_lock. + * + * TODOs + * ----- + * + * 0) Fast commit replay path hardening: Fast commit replay code should use + * journal handles to make sure all the updates it does during the replay + * path are atomic. With that if we crash during fast commit replay, after + * trying to do recovery again, we will find a file system where fast commit + * area is invalid (because new full commit would be found). In order to deal + * with that, fast commit replay code should ensure that the "FC_REPLAY" + * superblock state is persisted before starting the replay, so that after + * the crash, fast commit recovery code can look at that flag and perform + * fast commit recovery even if that area is invalidated by later full + * commits. + * + * 1) Handle more ineligible cases. + * + * 2) Change ext4_fc_commit() to lookup logical to physical mapping using extent + * status tree. This would get rid of the need to call ext4_fc_track_inode() + * before acquiring i_data_sem. To do that we would need to ensure that + * modified extents from the extent status tree are not evicted from memory. + */ + +#include <trace/events/ext4.h> +static struct kmem_cache *ext4_fc_dentry_cachep; + +static void ext4_end_buffer_io_sync(struct buffer_head *bh, int uptodate) +{ + BUFFER_TRACE(bh, ""); + if (uptodate) { + ext4_debug("%s: Block %lld up-to-date", + __func__, bh->b_blocknr); + set_buffer_uptodate(bh); + } else { + ext4_debug("%s: Block %lld not up-to-date", + __func__, bh->b_blocknr); + clear_buffer_uptodate(bh); + } + + unlock_buffer(bh); +} + +static inline void ext4_fc_reset_inode(struct inode *inode) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + + ei->i_fc_lblk_start = 0; + ei->i_fc_lblk_len = 0; +} + +void ext4_fc_init_inode(struct inode *inode) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + + ext4_fc_reset_inode(inode); + ext4_clear_inode_state(inode, EXT4_STATE_FC_COMMITTING); + INIT_LIST_HEAD(&ei->i_fc_list); + INIT_LIST_HEAD(&ei->i_fc_dilist); + init_waitqueue_head(&ei->i_fc_wait); +} + +static bool ext4_fc_disabled(struct super_block *sb) +{ + return (!test_opt2(sb, JOURNAL_FAST_COMMIT) || + (EXT4_SB(sb)->s_mount_state & EXT4_FC_REPLAY)); +} + +/* + * Remove inode from fast commit list. If the inode is being committed + * we wait until inode commit is done. + */ +void ext4_fc_del(struct inode *inode) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + struct ext4_fc_dentry_update *fc_dentry; + wait_queue_head_t *wq; + + if (ext4_fc_disabled(inode->i_sb)) + return; + + mutex_lock(&sbi->s_fc_lock); + if (list_empty(&ei->i_fc_list) && list_empty(&ei->i_fc_dilist)) { + mutex_unlock(&sbi->s_fc_lock); + return; + } + + /* + * Since ext4_fc_del is called from ext4_evict_inode while having a + * handle open, there is no need for us to wait here even if a fast + * commit is going on. That is because, if this inode is being + * committed, ext4_mark_inode_dirty would have waited for inode commit + * operation to finish before we come here. So, by the time we come + * here, inode's EXT4_STATE_FC_COMMITTING would have been cleared. So, + * we shouldn't see EXT4_STATE_FC_COMMITTING to be set on this inode + * here. + * + * We may come here without any handles open in the "no_delete" case of + * ext4_evict_inode as well. However, if that happens, we first mark the + * file system as fast commit ineligible anyway. So, even in that case, + * it is okay to remove the inode from the fc list. + */ + WARN_ON(ext4_test_inode_state(inode, EXT4_STATE_FC_COMMITTING) + && !ext4_test_mount_flag(inode->i_sb, EXT4_MF_FC_INELIGIBLE)); + while (ext4_test_inode_state(inode, EXT4_STATE_FC_FLUSHING_DATA)) { +#if (BITS_PER_LONG < 64) + DEFINE_WAIT_BIT(wait, &ei->i_state_flags, + EXT4_STATE_FC_FLUSHING_DATA); + wq = bit_waitqueue(&ei->i_state_flags, + EXT4_STATE_FC_FLUSHING_DATA); +#else + DEFINE_WAIT_BIT(wait, &ei->i_flags, + EXT4_STATE_FC_FLUSHING_DATA); + wq = bit_waitqueue(&ei->i_flags, + EXT4_STATE_FC_FLUSHING_DATA); +#endif + prepare_to_wait(wq, &wait.wq_entry, TASK_UNINTERRUPTIBLE); + if (ext4_test_inode_state(inode, EXT4_STATE_FC_FLUSHING_DATA)) { + mutex_unlock(&sbi->s_fc_lock); + schedule(); + mutex_lock(&sbi->s_fc_lock); + } + finish_wait(wq, &wait.wq_entry); + } + list_del_init(&ei->i_fc_list); + + /* + * Since this inode is getting removed, let's also remove all FC + * dentry create references, since it is not needed to log it anyways. + */ + if (list_empty(&ei->i_fc_dilist)) { + mutex_unlock(&sbi->s_fc_lock); + return; + } + + fc_dentry = list_first_entry(&ei->i_fc_dilist, struct ext4_fc_dentry_update, fcd_dilist); + WARN_ON(fc_dentry->fcd_op != EXT4_FC_TAG_CREAT); + list_del_init(&fc_dentry->fcd_list); + list_del_init(&fc_dentry->fcd_dilist); + + WARN_ON(!list_empty(&ei->i_fc_dilist)); + mutex_unlock(&sbi->s_fc_lock); + + release_dentry_name_snapshot(&fc_dentry->fcd_name); + kmem_cache_free(ext4_fc_dentry_cachep, fc_dentry); +} + +/* + * Mark file system as fast commit ineligible, and record latest + * ineligible transaction tid. This means until the recorded + * transaction, commit operation would result in a full jbd2 commit. + */ +void ext4_fc_mark_ineligible(struct super_block *sb, int reason, handle_t *handle) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + tid_t tid; + bool has_transaction = true; + bool is_ineligible; + + if (ext4_fc_disabled(sb)) + return; + + if (handle && !IS_ERR(handle)) + tid = handle->h_transaction->t_tid; + else { + read_lock(&sbi->s_journal->j_state_lock); + if (sbi->s_journal->j_running_transaction) + tid = sbi->s_journal->j_running_transaction->t_tid; + else + has_transaction = false; + read_unlock(&sbi->s_journal->j_state_lock); + } + mutex_lock(&sbi->s_fc_lock); + is_ineligible = ext4_test_mount_flag(sb, EXT4_MF_FC_INELIGIBLE); + if (has_transaction && (!is_ineligible || tid_gt(tid, sbi->s_fc_ineligible_tid))) + sbi->s_fc_ineligible_tid = tid; + ext4_set_mount_flag(sb, EXT4_MF_FC_INELIGIBLE); + mutex_unlock(&sbi->s_fc_lock); + WARN_ON(reason >= EXT4_FC_REASON_MAX); + sbi->s_fc_stats.fc_ineligible_reason_count[reason]++; +} + +/* + * Generic fast commit tracking function. If this is the first time this we are + * called after a full commit, we initialize fast commit fields and then call + * __fc_track_fn() with update = 0. If we have already been called after a full + * commit, we pass update = 1. Based on that, the track function can determine + * if it needs to track a field for the first time or if it needs to just + * update the previously tracked value. + * + * If enqueue is set, this function enqueues the inode in fast commit list. + */ +static int ext4_fc_track_template( + handle_t *handle, struct inode *inode, + int (*__fc_track_fn)(handle_t *handle, struct inode *, void *, bool), + void *args, int enqueue) +{ + bool update = false; + struct ext4_inode_info *ei = EXT4_I(inode); + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + tid_t tid = 0; + int ret; + + tid = handle->h_transaction->t_tid; + spin_lock(&ei->i_fc_lock); + if (tid == ei->i_sync_tid) { + update = true; + } else { + ext4_fc_reset_inode(inode); + ei->i_sync_tid = tid; + } + ret = __fc_track_fn(handle, inode, args, update); + spin_unlock(&ei->i_fc_lock); + if (!enqueue) + return ret; + + mutex_lock(&sbi->s_fc_lock); + if (list_empty(&EXT4_I(inode)->i_fc_list)) + list_add_tail(&EXT4_I(inode)->i_fc_list, + (sbi->s_journal->j_flags & JBD2_FULL_COMMIT_ONGOING || + sbi->s_journal->j_flags & JBD2_FAST_COMMIT_ONGOING) ? + &sbi->s_fc_q[FC_Q_STAGING] : + &sbi->s_fc_q[FC_Q_MAIN]); + mutex_unlock(&sbi->s_fc_lock); + + return ret; +} + +struct __track_dentry_update_args { + struct dentry *dentry; + int op; +}; + +/* __track_fn for directory entry updates. Called with ei->i_fc_lock. */ +static int __track_dentry_update(handle_t *handle, struct inode *inode, + void *arg, bool update) +{ + struct ext4_fc_dentry_update *node; + struct ext4_inode_info *ei = EXT4_I(inode); + struct __track_dentry_update_args *dentry_update = + (struct __track_dentry_update_args *)arg; + struct dentry *dentry = dentry_update->dentry; + struct inode *dir = dentry->d_parent->d_inode; + struct super_block *sb = inode->i_sb; + struct ext4_sb_info *sbi = EXT4_SB(sb); + + spin_unlock(&ei->i_fc_lock); + + if (IS_ENCRYPTED(dir)) { + ext4_fc_mark_ineligible(sb, EXT4_FC_REASON_ENCRYPTED_FILENAME, + handle); + spin_lock(&ei->i_fc_lock); + return -EOPNOTSUPP; + } + + node = kmem_cache_alloc(ext4_fc_dentry_cachep, GFP_NOFS); + if (!node) { + ext4_fc_mark_ineligible(sb, EXT4_FC_REASON_NOMEM, handle); + spin_lock(&ei->i_fc_lock); + return -ENOMEM; + } + + node->fcd_op = dentry_update->op; + node->fcd_parent = dir->i_ino; + node->fcd_ino = inode->i_ino; + take_dentry_name_snapshot(&node->fcd_name, dentry); + INIT_LIST_HEAD(&node->fcd_dilist); + INIT_LIST_HEAD(&node->fcd_list); + mutex_lock(&sbi->s_fc_lock); + if (sbi->s_journal->j_flags & JBD2_FULL_COMMIT_ONGOING || + sbi->s_journal->j_flags & JBD2_FAST_COMMIT_ONGOING) + list_add_tail(&node->fcd_list, + &sbi->s_fc_dentry_q[FC_Q_STAGING]); + else + list_add_tail(&node->fcd_list, &sbi->s_fc_dentry_q[FC_Q_MAIN]); + + /* + * This helps us keep a track of all fc_dentry updates which is part of + * this ext4 inode. So in case the inode is getting unlinked, before + * even we get a chance to fsync, we could remove all fc_dentry + * references while evicting the inode in ext4_fc_del(). + * Also with this, we don't need to loop over all the inodes in + * sbi->s_fc_q to get the corresponding inode in + * ext4_fc_commit_dentry_updates(). + */ + if (dentry_update->op == EXT4_FC_TAG_CREAT) { + WARN_ON(!list_empty(&ei->i_fc_dilist)); + list_add_tail(&node->fcd_dilist, &ei->i_fc_dilist); + } + mutex_unlock(&sbi->s_fc_lock); + spin_lock(&ei->i_fc_lock); + + return 0; +} + +void __ext4_fc_track_unlink(handle_t *handle, + struct inode *inode, struct dentry *dentry) +{ + struct __track_dentry_update_args args; + int ret; + + args.dentry = dentry; + args.op = EXT4_FC_TAG_UNLINK; + + ret = ext4_fc_track_template(handle, inode, __track_dentry_update, + (void *)&args, 0); + trace_ext4_fc_track_unlink(handle, inode, dentry, ret); +} + +void ext4_fc_track_unlink(handle_t *handle, struct dentry *dentry) +{ + struct inode *inode = d_inode(dentry); + + if (ext4_fc_disabled(inode->i_sb)) + return; + + if (ext4_test_mount_flag(inode->i_sb, EXT4_MF_FC_INELIGIBLE)) + return; + + __ext4_fc_track_unlink(handle, inode, dentry); +} + +void __ext4_fc_track_link(handle_t *handle, + struct inode *inode, struct dentry *dentry) +{ + struct __track_dentry_update_args args; + int ret; + + args.dentry = dentry; + args.op = EXT4_FC_TAG_LINK; + + ret = ext4_fc_track_template(handle, inode, __track_dentry_update, + (void *)&args, 0); + trace_ext4_fc_track_link(handle, inode, dentry, ret); +} + +void ext4_fc_track_link(handle_t *handle, struct dentry *dentry) +{ + struct inode *inode = d_inode(dentry); + + if (ext4_fc_disabled(inode->i_sb)) + return; + + if (ext4_test_mount_flag(inode->i_sb, EXT4_MF_FC_INELIGIBLE)) + return; + + __ext4_fc_track_link(handle, inode, dentry); +} + +void __ext4_fc_track_create(handle_t *handle, struct inode *inode, + struct dentry *dentry) +{ + struct __track_dentry_update_args args; + int ret; + + args.dentry = dentry; + args.op = EXT4_FC_TAG_CREAT; + + ret = ext4_fc_track_template(handle, inode, __track_dentry_update, + (void *)&args, 0); + trace_ext4_fc_track_create(handle, inode, dentry, ret); +} + +void ext4_fc_track_create(handle_t *handle, struct dentry *dentry) +{ + struct inode *inode = d_inode(dentry); + + if (ext4_fc_disabled(inode->i_sb)) + return; + + if (ext4_test_mount_flag(inode->i_sb, EXT4_MF_FC_INELIGIBLE)) + return; + + __ext4_fc_track_create(handle, inode, dentry); +} + +/* __track_fn for inode tracking */ +static int __track_inode(handle_t *handle, struct inode *inode, void *arg, + bool update) +{ + if (update) + return -EEXIST; + + EXT4_I(inode)->i_fc_lblk_len = 0; + + return 0; +} + +void ext4_fc_track_inode(handle_t *handle, struct inode *inode) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + wait_queue_head_t *wq; + int ret; + + if (S_ISDIR(inode->i_mode)) + return; + + if (ext4_fc_disabled(inode->i_sb)) + return; + + if (ext4_should_journal_data(inode)) { + ext4_fc_mark_ineligible(inode->i_sb, + EXT4_FC_REASON_INODE_JOURNAL_DATA, handle); + return; + } + + if (ext4_test_mount_flag(inode->i_sb, EXT4_MF_FC_INELIGIBLE)) + return; + + /* + * If we come here, we may sleep while waiting for the inode to + * commit. We shouldn't be holding i_data_sem when we go to sleep since + * the commit path needs to grab the lock while committing the inode. + */ + lockdep_assert_not_held(&ei->i_data_sem); + + while (ext4_test_inode_state(inode, EXT4_STATE_FC_COMMITTING)) { +#if (BITS_PER_LONG < 64) + DEFINE_WAIT_BIT(wait, &ei->i_state_flags, + EXT4_STATE_FC_COMMITTING); + wq = bit_waitqueue(&ei->i_state_flags, + EXT4_STATE_FC_COMMITTING); +#else + DEFINE_WAIT_BIT(wait, &ei->i_flags, + EXT4_STATE_FC_COMMITTING); + wq = bit_waitqueue(&ei->i_flags, + EXT4_STATE_FC_COMMITTING); +#endif + prepare_to_wait(wq, &wait.wq_entry, TASK_UNINTERRUPTIBLE); + if (ext4_test_inode_state(inode, EXT4_STATE_FC_COMMITTING)) + schedule(); + finish_wait(wq, &wait.wq_entry); + } + + /* + * From this point on, this inode will not be committed either + * by fast or full commit as long as the handle is open. + */ + ret = ext4_fc_track_template(handle, inode, __track_inode, NULL, 1); + trace_ext4_fc_track_inode(handle, inode, ret); +} + +struct __track_range_args { + ext4_lblk_t start, end; +}; + +/* __track_fn for tracking data updates */ +static int __track_range(handle_t *handle, struct inode *inode, void *arg, + bool update) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + ext4_lblk_t oldstart; + struct __track_range_args *__arg = + (struct __track_range_args *)arg; + + if (inode->i_ino < EXT4_FIRST_INO(inode->i_sb)) { + ext4_debug("Special inode %ld being modified\n", inode->i_ino); + return -ECANCELED; + } + + oldstart = ei->i_fc_lblk_start; + + if (update && ei->i_fc_lblk_len > 0) { + ei->i_fc_lblk_start = min(ei->i_fc_lblk_start, __arg->start); + ei->i_fc_lblk_len = + max(oldstart + ei->i_fc_lblk_len - 1, __arg->end) - + ei->i_fc_lblk_start + 1; + } else { + ei->i_fc_lblk_start = __arg->start; + ei->i_fc_lblk_len = __arg->end - __arg->start + 1; + } + + return 0; +} + +void ext4_fc_track_range(handle_t *handle, struct inode *inode, ext4_lblk_t start, + ext4_lblk_t end) +{ + struct __track_range_args args; + int ret; + + if (S_ISDIR(inode->i_mode)) + return; + + if (ext4_fc_disabled(inode->i_sb)) + return; + + if (ext4_test_mount_flag(inode->i_sb, EXT4_MF_FC_INELIGIBLE)) + return; + + if (ext4_has_inline_data(inode)) { + ext4_fc_mark_ineligible(inode->i_sb, EXT4_FC_REASON_XATTR, + handle); + return; + } + + args.start = start; + args.end = end; + + ret = ext4_fc_track_template(handle, inode, __track_range, &args, 1); + + trace_ext4_fc_track_range(handle, inode, start, end, ret); +} + +static void ext4_fc_submit_bh(struct super_block *sb, bool is_tail) +{ + blk_opf_t write_flags = JBD2_JOURNAL_REQ_FLAGS; + struct buffer_head *bh = EXT4_SB(sb)->s_fc_bh; + + /* Add REQ_FUA | REQ_PREFLUSH only its tail */ + if (test_opt(sb, BARRIER) && is_tail) + write_flags |= REQ_FUA | REQ_PREFLUSH; + lock_buffer(bh); + set_buffer_dirty(bh); + set_buffer_uptodate(bh); + bh->b_end_io = ext4_end_buffer_io_sync; + submit_bh(REQ_OP_WRITE | write_flags, bh); + EXT4_SB(sb)->s_fc_bh = NULL; +} + +/* Ext4 commit path routines */ + +/* + * Allocate len bytes on a fast commit buffer. + * + * During the commit time this function is used to manage fast commit + * block space. We don't split a fast commit log onto different + * blocks. So this function makes sure that if there's not enough space + * on the current block, the remaining space in the current block is + * marked as unused by adding EXT4_FC_TAG_PAD tag. In that case, + * new block is from jbd2 and CRC is updated to reflect the padding + * we added. + */ +static u8 *ext4_fc_reserve_space(struct super_block *sb, int len, u32 *crc) +{ + struct ext4_fc_tl tl; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct buffer_head *bh; + int bsize = sbi->s_journal->j_blocksize; + int ret, off = sbi->s_fc_bytes % bsize; + int remaining; + u8 *dst; + + /* + * If 'len' is too long to fit in any block alongside a PAD tlv, then we + * cannot fulfill the request. + */ + if (len > bsize - EXT4_FC_TAG_BASE_LEN) + return NULL; + + if (!sbi->s_fc_bh) { + ret = jbd2_fc_get_buf(EXT4_SB(sb)->s_journal, &bh); + if (ret) + return NULL; + sbi->s_fc_bh = bh; + } + dst = sbi->s_fc_bh->b_data + off; + + /* + * Allocate the bytes in the current block if we can do so while still + * leaving enough space for a PAD tlv. + */ + remaining = bsize - EXT4_FC_TAG_BASE_LEN - off; + if (len <= remaining) { + sbi->s_fc_bytes += len; + return dst; + } + + /* + * Else, terminate the current block with a PAD tlv, then allocate a new + * block and allocate the bytes at the start of that new block. + */ + + tl.fc_tag = cpu_to_le16(EXT4_FC_TAG_PAD); + tl.fc_len = cpu_to_le16(remaining); + memcpy(dst, &tl, EXT4_FC_TAG_BASE_LEN); + memset(dst + EXT4_FC_TAG_BASE_LEN, 0, remaining); + *crc = ext4_chksum(*crc, sbi->s_fc_bh->b_data, bsize); + + ext4_fc_submit_bh(sb, false); + + ret = jbd2_fc_get_buf(EXT4_SB(sb)->s_journal, &bh); + if (ret) + return NULL; + sbi->s_fc_bh = bh; + sbi->s_fc_bytes += bsize - off + len; + return sbi->s_fc_bh->b_data; +} + +/* + * Complete a fast commit by writing tail tag. + * + * Writing tail tag marks the end of a fast commit. In order to guarantee + * atomicity, after writing tail tag, even if there's space remaining + * in the block, next commit shouldn't use it. That's why tail tag + * has the length as that of the remaining space on the block. + */ +static int ext4_fc_write_tail(struct super_block *sb, u32 crc) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_fc_tl tl; + struct ext4_fc_tail tail; + int off, bsize = sbi->s_journal->j_blocksize; + u8 *dst; + + /* + * ext4_fc_reserve_space takes care of allocating an extra block if + * there's no enough space on this block for accommodating this tail. + */ + dst = ext4_fc_reserve_space(sb, EXT4_FC_TAG_BASE_LEN + sizeof(tail), &crc); + if (!dst) + return -ENOSPC; + + off = sbi->s_fc_bytes % bsize; + + tl.fc_tag = cpu_to_le16(EXT4_FC_TAG_TAIL); + tl.fc_len = cpu_to_le16(bsize - off + sizeof(struct ext4_fc_tail)); + sbi->s_fc_bytes = round_up(sbi->s_fc_bytes, bsize); + + memcpy(dst, &tl, EXT4_FC_TAG_BASE_LEN); + dst += EXT4_FC_TAG_BASE_LEN; + tail.fc_tid = cpu_to_le32(sbi->s_journal->j_running_transaction->t_tid); + memcpy(dst, &tail.fc_tid, sizeof(tail.fc_tid)); + dst += sizeof(tail.fc_tid); + crc = ext4_chksum(crc, sbi->s_fc_bh->b_data, + dst - (u8 *)sbi->s_fc_bh->b_data); + tail.fc_crc = cpu_to_le32(crc); + memcpy(dst, &tail.fc_crc, sizeof(tail.fc_crc)); + dst += sizeof(tail.fc_crc); + memset(dst, 0, bsize - off); /* Don't leak uninitialized memory. */ + + ext4_fc_submit_bh(sb, true); + + return 0; +} + +/* + * Adds tag, length, value and updates CRC. Returns true if tlv was added. + * Returns false if there's not enough space. + */ +static bool ext4_fc_add_tlv(struct super_block *sb, u16 tag, u16 len, u8 *val, + u32 *crc) +{ + struct ext4_fc_tl tl; + u8 *dst; + + dst = ext4_fc_reserve_space(sb, EXT4_FC_TAG_BASE_LEN + len, crc); + if (!dst) + return false; + + tl.fc_tag = cpu_to_le16(tag); + tl.fc_len = cpu_to_le16(len); + + memcpy(dst, &tl, EXT4_FC_TAG_BASE_LEN); + memcpy(dst + EXT4_FC_TAG_BASE_LEN, val, len); + + return true; +} + +/* Same as above, but adds dentry tlv. */ +static bool ext4_fc_add_dentry_tlv(struct super_block *sb, u32 *crc, + struct ext4_fc_dentry_update *fc_dentry) +{ + struct ext4_fc_dentry_info fcd; + struct ext4_fc_tl tl; + int dlen = fc_dentry->fcd_name.name.len; + u8 *dst = ext4_fc_reserve_space(sb, + EXT4_FC_TAG_BASE_LEN + sizeof(fcd) + dlen, crc); + + if (!dst) + return false; + + fcd.fc_parent_ino = cpu_to_le32(fc_dentry->fcd_parent); + fcd.fc_ino = cpu_to_le32(fc_dentry->fcd_ino); + tl.fc_tag = cpu_to_le16(fc_dentry->fcd_op); + tl.fc_len = cpu_to_le16(sizeof(fcd) + dlen); + memcpy(dst, &tl, EXT4_FC_TAG_BASE_LEN); + dst += EXT4_FC_TAG_BASE_LEN; + memcpy(dst, &fcd, sizeof(fcd)); + dst += sizeof(fcd); + memcpy(dst, fc_dentry->fcd_name.name.name, dlen); + + return true; +} + +/* + * Writes inode in the fast commit space under TLV with tag @tag. + * Returns 0 on success, error on failure. + */ +static int ext4_fc_write_inode(struct inode *inode, u32 *crc) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + int inode_len = EXT4_GOOD_OLD_INODE_SIZE; + int ret; + struct ext4_iloc iloc; + struct ext4_fc_inode fc_inode; + struct ext4_fc_tl tl; + u8 *dst; + + ret = ext4_get_inode_loc(inode, &iloc); + if (ret) + return ret; + + if (ext4_test_inode_flag(inode, EXT4_INODE_INLINE_DATA)) + inode_len = EXT4_INODE_SIZE(inode->i_sb); + else if (EXT4_INODE_SIZE(inode->i_sb) > EXT4_GOOD_OLD_INODE_SIZE) + inode_len += ei->i_extra_isize; + + fc_inode.fc_ino = cpu_to_le32(inode->i_ino); + tl.fc_tag = cpu_to_le16(EXT4_FC_TAG_INODE); + tl.fc_len = cpu_to_le16(inode_len + sizeof(fc_inode.fc_ino)); + + ret = -ECANCELED; + dst = ext4_fc_reserve_space(inode->i_sb, + EXT4_FC_TAG_BASE_LEN + inode_len + sizeof(fc_inode.fc_ino), crc); + if (!dst) + goto err; + + memcpy(dst, &tl, EXT4_FC_TAG_BASE_LEN); + dst += EXT4_FC_TAG_BASE_LEN; + memcpy(dst, &fc_inode, sizeof(fc_inode)); + dst += sizeof(fc_inode); + memcpy(dst, (u8 *)ext4_raw_inode(&iloc), inode_len); + ret = 0; +err: + brelse(iloc.bh); + return ret; +} + +/* + * Writes updated data ranges for the inode in question. Updates CRC. + * Returns 0 on success, error otherwise. + */ +static int ext4_fc_write_inode_data(struct inode *inode, u32 *crc) +{ + ext4_lblk_t old_blk_size, cur_lblk_off, new_blk_size; + struct ext4_inode_info *ei = EXT4_I(inode); + struct ext4_map_blocks map; + struct ext4_fc_add_range fc_ext; + struct ext4_fc_del_range lrange; + struct ext4_extent *ex; + int ret; + + spin_lock(&ei->i_fc_lock); + if (ei->i_fc_lblk_len == 0) { + spin_unlock(&ei->i_fc_lock); + return 0; + } + old_blk_size = ei->i_fc_lblk_start; + new_blk_size = ei->i_fc_lblk_start + ei->i_fc_lblk_len - 1; + ei->i_fc_lblk_len = 0; + spin_unlock(&ei->i_fc_lock); + + cur_lblk_off = old_blk_size; + ext4_debug("will try writing %d to %d for inode %ld\n", + cur_lblk_off, new_blk_size, inode->i_ino); + + while (cur_lblk_off <= new_blk_size) { + map.m_lblk = cur_lblk_off; + map.m_len = new_blk_size - cur_lblk_off + 1; + ret = ext4_map_blocks(NULL, inode, &map, + EXT4_GET_BLOCKS_IO_SUBMIT | + EXT4_EX_NOCACHE); + if (ret < 0) + return -ECANCELED; + + if (map.m_len == 0) { + cur_lblk_off++; + continue; + } + + if (ret == 0) { + lrange.fc_ino = cpu_to_le32(inode->i_ino); + lrange.fc_lblk = cpu_to_le32(map.m_lblk); + lrange.fc_len = cpu_to_le32(map.m_len); + if (!ext4_fc_add_tlv(inode->i_sb, EXT4_FC_TAG_DEL_RANGE, + sizeof(lrange), (u8 *)&lrange, crc)) + return -ENOSPC; + } else { + unsigned int max = (map.m_flags & EXT4_MAP_UNWRITTEN) ? + EXT_UNWRITTEN_MAX_LEN : EXT_INIT_MAX_LEN; + + /* Limit the number of blocks in one extent */ + map.m_len = min(max, map.m_len); + + fc_ext.fc_ino = cpu_to_le32(inode->i_ino); + ex = (struct ext4_extent *)&fc_ext.fc_ex; + ex->ee_block = cpu_to_le32(map.m_lblk); + ex->ee_len = cpu_to_le16(map.m_len); + ext4_ext_store_pblock(ex, map.m_pblk); + if (map.m_flags & EXT4_MAP_UNWRITTEN) + ext4_ext_mark_unwritten(ex); + else + ext4_ext_mark_initialized(ex); + if (!ext4_fc_add_tlv(inode->i_sb, EXT4_FC_TAG_ADD_RANGE, + sizeof(fc_ext), (u8 *)&fc_ext, crc)) + return -ENOSPC; + } + + cur_lblk_off += map.m_len; + } + + return 0; +} + + +/* Flushes data of all the inodes in the commit queue. */ +static int ext4_fc_flush_data(journal_t *journal) +{ + struct super_block *sb = journal->j_private; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_inode_info *ei; + int ret = 0; + + list_for_each_entry(ei, &sbi->s_fc_q[FC_Q_MAIN], i_fc_list) { + ret = jbd2_submit_inode_data(journal, ei->jinode); + if (ret) + return ret; + } + + list_for_each_entry(ei, &sbi->s_fc_q[FC_Q_MAIN], i_fc_list) { + ret = jbd2_wait_inode_data(journal, ei->jinode); + if (ret) + return ret; + } + + return 0; +} + +/* Commit all the directory entry updates */ +static int ext4_fc_commit_dentry_updates(journal_t *journal, u32 *crc) +{ + struct super_block *sb = journal->j_private; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_fc_dentry_update *fc_dentry, *fc_dentry_n; + struct inode *inode; + struct ext4_inode_info *ei; + int ret; + + if (list_empty(&sbi->s_fc_dentry_q[FC_Q_MAIN])) + return 0; + list_for_each_entry_safe(fc_dentry, fc_dentry_n, + &sbi->s_fc_dentry_q[FC_Q_MAIN], fcd_list) { + if (fc_dentry->fcd_op != EXT4_FC_TAG_CREAT) { + if (!ext4_fc_add_dentry_tlv(sb, crc, fc_dentry)) + return -ENOSPC; + continue; + } + /* + * With fcd_dilist we need not loop in sbi->s_fc_q to get the + * corresponding inode. Also, the corresponding inode could have been + * deleted, in which case, we don't need to do anything. + */ + if (list_empty(&fc_dentry->fcd_dilist)) + continue; + ei = list_first_entry(&fc_dentry->fcd_dilist, + struct ext4_inode_info, i_fc_dilist); + inode = &ei->vfs_inode; + WARN_ON(inode->i_ino != fc_dentry->fcd_ino); + + /* + * We first write the inode and then the create dirent. This + * allows the recovery code to create an unnamed inode first + * and then link it to a directory entry. This allows us + * to use namei.c routines almost as is and simplifies + * the recovery code. + */ + ret = ext4_fc_write_inode(inode, crc); + if (ret) + return ret; + ret = ext4_fc_write_inode_data(inode, crc); + if (ret) + return ret; + if (!ext4_fc_add_dentry_tlv(sb, crc, fc_dentry)) + return -ENOSPC; + } + return 0; +} + +static int ext4_fc_perform_commit(journal_t *journal) +{ + struct super_block *sb = journal->j_private; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_inode_info *iter; + struct ext4_fc_head head; + struct inode *inode; + struct blk_plug plug; + int ret = 0; + u32 crc = 0; + + /* + * Step 1: Mark all inodes on s_fc_q[MAIN] with + * EXT4_STATE_FC_FLUSHING_DATA. This prevents these inodes from being + * freed until the data flush is over. + */ + mutex_lock(&sbi->s_fc_lock); + list_for_each_entry(iter, &sbi->s_fc_q[FC_Q_MAIN], i_fc_list) { + ext4_set_inode_state(&iter->vfs_inode, + EXT4_STATE_FC_FLUSHING_DATA); + } + mutex_unlock(&sbi->s_fc_lock); + + /* Step 2: Flush data for all the eligible inodes. */ + ret = ext4_fc_flush_data(journal); + + /* + * Step 3: Clear EXT4_STATE_FC_FLUSHING_DATA flag, before returning + * any error from step 2. This ensures that waiters waiting on + * EXT4_STATE_FC_FLUSHING_DATA can resume. + */ + mutex_lock(&sbi->s_fc_lock); + list_for_each_entry(iter, &sbi->s_fc_q[FC_Q_MAIN], i_fc_list) { + ext4_clear_inode_state(&iter->vfs_inode, + EXT4_STATE_FC_FLUSHING_DATA); +#if (BITS_PER_LONG < 64) + wake_up_bit(&iter->i_state_flags, EXT4_STATE_FC_FLUSHING_DATA); +#else + wake_up_bit(&iter->i_flags, EXT4_STATE_FC_FLUSHING_DATA); +#endif + } + + /* + * Make sure clearing of EXT4_STATE_FC_FLUSHING_DATA is visible before + * the waiter checks the bit. Pairs with implicit barrier in + * prepare_to_wait() in ext4_fc_del(). + */ + smp_mb(); + mutex_unlock(&sbi->s_fc_lock); + + /* + * If we encountered error in Step 2, return it now after clearing + * EXT4_STATE_FC_FLUSHING_DATA bit. + */ + if (ret) + return ret; + + + /* Step 4: Mark all inodes as being committed. */ + jbd2_journal_lock_updates(journal); + /* + * The journal is now locked. No more handles can start and all the + * previous handles are now drained. We now mark the inodes on the + * commit queue as being committed. + */ + mutex_lock(&sbi->s_fc_lock); + list_for_each_entry(iter, &sbi->s_fc_q[FC_Q_MAIN], i_fc_list) { + ext4_set_inode_state(&iter->vfs_inode, + EXT4_STATE_FC_COMMITTING); + } + mutex_unlock(&sbi->s_fc_lock); + jbd2_journal_unlock_updates(journal); + + /* + * Step 5: If file system device is different from journal device, + * issue a cache flush before we start writing fast commit blocks. + */ + if (journal->j_fs_dev != journal->j_dev) + blkdev_issue_flush(journal->j_fs_dev); + + blk_start_plug(&plug); + /* Step 6: Write fast commit blocks to disk. */ + if (sbi->s_fc_bytes == 0) { + /* + * Step 6.1: Add a head tag only if this is the first fast + * commit in this TID. + */ + head.fc_features = cpu_to_le32(EXT4_FC_SUPPORTED_FEATURES); + head.fc_tid = cpu_to_le32( + sbi->s_journal->j_running_transaction->t_tid); + if (!ext4_fc_add_tlv(sb, EXT4_FC_TAG_HEAD, sizeof(head), + (u8 *)&head, &crc)) { + ret = -ENOSPC; + goto out; + } + } + + /* Step 6.2: Now write all the dentry updates. */ + mutex_lock(&sbi->s_fc_lock); + ret = ext4_fc_commit_dentry_updates(journal, &crc); + if (ret) + goto out; + + /* Step 6.3: Now write all the changed inodes to disk. */ + list_for_each_entry(iter, &sbi->s_fc_q[FC_Q_MAIN], i_fc_list) { + inode = &iter->vfs_inode; + if (!ext4_test_inode_state(inode, EXT4_STATE_FC_COMMITTING)) + continue; + + ret = ext4_fc_write_inode_data(inode, &crc); + if (ret) + goto out; + ret = ext4_fc_write_inode(inode, &crc); + if (ret) + goto out; + } + /* Step 6.4: Finally write tail tag to conclude this fast commit. */ + ret = ext4_fc_write_tail(sb, crc); + +out: + mutex_unlock(&sbi->s_fc_lock); + blk_finish_plug(&plug); + return ret; +} + +static void ext4_fc_update_stats(struct super_block *sb, int status, + u64 commit_time, int nblks, tid_t commit_tid) +{ + struct ext4_fc_stats *stats = &EXT4_SB(sb)->s_fc_stats; + + ext4_debug("Fast commit ended with status = %d for tid %u", + status, commit_tid); + if (status == EXT4_FC_STATUS_OK) { + stats->fc_num_commits++; + stats->fc_numblks += nblks; + if (likely(stats->s_fc_avg_commit_time)) + stats->s_fc_avg_commit_time = + (commit_time + + stats->s_fc_avg_commit_time * 3) / 4; + else + stats->s_fc_avg_commit_time = commit_time; + } else if (status == EXT4_FC_STATUS_FAILED || + status == EXT4_FC_STATUS_INELIGIBLE) { + if (status == EXT4_FC_STATUS_FAILED) + stats->fc_failed_commits++; + stats->fc_ineligible_commits++; + } else { + stats->fc_skipped_commits++; + } + trace_ext4_fc_commit_stop(sb, nblks, status, commit_tid); +} + +/* + * The main commit entry point. Performs a fast commit for transaction + * commit_tid if needed. If it's not possible to perform a fast commit + * due to various reasons, we fall back to full commit. Returns 0 + * on success, error otherwise. + */ +int ext4_fc_commit(journal_t *journal, tid_t commit_tid) +{ + struct super_block *sb = journal->j_private; + struct ext4_sb_info *sbi = EXT4_SB(sb); + int nblks = 0, ret, bsize = journal->j_blocksize; + int subtid = atomic_read(&sbi->s_fc_subtid); + int status = EXT4_FC_STATUS_OK, fc_bufs_before = 0; + ktime_t start_time, commit_time; + int old_ioprio, journal_ioprio; + + if (!test_opt2(sb, JOURNAL_FAST_COMMIT)) + return jbd2_complete_transaction(journal, commit_tid); + + trace_ext4_fc_commit_start(sb, commit_tid); + + start_time = ktime_get(); + old_ioprio = get_current_ioprio(); + +restart_fc: + ret = jbd2_fc_begin_commit(journal, commit_tid); + if (ret == -EALREADY) { + /* There was an ongoing commit, check if we need to restart */ + if (atomic_read(&sbi->s_fc_subtid) <= subtid && + tid_gt(commit_tid, journal->j_commit_sequence)) + goto restart_fc; + ext4_fc_update_stats(sb, EXT4_FC_STATUS_SKIPPED, 0, 0, + commit_tid); + return 0; + } else if (ret) { + /* + * Commit couldn't start. Just update stats and perform a + * full commit. + */ + ext4_fc_update_stats(sb, EXT4_FC_STATUS_FAILED, 0, 0, + commit_tid); + return jbd2_complete_transaction(journal, commit_tid); + } + + /* + * After establishing journal barrier via jbd2_fc_begin_commit(), check + * if we are fast commit ineligible. + */ + if (ext4_test_mount_flag(sb, EXT4_MF_FC_INELIGIBLE)) { + status = EXT4_FC_STATUS_INELIGIBLE; + goto fallback; + } + + /* + * Now that we know that this thread is going to do a fast commit, + * elevate the priority to match that of the journal thread. + */ + if (journal->j_task->io_context) + journal_ioprio = sbi->s_journal->j_task->io_context->ioprio; + else + journal_ioprio = EXT4_DEF_JOURNAL_IOPRIO; + set_task_ioprio(current, journal_ioprio); + fc_bufs_before = (sbi->s_fc_bytes + bsize - 1) / bsize; + ret = ext4_fc_perform_commit(journal); + if (ret < 0) { + status = EXT4_FC_STATUS_FAILED; + goto fallback; + } + nblks = (sbi->s_fc_bytes + bsize - 1) / bsize - fc_bufs_before; + ret = jbd2_fc_wait_bufs(journal, nblks); + if (ret < 0) { + status = EXT4_FC_STATUS_FAILED; + goto fallback; + } + atomic_inc(&sbi->s_fc_subtid); + ret = jbd2_fc_end_commit(journal); + set_task_ioprio(current, old_ioprio); + /* + * weight the commit time higher than the average time so we + * don't react too strongly to vast changes in the commit time + */ + commit_time = ktime_to_ns(ktime_sub(ktime_get(), start_time)); + ext4_fc_update_stats(sb, status, commit_time, nblks, commit_tid); + return ret; + +fallback: + set_task_ioprio(current, old_ioprio); + ret = jbd2_fc_end_commit_fallback(journal); + ext4_fc_update_stats(sb, status, 0, 0, commit_tid); + return ret; +} + +/* + * Fast commit cleanup routine. This is called after every fast commit and + * full commit. full is true if we are called after a full commit. + */ +static void ext4_fc_cleanup(journal_t *journal, int full, tid_t tid) +{ + struct super_block *sb = journal->j_private; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_inode_info *ei; + struct ext4_fc_dentry_update *fc_dentry; + + if (full && sbi->s_fc_bh) + sbi->s_fc_bh = NULL; + + trace_ext4_fc_cleanup(journal, full, tid); + jbd2_fc_release_bufs(journal); + + mutex_lock(&sbi->s_fc_lock); + while (!list_empty(&sbi->s_fc_q[FC_Q_MAIN])) { + ei = list_first_entry(&sbi->s_fc_q[FC_Q_MAIN], + struct ext4_inode_info, + i_fc_list); + list_del_init(&ei->i_fc_list); + ext4_clear_inode_state(&ei->vfs_inode, + EXT4_STATE_FC_COMMITTING); + if (tid_geq(tid, ei->i_sync_tid)) { + ext4_fc_reset_inode(&ei->vfs_inode); + } else if (full) { + /* + * We are called after a full commit, inode has been + * modified while the commit was running. Re-enqueue + * the inode into STAGING, which will then be splice + * back into MAIN. This cannot happen during + * fastcommit because the journal is locked all the + * time in that case (and tid doesn't increase so + * tid check above isn't reliable). + */ + list_add_tail(&ei->i_fc_list, + &sbi->s_fc_q[FC_Q_STAGING]); + } + /* + * Make sure clearing of EXT4_STATE_FC_COMMITTING is + * visible before we send the wakeup. Pairs with implicit + * barrier in prepare_to_wait() in ext4_fc_track_inode(). + */ + smp_mb(); +#if (BITS_PER_LONG < 64) + wake_up_bit(&ei->i_state_flags, EXT4_STATE_FC_COMMITTING); +#else + wake_up_bit(&ei->i_flags, EXT4_STATE_FC_COMMITTING); +#endif + } + + while (!list_empty(&sbi->s_fc_dentry_q[FC_Q_MAIN])) { + fc_dentry = list_first_entry(&sbi->s_fc_dentry_q[FC_Q_MAIN], + struct ext4_fc_dentry_update, + fcd_list); + list_del_init(&fc_dentry->fcd_list); + list_del_init(&fc_dentry->fcd_dilist); + + release_dentry_name_snapshot(&fc_dentry->fcd_name); + kmem_cache_free(ext4_fc_dentry_cachep, fc_dentry); + } + + list_splice_init(&sbi->s_fc_dentry_q[FC_Q_STAGING], + &sbi->s_fc_dentry_q[FC_Q_MAIN]); + list_splice_init(&sbi->s_fc_q[FC_Q_STAGING], + &sbi->s_fc_q[FC_Q_MAIN]); + + if (tid_geq(tid, sbi->s_fc_ineligible_tid)) { + sbi->s_fc_ineligible_tid = 0; + ext4_clear_mount_flag(sb, EXT4_MF_FC_INELIGIBLE); + } + + if (full) + sbi->s_fc_bytes = 0; + mutex_unlock(&sbi->s_fc_lock); + trace_ext4_fc_stats(sb); +} + +/* Ext4 Replay Path Routines */ + +/* Helper struct for dentry replay routines */ +struct dentry_info_args { + int parent_ino, dname_len, ino, inode_len; + char *dname; +}; + +/* Same as struct ext4_fc_tl, but uses native endianness fields */ +struct ext4_fc_tl_mem { + u16 fc_tag; + u16 fc_len; +}; + +static inline void tl_to_darg(struct dentry_info_args *darg, + struct ext4_fc_tl_mem *tl, u8 *val) +{ + struct ext4_fc_dentry_info fcd; + + memcpy(&fcd, val, sizeof(fcd)); + + darg->parent_ino = le32_to_cpu(fcd.fc_parent_ino); + darg->ino = le32_to_cpu(fcd.fc_ino); + darg->dname = val + offsetof(struct ext4_fc_dentry_info, fc_dname); + darg->dname_len = tl->fc_len - sizeof(struct ext4_fc_dentry_info); +} + +static inline void ext4_fc_get_tl(struct ext4_fc_tl_mem *tl, u8 *val) +{ + struct ext4_fc_tl tl_disk; + + memcpy(&tl_disk, val, EXT4_FC_TAG_BASE_LEN); + tl->fc_len = le16_to_cpu(tl_disk.fc_len); + tl->fc_tag = le16_to_cpu(tl_disk.fc_tag); +} + +/* Unlink replay function */ +static int ext4_fc_replay_unlink(struct super_block *sb, + struct ext4_fc_tl_mem *tl, u8 *val) +{ + struct inode *inode, *old_parent; + struct qstr entry; + struct dentry_info_args darg; + int ret = 0; + + tl_to_darg(&darg, tl, val); + + trace_ext4_fc_replay(sb, EXT4_FC_TAG_UNLINK, darg.ino, + darg.parent_ino, darg.dname_len); + + entry.name = darg.dname; + entry.len = darg.dname_len; + inode = ext4_iget(sb, darg.ino, EXT4_IGET_NORMAL); + + if (IS_ERR(inode)) { + ext4_debug("Inode %d not found", darg.ino); + return 0; + } + + old_parent = ext4_iget(sb, darg.parent_ino, + EXT4_IGET_NORMAL); + if (IS_ERR(old_parent)) { + ext4_debug("Dir with inode %d not found", darg.parent_ino); + iput(inode); + return 0; + } + + ret = __ext4_unlink(old_parent, &entry, inode, NULL); + /* -ENOENT ok coz it might not exist anymore. */ + if (ret == -ENOENT) + ret = 0; + iput(old_parent); + iput(inode); + return ret; +} + +static int ext4_fc_replay_link_internal(struct super_block *sb, + struct dentry_info_args *darg, + struct inode *inode) +{ + struct inode *dir = NULL; + struct dentry *dentry_dir = NULL, *dentry_inode = NULL; + struct qstr qstr_dname = QSTR_INIT(darg->dname, darg->dname_len); + int ret = 0; + + dir = ext4_iget(sb, darg->parent_ino, EXT4_IGET_NORMAL); + if (IS_ERR(dir)) { + ext4_debug("Dir with inode %d not found.", darg->parent_ino); + dir = NULL; + goto out; + } + + dentry_dir = d_obtain_alias(dir); + if (IS_ERR(dentry_dir)) { + ext4_debug("Failed to obtain dentry"); + dentry_dir = NULL; + goto out; + } + + dentry_inode = d_alloc(dentry_dir, &qstr_dname); + if (!dentry_inode) { + ext4_debug("Inode dentry not created."); + ret = -ENOMEM; + goto out; + } + + ret = __ext4_link(dir, inode, dentry_inode); + /* + * It's possible that link already existed since data blocks + * for the dir in question got persisted before we crashed OR + * we replayed this tag and crashed before the entire replay + * could complete. + */ + if (ret && ret != -EEXIST) { + ext4_debug("Failed to link\n"); + goto out; + } + + ret = 0; +out: + if (dentry_dir) { + d_drop(dentry_dir); + dput(dentry_dir); + } else if (dir) { + iput(dir); + } + if (dentry_inode) { + d_drop(dentry_inode); + dput(dentry_inode); + } + + return ret; +} + +/* Link replay function */ +static int ext4_fc_replay_link(struct super_block *sb, + struct ext4_fc_tl_mem *tl, u8 *val) +{ + struct inode *inode; + struct dentry_info_args darg; + int ret = 0; + + tl_to_darg(&darg, tl, val); + trace_ext4_fc_replay(sb, EXT4_FC_TAG_LINK, darg.ino, + darg.parent_ino, darg.dname_len); + + inode = ext4_iget(sb, darg.ino, EXT4_IGET_NORMAL); + if (IS_ERR(inode)) { + ext4_debug("Inode not found."); + return 0; + } + + ret = ext4_fc_replay_link_internal(sb, &darg, inode); + iput(inode); + return ret; +} + +/* + * Record all the modified inodes during replay. We use this later to setup + * block bitmaps correctly. + */ +static int ext4_fc_record_modified_inode(struct super_block *sb, int ino) +{ + struct ext4_fc_replay_state *state; + int i; + + state = &EXT4_SB(sb)->s_fc_replay_state; + for (i = 0; i < state->fc_modified_inodes_used; i++) + if (state->fc_modified_inodes[i] == ino) + return 0; + if (state->fc_modified_inodes_used == state->fc_modified_inodes_size) { + int *fc_modified_inodes; + + fc_modified_inodes = krealloc(state->fc_modified_inodes, + sizeof(int) * (state->fc_modified_inodes_size + + EXT4_FC_REPLAY_REALLOC_INCREMENT), + GFP_KERNEL); + if (!fc_modified_inodes) + return -ENOMEM; + state->fc_modified_inodes = fc_modified_inodes; + state->fc_modified_inodes_size += + EXT4_FC_REPLAY_REALLOC_INCREMENT; + } + state->fc_modified_inodes[state->fc_modified_inodes_used++] = ino; + return 0; +} + +/* + * Inode replay function + */ +static int ext4_fc_replay_inode(struct super_block *sb, + struct ext4_fc_tl_mem *tl, u8 *val) +{ + struct ext4_fc_inode fc_inode; + struct ext4_inode *raw_inode; + struct ext4_inode *raw_fc_inode; + struct inode *inode = NULL; + struct ext4_iloc iloc; + int inode_len, ino, ret, tag = tl->fc_tag; + struct ext4_extent_header *eh; + size_t off_gen = offsetof(struct ext4_inode, i_generation); + + memcpy(&fc_inode, val, sizeof(fc_inode)); + + ino = le32_to_cpu(fc_inode.fc_ino); + trace_ext4_fc_replay(sb, tag, ino, 0, 0); + + inode = ext4_iget(sb, ino, EXT4_IGET_NORMAL); + if (!IS_ERR(inode)) { + ext4_ext_clear_bb(inode); + iput(inode); + } + inode = NULL; + + ret = ext4_fc_record_modified_inode(sb, ino); + if (ret) + goto out; + + raw_fc_inode = (struct ext4_inode *) + (val + offsetof(struct ext4_fc_inode, fc_raw_inode)); + ret = ext4_get_fc_inode_loc(sb, ino, &iloc); + if (ret) + goto out; + + inode_len = tl->fc_len - sizeof(struct ext4_fc_inode); + raw_inode = ext4_raw_inode(&iloc); + + memcpy(raw_inode, raw_fc_inode, offsetof(struct ext4_inode, i_block)); + memcpy((u8 *)raw_inode + off_gen, (u8 *)raw_fc_inode + off_gen, + inode_len - off_gen); + if (le32_to_cpu(raw_inode->i_flags) & EXT4_EXTENTS_FL) { + eh = (struct ext4_extent_header *)(&raw_inode->i_block[0]); + if (eh->eh_magic != EXT4_EXT_MAGIC) { + memset(eh, 0, sizeof(*eh)); + eh->eh_magic = EXT4_EXT_MAGIC; + eh->eh_max = cpu_to_le16( + (sizeof(raw_inode->i_block) - + sizeof(struct ext4_extent_header)) + / sizeof(struct ext4_extent)); + } + } else if (le32_to_cpu(raw_inode->i_flags) & EXT4_INLINE_DATA_FL) { + memcpy(raw_inode->i_block, raw_fc_inode->i_block, + sizeof(raw_inode->i_block)); + } + + /* Immediately update the inode on disk. */ + ret = ext4_handle_dirty_metadata(NULL, NULL, iloc.bh); + if (ret) + goto out; + ret = sync_dirty_buffer(iloc.bh); + if (ret) + goto out; + ret = ext4_mark_inode_used(sb, ino); + if (ret) + goto out; + + /* Given that we just wrote the inode on disk, this SHOULD succeed. */ + inode = ext4_iget(sb, ino, EXT4_IGET_NORMAL); + if (IS_ERR(inode)) { + ext4_debug("Inode not found."); + return -EFSCORRUPTED; + } + + /* + * Our allocator could have made different decisions than before + * crashing. This should be fixed but until then, we calculate + * the number of blocks the inode. + */ + if (!ext4_test_inode_flag(inode, EXT4_INODE_INLINE_DATA)) + ext4_ext_replay_set_iblocks(inode); + + inode->i_generation = le32_to_cpu(ext4_raw_inode(&iloc)->i_generation); + ext4_reset_inode_seed(inode); + + ext4_inode_csum_set(inode, ext4_raw_inode(&iloc), EXT4_I(inode)); + ret = ext4_handle_dirty_metadata(NULL, NULL, iloc.bh); + sync_dirty_buffer(iloc.bh); + brelse(iloc.bh); +out: + iput(inode); + if (!ret) + blkdev_issue_flush(sb->s_bdev); + + return 0; +} + +/* + * Dentry create replay function. + * + * EXT4_FC_TAG_CREAT is preceded by EXT4_FC_TAG_INODE_FULL. Which means, the + * inode for which we are trying to create a dentry here, should already have + * been replayed before we start here. + */ +static int ext4_fc_replay_create(struct super_block *sb, + struct ext4_fc_tl_mem *tl, u8 *val) +{ + int ret = 0; + struct inode *inode = NULL; + struct inode *dir = NULL; + struct dentry_info_args darg; + + tl_to_darg(&darg, tl, val); + + trace_ext4_fc_replay(sb, EXT4_FC_TAG_CREAT, darg.ino, + darg.parent_ino, darg.dname_len); + + /* This takes care of update group descriptor and other metadata */ + ret = ext4_mark_inode_used(sb, darg.ino); + if (ret) + goto out; + + inode = ext4_iget(sb, darg.ino, EXT4_IGET_NORMAL); + if (IS_ERR(inode)) { + ext4_debug("inode %d not found.", darg.ino); + inode = NULL; + ret = -EINVAL; + goto out; + } + + if (S_ISDIR(inode->i_mode)) { + /* + * If we are creating a directory, we need to make sure that the + * dot and dot dot dirents are setup properly. + */ + dir = ext4_iget(sb, darg.parent_ino, EXT4_IGET_NORMAL); + if (IS_ERR(dir)) { + ext4_debug("Dir %d not found.", darg.ino); + goto out; + } + ret = ext4_init_new_dir(NULL, dir, inode); + iput(dir); + if (ret) { + ret = 0; + goto out; + } + } + ret = ext4_fc_replay_link_internal(sb, &darg, inode); + if (ret) + goto out; + set_nlink(inode, 1); + ext4_mark_inode_dirty(NULL, inode); +out: + iput(inode); + return ret; +} + +/* + * Record physical disk regions which are in use as per fast commit area, + * and used by inodes during replay phase. Our simple replay phase + * allocator excludes these regions from allocation. + */ +int ext4_fc_record_regions(struct super_block *sb, int ino, + ext4_lblk_t lblk, ext4_fsblk_t pblk, int len, int replay) +{ + struct ext4_fc_replay_state *state; + struct ext4_fc_alloc_region *region; + + state = &EXT4_SB(sb)->s_fc_replay_state; + /* + * during replay phase, the fc_regions_valid may not same as + * fc_regions_used, update it when do new additions. + */ + if (replay && state->fc_regions_used != state->fc_regions_valid) + state->fc_regions_used = state->fc_regions_valid; + if (state->fc_regions_used == state->fc_regions_size) { + struct ext4_fc_alloc_region *fc_regions; + + fc_regions = krealloc(state->fc_regions, + sizeof(struct ext4_fc_alloc_region) * + (state->fc_regions_size + + EXT4_FC_REPLAY_REALLOC_INCREMENT), + GFP_KERNEL); + if (!fc_regions) + return -ENOMEM; + state->fc_regions_size += + EXT4_FC_REPLAY_REALLOC_INCREMENT; + state->fc_regions = fc_regions; + } + region = &state->fc_regions[state->fc_regions_used++]; + region->ino = ino; + region->lblk = lblk; + region->pblk = pblk; + region->len = len; + + if (replay) + state->fc_regions_valid++; + + return 0; +} + +/* Replay add range tag */ +static int ext4_fc_replay_add_range(struct super_block *sb, + struct ext4_fc_tl_mem *tl, u8 *val) +{ + struct ext4_fc_add_range fc_add_ex; + struct ext4_extent newex, *ex; + struct inode *inode; + ext4_lblk_t start, cur; + int remaining, len; + ext4_fsblk_t start_pblk; + struct ext4_map_blocks map; + struct ext4_ext_path *path = NULL; + int ret; + + memcpy(&fc_add_ex, val, sizeof(fc_add_ex)); + ex = (struct ext4_extent *)&fc_add_ex.fc_ex; + + trace_ext4_fc_replay(sb, EXT4_FC_TAG_ADD_RANGE, + le32_to_cpu(fc_add_ex.fc_ino), le32_to_cpu(ex->ee_block), + ext4_ext_get_actual_len(ex)); + + inode = ext4_iget(sb, le32_to_cpu(fc_add_ex.fc_ino), EXT4_IGET_NORMAL); + if (IS_ERR(inode)) { + ext4_debug("Inode not found."); + return 0; + } + + ret = ext4_fc_record_modified_inode(sb, inode->i_ino); + if (ret) + goto out; + + start = le32_to_cpu(ex->ee_block); + start_pblk = ext4_ext_pblock(ex); + len = ext4_ext_get_actual_len(ex); + + cur = start; + remaining = len; + ext4_debug("ADD_RANGE, lblk %d, pblk %lld, len %d, unwritten %d, inode %ld\n", + start, start_pblk, len, ext4_ext_is_unwritten(ex), + inode->i_ino); + + while (remaining > 0) { + map.m_lblk = cur; + map.m_len = remaining; + map.m_pblk = 0; + ret = ext4_map_blocks(NULL, inode, &map, 0); + + if (ret < 0) + goto out; + + if (ret == 0) { + /* Range is not mapped */ + path = ext4_find_extent(inode, cur, path, 0); + if (IS_ERR(path)) + goto out; + memset(&newex, 0, sizeof(newex)); + newex.ee_block = cpu_to_le32(cur); + ext4_ext_store_pblock( + &newex, start_pblk + cur - start); + newex.ee_len = cpu_to_le16(map.m_len); + if (ext4_ext_is_unwritten(ex)) + ext4_ext_mark_unwritten(&newex); + down_write(&EXT4_I(inode)->i_data_sem); + path = ext4_ext_insert_extent(NULL, inode, + path, &newex, 0); + up_write((&EXT4_I(inode)->i_data_sem)); + if (IS_ERR(path)) + goto out; + goto next; + } + + if (start_pblk + cur - start != map.m_pblk) { + /* + * Logical to physical mapping changed. This can happen + * if this range was removed and then reallocated to + * map to new physical blocks during a fast commit. + */ + ret = ext4_ext_replay_update_ex(inode, cur, map.m_len, + ext4_ext_is_unwritten(ex), + start_pblk + cur - start); + if (ret) + goto out; + /* + * Mark the old blocks as free since they aren't used + * anymore. We maintain an array of all the modified + * inodes. In case these blocks are still used at either + * a different logical range in the same inode or in + * some different inode, we will mark them as allocated + * at the end of the FC replay using our array of + * modified inodes. + */ + ext4_mb_mark_bb(inode->i_sb, map.m_pblk, map.m_len, false); + goto next; + } + + /* Range is mapped and needs a state change */ + ext4_debug("Converting from %ld to %d %lld", + map.m_flags & EXT4_MAP_UNWRITTEN, + ext4_ext_is_unwritten(ex), map.m_pblk); + ret = ext4_ext_replay_update_ex(inode, cur, map.m_len, + ext4_ext_is_unwritten(ex), map.m_pblk); + if (ret) + goto out; + /* + * We may have split the extent tree while toggling the state. + * Try to shrink the extent tree now. + */ + ext4_ext_replay_shrink_inode(inode, start + len); +next: + cur += map.m_len; + remaining -= map.m_len; + } + ext4_ext_replay_shrink_inode(inode, i_size_read(inode) >> + sb->s_blocksize_bits); +out: + ext4_free_ext_path(path); + iput(inode); + return 0; +} + +/* Replay DEL_RANGE tag */ +static int +ext4_fc_replay_del_range(struct super_block *sb, + struct ext4_fc_tl_mem *tl, u8 *val) +{ + struct inode *inode; + struct ext4_fc_del_range lrange; + struct ext4_map_blocks map; + ext4_lblk_t cur, remaining; + int ret; + + memcpy(&lrange, val, sizeof(lrange)); + cur = le32_to_cpu(lrange.fc_lblk); + remaining = le32_to_cpu(lrange.fc_len); + + trace_ext4_fc_replay(sb, EXT4_FC_TAG_DEL_RANGE, + le32_to_cpu(lrange.fc_ino), cur, remaining); + + inode = ext4_iget(sb, le32_to_cpu(lrange.fc_ino), EXT4_IGET_NORMAL); + if (IS_ERR(inode)) { + ext4_debug("Inode %d not found", le32_to_cpu(lrange.fc_ino)); + return 0; + } + + ret = ext4_fc_record_modified_inode(sb, inode->i_ino); + if (ret) + goto out; + + ext4_debug("DEL_RANGE, inode %ld, lblk %d, len %d\n", + inode->i_ino, le32_to_cpu(lrange.fc_lblk), + le32_to_cpu(lrange.fc_len)); + while (remaining > 0) { + map.m_lblk = cur; + map.m_len = remaining; + + ret = ext4_map_blocks(NULL, inode, &map, 0); + if (ret < 0) + goto out; + if (ret > 0) { + remaining -= ret; + cur += ret; + ext4_mb_mark_bb(inode->i_sb, map.m_pblk, map.m_len, false); + } else { + remaining -= map.m_len; + cur += map.m_len; + } + } + + down_write(&EXT4_I(inode)->i_data_sem); + ret = ext4_ext_remove_space(inode, le32_to_cpu(lrange.fc_lblk), + le32_to_cpu(lrange.fc_lblk) + + le32_to_cpu(lrange.fc_len) - 1); + up_write(&EXT4_I(inode)->i_data_sem); + if (ret) + goto out; + ext4_ext_replay_shrink_inode(inode, + i_size_read(inode) >> sb->s_blocksize_bits); + ext4_mark_inode_dirty(NULL, inode); +out: + iput(inode); + return 0; +} + +static void ext4_fc_set_bitmaps_and_counters(struct super_block *sb) +{ + struct ext4_fc_replay_state *state; + struct inode *inode; + struct ext4_ext_path *path = NULL; + struct ext4_map_blocks map; + int i, ret, j; + ext4_lblk_t cur, end; + + state = &EXT4_SB(sb)->s_fc_replay_state; + for (i = 0; i < state->fc_modified_inodes_used; i++) { + inode = ext4_iget(sb, state->fc_modified_inodes[i], + EXT4_IGET_NORMAL); + if (IS_ERR(inode)) { + ext4_debug("Inode %d not found.", + state->fc_modified_inodes[i]); + continue; + } + cur = 0; + end = EXT_MAX_BLOCKS; + if (ext4_test_inode_flag(inode, EXT4_INODE_INLINE_DATA)) { + iput(inode); + continue; + } + while (cur < end) { + map.m_lblk = cur; + map.m_len = end - cur; + + ret = ext4_map_blocks(NULL, inode, &map, 0); + if (ret < 0) + break; + + if (ret > 0) { + path = ext4_find_extent(inode, map.m_lblk, path, 0); + if (!IS_ERR(path)) { + for (j = 0; j < path->p_depth; j++) + ext4_mb_mark_bb(inode->i_sb, + path[j].p_block, 1, true); + } else { + path = NULL; + } + cur += ret; + ext4_mb_mark_bb(inode->i_sb, map.m_pblk, + map.m_len, true); + } else { + cur = cur + (map.m_len ? map.m_len : 1); + } + } + iput(inode); + } + + ext4_free_ext_path(path); +} + +/* + * Check if block is in excluded regions for block allocation. The simple + * allocator that runs during replay phase is calls this function to see + * if it is okay to use a block. + */ +bool ext4_fc_replay_check_excluded(struct super_block *sb, ext4_fsblk_t blk) +{ + int i; + struct ext4_fc_replay_state *state; + + state = &EXT4_SB(sb)->s_fc_replay_state; + for (i = 0; i < state->fc_regions_valid; i++) { + if (state->fc_regions[i].ino == 0 || + state->fc_regions[i].len == 0) + continue; + if (in_range(blk, state->fc_regions[i].pblk, + state->fc_regions[i].len)) + return true; + } + return false; +} + +/* Cleanup function called after replay */ +void ext4_fc_replay_cleanup(struct super_block *sb) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + + sbi->s_mount_state &= ~EXT4_FC_REPLAY; + kfree(sbi->s_fc_replay_state.fc_regions); + kfree(sbi->s_fc_replay_state.fc_modified_inodes); +} + +static bool ext4_fc_value_len_isvalid(struct ext4_sb_info *sbi, + int tag, int len) +{ + switch (tag) { + case EXT4_FC_TAG_ADD_RANGE: + return len == sizeof(struct ext4_fc_add_range); + case EXT4_FC_TAG_DEL_RANGE: + return len == sizeof(struct ext4_fc_del_range); + case EXT4_FC_TAG_CREAT: + case EXT4_FC_TAG_LINK: + case EXT4_FC_TAG_UNLINK: + len -= sizeof(struct ext4_fc_dentry_info); + return len >= 1 && len <= EXT4_NAME_LEN; + case EXT4_FC_TAG_INODE: + len -= sizeof(struct ext4_fc_inode); + return len >= EXT4_GOOD_OLD_INODE_SIZE && + len <= sbi->s_inode_size; + case EXT4_FC_TAG_PAD: + return true; /* padding can have any length */ + case EXT4_FC_TAG_TAIL: + return len >= sizeof(struct ext4_fc_tail); + case EXT4_FC_TAG_HEAD: + return len == sizeof(struct ext4_fc_head); + } + return false; +} + +/* + * Recovery Scan phase handler + * + * This function is called during the scan phase and is responsible + * for doing following things: + * - Make sure the fast commit area has valid tags for replay + * - Count number of tags that need to be replayed by the replay handler + * - Verify CRC + * - Create a list of excluded blocks for allocation during replay phase + * + * This function returns JBD2_FC_REPLAY_CONTINUE to indicate that SCAN is + * incomplete and JBD2 should send more blocks. It returns JBD2_FC_REPLAY_STOP + * to indicate that scan has finished and JBD2 can now start replay phase. + * It returns a negative error to indicate that there was an error. At the end + * of a successful scan phase, sbi->s_fc_replay_state.fc_replay_num_tags is set + * to indicate the number of tags that need to replayed during the replay phase. + */ +static int ext4_fc_replay_scan(journal_t *journal, + struct buffer_head *bh, int off, + tid_t expected_tid) +{ + struct super_block *sb = journal->j_private; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_fc_replay_state *state; + int ret = JBD2_FC_REPLAY_CONTINUE; + struct ext4_fc_add_range ext; + struct ext4_fc_tl_mem tl; + struct ext4_fc_tail tail; + __u8 *start, *end, *cur, *val; + struct ext4_fc_head head; + struct ext4_extent *ex; + + state = &sbi->s_fc_replay_state; + + start = (u8 *)bh->b_data; + end = start + journal->j_blocksize; + + if (state->fc_replay_expected_off == 0) { + state->fc_cur_tag = 0; + state->fc_replay_num_tags = 0; + state->fc_crc = 0; + state->fc_regions = NULL; + state->fc_regions_valid = state->fc_regions_used = + state->fc_regions_size = 0; + /* Check if we can stop early */ + if (le16_to_cpu(((struct ext4_fc_tl *)start)->fc_tag) + != EXT4_FC_TAG_HEAD) + return 0; + } + + if (off != state->fc_replay_expected_off) { + ret = -EFSCORRUPTED; + goto out_err; + } + + state->fc_replay_expected_off++; + for (cur = start; cur <= end - EXT4_FC_TAG_BASE_LEN; + cur = cur + EXT4_FC_TAG_BASE_LEN + tl.fc_len) { + ext4_fc_get_tl(&tl, cur); + val = cur + EXT4_FC_TAG_BASE_LEN; + if (tl.fc_len > end - val || + !ext4_fc_value_len_isvalid(sbi, tl.fc_tag, tl.fc_len)) { + ret = state->fc_replay_num_tags ? + JBD2_FC_REPLAY_STOP : -ECANCELED; + goto out_err; + } + ext4_debug("Scan phase, tag:%s, blk %lld\n", + tag2str(tl.fc_tag), bh->b_blocknr); + switch (tl.fc_tag) { + case EXT4_FC_TAG_ADD_RANGE: + memcpy(&ext, val, sizeof(ext)); + ex = (struct ext4_extent *)&ext.fc_ex; + ret = ext4_fc_record_regions(sb, + le32_to_cpu(ext.fc_ino), + le32_to_cpu(ex->ee_block), ext4_ext_pblock(ex), + ext4_ext_get_actual_len(ex), 0); + if (ret < 0) + break; + ret = JBD2_FC_REPLAY_CONTINUE; + fallthrough; + case EXT4_FC_TAG_DEL_RANGE: + case EXT4_FC_TAG_LINK: + case EXT4_FC_TAG_UNLINK: + case EXT4_FC_TAG_CREAT: + case EXT4_FC_TAG_INODE: + case EXT4_FC_TAG_PAD: + state->fc_cur_tag++; + state->fc_crc = ext4_chksum(state->fc_crc, cur, + EXT4_FC_TAG_BASE_LEN + tl.fc_len); + break; + case EXT4_FC_TAG_TAIL: + state->fc_cur_tag++; + memcpy(&tail, val, sizeof(tail)); + state->fc_crc = ext4_chksum(state->fc_crc, cur, + EXT4_FC_TAG_BASE_LEN + + offsetof(struct ext4_fc_tail, + fc_crc)); + if (le32_to_cpu(tail.fc_tid) == expected_tid && + le32_to_cpu(tail.fc_crc) == state->fc_crc) { + state->fc_replay_num_tags = state->fc_cur_tag; + state->fc_regions_valid = + state->fc_regions_used; + } else { + ret = state->fc_replay_num_tags ? + JBD2_FC_REPLAY_STOP : -EFSBADCRC; + } + state->fc_crc = 0; + break; + case EXT4_FC_TAG_HEAD: + memcpy(&head, val, sizeof(head)); + if (le32_to_cpu(head.fc_features) & + ~EXT4_FC_SUPPORTED_FEATURES) { + ret = -EOPNOTSUPP; + break; + } + if (le32_to_cpu(head.fc_tid) != expected_tid) { + ret = JBD2_FC_REPLAY_STOP; + break; + } + state->fc_cur_tag++; + state->fc_crc = ext4_chksum(state->fc_crc, cur, + EXT4_FC_TAG_BASE_LEN + tl.fc_len); + break; + default: + ret = state->fc_replay_num_tags ? + JBD2_FC_REPLAY_STOP : -ECANCELED; + } + if (ret < 0 || ret == JBD2_FC_REPLAY_STOP) + break; + } + +out_err: + trace_ext4_fc_replay_scan(sb, ret, off); + return ret; +} + +/* + * Main recovery path entry point. + * The meaning of return codes is similar as above. + */ +static int ext4_fc_replay(journal_t *journal, struct buffer_head *bh, + enum passtype pass, int off, tid_t expected_tid) +{ + struct super_block *sb = journal->j_private; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_fc_tl_mem tl; + __u8 *start, *end, *cur, *val; + int ret = JBD2_FC_REPLAY_CONTINUE; + struct ext4_fc_replay_state *state = &sbi->s_fc_replay_state; + struct ext4_fc_tail tail; + + if (pass == PASS_SCAN) { + state->fc_current_pass = PASS_SCAN; + return ext4_fc_replay_scan(journal, bh, off, expected_tid); + } + + if (state->fc_current_pass != pass) { + state->fc_current_pass = pass; + sbi->s_mount_state |= EXT4_FC_REPLAY; + } + if (!sbi->s_fc_replay_state.fc_replay_num_tags) { + ext4_debug("Replay stops\n"); + ext4_fc_set_bitmaps_and_counters(sb); + return 0; + } + +#ifdef CONFIG_EXT4_DEBUG + if (sbi->s_fc_debug_max_replay && off >= sbi->s_fc_debug_max_replay) { + pr_warn("Dropping fc block %d because max_replay set\n", off); + return JBD2_FC_REPLAY_STOP; + } +#endif + + start = (u8 *)bh->b_data; + end = start + journal->j_blocksize; + + for (cur = start; cur <= end - EXT4_FC_TAG_BASE_LEN; + cur = cur + EXT4_FC_TAG_BASE_LEN + tl.fc_len) { + ext4_fc_get_tl(&tl, cur); + val = cur + EXT4_FC_TAG_BASE_LEN; + + if (state->fc_replay_num_tags == 0) { + ret = JBD2_FC_REPLAY_STOP; + ext4_fc_set_bitmaps_and_counters(sb); + break; + } + + ext4_debug("Replay phase, tag:%s\n", tag2str(tl.fc_tag)); + state->fc_replay_num_tags--; + switch (tl.fc_tag) { + case EXT4_FC_TAG_LINK: + ret = ext4_fc_replay_link(sb, &tl, val); + break; + case EXT4_FC_TAG_UNLINK: + ret = ext4_fc_replay_unlink(sb, &tl, val); + break; + case EXT4_FC_TAG_ADD_RANGE: + ret = ext4_fc_replay_add_range(sb, &tl, val); + break; + case EXT4_FC_TAG_CREAT: + ret = ext4_fc_replay_create(sb, &tl, val); + break; + case EXT4_FC_TAG_DEL_RANGE: + ret = ext4_fc_replay_del_range(sb, &tl, val); + break; + case EXT4_FC_TAG_INODE: + ret = ext4_fc_replay_inode(sb, &tl, val); + break; + case EXT4_FC_TAG_PAD: + trace_ext4_fc_replay(sb, EXT4_FC_TAG_PAD, 0, + tl.fc_len, 0); + break; + case EXT4_FC_TAG_TAIL: + trace_ext4_fc_replay(sb, EXT4_FC_TAG_TAIL, + 0, tl.fc_len, 0); + memcpy(&tail, val, sizeof(tail)); + WARN_ON(le32_to_cpu(tail.fc_tid) != expected_tid); + break; + case EXT4_FC_TAG_HEAD: + break; + default: + trace_ext4_fc_replay(sb, tl.fc_tag, 0, tl.fc_len, 0); + ret = -ECANCELED; + break; + } + if (ret < 0) + break; + ret = JBD2_FC_REPLAY_CONTINUE; + } + return ret; +} + +void ext4_fc_init(struct super_block *sb, journal_t *journal) +{ + /* + * We set replay callback even if fast commit disabled because we may + * could still have fast commit blocks that need to be replayed even if + * fast commit has now been turned off. + */ + journal->j_fc_replay_callback = ext4_fc_replay; + if (!test_opt2(sb, JOURNAL_FAST_COMMIT)) + return; + journal->j_fc_cleanup_callback = ext4_fc_cleanup; +} + +static const char * const fc_ineligible_reasons[] = { + [EXT4_FC_REASON_XATTR] = "Extended attributes changed", + [EXT4_FC_REASON_CROSS_RENAME] = "Cross rename", + [EXT4_FC_REASON_JOURNAL_FLAG_CHANGE] = "Journal flag changed", + [EXT4_FC_REASON_NOMEM] = "Insufficient memory", + [EXT4_FC_REASON_SWAP_BOOT] = "Swap boot", + [EXT4_FC_REASON_RESIZE] = "Resize", + [EXT4_FC_REASON_RENAME_DIR] = "Dir renamed", + [EXT4_FC_REASON_FALLOC_RANGE] = "Falloc range op", + [EXT4_FC_REASON_INODE_JOURNAL_DATA] = "Data journalling", + [EXT4_FC_REASON_ENCRYPTED_FILENAME] = "Encrypted filename", +}; + +int ext4_fc_info_show(struct seq_file *seq, void *v) +{ + struct ext4_sb_info *sbi = EXT4_SB((struct super_block *)seq->private); + struct ext4_fc_stats *stats = &sbi->s_fc_stats; + int i; + + if (v != SEQ_START_TOKEN) + return 0; + + seq_printf(seq, + "fc stats:\n%ld commits\n%ld ineligible\n%ld numblks\n%lluus avg_commit_time\n", + stats->fc_num_commits, stats->fc_ineligible_commits, + stats->fc_numblks, + div_u64(stats->s_fc_avg_commit_time, 1000)); + seq_puts(seq, "Ineligible reasons:\n"); + for (i = 0; i < EXT4_FC_REASON_MAX; i++) + seq_printf(seq, "\"%s\":\t%d\n", fc_ineligible_reasons[i], + stats->fc_ineligible_reason_count[i]); + + return 0; +} + +int __init ext4_fc_init_dentry_cache(void) +{ + ext4_fc_dentry_cachep = KMEM_CACHE(ext4_fc_dentry_update, + SLAB_RECLAIM_ACCOUNT); + + if (ext4_fc_dentry_cachep == NULL) + return -ENOMEM; + + return 0; +} + +void ext4_fc_destroy_dentry_cache(void) +{ + kmem_cache_destroy(ext4_fc_dentry_cachep); +} diff --git a/fs/ext4l/fast_commit.h b/fs/ext4l/fast_commit.h new file mode 100644 index 00000000000..3bd534e4dbb --- /dev/null +++ b/fs/ext4l/fast_commit.h @@ -0,0 +1,186 @@ +/* SPDX-License-Identifier: GPL-2.0 */ + +#ifndef __FAST_COMMIT_H__ +#define __FAST_COMMIT_H__ + +/* + * Note this file is present in e2fsprogs/lib/ext2fs/fast_commit.h and + * linux/fs/ext4/fast_commit.h. These file should always be byte identical. + */ + +/* Fast commit tags */ +#define EXT4_FC_TAG_ADD_RANGE 0x0001 +#define EXT4_FC_TAG_DEL_RANGE 0x0002 +#define EXT4_FC_TAG_CREAT 0x0003 +#define EXT4_FC_TAG_LINK 0x0004 +#define EXT4_FC_TAG_UNLINK 0x0005 +#define EXT4_FC_TAG_INODE 0x0006 +#define EXT4_FC_TAG_PAD 0x0007 +#define EXT4_FC_TAG_TAIL 0x0008 +#define EXT4_FC_TAG_HEAD 0x0009 + +#define EXT4_FC_SUPPORTED_FEATURES 0x0 + +/* On disk fast commit tlv value structures */ + +/* Fast commit on disk tag length structure */ +struct ext4_fc_tl { + __le16 fc_tag; + __le16 fc_len; +}; + +/* Value structure for tag EXT4_FC_TAG_HEAD. */ +struct ext4_fc_head { + __le32 fc_features; + __le32 fc_tid; +}; + +/* Value structure for EXT4_FC_TAG_ADD_RANGE. */ +struct ext4_fc_add_range { + __le32 fc_ino; + __u8 fc_ex[12]; +}; + +/* Value structure for tag EXT4_FC_TAG_DEL_RANGE. */ +struct ext4_fc_del_range { + __le32 fc_ino; + __le32 fc_lblk; + __le32 fc_len; +}; + +/* + * This is the value structure for tags EXT4_FC_TAG_CREAT, EXT4_FC_TAG_LINK + * and EXT4_FC_TAG_UNLINK. + */ +struct ext4_fc_dentry_info { + __le32 fc_parent_ino; + __le32 fc_ino; + __u8 fc_dname[]; +}; + +/* Value structure for EXT4_FC_TAG_INODE. */ +struct ext4_fc_inode { + __le32 fc_ino; + __u8 fc_raw_inode[]; +}; + +/* Value structure for tag EXT4_FC_TAG_TAIL. */ +struct ext4_fc_tail { + __le32 fc_tid; + __le32 fc_crc; +}; + +/* Tag base length */ +#define EXT4_FC_TAG_BASE_LEN (sizeof(struct ext4_fc_tl)) + +/* + * Fast commit status codes + */ +enum { + EXT4_FC_STATUS_OK = 0, + EXT4_FC_STATUS_INELIGIBLE, + EXT4_FC_STATUS_SKIPPED, + EXT4_FC_STATUS_FAILED, +}; + +/* + * Fast commit ineligiblity reasons: + */ +enum { + EXT4_FC_REASON_XATTR = 0, + EXT4_FC_REASON_CROSS_RENAME, + EXT4_FC_REASON_JOURNAL_FLAG_CHANGE, + EXT4_FC_REASON_NOMEM, + EXT4_FC_REASON_SWAP_BOOT, + EXT4_FC_REASON_RESIZE, + EXT4_FC_REASON_RENAME_DIR, + EXT4_FC_REASON_FALLOC_RANGE, + EXT4_FC_REASON_INODE_JOURNAL_DATA, + EXT4_FC_REASON_ENCRYPTED_FILENAME, + EXT4_FC_REASON_MAX +}; + +#ifdef __KERNEL__ +/* + * In memory list of dentry updates that are performed on the file + * system used by fast commit code. + */ +struct ext4_fc_dentry_update { + int fcd_op; /* Type of update create / unlink / link */ + int fcd_parent; /* Parent inode number */ + int fcd_ino; /* Inode number */ + struct name_snapshot fcd_name; /* Dirent name */ + struct list_head fcd_list; + struct list_head fcd_dilist; +}; + +struct ext4_fc_stats { + unsigned int fc_ineligible_reason_count[EXT4_FC_REASON_MAX]; + unsigned long fc_num_commits; + unsigned long fc_ineligible_commits; + unsigned long fc_failed_commits; + unsigned long fc_skipped_commits; + unsigned long fc_numblks; + u64 s_fc_avg_commit_time; +}; + +#define EXT4_FC_REPLAY_REALLOC_INCREMENT 4 + +/* + * Physical block regions added to different inodes due to fast commit + * recovery. These are set during the SCAN phase. During the replay phase, + * our allocator excludes these from its allocation. This ensures that + * we don't accidentally allocating a block that is going to be used by + * another inode. + */ +struct ext4_fc_alloc_region { + ext4_lblk_t lblk; + ext4_fsblk_t pblk; + int ino, len; +}; + +/* + * Fast commit replay state. + */ +struct ext4_fc_replay_state { + int fc_replay_num_tags; + int fc_replay_expected_off; + int fc_current_pass; + int fc_cur_tag; + int fc_crc; + struct ext4_fc_alloc_region *fc_regions; + int fc_regions_size, fc_regions_used, fc_regions_valid; + int *fc_modified_inodes; + int fc_modified_inodes_used, fc_modified_inodes_size; +}; + +#define region_last(__region) (((__region)->lblk) + ((__region)->len) - 1) +#endif + +static inline const char *tag2str(__u16 tag) +{ + switch (tag) { + case EXT4_FC_TAG_LINK: + return "ADD_ENTRY"; + case EXT4_FC_TAG_UNLINK: + return "DEL_ENTRY"; + case EXT4_FC_TAG_ADD_RANGE: + return "ADD_RANGE"; + case EXT4_FC_TAG_CREAT: + return "CREAT_DENTRY"; + case EXT4_FC_TAG_DEL_RANGE: + return "DEL_RANGE"; + case EXT4_FC_TAG_INODE: + return "INODE"; + case EXT4_FC_TAG_PAD: + return "PAD"; + case EXT4_FC_TAG_TAIL: + return "TAIL"; + case EXT4_FC_TAG_HEAD: + return "HEAD"; + default: + return "ERROR"; + } +} + +#endif /* __FAST_COMMIT_H__ */ -- 2.43.0

From: Simon Glass <simon.glass@canonical.com> Copy indirect.c and inline.c from Linux v6.18 fs/ext4 directory. - indirect: indirect block mapping for ext2/ext3 compatibility - inline: inline data support for small files stored in inode Co-developed-by: Claude Opus 4.5 <noreply@anthropic.com> --- fs/ext4l/indirect.c | 1474 ++++++++++++++++++++++++++++++++ fs/ext4l/inline.c | 1982 +++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 3456 insertions(+) create mode 100644 fs/ext4l/indirect.c create mode 100644 fs/ext4l/inline.c diff --git a/fs/ext4l/indirect.c b/fs/ext4l/indirect.c new file mode 100644 index 00000000000..da76353b3a5 --- /dev/null +++ b/fs/ext4l/indirect.c @@ -0,0 +1,1474 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * linux/fs/ext4/indirect.c + * + * from + * + * linux/fs/ext4/inode.c + * + * Copyright (C) 1992, 1993, 1994, 1995 + * Remy Card (card@masi.ibp.fr) + * Laboratoire MASI - Institut Blaise Pascal + * Universite Pierre et Marie Curie (Paris VI) + * + * from + * + * linux/fs/minix/inode.c + * + * Copyright (C) 1991, 1992 Linus Torvalds + * + * Goal-directed block allocation by Stephen Tweedie + * (sct@redhat.com), 1993, 1998 + */ + +#include "ext4_jbd2.h" +#include "truncate.h" +#include <linux/dax.h> +#include <linux/uio.h> + +#include <trace/events/ext4.h> + +typedef struct { + __le32 *p; + __le32 key; + struct buffer_head *bh; +} Indirect; + +static inline void add_chain(Indirect *p, struct buffer_head *bh, __le32 *v) +{ + p->key = *(p->p = v); + p->bh = bh; +} + +/** + * ext4_block_to_path - parse the block number into array of offsets + * @inode: inode in question (we are only interested in its superblock) + * @i_block: block number to be parsed + * @offsets: array to store the offsets in + * @boundary: set this non-zero if the referred-to block is likely to be + * followed (on disk) by an indirect block. + * + * To store the locations of file's data ext4 uses a data structure common + * for UNIX filesystems - tree of pointers anchored in the inode, with + * data blocks at leaves and indirect blocks in intermediate nodes. + * This function translates the block number into path in that tree - + * return value is the path length and @offsets[n] is the offset of + * pointer to (n+1)th node in the nth one. If @block is out of range + * (negative or too large) warning is printed and zero returned. + * + * Note: function doesn't find node addresses, so no IO is needed. All + * we need to know is the capacity of indirect blocks (taken from the + * inode->i_sb). + */ + +/* + * Portability note: the last comparison (check that we fit into triple + * indirect block) is spelled differently, because otherwise on an + * architecture with 32-bit longs and 8Kb pages we might get into trouble + * if our filesystem had 8Kb blocks. We might use long long, but that would + * kill us on x86. Oh, well, at least the sign propagation does not matter - + * i_block would have to be negative in the very beginning, so we would not + * get there at all. + */ + +static int ext4_block_to_path(struct inode *inode, + ext4_lblk_t i_block, + ext4_lblk_t offsets[4], int *boundary) +{ + int ptrs = EXT4_ADDR_PER_BLOCK(inode->i_sb); + int ptrs_bits = EXT4_ADDR_PER_BLOCK_BITS(inode->i_sb); + const long direct_blocks = EXT4_NDIR_BLOCKS, + indirect_blocks = ptrs, + double_blocks = (1 << (ptrs_bits * 2)); + int n = 0; + int final = 0; + + if (i_block < direct_blocks) { + offsets[n++] = i_block; + final = direct_blocks; + } else if ((i_block -= direct_blocks) < indirect_blocks) { + offsets[n++] = EXT4_IND_BLOCK; + offsets[n++] = i_block; + final = ptrs; + } else if ((i_block -= indirect_blocks) < double_blocks) { + offsets[n++] = EXT4_DIND_BLOCK; + offsets[n++] = i_block >> ptrs_bits; + offsets[n++] = i_block & (ptrs - 1); + final = ptrs; + } else if (((i_block -= double_blocks) >> (ptrs_bits * 2)) < ptrs) { + offsets[n++] = EXT4_TIND_BLOCK; + offsets[n++] = i_block >> (ptrs_bits * 2); + offsets[n++] = (i_block >> ptrs_bits) & (ptrs - 1); + offsets[n++] = i_block & (ptrs - 1); + final = ptrs; + } else { + ext4_warning(inode->i_sb, "block %lu > max in inode %lu", + i_block + direct_blocks + + indirect_blocks + double_blocks, inode->i_ino); + } + if (boundary) + *boundary = final - 1 - (i_block & (ptrs - 1)); + return n; +} + +/** + * ext4_get_branch - read the chain of indirect blocks leading to data + * @inode: inode in question + * @depth: depth of the chain (1 - direct pointer, etc.) + * @offsets: offsets of pointers in inode/indirect blocks + * @chain: place to store the result + * @err: here we store the error value + * + * Function fills the array of triples <key, p, bh> and returns %NULL + * if everything went OK or the pointer to the last filled triple + * (incomplete one) otherwise. Upon the return chain[i].key contains + * the number of (i+1)-th block in the chain (as it is stored in memory, + * i.e. little-endian 32-bit), chain[i].p contains the address of that + * number (it points into struct inode for i==0 and into the bh->b_data + * for i>0) and chain[i].bh points to the buffer_head of i-th indirect + * block for i>0 and NULL for i==0. In other words, it holds the block + * numbers of the chain, addresses they were taken from (and where we can + * verify that chain did not change) and buffer_heads hosting these + * numbers. + * + * Function stops when it stumbles upon zero pointer (absent block) + * (pointer to last triple returned, *@err == 0) + * or when it gets an IO error reading an indirect block + * (ditto, *@err == -EIO) + * or when it reads all @depth-1 indirect blocks successfully and finds + * the whole chain, all way to the data (returns %NULL, *err == 0). + * + * Need to be called with + * down_read(&EXT4_I(inode)->i_data_sem) + */ +static Indirect *ext4_get_branch(struct inode *inode, int depth, + ext4_lblk_t *offsets, + Indirect chain[4], int *err) +{ + struct super_block *sb = inode->i_sb; + Indirect *p = chain; + struct buffer_head *bh; + unsigned int key; + int ret = -EIO; + + *err = 0; + /* i_data is not going away, no lock needed */ + add_chain(chain, NULL, EXT4_I(inode)->i_data + *offsets); + if (!p->key) + goto no_block; + while (--depth) { + key = le32_to_cpu(p->key); + if (key > ext4_blocks_count(EXT4_SB(sb)->s_es)) { + /* the block was out of range */ + ret = -EFSCORRUPTED; + goto failure; + } + bh = sb_getblk(sb, key); + if (unlikely(!bh)) { + ret = -ENOMEM; + goto failure; + } + + if (!bh_uptodate_or_lock(bh)) { + if (ext4_read_bh(bh, 0, NULL, false) < 0) { + put_bh(bh); + goto failure; + } + /* validate block references */ + if (ext4_check_indirect_blockref(inode, bh)) { + put_bh(bh); + goto failure; + } + } + + add_chain(++p, bh, (__le32 *)bh->b_data + *++offsets); + /* Reader: end */ + if (!p->key) + goto no_block; + } + return NULL; + +failure: + *err = ret; +no_block: + return p; +} + +/** + * ext4_find_near - find a place for allocation with sufficient locality + * @inode: owner + * @ind: descriptor of indirect block. + * + * This function returns the preferred place for block allocation. + * It is used when heuristic for sequential allocation fails. + * Rules are: + * + if there is a block to the left of our position - allocate near it. + * + if pointer will live in indirect block - allocate near that block. + * + if pointer will live in inode - allocate in the same + * cylinder group. + * + * In the latter case we colour the starting block by the callers PID to + * prevent it from clashing with concurrent allocations for a different inode + * in the same block group. The PID is used here so that functionally related + * files will be close-by on-disk. + * + * Caller must make sure that @ind is valid and will stay that way. + */ +static ext4_fsblk_t ext4_find_near(struct inode *inode, Indirect *ind) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + __le32 *start = ind->bh ? (__le32 *) ind->bh->b_data : ei->i_data; + __le32 *p; + + /* Try to find previous block */ + for (p = ind->p - 1; p >= start; p--) { + if (*p) + return le32_to_cpu(*p); + } + + /* No such thing, so let's try location of indirect block */ + if (ind->bh) + return ind->bh->b_blocknr; + + /* + * It is going to be referred to from the inode itself? OK, just put it + * into the same cylinder group then. + */ + return ext4_inode_to_goal_block(inode); +} + +/** + * ext4_find_goal - find a preferred place for allocation. + * @inode: owner + * @block: block we want + * @partial: pointer to the last triple within a chain + * + * Normally this function find the preferred place for block allocation, + * returns it. + * Because this is only used for non-extent files, we limit the block nr + * to 32 bits. + */ +static ext4_fsblk_t ext4_find_goal(struct inode *inode, ext4_lblk_t block, + Indirect *partial) +{ + ext4_fsblk_t goal; + + /* + * XXX need to get goal block from mballoc's data structures + */ + + goal = ext4_find_near(inode, partial); + goal = goal & EXT4_MAX_BLOCK_FILE_PHYS; + return goal; +} + +/** + * ext4_blks_to_allocate - Look up the block map and count the number + * of direct blocks need to be allocated for the given branch. + * + * @branch: chain of indirect blocks + * @k: number of blocks need for indirect blocks + * @blks: number of data blocks to be mapped. + * @blocks_to_boundary: the offset in the indirect block + * + * return the total number of blocks to be allocate, including the + * direct and indirect blocks. + */ +static int ext4_blks_to_allocate(Indirect *branch, int k, unsigned int blks, + int blocks_to_boundary) +{ + unsigned int count = 0; + + /* + * Simple case, [t,d]Indirect block(s) has not allocated yet + * then it's clear blocks on that path have not allocated + */ + if (k > 0) { + /* right now we don't handle cross boundary allocation */ + if (blks < blocks_to_boundary + 1) + count += blks; + else + count += blocks_to_boundary + 1; + return count; + } + + count++; + while (count < blks && count <= blocks_to_boundary && + le32_to_cpu(*(branch[0].p + count)) == 0) { + count++; + } + return count; +} + +/** + * ext4_alloc_branch() - allocate and set up a chain of blocks + * @handle: handle for this transaction + * @ar: structure describing the allocation request + * @indirect_blks: number of allocated indirect blocks + * @offsets: offsets (in the blocks) to store the pointers to next. + * @branch: place to store the chain in. + * + * This function allocates blocks, zeroes out all but the last one, + * links them into chain and (if we are synchronous) writes them to disk. + * In other words, it prepares a branch that can be spliced onto the + * inode. It stores the information about that chain in the branch[], in + * the same format as ext4_get_branch() would do. We are calling it after + * we had read the existing part of chain and partial points to the last + * triple of that (one with zero ->key). Upon the exit we have the same + * picture as after the successful ext4_get_block(), except that in one + * place chain is disconnected - *branch->p is still zero (we did not + * set the last link), but branch->key contains the number that should + * be placed into *branch->p to fill that gap. + * + * If allocation fails we free all blocks we've allocated (and forget + * their buffer_heads) and return the error value the from failed + * ext4_alloc_block() (normally -ENOSPC). Otherwise we set the chain + * as described above and return 0. + */ +static int ext4_alloc_branch(handle_t *handle, + struct ext4_allocation_request *ar, + int indirect_blks, ext4_lblk_t *offsets, + Indirect *branch) +{ + struct buffer_head * bh; + ext4_fsblk_t b, new_blocks[4]; + __le32 *p; + int i, j, err, len = 1; + + for (i = 0; i <= indirect_blks; i++) { + if (i == indirect_blks) { + new_blocks[i] = ext4_mb_new_blocks(handle, ar, &err); + } else { + ar->goal = new_blocks[i] = ext4_new_meta_blocks(handle, + ar->inode, ar->goal, + ar->flags & EXT4_MB_DELALLOC_RESERVED, + NULL, &err); + /* Simplify error cleanup... */ + branch[i+1].bh = NULL; + } + if (err) { + i--; + goto failed; + } + branch[i].key = cpu_to_le32(new_blocks[i]); + if (i == 0) + continue; + + bh = branch[i].bh = sb_getblk(ar->inode->i_sb, new_blocks[i-1]); + if (unlikely(!bh)) { + err = -ENOMEM; + goto failed; + } + lock_buffer(bh); + BUFFER_TRACE(bh, "call get_create_access"); + err = ext4_journal_get_create_access(handle, ar->inode->i_sb, + bh, EXT4_JTR_NONE); + if (err) { + unlock_buffer(bh); + goto failed; + } + + memset(bh->b_data, 0, bh->b_size); + p = branch[i].p = (__le32 *) bh->b_data + offsets[i]; + b = new_blocks[i]; + + if (i == indirect_blks) + len = ar->len; + for (j = 0; j < len; j++) + *p++ = cpu_to_le32(b++); + + BUFFER_TRACE(bh, "marking uptodate"); + set_buffer_uptodate(bh); + unlock_buffer(bh); + + BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata"); + err = ext4_handle_dirty_metadata(handle, ar->inode, bh); + if (err) + goto failed; + } + return 0; +failed: + if (i == indirect_blks) { + /* Free data blocks */ + ext4_free_blocks(handle, ar->inode, NULL, new_blocks[i], + ar->len, 0); + i--; + } + for (; i >= 0; i--) { + /* + * We want to ext4_forget() only freshly allocated indirect + * blocks. Buffer for new_blocks[i] is at branch[i+1].bh + * (buffer at branch[0].bh is indirect block / inode already + * existing before ext4_alloc_branch() was called). Also + * because blocks are freshly allocated, we don't need to + * revoke them which is why we don't set + * EXT4_FREE_BLOCKS_METADATA. + */ + ext4_free_blocks(handle, ar->inode, branch[i+1].bh, + new_blocks[i], 1, + branch[i+1].bh ? EXT4_FREE_BLOCKS_FORGET : 0); + } + return err; +} + +/** + * ext4_splice_branch() - splice the allocated branch onto inode. + * @handle: handle for this transaction + * @ar: structure describing the allocation request + * @where: location of missing link + * @num: number of indirect blocks we are adding + * + * This function fills the missing link and does all housekeeping needed in + * inode (->i_blocks, etc.). In case of success we end up with the full + * chain to new block and return 0. + */ +static int ext4_splice_branch(handle_t *handle, + struct ext4_allocation_request *ar, + Indirect *where, int num) +{ + int i; + int err = 0; + ext4_fsblk_t current_block; + + /* + * If we're splicing into a [td]indirect block (as opposed to the + * inode) then we need to get write access to the [td]indirect block + * before the splice. + */ + if (where->bh) { + BUFFER_TRACE(where->bh, "get_write_access"); + err = ext4_journal_get_write_access(handle, ar->inode->i_sb, + where->bh, EXT4_JTR_NONE); + if (err) + goto err_out; + } + /* That's it */ + + *where->p = where->key; + + /* + * Update the host buffer_head or inode to point to more just allocated + * direct blocks blocks + */ + if (num == 0 && ar->len > 1) { + current_block = le32_to_cpu(where->key) + 1; + for (i = 1; i < ar->len; i++) + *(where->p + i) = cpu_to_le32(current_block++); + } + + /* We are done with atomic stuff, now do the rest of housekeeping */ + /* had we spliced it onto indirect block? */ + if (where->bh) { + /* + * If we spliced it onto an indirect block, we haven't + * altered the inode. Note however that if it is being spliced + * onto an indirect block at the very end of the file (the + * file is growing) then we *will* alter the inode to reflect + * the new i_size. But that is not done here - it is done in + * generic_commit_write->__mark_inode_dirty->ext4_dirty_inode. + */ + ext4_debug("splicing indirect only\n"); + BUFFER_TRACE(where->bh, "call ext4_handle_dirty_metadata"); + err = ext4_handle_dirty_metadata(handle, ar->inode, where->bh); + if (err) + goto err_out; + } else { + /* + * OK, we spliced it into the inode itself on a direct block. + */ + err = ext4_mark_inode_dirty(handle, ar->inode); + if (unlikely(err)) + goto err_out; + ext4_debug("splicing direct\n"); + } + return err; + +err_out: + for (i = 1; i <= num; i++) { + /* + * branch[i].bh is newly allocated, so there is no + * need to revoke the block, which is why we don't + * need to set EXT4_FREE_BLOCKS_METADATA. + */ + ext4_free_blocks(handle, ar->inode, where[i].bh, 0, 1, + EXT4_FREE_BLOCKS_FORGET); + } + ext4_free_blocks(handle, ar->inode, NULL, le32_to_cpu(where[num].key), + ar->len, 0); + + return err; +} + +/* + * The ext4_ind_map_blocks() function handles non-extents inodes + * (i.e., using the traditional indirect/double-indirect i_blocks + * scheme) for ext4_map_blocks(). + * + * Allocation strategy is simple: if we have to allocate something, we will + * have to go the whole way to leaf. So let's do it before attaching anything + * to tree, set linkage between the newborn blocks, write them if sync is + * required, recheck the path, free and repeat if check fails, otherwise + * set the last missing link (that will protect us from any truncate-generated + * removals - all blocks on the path are immune now) and possibly force the + * write on the parent block. + * That has a nice additional property: no special recovery from the failed + * allocations is needed - we simply release blocks and do not touch anything + * reachable from inode. + * + * `handle' can be NULL if create == 0. + * + * return > 0, # of blocks mapped or allocated. + * return = 0, if plain lookup failed. + * return < 0, error case. + * + * The ext4_ind_get_blocks() function should be called with + * down_write(&EXT4_I(inode)->i_data_sem) if allocating filesystem + * blocks (i.e., flags has EXT4_GET_BLOCKS_CREATE set) or + * down_read(&EXT4_I(inode)->i_data_sem) if not allocating file system + * blocks. + */ +int ext4_ind_map_blocks(handle_t *handle, struct inode *inode, + struct ext4_map_blocks *map, + int flags) +{ + struct ext4_allocation_request ar; + int err = -EIO; + ext4_lblk_t offsets[4]; + Indirect chain[4]; + Indirect *partial; + int indirect_blks; + int blocks_to_boundary = 0; + int depth; + u64 count = 0; + ext4_fsblk_t first_block = 0; + + trace_ext4_ind_map_blocks_enter(inode, map->m_lblk, map->m_len, flags); + ASSERT(!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))); + ASSERT(handle != NULL || (flags & EXT4_GET_BLOCKS_CREATE) == 0); + depth = ext4_block_to_path(inode, map->m_lblk, offsets, + &blocks_to_boundary); + + if (depth == 0) + goto out; + + partial = ext4_get_branch(inode, depth, offsets, chain, &err); + + /* Simplest case - block found, no allocation needed */ + if (!partial) { + first_block = le32_to_cpu(chain[depth - 1].key); + count++; + /*map more blocks*/ + while (count < map->m_len && count <= blocks_to_boundary) { + ext4_fsblk_t blk; + + blk = le32_to_cpu(*(chain[depth-1].p + count)); + + if (blk == first_block + count) + count++; + else + break; + } + goto got_it; + } + + /* Next simple case - plain lookup failed */ + if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) { + unsigned epb = inode->i_sb->s_blocksize / sizeof(u32); + int i; + + /* + * Count number blocks in a subtree under 'partial'. At each + * level we count number of complete empty subtrees beyond + * current offset and then descend into the subtree only + * partially beyond current offset. + */ + count = 0; + for (i = partial - chain + 1; i < depth; i++) + count = count * epb + (epb - offsets[i] - 1); + count++; + /* Fill in size of a hole we found */ + map->m_pblk = 0; + map->m_len = umin(map->m_len, count); + goto cleanup; + } + + /* Failed read of indirect block */ + if (err == -EIO) + goto cleanup; + + /* + * Okay, we need to do block allocation. + */ + if (ext4_has_feature_bigalloc(inode->i_sb)) { + EXT4_ERROR_INODE(inode, "Can't allocate blocks for " + "non-extent mapped inodes with bigalloc"); + err = -EFSCORRUPTED; + goto out; + } + + /* Set up for the direct block allocation */ + memset(&ar, 0, sizeof(ar)); + ar.inode = inode; + ar.logical = map->m_lblk; + if (S_ISREG(inode->i_mode)) + ar.flags = EXT4_MB_HINT_DATA; + if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) + ar.flags |= EXT4_MB_DELALLOC_RESERVED; + if (flags & EXT4_GET_BLOCKS_METADATA_NOFAIL) + ar.flags |= EXT4_MB_USE_RESERVED; + + ar.goal = ext4_find_goal(inode, map->m_lblk, partial); + + /* the number of blocks need to allocate for [d,t]indirect blocks */ + indirect_blks = (chain + depth) - partial - 1; + + /* + * Next look up the indirect map to count the totoal number of + * direct blocks to allocate for this branch. + */ + ar.len = ext4_blks_to_allocate(partial, indirect_blks, + map->m_len, blocks_to_boundary); + + /* + * Block out ext4_truncate while we alter the tree + */ + err = ext4_alloc_branch(handle, &ar, indirect_blks, + offsets + (partial - chain), partial); + + /* + * The ext4_splice_branch call will free and forget any buffers + * on the new chain if there is a failure, but that risks using + * up transaction credits, especially for bitmaps where the + * credits cannot be returned. Can we handle this somehow? We + * may need to return -EAGAIN upwards in the worst case. --sct + */ + if (!err) + err = ext4_splice_branch(handle, &ar, partial, indirect_blks); + if (err) + goto cleanup; + + map->m_flags |= EXT4_MAP_NEW; + + ext4_update_inode_fsync_trans(handle, inode, 1); + count = ar.len; + +got_it: + map->m_flags |= EXT4_MAP_MAPPED; + map->m_pblk = le32_to_cpu(chain[depth-1].key); + map->m_len = count; + if (count > blocks_to_boundary) + map->m_flags |= EXT4_MAP_BOUNDARY; + err = count; + /* Clean up and exit */ + partial = chain + depth - 1; /* the whole chain */ +cleanup: + while (partial > chain) { + BUFFER_TRACE(partial->bh, "call brelse"); + brelse(partial->bh); + partial--; + } +out: + trace_ext4_ind_map_blocks_exit(inode, flags, map, err); + return err; +} + +/* + * Calculate number of indirect blocks touched by mapping @nrblocks logically + * contiguous blocks + */ +int ext4_ind_trans_blocks(struct inode *inode, int nrblocks) +{ + /* + * With N contiguous data blocks, we need at most + * N/EXT4_ADDR_PER_BLOCK(inode->i_sb) + 1 indirect blocks, + * 2 dindirect blocks, and 1 tindirect block + */ + return DIV_ROUND_UP(nrblocks, EXT4_ADDR_PER_BLOCK(inode->i_sb)) + 4; +} + +static int ext4_ind_trunc_restart_fn(handle_t *handle, struct inode *inode, + struct buffer_head *bh, int *dropped) +{ + int err; + + if (bh) { + BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata"); + err = ext4_handle_dirty_metadata(handle, inode, bh); + if (unlikely(err)) + return err; + } + err = ext4_mark_inode_dirty(handle, inode); + if (unlikely(err)) + return err; + /* + * Drop i_data_sem to avoid deadlock with ext4_map_blocks. At this + * moment, get_block can be called only for blocks inside i_size since + * page cache has been already dropped and writes are blocked by + * i_rwsem. So we can safely drop the i_data_sem here. + */ + BUG_ON(EXT4_JOURNAL(inode) == NULL); + ext4_discard_preallocations(inode); + up_write(&EXT4_I(inode)->i_data_sem); + *dropped = 1; + return 0; +} + +/* + * Truncate transactions can be complex and absolutely huge. So we need to + * be able to restart the transaction at a convenient checkpoint to make + * sure we don't overflow the journal. + * + * Try to extend this transaction for the purposes of truncation. If + * extend fails, we restart transaction. + */ +static int ext4_ind_truncate_ensure_credits(handle_t *handle, + struct inode *inode, + struct buffer_head *bh, + int revoke_creds) +{ + int ret; + int dropped = 0; + + ret = ext4_journal_ensure_credits_fn(handle, EXT4_RESERVE_TRANS_BLOCKS, + ext4_blocks_for_truncate(inode), revoke_creds, + ext4_ind_trunc_restart_fn(handle, inode, bh, &dropped)); + if (dropped) + down_write(&EXT4_I(inode)->i_data_sem); + if (ret <= 0) + return ret; + if (bh) { + BUFFER_TRACE(bh, "retaking write access"); + ret = ext4_journal_get_write_access(handle, inode->i_sb, bh, + EXT4_JTR_NONE); + if (unlikely(ret)) + return ret; + } + return 0; +} + +/* + * Probably it should be a library function... search for first non-zero word + * or memcmp with zero_page, whatever is better for particular architecture. + * Linus? + */ +static inline int all_zeroes(__le32 *p, __le32 *q) +{ + while (p < q) + if (*p++) + return 0; + return 1; +} + +/** + * ext4_find_shared - find the indirect blocks for partial truncation. + * @inode: inode in question + * @depth: depth of the affected branch + * @offsets: offsets of pointers in that branch (see ext4_block_to_path) + * @chain: place to store the pointers to partial indirect blocks + * @top: place to the (detached) top of branch + * + * This is a helper function used by ext4_truncate(). + * + * When we do truncate() we may have to clean the ends of several + * indirect blocks but leave the blocks themselves alive. Block is + * partially truncated if some data below the new i_size is referred + * from it (and it is on the path to the first completely truncated + * data block, indeed). We have to free the top of that path along + * with everything to the right of the path. Since no allocation + * past the truncation point is possible until ext4_truncate() + * finishes, we may safely do the latter, but top of branch may + * require special attention - pageout below the truncation point + * might try to populate it. + * + * We atomically detach the top of branch from the tree, store the + * block number of its root in *@top, pointers to buffer_heads of + * partially truncated blocks - in @chain[].bh and pointers to + * their last elements that should not be removed - in + * @chain[].p. Return value is the pointer to last filled element + * of @chain. + * + * The work left to caller to do the actual freeing of subtrees: + * a) free the subtree starting from *@top + * b) free the subtrees whose roots are stored in + * (@chain[i].p+1 .. end of @chain[i].bh->b_data) + * c) free the subtrees growing from the inode past the @chain[0]. + * (no partially truncated stuff there). */ + +static Indirect *ext4_find_shared(struct inode *inode, int depth, + ext4_lblk_t offsets[4], Indirect chain[4], + __le32 *top) +{ + Indirect *partial, *p; + int k, err; + + *top = 0; + /* Make k index the deepest non-null offset + 1 */ + for (k = depth; k > 1 && !offsets[k-1]; k--) + ; + partial = ext4_get_branch(inode, k, offsets, chain, &err); + /* Writer: pointers */ + if (!partial) + partial = chain + k-1; + /* + * If the branch acquired continuation since we've looked at it - + * fine, it should all survive and (new) top doesn't belong to us. + */ + if (!partial->key && *partial->p) + /* Writer: end */ + goto no_top; + for (p = partial; (p > chain) && all_zeroes((__le32 *) p->bh->b_data, p->p); p--) + ; + /* + * OK, we've found the last block that must survive. The rest of our + * branch should be detached before unlocking. However, if that rest + * of branch is all ours and does not grow immediately from the inode + * it's easier to cheat and just decrement partial->p. + */ + if (p == chain + k - 1 && p > chain) { + p->p--; + } else { + *top = *p->p; + /* Nope, don't do this in ext4. Must leave the tree intact */ +#if 0 + *p->p = 0; +#endif + } + /* Writer: end */ + + while (partial > p) { + brelse(partial->bh); + partial--; + } +no_top: + return partial; +} + +/* + * Zero a number of block pointers in either an inode or an indirect block. + * If we restart the transaction we must again get write access to the + * indirect block for further modification. + * + * We release `count' blocks on disk, but (last - first) may be greater + * than `count' because there can be holes in there. + * + * Return 0 on success, 1 on invalid block range + * and < 0 on fatal error. + */ +static int ext4_clear_blocks(handle_t *handle, struct inode *inode, + struct buffer_head *bh, + ext4_fsblk_t block_to_free, + unsigned long count, __le32 *first, + __le32 *last) +{ + __le32 *p; + int flags = EXT4_FREE_BLOCKS_VALIDATED; + int err; + + if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode) || + ext4_test_inode_flag(inode, EXT4_INODE_EA_INODE)) + flags |= EXT4_FREE_BLOCKS_FORGET | EXT4_FREE_BLOCKS_METADATA; + else if (ext4_should_journal_data(inode)) + flags |= EXT4_FREE_BLOCKS_FORGET; + + if (!ext4_inode_block_valid(inode, block_to_free, count)) { + EXT4_ERROR_INODE(inode, "attempt to clear invalid " + "blocks %llu len %lu", + (unsigned long long) block_to_free, count); + return 1; + } + + err = ext4_ind_truncate_ensure_credits(handle, inode, bh, + ext4_free_data_revoke_credits(inode, count)); + if (err < 0) + goto out_err; + + for (p = first; p < last; p++) + *p = 0; + + ext4_free_blocks(handle, inode, NULL, block_to_free, count, flags); + return 0; +out_err: + ext4_std_error(inode->i_sb, err); + return err; +} + +/** + * ext4_free_data - free a list of data blocks + * @handle: handle for this transaction + * @inode: inode we are dealing with + * @this_bh: indirect buffer_head which contains *@first and *@last + * @first: array of block numbers + * @last: points immediately past the end of array + * + * We are freeing all blocks referred from that array (numbers are stored as + * little-endian 32-bit) and updating @inode->i_blocks appropriately. + * + * We accumulate contiguous runs of blocks to free. Conveniently, if these + * blocks are contiguous then releasing them at one time will only affect one + * or two bitmap blocks (+ group descriptor(s) and superblock) and we won't + * actually use a lot of journal space. + * + * @this_bh will be %NULL if @first and @last point into the inode's direct + * block pointers. + */ +static void ext4_free_data(handle_t *handle, struct inode *inode, + struct buffer_head *this_bh, + __le32 *first, __le32 *last) +{ + ext4_fsblk_t block_to_free = 0; /* Starting block # of a run */ + unsigned long count = 0; /* Number of blocks in the run */ + __le32 *block_to_free_p = NULL; /* Pointer into inode/ind + corresponding to + block_to_free */ + ext4_fsblk_t nr; /* Current block # */ + __le32 *p; /* Pointer into inode/ind + for current block */ + int err = 0; + + if (this_bh) { /* For indirect block */ + BUFFER_TRACE(this_bh, "get_write_access"); + err = ext4_journal_get_write_access(handle, inode->i_sb, + this_bh, EXT4_JTR_NONE); + /* Important: if we can't update the indirect pointers + * to the blocks, we can't free them. */ + if (err) + return; + } + + for (p = first; p < last; p++) { + nr = le32_to_cpu(*p); + if (nr) { + /* accumulate blocks to free if they're contiguous */ + if (count == 0) { + block_to_free = nr; + block_to_free_p = p; + count = 1; + } else if (nr == block_to_free + count) { + count++; + } else { + err = ext4_clear_blocks(handle, inode, this_bh, + block_to_free, count, + block_to_free_p, p); + if (err) + break; + block_to_free = nr; + block_to_free_p = p; + count = 1; + } + } + } + + if (!err && count > 0) + err = ext4_clear_blocks(handle, inode, this_bh, block_to_free, + count, block_to_free_p, p); + if (err < 0) + /* fatal error */ + return; + + if (this_bh) { + BUFFER_TRACE(this_bh, "call ext4_handle_dirty_metadata"); + + /* + * The buffer head should have an attached journal head at this + * point. However, if the data is corrupted and an indirect + * block pointed to itself, it would have been detached when + * the block was cleared. Check for this instead of OOPSing. + */ + if ((EXT4_JOURNAL(inode) == NULL) || bh2jh(this_bh)) + ext4_handle_dirty_metadata(handle, inode, this_bh); + else + EXT4_ERROR_INODE(inode, + "circular indirect block detected at " + "block %llu", + (unsigned long long) this_bh->b_blocknr); + } +} + +/** + * ext4_free_branches - free an array of branches + * @handle: JBD handle for this transaction + * @inode: inode we are dealing with + * @parent_bh: the buffer_head which contains *@first and *@last + * @first: array of block numbers + * @last: pointer immediately past the end of array + * @depth: depth of the branches to free + * + * We are freeing all blocks referred from these branches (numbers are + * stored as little-endian 32-bit) and updating @inode->i_blocks + * appropriately. + */ +static void ext4_free_branches(handle_t *handle, struct inode *inode, + struct buffer_head *parent_bh, + __le32 *first, __le32 *last, int depth) +{ + ext4_fsblk_t nr; + __le32 *p; + + if (ext4_handle_is_aborted(handle)) + return; + + if (depth--) { + struct buffer_head *bh; + int addr_per_block = EXT4_ADDR_PER_BLOCK(inode->i_sb); + p = last; + while (--p >= first) { + nr = le32_to_cpu(*p); + if (!nr) + continue; /* A hole */ + + if (!ext4_inode_block_valid(inode, nr, 1)) { + EXT4_ERROR_INODE(inode, + "invalid indirect mapped " + "block %lu (level %d)", + (unsigned long) nr, depth); + break; + } + + /* Go read the buffer for the next level down */ + bh = ext4_sb_bread_nofail(inode->i_sb, nr); + + /* + * A read failure? Report error and clear slot + * (should be rare). + */ + if (IS_ERR(bh)) { + ext4_error_inode_block(inode, nr, -PTR_ERR(bh), + "Read failure"); + continue; + } + + /* This zaps the entire block. Bottom up. */ + BUFFER_TRACE(bh, "free child branches"); + ext4_free_branches(handle, inode, bh, + (__le32 *) bh->b_data, + (__le32 *) bh->b_data + addr_per_block, + depth); + brelse(bh); + + /* + * Everything below this pointer has been + * released. Now let this top-of-subtree go. + * + * We want the freeing of this indirect block to be + * atomic in the journal with the updating of the + * bitmap block which owns it. So make some room in + * the journal. + * + * We zero the parent pointer *after* freeing its + * pointee in the bitmaps, so if extend_transaction() + * for some reason fails to put the bitmap changes and + * the release into the same transaction, recovery + * will merely complain about releasing a free block, + * rather than leaking blocks. + */ + if (ext4_handle_is_aborted(handle)) + return; + if (ext4_ind_truncate_ensure_credits(handle, inode, + NULL, + ext4_free_metadata_revoke_credits( + inode->i_sb, 1)) < 0) + return; + + /* + * The forget flag here is critical because if + * we are journaling (and not doing data + * journaling), we have to make sure a revoke + * record is written to prevent the journal + * replay from overwriting the (former) + * indirect block if it gets reallocated as a + * data block. This must happen in the same + * transaction where the data blocks are + * actually freed. + */ + ext4_free_blocks(handle, inode, NULL, nr, 1, + EXT4_FREE_BLOCKS_METADATA| + EXT4_FREE_BLOCKS_FORGET); + + if (parent_bh) { + /* + * The block which we have just freed is + * pointed to by an indirect block: journal it + */ + BUFFER_TRACE(parent_bh, "get_write_access"); + if (!ext4_journal_get_write_access(handle, + inode->i_sb, parent_bh, + EXT4_JTR_NONE)) { + *p = 0; + BUFFER_TRACE(parent_bh, + "call ext4_handle_dirty_metadata"); + ext4_handle_dirty_metadata(handle, + inode, + parent_bh); + } + } + } + } else { + /* We have reached the bottom of the tree. */ + BUFFER_TRACE(parent_bh, "free data blocks"); + ext4_free_data(handle, inode, parent_bh, first, last); + } +} + +void ext4_ind_truncate(handle_t *handle, struct inode *inode) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + __le32 *i_data = ei->i_data; + int addr_per_block = EXT4_ADDR_PER_BLOCK(inode->i_sb); + ext4_lblk_t offsets[4]; + Indirect chain[4]; + Indirect *partial; + __le32 nr = 0; + int n = 0; + ext4_lblk_t last_block, max_block; + unsigned blocksize = inode->i_sb->s_blocksize; + + last_block = (inode->i_size + blocksize-1) + >> EXT4_BLOCK_SIZE_BITS(inode->i_sb); + max_block = (EXT4_SB(inode->i_sb)->s_bitmap_maxbytes + blocksize-1) + >> EXT4_BLOCK_SIZE_BITS(inode->i_sb); + + if (last_block != max_block) { + n = ext4_block_to_path(inode, last_block, offsets, NULL); + if (n == 0) + return; + } + + ext4_es_remove_extent(inode, last_block, EXT_MAX_BLOCKS - last_block); + + /* + * The orphan list entry will now protect us from any crash which + * occurs before the truncate completes, so it is now safe to propagate + * the new, shorter inode size (held for now in i_size) into the + * on-disk inode. We do this via i_disksize, which is the value which + * ext4 *really* writes onto the disk inode. + */ + ei->i_disksize = inode->i_size; + + if (last_block == max_block) { + /* + * It is unnecessary to free any data blocks if last_block is + * equal to the indirect block limit. + */ + return; + } else if (n == 1) { /* direct blocks */ + ext4_free_data(handle, inode, NULL, i_data+offsets[0], + i_data + EXT4_NDIR_BLOCKS); + goto do_indirects; + } + + partial = ext4_find_shared(inode, n, offsets, chain, &nr); + /* Kill the top of shared branch (not detached) */ + if (nr) { + if (partial == chain) { + /* Shared branch grows from the inode */ + ext4_free_branches(handle, inode, NULL, + &nr, &nr+1, (chain+n-1) - partial); + *partial->p = 0; + /* + * We mark the inode dirty prior to restart, + * and prior to stop. No need for it here. + */ + } else { + /* Shared branch grows from an indirect block */ + BUFFER_TRACE(partial->bh, "get_write_access"); + ext4_free_branches(handle, inode, partial->bh, + partial->p, + partial->p+1, (chain+n-1) - partial); + } + } + /* Clear the ends of indirect blocks on the shared branch */ + while (partial > chain) { + ext4_free_branches(handle, inode, partial->bh, partial->p + 1, + (__le32*)partial->bh->b_data+addr_per_block, + (chain+n-1) - partial); + BUFFER_TRACE(partial->bh, "call brelse"); + brelse(partial->bh); + partial--; + } +do_indirects: + /* Kill the remaining (whole) subtrees */ + switch (offsets[0]) { + default: + nr = i_data[EXT4_IND_BLOCK]; + if (nr) { + ext4_free_branches(handle, inode, NULL, &nr, &nr+1, 1); + i_data[EXT4_IND_BLOCK] = 0; + } + fallthrough; + case EXT4_IND_BLOCK: + nr = i_data[EXT4_DIND_BLOCK]; + if (nr) { + ext4_free_branches(handle, inode, NULL, &nr, &nr+1, 2); + i_data[EXT4_DIND_BLOCK] = 0; + } + fallthrough; + case EXT4_DIND_BLOCK: + nr = i_data[EXT4_TIND_BLOCK]; + if (nr) { + ext4_free_branches(handle, inode, NULL, &nr, &nr+1, 3); + i_data[EXT4_TIND_BLOCK] = 0; + } + fallthrough; + case EXT4_TIND_BLOCK: + ; + } +} + +/** + * ext4_ind_remove_space - remove space from the range + * @handle: JBD handle for this transaction + * @inode: inode we are dealing with + * @start: First block to remove + * @end: One block after the last block to remove (exclusive) + * + * Free the blocks in the defined range (end is exclusive endpoint of + * range). This is used by ext4_punch_hole(). + */ +int ext4_ind_remove_space(handle_t *handle, struct inode *inode, + ext4_lblk_t start, ext4_lblk_t end) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + __le32 *i_data = ei->i_data; + int addr_per_block = EXT4_ADDR_PER_BLOCK(inode->i_sb); + ext4_lblk_t offsets[4], offsets2[4]; + Indirect chain[4], chain2[4]; + Indirect *partial, *partial2; + Indirect *p = NULL, *p2 = NULL; + ext4_lblk_t max_block; + __le32 nr = 0, nr2 = 0; + int n = 0, n2 = 0; + unsigned blocksize = inode->i_sb->s_blocksize; + + max_block = (EXT4_SB(inode->i_sb)->s_bitmap_maxbytes + blocksize-1) + >> EXT4_BLOCK_SIZE_BITS(inode->i_sb); + if (end >= max_block) + end = max_block; + if ((start >= end) || (start > max_block)) + return 0; + + n = ext4_block_to_path(inode, start, offsets, NULL); + n2 = ext4_block_to_path(inode, end, offsets2, NULL); + + BUG_ON(n > n2); + + if ((n == 1) && (n == n2)) { + /* We're punching only within direct block range */ + ext4_free_data(handle, inode, NULL, i_data + offsets[0], + i_data + offsets2[0]); + return 0; + } else if (n2 > n) { + /* + * Start and end are on a different levels so we're going to + * free partial block at start, and partial block at end of + * the range. If there are some levels in between then + * do_indirects label will take care of that. + */ + + if (n == 1) { + /* + * Start is at the direct block level, free + * everything to the end of the level. + */ + ext4_free_data(handle, inode, NULL, i_data + offsets[0], + i_data + EXT4_NDIR_BLOCKS); + goto end_range; + } + + + partial = p = ext4_find_shared(inode, n, offsets, chain, &nr); + if (nr) { + if (partial == chain) { + /* Shared branch grows from the inode */ + ext4_free_branches(handle, inode, NULL, + &nr, &nr+1, (chain+n-1) - partial); + *partial->p = 0; + } else { + /* Shared branch grows from an indirect block */ + BUFFER_TRACE(partial->bh, "get_write_access"); + ext4_free_branches(handle, inode, partial->bh, + partial->p, + partial->p+1, (chain+n-1) - partial); + } + } + + /* + * Clear the ends of indirect blocks on the shared branch + * at the start of the range + */ + while (partial > chain) { + ext4_free_branches(handle, inode, partial->bh, + partial->p + 1, + (__le32 *)partial->bh->b_data+addr_per_block, + (chain+n-1) - partial); + partial--; + } + +end_range: + partial2 = p2 = ext4_find_shared(inode, n2, offsets2, chain2, &nr2); + if (nr2) { + if (partial2 == chain2) { + /* + * Remember, end is exclusive so here we're at + * the start of the next level we're not going + * to free. Everything was covered by the start + * of the range. + */ + goto do_indirects; + } + } else { + /* + * ext4_find_shared returns Indirect structure which + * points to the last element which should not be + * removed by truncate. But this is end of the range + * in punch_hole so we need to point to the next element + */ + partial2->p++; + } + + /* + * Clear the ends of indirect blocks on the shared branch + * at the end of the range + */ + while (partial2 > chain2) { + ext4_free_branches(handle, inode, partial2->bh, + (__le32 *)partial2->bh->b_data, + partial2->p, + (chain2+n2-1) - partial2); + partial2--; + } + goto do_indirects; + } + + /* Punch happened within the same level (n == n2) */ + partial = p = ext4_find_shared(inode, n, offsets, chain, &nr); + partial2 = p2 = ext4_find_shared(inode, n2, offsets2, chain2, &nr2); + + /* Free top, but only if partial2 isn't its subtree. */ + if (nr) { + int level = min(partial - chain, partial2 - chain2); + int i; + int subtree = 1; + + for (i = 0; i <= level; i++) { + if (offsets[i] != offsets2[i]) { + subtree = 0; + break; + } + } + + if (!subtree) { + if (partial == chain) { + /* Shared branch grows from the inode */ + ext4_free_branches(handle, inode, NULL, + &nr, &nr+1, + (chain+n-1) - partial); + *partial->p = 0; + } else { + /* Shared branch grows from an indirect block */ + BUFFER_TRACE(partial->bh, "get_write_access"); + ext4_free_branches(handle, inode, partial->bh, + partial->p, + partial->p+1, + (chain+n-1) - partial); + } + } + } + + if (!nr2) { + /* + * ext4_find_shared returns Indirect structure which + * points to the last element which should not be + * removed by truncate. But this is end of the range + * in punch_hole so we need to point to the next element + */ + partial2->p++; + } + + while (partial > chain || partial2 > chain2) { + int depth = (chain+n-1) - partial; + int depth2 = (chain2+n2-1) - partial2; + + if (partial > chain && partial2 > chain2 && + partial->bh->b_blocknr == partial2->bh->b_blocknr) { + /* + * We've converged on the same block. Clear the range, + * then we're done. + */ + ext4_free_branches(handle, inode, partial->bh, + partial->p + 1, + partial2->p, + (chain+n-1) - partial); + goto cleanup; + } + + /* + * The start and end partial branches may not be at the same + * level even though the punch happened within one level. So, we + * give them a chance to arrive at the same level, then walk + * them in step with each other until we converge on the same + * block. + */ + if (partial > chain && depth <= depth2) { + ext4_free_branches(handle, inode, partial->bh, + partial->p + 1, + (__le32 *)partial->bh->b_data+addr_per_block, + (chain+n-1) - partial); + partial--; + } + if (partial2 > chain2 && depth2 <= depth) { + ext4_free_branches(handle, inode, partial2->bh, + (__le32 *)partial2->bh->b_data, + partial2->p, + (chain2+n2-1) - partial2); + partial2--; + } + } + +cleanup: + while (p && p > chain) { + BUFFER_TRACE(p->bh, "call brelse"); + brelse(p->bh); + p--; + } + while (p2 && p2 > chain2) { + BUFFER_TRACE(p2->bh, "call brelse"); + brelse(p2->bh); + p2--; + } + return 0; + +do_indirects: + /* Kill the remaining (whole) subtrees */ + switch (offsets[0]) { + default: + if (++n >= n2) + break; + nr = i_data[EXT4_IND_BLOCK]; + if (nr) { + ext4_free_branches(handle, inode, NULL, &nr, &nr+1, 1); + i_data[EXT4_IND_BLOCK] = 0; + } + fallthrough; + case EXT4_IND_BLOCK: + if (++n >= n2) + break; + nr = i_data[EXT4_DIND_BLOCK]; + if (nr) { + ext4_free_branches(handle, inode, NULL, &nr, &nr+1, 2); + i_data[EXT4_DIND_BLOCK] = 0; + } + fallthrough; + case EXT4_DIND_BLOCK: + if (++n >= n2) + break; + nr = i_data[EXT4_TIND_BLOCK]; + if (nr) { + ext4_free_branches(handle, inode, NULL, &nr, &nr+1, 3); + i_data[EXT4_TIND_BLOCK] = 0; + } + fallthrough; + case EXT4_TIND_BLOCK: + ; + } + goto cleanup; +} diff --git a/fs/ext4l/inline.c b/fs/ext4l/inline.c new file mode 100644 index 00000000000..1b094a4f386 --- /dev/null +++ b/fs/ext4l/inline.c @@ -0,0 +1,1982 @@ +// SPDX-License-Identifier: LGPL-2.1 +/* + * Copyright (c) 2012 Taobao. + * Written by Tao Ma <boyu.mt@taobao.com> + */ + +#include <linux/iomap.h> +#include <linux/fiemap.h> +#include <linux/namei.h> +#include <linux/iversion.h> +#include <linux/sched/mm.h> + +#include "ext4_jbd2.h" +#include "ext4.h" +#include "xattr.h" +#include "truncate.h" + +#define EXT4_XATTR_SYSTEM_DATA "data" +#define EXT4_MIN_INLINE_DATA_SIZE ((sizeof(__le32) * EXT4_N_BLOCKS)) +#define EXT4_INLINE_DOTDOT_OFFSET 2 +#define EXT4_INLINE_DOTDOT_SIZE 4 + + +static int ext4_da_convert_inline_data_to_extent(struct address_space *mapping, + struct inode *inode, + void **fsdata); + +static int ext4_get_inline_size(struct inode *inode) +{ + if (EXT4_I(inode)->i_inline_off) + return EXT4_I(inode)->i_inline_size; + + return 0; +} + +static int get_max_inline_xattr_value_size(struct inode *inode, + struct ext4_iloc *iloc) +{ + struct ext4_xattr_ibody_header *header; + struct ext4_xattr_entry *entry; + struct ext4_inode *raw_inode; + void *end; + int free, min_offs; + + if (!EXT4_INODE_HAS_XATTR_SPACE(inode)) + return 0; + + min_offs = EXT4_SB(inode->i_sb)->s_inode_size - + EXT4_GOOD_OLD_INODE_SIZE - + EXT4_I(inode)->i_extra_isize - + sizeof(struct ext4_xattr_ibody_header); + + /* + * We need to subtract another sizeof(__u32) since an in-inode xattr + * needs an empty 4 bytes to indicate the gap between the xattr entry + * and the name/value pair. + */ + if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR)) + return EXT4_XATTR_SIZE(min_offs - + EXT4_XATTR_LEN(strlen(EXT4_XATTR_SYSTEM_DATA)) - + EXT4_XATTR_ROUND - sizeof(__u32)); + + raw_inode = ext4_raw_inode(iloc); + header = IHDR(inode, raw_inode); + entry = IFIRST(header); + end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size; + + /* Compute min_offs. */ + while (!IS_LAST_ENTRY(entry)) { + void *next = EXT4_XATTR_NEXT(entry); + + if (next >= end) { + EXT4_ERROR_INODE(inode, + "corrupt xattr in inline inode"); + return 0; + } + if (!entry->e_value_inum && entry->e_value_size) { + size_t offs = le16_to_cpu(entry->e_value_offs); + if (offs < min_offs) + min_offs = offs; + } + entry = next; + } + free = min_offs - + ((void *)entry - (void *)IFIRST(header)) - sizeof(__u32); + + if (EXT4_I(inode)->i_inline_off) { + entry = (struct ext4_xattr_entry *) + ((void *)raw_inode + EXT4_I(inode)->i_inline_off); + + free += EXT4_XATTR_SIZE(le32_to_cpu(entry->e_value_size)); + goto out; + } + + free -= EXT4_XATTR_LEN(strlen(EXT4_XATTR_SYSTEM_DATA)); + + if (free > EXT4_XATTR_ROUND) + free = EXT4_XATTR_SIZE(free - EXT4_XATTR_ROUND); + else + free = 0; + +out: + return free; +} + +/* + * Get the maximum size we now can store in an inode. + * If we can't find the space for a xattr entry, don't use the space + * of the extents since we have no space to indicate the inline data. + */ +int ext4_get_max_inline_size(struct inode *inode) +{ + int error, max_inline_size; + struct ext4_iloc iloc; + + if (EXT4_I(inode)->i_extra_isize == 0) + return 0; + + error = ext4_get_inode_loc(inode, &iloc); + if (error) { + ext4_error_inode_err(inode, __func__, __LINE__, 0, -error, + "can't get inode location %lu", + inode->i_ino); + return 0; + } + + down_read(&EXT4_I(inode)->xattr_sem); + max_inline_size = get_max_inline_xattr_value_size(inode, &iloc); + up_read(&EXT4_I(inode)->xattr_sem); + + brelse(iloc.bh); + + if (!max_inline_size) + return 0; + + return max_inline_size + EXT4_MIN_INLINE_DATA_SIZE; +} + +/* + * this function does not take xattr_sem, which is OK because it is + * currently only used in a code path coming form ext4_iget, before + * the new inode has been unlocked + */ +int ext4_find_inline_data_nolock(struct inode *inode) +{ + struct ext4_xattr_ibody_find is = { + .s = { .not_found = -ENODATA, }, + }; + struct ext4_xattr_info i = { + .name_index = EXT4_XATTR_INDEX_SYSTEM, + .name = EXT4_XATTR_SYSTEM_DATA, + }; + int error; + + if (EXT4_I(inode)->i_extra_isize == 0) + return 0; + + error = ext4_get_inode_loc(inode, &is.iloc); + if (error) + return error; + + error = ext4_xattr_ibody_find(inode, &i, &is); + if (error) + goto out; + + if (!is.s.not_found) { + if (is.s.here->e_value_inum) { + EXT4_ERROR_INODE(inode, "inline data xattr refers " + "to an external xattr inode"); + error = -EFSCORRUPTED; + goto out; + } + EXT4_I(inode)->i_inline_off = (u16)((void *)is.s.here - + (void *)ext4_raw_inode(&is.iloc)); + EXT4_I(inode)->i_inline_size = EXT4_MIN_INLINE_DATA_SIZE + + le32_to_cpu(is.s.here->e_value_size); + } +out: + brelse(is.iloc.bh); + return error; +} + +static int ext4_read_inline_data(struct inode *inode, void *buffer, + unsigned int len, + struct ext4_iloc *iloc) +{ + struct ext4_xattr_entry *entry; + struct ext4_xattr_ibody_header *header; + int cp_len = 0; + struct ext4_inode *raw_inode; + + if (!len) + return 0; + + BUG_ON(len > EXT4_I(inode)->i_inline_size); + + cp_len = min_t(unsigned int, len, EXT4_MIN_INLINE_DATA_SIZE); + + raw_inode = ext4_raw_inode(iloc); + memcpy(buffer, (void *)(raw_inode->i_block), cp_len); + + len -= cp_len; + buffer += cp_len; + + if (!len) + goto out; + + header = IHDR(inode, raw_inode); + entry = (struct ext4_xattr_entry *)((void *)raw_inode + + EXT4_I(inode)->i_inline_off); + len = min_t(unsigned int, len, + (unsigned int)le32_to_cpu(entry->e_value_size)); + + memcpy(buffer, + (void *)IFIRST(header) + le16_to_cpu(entry->e_value_offs), len); + cp_len += len; + +out: + return cp_len; +} + +/* + * write the buffer to the inline inode. + * If 'create' is set, we don't need to do the extra copy in the xattr + * value since it is already handled by ext4_xattr_ibody_set. + * That saves us one memcpy. + */ +static void ext4_write_inline_data(struct inode *inode, struct ext4_iloc *iloc, + void *buffer, loff_t pos, unsigned int len) +{ + struct ext4_xattr_entry *entry; + struct ext4_xattr_ibody_header *header; + struct ext4_inode *raw_inode; + int cp_len = 0; + + if (unlikely(ext4_emergency_state(inode->i_sb))) + return; + + BUG_ON(!EXT4_I(inode)->i_inline_off); + BUG_ON(pos + len > EXT4_I(inode)->i_inline_size); + + raw_inode = ext4_raw_inode(iloc); + buffer += pos; + + if (pos < EXT4_MIN_INLINE_DATA_SIZE) { + cp_len = pos + len > EXT4_MIN_INLINE_DATA_SIZE ? + EXT4_MIN_INLINE_DATA_SIZE - pos : len; + memcpy((void *)raw_inode->i_block + pos, buffer, cp_len); + + len -= cp_len; + buffer += cp_len; + pos += cp_len; + } + + if (!len) + return; + + pos -= EXT4_MIN_INLINE_DATA_SIZE; + header = IHDR(inode, raw_inode); + entry = (struct ext4_xattr_entry *)((void *)raw_inode + + EXT4_I(inode)->i_inline_off); + + memcpy((void *)IFIRST(header) + le16_to_cpu(entry->e_value_offs) + pos, + buffer, len); +} + +static int ext4_create_inline_data(handle_t *handle, + struct inode *inode, unsigned len) +{ + int error; + void *value = NULL; + struct ext4_xattr_ibody_find is = { + .s = { .not_found = -ENODATA, }, + }; + struct ext4_xattr_info i = { + .name_index = EXT4_XATTR_INDEX_SYSTEM, + .name = EXT4_XATTR_SYSTEM_DATA, + }; + + error = ext4_get_inode_loc(inode, &is.iloc); + if (error) + return error; + + BUFFER_TRACE(is.iloc.bh, "get_write_access"); + error = ext4_journal_get_write_access(handle, inode->i_sb, is.iloc.bh, + EXT4_JTR_NONE); + if (error) + goto out; + + if (len > EXT4_MIN_INLINE_DATA_SIZE) { + value = EXT4_ZERO_XATTR_VALUE; + len -= EXT4_MIN_INLINE_DATA_SIZE; + } else { + value = ""; + len = 0; + } + + /* Insert the xttr entry. */ + i.value = value; + i.value_len = len; + + error = ext4_xattr_ibody_find(inode, &i, &is); + if (error) + goto out; + + if (!is.s.not_found) { + EXT4_ERROR_INODE(inode, "unexpected inline data xattr"); + error = -EFSCORRUPTED; + goto out; + } + + error = ext4_xattr_ibody_set(handle, inode, &i, &is); + if (error) { + if (error == -ENOSPC) + ext4_clear_inode_state(inode, + EXT4_STATE_MAY_INLINE_DATA); + goto out; + } + + memset((void *)ext4_raw_inode(&is.iloc)->i_block, + 0, EXT4_MIN_INLINE_DATA_SIZE); + + EXT4_I(inode)->i_inline_off = (u16)((void *)is.s.here - + (void *)ext4_raw_inode(&is.iloc)); + EXT4_I(inode)->i_inline_size = len + EXT4_MIN_INLINE_DATA_SIZE; + ext4_clear_inode_flag(inode, EXT4_INODE_EXTENTS); + ext4_set_inode_flag(inode, EXT4_INODE_INLINE_DATA); + get_bh(is.iloc.bh); + error = ext4_mark_iloc_dirty(handle, inode, &is.iloc); + +out: + brelse(is.iloc.bh); + return error; +} + +static int ext4_update_inline_data(handle_t *handle, struct inode *inode, + unsigned int len) +{ + int error; + void *value = NULL; + struct ext4_xattr_ibody_find is = { + .s = { .not_found = -ENODATA, }, + }; + struct ext4_xattr_info i = { + .name_index = EXT4_XATTR_INDEX_SYSTEM, + .name = EXT4_XATTR_SYSTEM_DATA, + }; + + /* If the old space is ok, write the data directly. */ + if (len <= EXT4_I(inode)->i_inline_size) + return 0; + + error = ext4_get_inode_loc(inode, &is.iloc); + if (error) + return error; + + error = ext4_xattr_ibody_find(inode, &i, &is); + if (error) + goto out; + + if (is.s.not_found) { + EXT4_ERROR_INODE(inode, "missing inline data xattr"); + error = -EFSCORRUPTED; + goto out; + } + + len -= EXT4_MIN_INLINE_DATA_SIZE; + value = kzalloc(len, GFP_NOFS); + if (!value) { + error = -ENOMEM; + goto out; + } + + error = ext4_xattr_ibody_get(inode, i.name_index, i.name, + value, len); + if (error < 0) + goto out; + + BUFFER_TRACE(is.iloc.bh, "get_write_access"); + error = ext4_journal_get_write_access(handle, inode->i_sb, is.iloc.bh, + EXT4_JTR_NONE); + if (error) + goto out; + + /* Update the xattr entry. */ + i.value = value; + i.value_len = len; + + error = ext4_xattr_ibody_set(handle, inode, &i, &is); + if (error) + goto out; + + EXT4_I(inode)->i_inline_off = (u16)((void *)is.s.here - + (void *)ext4_raw_inode(&is.iloc)); + EXT4_I(inode)->i_inline_size = EXT4_MIN_INLINE_DATA_SIZE + + le32_to_cpu(is.s.here->e_value_size); + ext4_set_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA); + get_bh(is.iloc.bh); + error = ext4_mark_iloc_dirty(handle, inode, &is.iloc); + +out: + kfree(value); + brelse(is.iloc.bh); + return error; +} + +static int ext4_prepare_inline_data(handle_t *handle, struct inode *inode, + loff_t len) +{ + int ret, size, no_expand; + struct ext4_inode_info *ei = EXT4_I(inode); + + if (!ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA)) + return -ENOSPC; + + size = ext4_get_max_inline_size(inode); + if (size < len) + return -ENOSPC; + + ext4_write_lock_xattr(inode, &no_expand); + + if (ei->i_inline_off) + ret = ext4_update_inline_data(handle, inode, len); + else + ret = ext4_create_inline_data(handle, inode, len); + + ext4_write_unlock_xattr(inode, &no_expand); + return ret; +} + +static int ext4_destroy_inline_data_nolock(handle_t *handle, + struct inode *inode) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + struct ext4_xattr_ibody_find is = { + .s = { .not_found = 0, }, + }; + struct ext4_xattr_info i = { + .name_index = EXT4_XATTR_INDEX_SYSTEM, + .name = EXT4_XATTR_SYSTEM_DATA, + .value = NULL, + .value_len = 0, + }; + int error; + + if (!ei->i_inline_off) + return 0; + + error = ext4_get_inode_loc(inode, &is.iloc); + if (error) + return error; + + error = ext4_xattr_ibody_find(inode, &i, &is); + if (error) + goto out; + + BUFFER_TRACE(is.iloc.bh, "get_write_access"); + error = ext4_journal_get_write_access(handle, inode->i_sb, is.iloc.bh, + EXT4_JTR_NONE); + if (error) + goto out; + + error = ext4_xattr_ibody_set(handle, inode, &i, &is); + if (error) + goto out; + + memset((void *)ext4_raw_inode(&is.iloc)->i_block, + 0, EXT4_MIN_INLINE_DATA_SIZE); + memset(ei->i_data, 0, EXT4_MIN_INLINE_DATA_SIZE); + + if (ext4_has_feature_extents(inode->i_sb)) { + if (S_ISDIR(inode->i_mode) || + S_ISREG(inode->i_mode) || S_ISLNK(inode->i_mode)) { + ext4_set_inode_flag(inode, EXT4_INODE_EXTENTS); + ext4_ext_tree_init(handle, inode); + } + } + ext4_clear_inode_flag(inode, EXT4_INODE_INLINE_DATA); + + get_bh(is.iloc.bh); + error = ext4_mark_iloc_dirty(handle, inode, &is.iloc); + + EXT4_I(inode)->i_inline_off = 0; + EXT4_I(inode)->i_inline_size = 0; + ext4_clear_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA); +out: + brelse(is.iloc.bh); + if (error == -ENODATA) + error = 0; + return error; +} + +static int ext4_read_inline_folio(struct inode *inode, struct folio *folio) +{ + void *kaddr; + int ret = 0; + size_t len; + struct ext4_iloc iloc; + + BUG_ON(!folio_test_locked(folio)); + BUG_ON(!ext4_has_inline_data(inode)); + BUG_ON(folio->index); + + if (!EXT4_I(inode)->i_inline_off) { + ext4_warning(inode->i_sb, "inode %lu doesn't have inline data.", + inode->i_ino); + goto out; + } + + ret = ext4_get_inode_loc(inode, &iloc); + if (ret) + goto out; + + len = min_t(size_t, ext4_get_inline_size(inode), i_size_read(inode)); + BUG_ON(len > PAGE_SIZE); + kaddr = kmap_local_folio(folio, 0); + ret = ext4_read_inline_data(inode, kaddr, len, &iloc); + kaddr = folio_zero_tail(folio, len, kaddr + len); + kunmap_local(kaddr); + folio_mark_uptodate(folio); + brelse(iloc.bh); + +out: + return ret; +} + +int ext4_readpage_inline(struct inode *inode, struct folio *folio) +{ + int ret = 0; + + down_read(&EXT4_I(inode)->xattr_sem); + if (!ext4_has_inline_data(inode)) { + up_read(&EXT4_I(inode)->xattr_sem); + return -EAGAIN; + } + + /* + * Current inline data can only exist in the 1st page, + * So for all the other pages, just set them uptodate. + */ + if (!folio->index) + ret = ext4_read_inline_folio(inode, folio); + else if (!folio_test_uptodate(folio)) { + folio_zero_segment(folio, 0, folio_size(folio)); + folio_mark_uptodate(folio); + } + + up_read(&EXT4_I(inode)->xattr_sem); + + folio_unlock(folio); + return ret >= 0 ? 0 : ret; +} + +static int ext4_convert_inline_data_to_extent(struct address_space *mapping, + struct inode *inode) +{ + int ret, needed_blocks, no_expand; + handle_t *handle = NULL; + int retries = 0, sem_held = 0; + struct folio *folio = NULL; + unsigned from, to; + struct ext4_iloc iloc; + + if (!ext4_has_inline_data(inode)) { + /* + * clear the flag so that no new write + * will trap here again. + */ + ext4_clear_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA); + return 0; + } + + needed_blocks = ext4_chunk_trans_extent(inode, 1); + + ret = ext4_get_inode_loc(inode, &iloc); + if (ret) + return ret; + +retry: + handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, needed_blocks); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + handle = NULL; + goto out; + } + + /* We cannot recurse into the filesystem as the transaction is already + * started */ + folio = __filemap_get_folio(mapping, 0, FGP_WRITEBEGIN | FGP_NOFS, + mapping_gfp_mask(mapping)); + if (IS_ERR(folio)) { + ret = PTR_ERR(folio); + goto out_nofolio; + } + + ext4_write_lock_xattr(inode, &no_expand); + sem_held = 1; + /* If some one has already done this for us, just exit. */ + if (!ext4_has_inline_data(inode)) { + ret = 0; + goto out; + } + + from = 0; + to = ext4_get_inline_size(inode); + if (!folio_test_uptodate(folio)) { + ret = ext4_read_inline_folio(inode, folio); + if (ret < 0) + goto out; + } + + ext4_fc_track_inode(handle, inode); + ret = ext4_destroy_inline_data_nolock(handle, inode); + if (ret) + goto out; + + if (ext4_should_dioread_nolock(inode)) { + ret = ext4_block_write_begin(handle, folio, from, to, + ext4_get_block_unwritten); + } else + ret = ext4_block_write_begin(handle, folio, from, to, + ext4_get_block); + clear_buffer_new(folio_buffers(folio)); + + if (!ret && ext4_should_journal_data(inode)) { + ret = ext4_walk_page_buffers(handle, inode, + folio_buffers(folio), from, to, + NULL, do_journal_get_write_access); + } + + if (ret) { + folio_unlock(folio); + folio_put(folio); + folio = NULL; + ext4_orphan_add(handle, inode); + ext4_write_unlock_xattr(inode, &no_expand); + sem_held = 0; + ext4_journal_stop(handle); + handle = NULL; + ext4_truncate_failed_write(inode); + /* + * If truncate failed early the inode might + * still be on the orphan list; we need to + * make sure the inode is removed from the + * orphan list in that case. + */ + if (inode->i_nlink) + ext4_orphan_del(NULL, inode); + } + + if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries)) + goto retry; + + if (folio) + block_commit_write(folio, from, to); +out: + if (folio) { + folio_unlock(folio); + folio_put(folio); + } +out_nofolio: + if (sem_held) + ext4_write_unlock_xattr(inode, &no_expand); + if (handle) + ext4_journal_stop(handle); + brelse(iloc.bh); + return ret; +} + +/* + * Prepare the write for the inline data. + * If the data can be written into the inode, we just read + * the page and make it uptodate, and start the journal. + * Otherwise read the page, makes it dirty so that it can be + * handle in writepages(the i_disksize update is left to the + * normal ext4_da_write_end). + */ +int ext4_generic_write_inline_data(struct address_space *mapping, + struct inode *inode, + loff_t pos, unsigned len, + struct folio **foliop, + void **fsdata, bool da) +{ + int ret; + handle_t *handle; + struct folio *folio; + struct ext4_iloc iloc; + int retries = 0; + + ret = ext4_get_inode_loc(inode, &iloc); + if (ret) + return ret; + +retry_journal: + handle = ext4_journal_start(inode, EXT4_HT_INODE, 1); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + goto out_release_bh; + } + + ret = ext4_prepare_inline_data(handle, inode, pos + len); + if (ret && ret != -ENOSPC) + goto out_stop_journal; + + if (ret == -ENOSPC) { + ext4_journal_stop(handle); + if (!da) { + brelse(iloc.bh); + /* Retry inside */ + return ext4_convert_inline_data_to_extent(mapping, inode); + } + + ret = ext4_da_convert_inline_data_to_extent(mapping, inode, fsdata); + if (ret == -ENOSPC && + ext4_should_retry_alloc(inode->i_sb, &retries)) + goto retry_journal; + goto out_release_bh; + } + + folio = __filemap_get_folio(mapping, 0, FGP_WRITEBEGIN | FGP_NOFS, + mapping_gfp_mask(mapping)); + if (IS_ERR(folio)) { + ret = PTR_ERR(folio); + goto out_stop_journal; + } + + down_read(&EXT4_I(inode)->xattr_sem); + /* Someone else had converted it to extent */ + if (!ext4_has_inline_data(inode)) { + ret = 0; + goto out_release_folio; + } + + if (!folio_test_uptodate(folio)) { + ret = ext4_read_inline_folio(inode, folio); + if (ret < 0) + goto out_release_folio; + } + + ret = ext4_journal_get_write_access(handle, inode->i_sb, iloc.bh, EXT4_JTR_NONE); + if (ret) + goto out_release_folio; + *foliop = folio; + up_read(&EXT4_I(inode)->xattr_sem); + brelse(iloc.bh); + return 1; + +out_release_folio: + up_read(&EXT4_I(inode)->xattr_sem); + folio_unlock(folio); + folio_put(folio); +out_stop_journal: + ext4_journal_stop(handle); +out_release_bh: + brelse(iloc.bh); + return ret; +} + +/* + * Try to write data in the inode. + * If the inode has inline data, check whether the new write can be + * in the inode also. If not, create the page the handle, move the data + * to the page make it update and let the later codes create extent for it. + */ +int ext4_try_to_write_inline_data(struct address_space *mapping, + struct inode *inode, + loff_t pos, unsigned len, + struct folio **foliop) +{ + if (pos + len > ext4_get_max_inline_size(inode)) + return ext4_convert_inline_data_to_extent(mapping, inode); + return ext4_generic_write_inline_data(mapping, inode, pos, len, + foliop, NULL, false); +} + +int ext4_write_inline_data_end(struct inode *inode, loff_t pos, unsigned len, + unsigned copied, struct folio *folio) +{ + handle_t *handle = ext4_journal_current_handle(); + int no_expand; + void *kaddr; + struct ext4_iloc iloc; + int ret = 0, ret2; + + if (unlikely(copied < len) && !folio_test_uptodate(folio)) + copied = 0; + + if (likely(copied)) { + ret = ext4_get_inode_loc(inode, &iloc); + if (ret) { + folio_unlock(folio); + folio_put(folio); + ext4_std_error(inode->i_sb, ret); + goto out; + } + ext4_write_lock_xattr(inode, &no_expand); + BUG_ON(!ext4_has_inline_data(inode)); + + /* + * ei->i_inline_off may have changed since + * ext4_write_begin() called + * ext4_try_to_write_inline_data() + */ + (void) ext4_find_inline_data_nolock(inode); + + kaddr = kmap_local_folio(folio, 0); + ext4_write_inline_data(inode, &iloc, kaddr, pos, copied); + kunmap_local(kaddr); + folio_mark_uptodate(folio); + /* clear dirty flag so that writepages wouldn't work for us. */ + folio_clear_dirty(folio); + + ext4_write_unlock_xattr(inode, &no_expand); + brelse(iloc.bh); + + /* + * It's important to update i_size while still holding folio + * lock: page writeout could otherwise come in and zero + * beyond i_size. + */ + ext4_update_inode_size(inode, pos + copied); + } + folio_unlock(folio); + folio_put(folio); + + /* + * Don't mark the inode dirty under folio lock. First, it unnecessarily + * makes the holding time of folio lock longer. Second, it forces lock + * ordering of folio lock and transaction start for journaling + * filesystems. + */ + if (likely(copied)) + mark_inode_dirty(inode); +out: + /* + * If we didn't copy as much data as expected, we need to trim back + * size of xattr containing inline data. + */ + if (pos + len > inode->i_size && ext4_can_truncate(inode)) + ext4_orphan_add(handle, inode); + + ret2 = ext4_journal_stop(handle); + if (!ret) + ret = ret2; + if (pos + len > inode->i_size) { + ext4_truncate_failed_write(inode); + /* + * If truncate failed early the inode might still be + * on the orphan list; we need to make sure the inode + * is removed from the orphan list in that case. + */ + if (inode->i_nlink) + ext4_orphan_del(NULL, inode); + } + return ret ? ret : copied; +} + +/* + * Try to make the page cache and handle ready for the inline data case. + * We can call this function in 2 cases: + * 1. The inode is created and the first write exceeds inline size. We can + * clear the inode state safely. + * 2. The inode has inline data, then we need to read the data, make it + * update and dirty so that ext4_da_writepages can handle it. We don't + * need to start the journal since the file's metadata isn't changed now. + */ +static int ext4_da_convert_inline_data_to_extent(struct address_space *mapping, + struct inode *inode, + void **fsdata) +{ + int ret = 0, inline_size; + struct folio *folio; + + folio = __filemap_get_folio(mapping, 0, FGP_WRITEBEGIN, + mapping_gfp_mask(mapping)); + if (IS_ERR(folio)) + return PTR_ERR(folio); + + down_read(&EXT4_I(inode)->xattr_sem); + if (!ext4_has_inline_data(inode)) { + ext4_clear_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA); + goto out; + } + + inline_size = ext4_get_inline_size(inode); + + if (!folio_test_uptodate(folio)) { + ret = ext4_read_inline_folio(inode, folio); + if (ret < 0) + goto out; + } + + ret = ext4_block_write_begin(NULL, folio, 0, inline_size, + ext4_da_get_block_prep); + if (ret) { + up_read(&EXT4_I(inode)->xattr_sem); + folio_unlock(folio); + folio_put(folio); + ext4_truncate_failed_write(inode); + return ret; + } + + clear_buffer_new(folio_buffers(folio)); + folio_mark_dirty(folio); + folio_mark_uptodate(folio); + ext4_clear_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA); + *fsdata = (void *)CONVERT_INLINE_DATA; + +out: + up_read(&EXT4_I(inode)->xattr_sem); + if (folio) { + folio_unlock(folio); + folio_put(folio); + } + return ret; +} + +#ifdef INLINE_DIR_DEBUG +void ext4_show_inline_dir(struct inode *dir, struct buffer_head *bh, + void *inline_start, int inline_size) +{ + int offset; + unsigned short de_len; + struct ext4_dir_entry_2 *de = inline_start; + void *dlimit = inline_start + inline_size; + + trace_printk("inode %lu\n", dir->i_ino); + offset = 0; + while ((void *)de < dlimit) { + de_len = ext4_rec_len_from_disk(de->rec_len, inline_size); + trace_printk("de: off %u rlen %u name %.*s nlen %u ino %u\n", + offset, de_len, de->name_len, de->name, + de->name_len, le32_to_cpu(de->inode)); + if (ext4_check_dir_entry(dir, NULL, de, bh, + inline_start, inline_size, offset)) + BUG(); + + offset += de_len; + de = (struct ext4_dir_entry_2 *) ((char *) de + de_len); + } +} +#else +#define ext4_show_inline_dir(dir, bh, inline_start, inline_size) +#endif + +/* + * Add a new entry into a inline dir. + * It will return -ENOSPC if no space is available, and -EIO + * and -EEXIST if directory entry already exists. + */ +static int ext4_add_dirent_to_inline(handle_t *handle, + struct ext4_filename *fname, + struct inode *dir, + struct inode *inode, + struct ext4_iloc *iloc, + void *inline_start, int inline_size) +{ + int err; + struct ext4_dir_entry_2 *de; + + err = ext4_find_dest_de(dir, iloc->bh, inline_start, + inline_size, fname, &de); + if (err) + return err; + + BUFFER_TRACE(iloc->bh, "get_write_access"); + err = ext4_journal_get_write_access(handle, dir->i_sb, iloc->bh, + EXT4_JTR_NONE); + if (err) + return err; + ext4_insert_dentry(dir, inode, de, inline_size, fname); + + ext4_show_inline_dir(dir, iloc->bh, inline_start, inline_size); + + /* + * XXX shouldn't update any times until successful + * completion of syscall, but too many callers depend + * on this. + * + * XXX similarly, too many callers depend on + * ext4_new_inode() setting the times, but error + * recovery deletes the inode, so the worst that can + * happen is that the times are slightly out of date + * and/or different from the directory change time. + */ + inode_set_mtime_to_ts(dir, inode_set_ctime_current(dir)); + ext4_update_dx_flag(dir); + inode_inc_iversion(dir); + return 1; +} + +static void *ext4_get_inline_xattr_pos(struct inode *inode, + struct ext4_iloc *iloc) +{ + struct ext4_xattr_entry *entry; + struct ext4_xattr_ibody_header *header; + + BUG_ON(!EXT4_I(inode)->i_inline_off); + + header = IHDR(inode, ext4_raw_inode(iloc)); + entry = (struct ext4_xattr_entry *)((void *)ext4_raw_inode(iloc) + + EXT4_I(inode)->i_inline_off); + + return (void *)IFIRST(header) + le16_to_cpu(entry->e_value_offs); +} + +/* Set the final de to cover the whole block. */ +void ext4_update_final_de(void *de_buf, int old_size, int new_size) +{ + struct ext4_dir_entry_2 *de, *prev_de; + void *limit; + int de_len; + + de = de_buf; + if (old_size) { + limit = de_buf + old_size; + do { + prev_de = de; + de_len = ext4_rec_len_from_disk(de->rec_len, old_size); + de_buf += de_len; + de = de_buf; + } while (de_buf < limit); + + prev_de->rec_len = ext4_rec_len_to_disk(de_len + new_size - + old_size, new_size); + } else { + /* this is just created, so create an empty entry. */ + de->inode = 0; + de->rec_len = ext4_rec_len_to_disk(new_size, new_size); + } +} + +static int ext4_update_inline_dir(handle_t *handle, struct inode *dir, + struct ext4_iloc *iloc) +{ + int ret; + int old_size = EXT4_I(dir)->i_inline_size - EXT4_MIN_INLINE_DATA_SIZE; + int new_size = get_max_inline_xattr_value_size(dir, iloc); + + if (new_size - old_size <= ext4_dir_rec_len(1, NULL)) + return -ENOSPC; + + ret = ext4_update_inline_data(handle, dir, + new_size + EXT4_MIN_INLINE_DATA_SIZE); + if (ret) + return ret; + + ext4_update_final_de(ext4_get_inline_xattr_pos(dir, iloc), old_size, + EXT4_I(dir)->i_inline_size - + EXT4_MIN_INLINE_DATA_SIZE); + dir->i_size = EXT4_I(dir)->i_disksize = EXT4_I(dir)->i_inline_size; + return 0; +} + +static void ext4_restore_inline_data(handle_t *handle, struct inode *inode, + struct ext4_iloc *iloc, + void *buf, int inline_size) +{ + int ret; + + ret = ext4_create_inline_data(handle, inode, inline_size); + if (ret) { + ext4_msg(inode->i_sb, KERN_EMERG, + "error restoring inline_data for inode -- potential data loss! (inode %lu, error %d)", + inode->i_ino, ret); + return; + } + ext4_write_inline_data(inode, iloc, buf, 0, inline_size); + ext4_set_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA); +} + +static int ext4_convert_inline_data_nolock(handle_t *handle, + struct inode *inode, + struct ext4_iloc *iloc) +{ + int error; + void *buf = NULL; + struct buffer_head *data_bh = NULL; + struct ext4_map_blocks map; + int inline_size; + + inline_size = ext4_get_inline_size(inode); + buf = kmalloc(inline_size, GFP_NOFS); + if (!buf) { + error = -ENOMEM; + goto out; + } + + error = ext4_read_inline_data(inode, buf, inline_size, iloc); + if (error < 0) + goto out; + + /* + * Make sure the inline directory entries pass checks before we try to + * convert them, so that we avoid touching stuff that needs fsck. + */ + if (S_ISDIR(inode->i_mode)) { + error = ext4_check_all_de(inode, iloc->bh, + buf + EXT4_INLINE_DOTDOT_SIZE, + inline_size - EXT4_INLINE_DOTDOT_SIZE); + if (error) + goto out; + } + + error = ext4_destroy_inline_data_nolock(handle, inode); + if (error) + goto out; + + map.m_lblk = 0; + map.m_len = 1; + map.m_flags = 0; + error = ext4_map_blocks(handle, inode, &map, EXT4_GET_BLOCKS_CREATE); + if (error < 0) + goto out_restore; + if (!(map.m_flags & EXT4_MAP_MAPPED)) { + error = -EIO; + goto out_restore; + } + + data_bh = sb_getblk(inode->i_sb, map.m_pblk); + if (!data_bh) { + error = -ENOMEM; + goto out_restore; + } + + lock_buffer(data_bh); + error = ext4_journal_get_create_access(handle, inode->i_sb, data_bh, + EXT4_JTR_NONE); + if (error) { + unlock_buffer(data_bh); + error = -EIO; + goto out_restore; + } + memset(data_bh->b_data, 0, inode->i_sb->s_blocksize); + + if (!S_ISDIR(inode->i_mode)) { + memcpy(data_bh->b_data, buf, inline_size); + set_buffer_uptodate(data_bh); + unlock_buffer(data_bh); + error = ext4_handle_dirty_metadata(handle, + inode, data_bh); + } else { + unlock_buffer(data_bh); + inode->i_size = inode->i_sb->s_blocksize; + i_size_write(inode, inode->i_sb->s_blocksize); + EXT4_I(inode)->i_disksize = inode->i_sb->s_blocksize; + + error = ext4_init_dirblock(handle, inode, data_bh, + le32_to_cpu(((struct ext4_dir_entry_2 *)buf)->inode), + buf + EXT4_INLINE_DOTDOT_SIZE, + inline_size - EXT4_INLINE_DOTDOT_SIZE); + if (!error) + error = ext4_mark_inode_dirty(handle, inode); + } + +out_restore: + if (error) + ext4_restore_inline_data(handle, inode, iloc, buf, inline_size); + +out: + brelse(data_bh); + kfree(buf); + return error; +} + +/* + * Try to add the new entry to the inline data. + * If succeeds, return 0. If not, extended the inline dir and copied data to + * the new created block. + */ +int ext4_try_add_inline_entry(handle_t *handle, struct ext4_filename *fname, + struct inode *dir, struct inode *inode) +{ + int ret, ret2, inline_size, no_expand; + void *inline_start; + struct ext4_iloc iloc; + + ret = ext4_get_inode_loc(dir, &iloc); + if (ret) + return ret; + + ext4_write_lock_xattr(dir, &no_expand); + if (!ext4_has_inline_data(dir)) + goto out; + + inline_start = (void *)ext4_raw_inode(&iloc)->i_block + + EXT4_INLINE_DOTDOT_SIZE; + inline_size = EXT4_MIN_INLINE_DATA_SIZE - EXT4_INLINE_DOTDOT_SIZE; + + ret = ext4_add_dirent_to_inline(handle, fname, dir, inode, &iloc, + inline_start, inline_size); + if (ret != -ENOSPC) + goto out; + + /* check whether it can be inserted to inline xattr space. */ + inline_size = EXT4_I(dir)->i_inline_size - + EXT4_MIN_INLINE_DATA_SIZE; + if (!inline_size) { + /* Try to use the xattr space.*/ + ret = ext4_update_inline_dir(handle, dir, &iloc); + if (ret && ret != -ENOSPC) + goto out; + + inline_size = EXT4_I(dir)->i_inline_size - + EXT4_MIN_INLINE_DATA_SIZE; + } + + if (inline_size) { + inline_start = ext4_get_inline_xattr_pos(dir, &iloc); + + ret = ext4_add_dirent_to_inline(handle, fname, dir, + inode, &iloc, inline_start, + inline_size); + + if (ret != -ENOSPC) + goto out; + } + + /* + * The inline space is filled up, so create a new block for it. + * As the extent tree will be created, we have to save the inline + * dir first. + */ + ret = ext4_convert_inline_data_nolock(handle, dir, &iloc); + +out: + ext4_write_unlock_xattr(dir, &no_expand); + ret2 = ext4_mark_inode_dirty(handle, dir); + if (unlikely(ret2 && !ret)) + ret = ret2; + brelse(iloc.bh); + return ret; +} + +/* + * This function fills a red-black tree with information from an + * inlined dir. It returns the number directory entries loaded + * into the tree. If there is an error it is returned in err. + */ +int ext4_inlinedir_to_tree(struct file *dir_file, + struct inode *dir, ext4_lblk_t block, + struct dx_hash_info *hinfo, + __u32 start_hash, __u32 start_minor_hash, + int *has_inline_data) +{ + int err = 0, count = 0; + unsigned int parent_ino; + int pos; + struct ext4_dir_entry_2 *de; + struct inode *inode = file_inode(dir_file); + int ret, inline_size = 0; + struct ext4_iloc iloc; + void *dir_buf = NULL; + struct ext4_dir_entry_2 fake; + struct fscrypt_str tmp_str; + + ret = ext4_get_inode_loc(inode, &iloc); + if (ret) + return ret; + + down_read(&EXT4_I(inode)->xattr_sem); + if (!ext4_has_inline_data(inode)) { + up_read(&EXT4_I(inode)->xattr_sem); + *has_inline_data = 0; + goto out; + } + + inline_size = ext4_get_inline_size(inode); + dir_buf = kmalloc(inline_size, GFP_NOFS); + if (!dir_buf) { + ret = -ENOMEM; + up_read(&EXT4_I(inode)->xattr_sem); + goto out; + } + + ret = ext4_read_inline_data(inode, dir_buf, inline_size, &iloc); + up_read(&EXT4_I(inode)->xattr_sem); + if (ret < 0) + goto out; + + pos = 0; + parent_ino = le32_to_cpu(((struct ext4_dir_entry_2 *)dir_buf)->inode); + while (pos < inline_size) { + /* + * As inlined dir doesn't store any information about '.' and + * only the inode number of '..' is stored, we have to handle + * them differently. + */ + if (pos == 0) { + fake.inode = cpu_to_le32(inode->i_ino); + fake.name_len = 1; + memcpy(fake.name, ".", 2); + fake.rec_len = ext4_rec_len_to_disk( + ext4_dir_rec_len(fake.name_len, NULL), + inline_size); + ext4_set_de_type(inode->i_sb, &fake, S_IFDIR); + de = &fake; + pos = EXT4_INLINE_DOTDOT_OFFSET; + } else if (pos == EXT4_INLINE_DOTDOT_OFFSET) { + fake.inode = cpu_to_le32(parent_ino); + fake.name_len = 2; + memcpy(fake.name, "..", 3); + fake.rec_len = ext4_rec_len_to_disk( + ext4_dir_rec_len(fake.name_len, NULL), + inline_size); + ext4_set_de_type(inode->i_sb, &fake, S_IFDIR); + de = &fake; + pos = EXT4_INLINE_DOTDOT_SIZE; + } else { + de = (struct ext4_dir_entry_2 *)(dir_buf + pos); + pos += ext4_rec_len_from_disk(de->rec_len, inline_size); + if (ext4_check_dir_entry(inode, dir_file, de, + iloc.bh, dir_buf, + inline_size, pos)) { + ret = count; + goto out; + } + } + + if (ext4_hash_in_dirent(dir)) { + hinfo->hash = EXT4_DIRENT_HASH(de); + hinfo->minor_hash = EXT4_DIRENT_MINOR_HASH(de); + } else { + err = ext4fs_dirhash(dir, de->name, de->name_len, hinfo); + if (err) { + ret = err; + goto out; + } + } + if ((hinfo->hash < start_hash) || + ((hinfo->hash == start_hash) && + (hinfo->minor_hash < start_minor_hash))) + continue; + if (de->inode == 0) + continue; + tmp_str.name = de->name; + tmp_str.len = de->name_len; + err = ext4_htree_store_dirent(dir_file, hinfo->hash, + hinfo->minor_hash, de, &tmp_str); + if (err) { + ret = err; + goto out; + } + count++; + } + ret = count; +out: + kfree(dir_buf); + brelse(iloc.bh); + return ret; +} + +/* + * So this function is called when the volume is mkfsed with + * dir_index disabled. In order to keep f_pos persistent + * after we convert from an inlined dir to a blocked based, + * we just pretend that we are a normal dir and return the + * offset as if '.' and '..' really take place. + * + */ +int ext4_read_inline_dir(struct file *file, + struct dir_context *ctx, + int *has_inline_data) +{ + unsigned int offset, parent_ino; + int i; + struct ext4_dir_entry_2 *de; + struct super_block *sb; + struct inode *inode = file_inode(file); + int ret, inline_size = 0; + struct ext4_iloc iloc; + void *dir_buf = NULL; + int dotdot_offset, dotdot_size, extra_offset, extra_size; + struct dir_private_info *info = file->private_data; + + ret = ext4_get_inode_loc(inode, &iloc); + if (ret) + return ret; + + down_read(&EXT4_I(inode)->xattr_sem); + if (!ext4_has_inline_data(inode)) { + up_read(&EXT4_I(inode)->xattr_sem); + *has_inline_data = 0; + goto out; + } + + inline_size = ext4_get_inline_size(inode); + dir_buf = kmalloc(inline_size, GFP_NOFS); + if (!dir_buf) { + ret = -ENOMEM; + up_read(&EXT4_I(inode)->xattr_sem); + goto out; + } + + ret = ext4_read_inline_data(inode, dir_buf, inline_size, &iloc); + up_read(&EXT4_I(inode)->xattr_sem); + if (ret < 0) + goto out; + + ret = 0; + sb = inode->i_sb; + parent_ino = le32_to_cpu(((struct ext4_dir_entry_2 *)dir_buf)->inode); + offset = ctx->pos; + + /* + * dotdot_offset and dotdot_size is the real offset and + * size for ".." and "." if the dir is block based while + * the real size for them are only EXT4_INLINE_DOTDOT_SIZE. + * So we will use extra_offset and extra_size to indicate them + * during the inline dir iteration. + */ + dotdot_offset = ext4_dir_rec_len(1, NULL); + dotdot_size = dotdot_offset + ext4_dir_rec_len(2, NULL); + extra_offset = dotdot_size - EXT4_INLINE_DOTDOT_SIZE; + extra_size = extra_offset + inline_size; + + /* + * If the cookie has changed since the last call to + * readdir(2), then we might be pointing to an invalid + * dirent right now. Scan from the start of the inline + * dir to make sure. + */ + if (!inode_eq_iversion(inode, info->cookie)) { + for (i = 0; i < extra_size && i < offset;) { + /* + * "." is with offset 0 and + * ".." is dotdot_offset. + */ + if (!i) { + i = dotdot_offset; + continue; + } else if (i == dotdot_offset) { + i = dotdot_size; + continue; + } + /* for other entry, the real offset in + * the buf has to be tuned accordingly. + */ + de = (struct ext4_dir_entry_2 *) + (dir_buf + i - extra_offset); + /* It's too expensive to do a full + * dirent test each time round this + * loop, but we do have to test at + * least that it is non-zero. A + * failure will be detected in the + * dirent test below. */ + if (ext4_rec_len_from_disk(de->rec_len, extra_size) + < ext4_dir_rec_len(1, NULL)) + break; + i += ext4_rec_len_from_disk(de->rec_len, + extra_size); + } + offset = i; + ctx->pos = offset; + info->cookie = inode_query_iversion(inode); + } + + while (ctx->pos < extra_size) { + if (ctx->pos == 0) { + if (!dir_emit(ctx, ".", 1, inode->i_ino, DT_DIR)) + goto out; + ctx->pos = dotdot_offset; + continue; + } + + if (ctx->pos == dotdot_offset) { + if (!dir_emit(ctx, "..", 2, parent_ino, DT_DIR)) + goto out; + ctx->pos = dotdot_size; + continue; + } + + de = (struct ext4_dir_entry_2 *) + (dir_buf + ctx->pos - extra_offset); + if (ext4_check_dir_entry(inode, file, de, iloc.bh, dir_buf, + extra_size, ctx->pos)) + goto out; + if (le32_to_cpu(de->inode)) { + if (!dir_emit(ctx, de->name, de->name_len, + le32_to_cpu(de->inode), + get_dtype(sb, de->file_type))) + goto out; + } + ctx->pos += ext4_rec_len_from_disk(de->rec_len, extra_size); + } +out: + kfree(dir_buf); + brelse(iloc.bh); + return ret; +} + +void *ext4_read_inline_link(struct inode *inode) +{ + struct ext4_iloc iloc; + int ret, inline_size; + void *link; + + ret = ext4_get_inode_loc(inode, &iloc); + if (ret) + return ERR_PTR(ret); + + ret = -ENOMEM; + inline_size = ext4_get_inline_size(inode); + link = kmalloc(inline_size + 1, GFP_NOFS); + if (!link) + goto out; + + ret = ext4_read_inline_data(inode, link, inline_size, &iloc); + if (ret < 0) { + kfree(link); + goto out; + } + nd_terminate_link(link, inode->i_size, ret); +out: + if (ret < 0) + link = ERR_PTR(ret); + brelse(iloc.bh); + return link; +} + +struct buffer_head *ext4_get_first_inline_block(struct inode *inode, + struct ext4_dir_entry_2 **parent_de, + int *retval) +{ + struct ext4_iloc iloc; + + *retval = ext4_get_inode_loc(inode, &iloc); + if (*retval) + return NULL; + + *parent_de = (struct ext4_dir_entry_2 *)ext4_raw_inode(&iloc)->i_block; + + return iloc.bh; +} + +/* + * Try to create the inline data for the new dir. + * If it succeeds, return 0, otherwise return the error. + * In case of ENOSPC, the caller should create the normal disk layout dir. + */ +int ext4_try_create_inline_dir(handle_t *handle, struct inode *parent, + struct inode *inode) +{ + int ret, inline_size = EXT4_MIN_INLINE_DATA_SIZE; + struct ext4_iloc iloc; + struct ext4_dir_entry_2 *de; + + ret = ext4_get_inode_loc(inode, &iloc); + if (ret) + return ret; + + ret = ext4_prepare_inline_data(handle, inode, inline_size); + if (ret) + goto out; + + /* + * For inline dir, we only save the inode information for the ".." + * and create a fake dentry to cover the left space. + */ + de = (struct ext4_dir_entry_2 *)ext4_raw_inode(&iloc)->i_block; + de->inode = cpu_to_le32(parent->i_ino); + de = (struct ext4_dir_entry_2 *)((void *)de + EXT4_INLINE_DOTDOT_SIZE); + de->inode = 0; + de->rec_len = ext4_rec_len_to_disk( + inline_size - EXT4_INLINE_DOTDOT_SIZE, + inline_size); + set_nlink(inode, 2); + inode->i_size = EXT4_I(inode)->i_disksize = inline_size; +out: + brelse(iloc.bh); + return ret; +} + +struct buffer_head *ext4_find_inline_entry(struct inode *dir, + struct ext4_filename *fname, + struct ext4_dir_entry_2 **res_dir, + int *has_inline_data) +{ + struct ext4_xattr_ibody_find is = { + .s = { .not_found = -ENODATA, }, + }; + struct ext4_xattr_info i = { + .name_index = EXT4_XATTR_INDEX_SYSTEM, + .name = EXT4_XATTR_SYSTEM_DATA, + }; + int ret; + void *inline_start; + int inline_size; + + ret = ext4_get_inode_loc(dir, &is.iloc); + if (ret) + return ERR_PTR(ret); + + down_read(&EXT4_I(dir)->xattr_sem); + + ret = ext4_xattr_ibody_find(dir, &i, &is); + if (ret) + goto out; + + if (!ext4_has_inline_data(dir)) { + *has_inline_data = 0; + goto out; + } + + inline_start = (void *)ext4_raw_inode(&is.iloc)->i_block + + EXT4_INLINE_DOTDOT_SIZE; + inline_size = EXT4_MIN_INLINE_DATA_SIZE - EXT4_INLINE_DOTDOT_SIZE; + ret = ext4_search_dir(is.iloc.bh, inline_start, inline_size, + dir, fname, 0, res_dir); + if (ret == 1) + goto out_find; + if (ret < 0) + goto out; + + if (ext4_get_inline_size(dir) == EXT4_MIN_INLINE_DATA_SIZE) + goto out; + + inline_start = ext4_get_inline_xattr_pos(dir, &is.iloc); + inline_size = ext4_get_inline_size(dir) - EXT4_MIN_INLINE_DATA_SIZE; + + ret = ext4_search_dir(is.iloc.bh, inline_start, inline_size, + dir, fname, 0, res_dir); + if (ret == 1) + goto out_find; + +out: + brelse(is.iloc.bh); + if (ret < 0) + is.iloc.bh = ERR_PTR(ret); + else + is.iloc.bh = NULL; +out_find: + up_read(&EXT4_I(dir)->xattr_sem); + return is.iloc.bh; +} + +int ext4_delete_inline_entry(handle_t *handle, + struct inode *dir, + struct ext4_dir_entry_2 *de_del, + struct buffer_head *bh, + int *has_inline_data) +{ + int err, inline_size, no_expand; + struct ext4_iloc iloc; + void *inline_start; + + err = ext4_get_inode_loc(dir, &iloc); + if (err) + return err; + + ext4_write_lock_xattr(dir, &no_expand); + if (!ext4_has_inline_data(dir)) { + *has_inline_data = 0; + goto out; + } + + if ((void *)de_del - ((void *)ext4_raw_inode(&iloc)->i_block) < + EXT4_MIN_INLINE_DATA_SIZE) { + inline_start = (void *)ext4_raw_inode(&iloc)->i_block + + EXT4_INLINE_DOTDOT_SIZE; + inline_size = EXT4_MIN_INLINE_DATA_SIZE - + EXT4_INLINE_DOTDOT_SIZE; + } else { + inline_start = ext4_get_inline_xattr_pos(dir, &iloc); + inline_size = ext4_get_inline_size(dir) - + EXT4_MIN_INLINE_DATA_SIZE; + } + + BUFFER_TRACE(bh, "get_write_access"); + err = ext4_journal_get_write_access(handle, dir->i_sb, bh, + EXT4_JTR_NONE); + if (err) + goto out; + + err = ext4_generic_delete_entry(dir, de_del, bh, + inline_start, inline_size, 0); + if (err) + goto out; + + ext4_show_inline_dir(dir, iloc.bh, inline_start, inline_size); +out: + ext4_write_unlock_xattr(dir, &no_expand); + if (likely(err == 0)) + err = ext4_mark_inode_dirty(handle, dir); + brelse(iloc.bh); + if (err != -ENOENT) + ext4_std_error(dir->i_sb, err); + return err; +} + +/* + * Get the inline dentry at offset. + */ +static inline struct ext4_dir_entry_2 * +ext4_get_inline_entry(struct inode *inode, + struct ext4_iloc *iloc, + unsigned int offset, + void **inline_start, + int *inline_size) +{ + void *inline_pos; + + BUG_ON(offset > ext4_get_inline_size(inode)); + + if (offset < EXT4_MIN_INLINE_DATA_SIZE) { + inline_pos = (void *)ext4_raw_inode(iloc)->i_block; + *inline_size = EXT4_MIN_INLINE_DATA_SIZE; + } else { + inline_pos = ext4_get_inline_xattr_pos(inode, iloc); + offset -= EXT4_MIN_INLINE_DATA_SIZE; + *inline_size = ext4_get_inline_size(inode) - + EXT4_MIN_INLINE_DATA_SIZE; + } + + if (inline_start) + *inline_start = inline_pos; + return (struct ext4_dir_entry_2 *)(inline_pos + offset); +} + +bool empty_inline_dir(struct inode *dir, int *has_inline_data) +{ + int err, inline_size; + struct ext4_iloc iloc; + size_t inline_len; + void *inline_pos; + unsigned int offset; + struct ext4_dir_entry_2 *de; + bool ret = false; + + err = ext4_get_inode_loc(dir, &iloc); + if (err) { + EXT4_ERROR_INODE_ERR(dir, -err, + "error %d getting inode %lu block", + err, dir->i_ino); + return false; + } + + down_read(&EXT4_I(dir)->xattr_sem); + if (!ext4_has_inline_data(dir)) { + *has_inline_data = 0; + ret = true; + goto out; + } + + de = (struct ext4_dir_entry_2 *)ext4_raw_inode(&iloc)->i_block; + if (!le32_to_cpu(de->inode)) { + ext4_warning(dir->i_sb, + "bad inline directory (dir #%lu) - no `..'", + dir->i_ino); + goto out; + } + + inline_len = ext4_get_inline_size(dir); + offset = EXT4_INLINE_DOTDOT_SIZE; + while (offset < inline_len) { + de = ext4_get_inline_entry(dir, &iloc, offset, + &inline_pos, &inline_size); + if (ext4_check_dir_entry(dir, NULL, de, + iloc.bh, inline_pos, + inline_size, offset)) { + ext4_warning(dir->i_sb, + "bad inline directory (dir #%lu) - " + "inode %u, rec_len %u, name_len %d" + "inline size %d", + dir->i_ino, le32_to_cpu(de->inode), + le16_to_cpu(de->rec_len), de->name_len, + inline_size); + goto out; + } + if (le32_to_cpu(de->inode)) { + goto out; + } + offset += ext4_rec_len_from_disk(de->rec_len, inline_size); + } + + ret = true; +out: + up_read(&EXT4_I(dir)->xattr_sem); + brelse(iloc.bh); + return ret; +} + +int ext4_destroy_inline_data(handle_t *handle, struct inode *inode) +{ + int ret, no_expand; + + ext4_write_lock_xattr(inode, &no_expand); + ret = ext4_destroy_inline_data_nolock(handle, inode); + ext4_write_unlock_xattr(inode, &no_expand); + + return ret; +} + +int ext4_inline_data_iomap(struct inode *inode, struct iomap *iomap) +{ + __u64 addr; + int error = -EAGAIN; + struct ext4_iloc iloc; + + down_read(&EXT4_I(inode)->xattr_sem); + if (!ext4_has_inline_data(inode)) + goto out; + + error = ext4_get_inode_loc(inode, &iloc); + if (error) + goto out; + + addr = (__u64)iloc.bh->b_blocknr << inode->i_sb->s_blocksize_bits; + addr += (char *)ext4_raw_inode(&iloc) - iloc.bh->b_data; + addr += offsetof(struct ext4_inode, i_block); + + brelse(iloc.bh); + + iomap->addr = addr; + iomap->offset = 0; + iomap->length = min_t(loff_t, ext4_get_inline_size(inode), + i_size_read(inode)); + iomap->type = IOMAP_INLINE; + iomap->flags = 0; + +out: + up_read(&EXT4_I(inode)->xattr_sem); + return error; +} + +int ext4_inline_data_truncate(struct inode *inode, int *has_inline) +{ + handle_t *handle; + int inline_size, value_len, needed_blocks, no_expand, err = 0; + size_t i_size; + void *value = NULL; + struct ext4_xattr_ibody_find is = { + .s = { .not_found = -ENODATA, }, + }; + struct ext4_xattr_info i = { + .name_index = EXT4_XATTR_INDEX_SYSTEM, + .name = EXT4_XATTR_SYSTEM_DATA, + }; + + + needed_blocks = ext4_chunk_trans_extent(inode, 1); + handle = ext4_journal_start(inode, EXT4_HT_INODE, needed_blocks); + if (IS_ERR(handle)) + return PTR_ERR(handle); + + ext4_write_lock_xattr(inode, &no_expand); + if (!ext4_has_inline_data(inode)) { + ext4_write_unlock_xattr(inode, &no_expand); + *has_inline = 0; + ext4_journal_stop(handle); + return 0; + } + + if ((err = ext4_orphan_add(handle, inode)) != 0) + goto out; + + if ((err = ext4_get_inode_loc(inode, &is.iloc)) != 0) + goto out; + + down_write(&EXT4_I(inode)->i_data_sem); + i_size = inode->i_size; + inline_size = ext4_get_inline_size(inode); + EXT4_I(inode)->i_disksize = i_size; + + if (i_size < inline_size) { + /* + * if there's inline data to truncate and this file was + * converted to extents after that inline data was written, + * the extent status cache must be cleared to avoid leaving + * behind stale delayed allocated extent entries + */ + if (!ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA)) + ext4_es_remove_extent(inode, 0, EXT_MAX_BLOCKS); + + /* Clear the content in the xattr space. */ + if (inline_size > EXT4_MIN_INLINE_DATA_SIZE) { + if ((err = ext4_xattr_ibody_find(inode, &i, &is)) != 0) + goto out_error; + + if (is.s.not_found) { + EXT4_ERROR_INODE(inode, + "missing inline data xattr"); + err = -EFSCORRUPTED; + goto out_error; + } + + value_len = le32_to_cpu(is.s.here->e_value_size); + value = kmalloc(value_len, GFP_NOFS); + if (!value) { + err = -ENOMEM; + goto out_error; + } + + err = ext4_xattr_ibody_get(inode, i.name_index, + i.name, value, value_len); + if (err <= 0) + goto out_error; + + i.value = value; + i.value_len = i_size > EXT4_MIN_INLINE_DATA_SIZE ? + i_size - EXT4_MIN_INLINE_DATA_SIZE : 0; + err = ext4_xattr_ibody_set(handle, inode, &i, &is); + if (err) + goto out_error; + } + + /* Clear the content within i_blocks. */ + if (i_size < EXT4_MIN_INLINE_DATA_SIZE) { + void *p = (void *) ext4_raw_inode(&is.iloc)->i_block; + memset(p + i_size, 0, + EXT4_MIN_INLINE_DATA_SIZE - i_size); + } + + EXT4_I(inode)->i_inline_size = i_size < + EXT4_MIN_INLINE_DATA_SIZE ? + EXT4_MIN_INLINE_DATA_SIZE : i_size; + } + +out_error: + up_write(&EXT4_I(inode)->i_data_sem); +out: + brelse(is.iloc.bh); + ext4_write_unlock_xattr(inode, &no_expand); + kfree(value); + if (inode->i_nlink) + ext4_orphan_del(handle, inode); + + if (err == 0) { + inode_set_mtime_to_ts(inode, inode_set_ctime_current(inode)); + err = ext4_mark_inode_dirty(handle, inode); + if (IS_SYNC(inode)) + ext4_handle_sync(handle); + } + ext4_journal_stop(handle); + return err; +} + +int ext4_convert_inline_data(struct inode *inode) +{ + int error, needed_blocks, no_expand; + handle_t *handle; + struct ext4_iloc iloc; + + if (!ext4_has_inline_data(inode)) { + ext4_clear_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA); + return 0; + } else if (!ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA)) { + /* + * Inode has inline data but EXT4_STATE_MAY_INLINE_DATA is + * cleared. This means we are in the middle of moving of + * inline data to delay allocated block. Just force writeout + * here to finish conversion. + */ + error = filemap_flush(inode->i_mapping); + if (error) + return error; + if (!ext4_has_inline_data(inode)) + return 0; + } + + needed_blocks = ext4_chunk_trans_extent(inode, 1); + + iloc.bh = NULL; + error = ext4_get_inode_loc(inode, &iloc); + if (error) + return error; + + handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, needed_blocks); + if (IS_ERR(handle)) { + error = PTR_ERR(handle); + goto out_free; + } + + ext4_write_lock_xattr(inode, &no_expand); + if (ext4_has_inline_data(inode)) + error = ext4_convert_inline_data_nolock(handle, inode, &iloc); + ext4_write_unlock_xattr(inode, &no_expand); + ext4_journal_stop(handle); +out_free: + brelse(iloc.bh); + return error; +} -- 2.43.0

From: Simon Glass <simon.glass@canonical.com> Copy migrate.c, mmp.c, move_extent.c, orphan.c, and page-io.c from Linux v6.18 fs/ext4 directory. - migrate: migration between indirect and extent mapping - mmp: multi-mount protection - move_extent: extent moving ioctl support - orphan: orphan inode handling - page-io: page I/O operations Co-developed-by: Claude Opus 4.5 <noreply@anthropic.com> --- fs/ext4l/migrate.c | 672 +++++++++++++++++++++++++++++++++++++++ fs/ext4l/mmp.c | 408 ++++++++++++++++++++++++ fs/ext4l/move_extent.c | 706 +++++++++++++++++++++++++++++++++++++++++ fs/ext4l/orphan.c | 659 ++++++++++++++++++++++++++++++++++++++ fs/ext4l/page-io.c | 593 ++++++++++++++++++++++++++++++++++ 5 files changed, 3038 insertions(+) create mode 100644 fs/ext4l/migrate.c create mode 100644 fs/ext4l/mmp.c create mode 100644 fs/ext4l/move_extent.c create mode 100644 fs/ext4l/orphan.c create mode 100644 fs/ext4l/page-io.c diff --git a/fs/ext4l/migrate.c b/fs/ext4l/migrate.c new file mode 100644 index 00000000000..1b0dfd963d3 --- /dev/null +++ b/fs/ext4l/migrate.c @@ -0,0 +1,672 @@ +// SPDX-License-Identifier: LGPL-2.1 +/* + * Copyright IBM Corporation, 2007 + * Author Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> + * + */ + +#include <linux/slab.h> +#include "ext4_jbd2.h" +#include "ext4_extents.h" + +/* + * The contiguous blocks details which can be + * represented by a single extent + */ +struct migrate_struct { + ext4_lblk_t first_block, last_block, curr_block; + ext4_fsblk_t first_pblock, last_pblock; +}; + +static int finish_range(handle_t *handle, struct inode *inode, + struct migrate_struct *lb) + +{ + int retval = 0, needed; + struct ext4_extent newext; + struct ext4_ext_path *path; + if (lb->first_pblock == 0) + return 0; + + /* Add the extent to temp inode*/ + newext.ee_block = cpu_to_le32(lb->first_block); + newext.ee_len = cpu_to_le16(lb->last_block - lb->first_block + 1); + ext4_ext_store_pblock(&newext, lb->first_pblock); + /* Locking only for convenience since we are operating on temp inode */ + down_write(&EXT4_I(inode)->i_data_sem); + path = ext4_find_extent(inode, lb->first_block, NULL, 0); + if (IS_ERR(path)) { + retval = PTR_ERR(path); + goto err_out; + } + + /* + * Calculate the credit needed to inserting this extent + * Since we are doing this in loop we may accumulate extra + * credit. But below we try to not accumulate too much + * of them by restarting the journal. + */ + needed = ext4_ext_calc_credits_for_single_extent(inode, + lb->last_block - lb->first_block + 1, path); + + retval = ext4_datasem_ensure_credits(handle, inode, needed, needed, 0); + if (retval < 0) + goto err_out; + path = ext4_ext_insert_extent(handle, inode, path, &newext, 0); + if (IS_ERR(path)) + retval = PTR_ERR(path); +err_out: + up_write((&EXT4_I(inode)->i_data_sem)); + ext4_free_ext_path(path); + lb->first_pblock = 0; + return retval; +} + +static int update_extent_range(handle_t *handle, struct inode *inode, + ext4_fsblk_t pblock, struct migrate_struct *lb) +{ + int retval; + /* + * See if we can add on to the existing range (if it exists) + */ + if (lb->first_pblock && + (lb->last_pblock+1 == pblock) && + (lb->last_block+1 == lb->curr_block)) { + lb->last_pblock = pblock; + lb->last_block = lb->curr_block; + lb->curr_block++; + return 0; + } + /* + * Start a new range. + */ + retval = finish_range(handle, inode, lb); + lb->first_pblock = lb->last_pblock = pblock; + lb->first_block = lb->last_block = lb->curr_block; + lb->curr_block++; + return retval; +} + +static int update_ind_extent_range(handle_t *handle, struct inode *inode, + ext4_fsblk_t pblock, + struct migrate_struct *lb) +{ + struct buffer_head *bh; + __le32 *i_data; + int i, retval = 0; + unsigned long max_entries = inode->i_sb->s_blocksize >> 2; + + bh = ext4_sb_bread(inode->i_sb, pblock, 0); + if (IS_ERR(bh)) + return PTR_ERR(bh); + + i_data = (__le32 *)bh->b_data; + for (i = 0; i < max_entries; i++) { + if (i_data[i]) { + retval = update_extent_range(handle, inode, + le32_to_cpu(i_data[i]), lb); + if (retval) + break; + } else { + lb->curr_block++; + } + } + put_bh(bh); + return retval; + +} + +static int update_dind_extent_range(handle_t *handle, struct inode *inode, + ext4_fsblk_t pblock, + struct migrate_struct *lb) +{ + struct buffer_head *bh; + __le32 *i_data; + int i, retval = 0; + unsigned long max_entries = inode->i_sb->s_blocksize >> 2; + + bh = ext4_sb_bread(inode->i_sb, pblock, 0); + if (IS_ERR(bh)) + return PTR_ERR(bh); + + i_data = (__le32 *)bh->b_data; + for (i = 0; i < max_entries; i++) { + if (i_data[i]) { + retval = update_ind_extent_range(handle, inode, + le32_to_cpu(i_data[i]), lb); + if (retval) + break; + } else { + /* Only update the file block number */ + lb->curr_block += max_entries; + } + } + put_bh(bh); + return retval; + +} + +static int update_tind_extent_range(handle_t *handle, struct inode *inode, + ext4_fsblk_t pblock, + struct migrate_struct *lb) +{ + struct buffer_head *bh; + __le32 *i_data; + int i, retval = 0; + unsigned long max_entries = inode->i_sb->s_blocksize >> 2; + + bh = ext4_sb_bread(inode->i_sb, pblock, 0); + if (IS_ERR(bh)) + return PTR_ERR(bh); + + i_data = (__le32 *)bh->b_data; + for (i = 0; i < max_entries; i++) { + if (i_data[i]) { + retval = update_dind_extent_range(handle, inode, + le32_to_cpu(i_data[i]), lb); + if (retval) + break; + } else { + /* Only update the file block number */ + lb->curr_block += max_entries * max_entries; + } + } + put_bh(bh); + return retval; + +} + +static int free_dind_blocks(handle_t *handle, + struct inode *inode, __le32 i_data) +{ + int i; + __le32 *tmp_idata; + struct buffer_head *bh; + struct super_block *sb = inode->i_sb; + unsigned long max_entries = inode->i_sb->s_blocksize >> 2; + int err; + + bh = ext4_sb_bread(sb, le32_to_cpu(i_data), 0); + if (IS_ERR(bh)) + return PTR_ERR(bh); + + tmp_idata = (__le32 *)bh->b_data; + for (i = 0; i < max_entries; i++) { + if (tmp_idata[i]) { + err = ext4_journal_ensure_credits(handle, + EXT4_RESERVE_TRANS_BLOCKS, + ext4_free_metadata_revoke_credits(sb, 1)); + if (err < 0) { + put_bh(bh); + return err; + } + ext4_free_blocks(handle, inode, NULL, + le32_to_cpu(tmp_idata[i]), 1, + EXT4_FREE_BLOCKS_METADATA | + EXT4_FREE_BLOCKS_FORGET); + } + } + put_bh(bh); + err = ext4_journal_ensure_credits(handle, EXT4_RESERVE_TRANS_BLOCKS, + ext4_free_metadata_revoke_credits(sb, 1)); + if (err < 0) + return err; + ext4_free_blocks(handle, inode, NULL, le32_to_cpu(i_data), 1, + EXT4_FREE_BLOCKS_METADATA | + EXT4_FREE_BLOCKS_FORGET); + return 0; +} + +static int free_tind_blocks(handle_t *handle, + struct inode *inode, __le32 i_data) +{ + int i, retval = 0; + __le32 *tmp_idata; + struct buffer_head *bh; + unsigned long max_entries = inode->i_sb->s_blocksize >> 2; + + bh = ext4_sb_bread(inode->i_sb, le32_to_cpu(i_data), 0); + if (IS_ERR(bh)) + return PTR_ERR(bh); + + tmp_idata = (__le32 *)bh->b_data; + for (i = 0; i < max_entries; i++) { + if (tmp_idata[i]) { + retval = free_dind_blocks(handle, + inode, tmp_idata[i]); + if (retval) { + put_bh(bh); + return retval; + } + } + } + put_bh(bh); + retval = ext4_journal_ensure_credits(handle, EXT4_RESERVE_TRANS_BLOCKS, + ext4_free_metadata_revoke_credits(inode->i_sb, 1)); + if (retval < 0) + return retval; + ext4_free_blocks(handle, inode, NULL, le32_to_cpu(i_data), 1, + EXT4_FREE_BLOCKS_METADATA | + EXT4_FREE_BLOCKS_FORGET); + return 0; +} + +static int free_ind_block(handle_t *handle, struct inode *inode, __le32 *i_data) +{ + int retval; + + /* ei->i_data[EXT4_IND_BLOCK] */ + if (i_data[0]) { + retval = ext4_journal_ensure_credits(handle, + EXT4_RESERVE_TRANS_BLOCKS, + ext4_free_metadata_revoke_credits(inode->i_sb, 1)); + if (retval < 0) + return retval; + ext4_free_blocks(handle, inode, NULL, + le32_to_cpu(i_data[0]), 1, + EXT4_FREE_BLOCKS_METADATA | + EXT4_FREE_BLOCKS_FORGET); + } + + /* ei->i_data[EXT4_DIND_BLOCK] */ + if (i_data[1]) { + retval = free_dind_blocks(handle, inode, i_data[1]); + if (retval) + return retval; + } + + /* ei->i_data[EXT4_TIND_BLOCK] */ + if (i_data[2]) { + retval = free_tind_blocks(handle, inode, i_data[2]); + if (retval) + return retval; + } + return 0; +} + +static int ext4_ext_swap_inode_data(handle_t *handle, struct inode *inode, + struct inode *tmp_inode) +{ + int retval, retval2 = 0; + __le32 i_data[3]; + struct ext4_inode_info *ei = EXT4_I(inode); + struct ext4_inode_info *tmp_ei = EXT4_I(tmp_inode); + + /* + * One credit accounted for writing the + * i_data field of the original inode + */ + retval = ext4_journal_ensure_credits(handle, 1, 0); + if (retval < 0) + goto err_out; + + i_data[0] = ei->i_data[EXT4_IND_BLOCK]; + i_data[1] = ei->i_data[EXT4_DIND_BLOCK]; + i_data[2] = ei->i_data[EXT4_TIND_BLOCK]; + + down_write(&EXT4_I(inode)->i_data_sem); + /* + * if EXT4_STATE_EXT_MIGRATE is cleared a block allocation + * happened after we started the migrate. We need to + * fail the migrate + */ + if (!ext4_test_inode_state(inode, EXT4_STATE_EXT_MIGRATE)) { + retval = -EAGAIN; + up_write(&EXT4_I(inode)->i_data_sem); + goto err_out; + } else + ext4_clear_inode_state(inode, EXT4_STATE_EXT_MIGRATE); + /* + * We have the extent map build with the tmp inode. + * Now copy the i_data across + */ + ext4_set_inode_flag(inode, EXT4_INODE_EXTENTS); + memcpy(ei->i_data, tmp_ei->i_data, sizeof(ei->i_data)); + + /* + * Update i_blocks with the new blocks that got + * allocated while adding extents for extent index + * blocks. + * + * While converting to extents we need not + * update the original inode i_blocks for extent blocks + * via quota APIs. The quota update happened via tmp_inode already. + */ + spin_lock(&inode->i_lock); + inode->i_blocks += tmp_inode->i_blocks; + spin_unlock(&inode->i_lock); + up_write(&EXT4_I(inode)->i_data_sem); + + /* + * We mark the inode dirty after, because we decrement the + * i_blocks when freeing the indirect meta-data blocks + */ + retval = free_ind_block(handle, inode, i_data); + retval2 = ext4_mark_inode_dirty(handle, inode); + if (unlikely(retval2 && !retval)) + retval = retval2; + +err_out: + return retval; +} + +static int free_ext_idx(handle_t *handle, struct inode *inode, + struct ext4_extent_idx *ix) +{ + int i, retval = 0; + ext4_fsblk_t block; + struct buffer_head *bh; + struct ext4_extent_header *eh; + + block = ext4_idx_pblock(ix); + bh = ext4_sb_bread(inode->i_sb, block, 0); + if (IS_ERR(bh)) + return PTR_ERR(bh); + + eh = (struct ext4_extent_header *)bh->b_data; + if (eh->eh_depth != 0) { + ix = EXT_FIRST_INDEX(eh); + for (i = 0; i < le16_to_cpu(eh->eh_entries); i++, ix++) { + retval = free_ext_idx(handle, inode, ix); + if (retval) { + put_bh(bh); + return retval; + } + } + } + put_bh(bh); + retval = ext4_journal_ensure_credits(handle, EXT4_RESERVE_TRANS_BLOCKS, + ext4_free_metadata_revoke_credits(inode->i_sb, 1)); + if (retval < 0) + return retval; + ext4_free_blocks(handle, inode, NULL, block, 1, + EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET); + return 0; +} + +/* + * Free the extent meta data blocks only + */ +static int free_ext_block(handle_t *handle, struct inode *inode) +{ + int i, retval = 0; + struct ext4_inode_info *ei = EXT4_I(inode); + struct ext4_extent_header *eh = (struct ext4_extent_header *)ei->i_data; + struct ext4_extent_idx *ix; + if (eh->eh_depth == 0) + /* + * No extra blocks allocated for extent meta data + */ + return 0; + ix = EXT_FIRST_INDEX(eh); + for (i = 0; i < le16_to_cpu(eh->eh_entries); i++, ix++) { + retval = free_ext_idx(handle, inode, ix); + if (retval) + return retval; + } + return retval; +} + +int ext4_ext_migrate(struct inode *inode) +{ + handle_t *handle; + int retval = 0, i; + __le32 *i_data; + struct ext4_inode_info *ei; + struct inode *tmp_inode = NULL; + struct migrate_struct lb; + unsigned long max_entries; + __u32 goal, tmp_csum_seed; + uid_t owner[2]; + int alloc_ctx; + + /* + * If the filesystem does not support extents, or the inode + * already is extent-based, error out. + */ + if (!ext4_has_feature_extents(inode->i_sb) || + ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS) || + ext4_has_inline_data(inode)) + return -EINVAL; + + if (S_ISLNK(inode->i_mode) && inode->i_blocks == 0) + /* + * don't migrate fast symlink + */ + return retval; + + alloc_ctx = ext4_writepages_down_write(inode->i_sb); + + /* + * Worst case we can touch the allocation bitmaps and a block + * group descriptor block. We do need to worry about + * credits for modifying the quota inode. + */ + handle = ext4_journal_start(inode, EXT4_HT_MIGRATE, + 3 + EXT4_MAXQUOTAS_TRANS_BLOCKS(inode->i_sb)); + + if (IS_ERR(handle)) { + retval = PTR_ERR(handle); + goto out_unlock; + } + goal = (((inode->i_ino - 1) / EXT4_INODES_PER_GROUP(inode->i_sb)) * + EXT4_INODES_PER_GROUP(inode->i_sb)) + 1; + owner[0] = i_uid_read(inode); + owner[1] = i_gid_read(inode); + tmp_inode = ext4_new_inode(handle, d_inode(inode->i_sb->s_root), + S_IFREG, NULL, goal, owner, 0); + if (IS_ERR(tmp_inode)) { + retval = PTR_ERR(tmp_inode); + ext4_journal_stop(handle); + goto out_unlock; + } + /* + * Use the correct seed for checksum (i.e. the seed from 'inode'). This + * is so that the metadata blocks will have the correct checksum after + * the migration. + */ + ei = EXT4_I(inode); + tmp_csum_seed = EXT4_I(tmp_inode)->i_csum_seed; + EXT4_I(tmp_inode)->i_csum_seed = ei->i_csum_seed; + i_size_write(tmp_inode, i_size_read(inode)); + /* + * Set the i_nlink to zero so it will be deleted later + * when we drop inode reference. + */ + clear_nlink(tmp_inode); + + ext4_ext_tree_init(handle, tmp_inode); + ext4_journal_stop(handle); + + /* + * start with one credit accounted for + * superblock modification. + * + * For the tmp_inode we already have committed the + * transaction that created the inode. Later as and + * when we add extents we extent the journal + */ + /* + * Even though we take i_rwsem we can still cause block + * allocation via mmap write to holes. If we have allocated + * new blocks we fail migrate. New block allocation will + * clear EXT4_STATE_EXT_MIGRATE flag. The flag is updated + * with i_data_sem held to prevent racing with block + * allocation. + */ + down_read(&EXT4_I(inode)->i_data_sem); + ext4_set_inode_state(inode, EXT4_STATE_EXT_MIGRATE); + up_read((&EXT4_I(inode)->i_data_sem)); + + handle = ext4_journal_start(inode, EXT4_HT_MIGRATE, 1); + if (IS_ERR(handle)) { + retval = PTR_ERR(handle); + goto out_tmp_inode; + } + + i_data = ei->i_data; + memset(&lb, 0, sizeof(lb)); + + /* 32 bit block address 4 bytes */ + max_entries = inode->i_sb->s_blocksize >> 2; + for (i = 0; i < EXT4_NDIR_BLOCKS; i++) { + if (i_data[i]) { + retval = update_extent_range(handle, tmp_inode, + le32_to_cpu(i_data[i]), &lb); + if (retval) + goto err_out; + } else + lb.curr_block++; + } + if (i_data[EXT4_IND_BLOCK]) { + retval = update_ind_extent_range(handle, tmp_inode, + le32_to_cpu(i_data[EXT4_IND_BLOCK]), &lb); + if (retval) + goto err_out; + } else + lb.curr_block += max_entries; + if (i_data[EXT4_DIND_BLOCK]) { + retval = update_dind_extent_range(handle, tmp_inode, + le32_to_cpu(i_data[EXT4_DIND_BLOCK]), &lb); + if (retval) + goto err_out; + } else + lb.curr_block += max_entries * max_entries; + if (i_data[EXT4_TIND_BLOCK]) { + retval = update_tind_extent_range(handle, tmp_inode, + le32_to_cpu(i_data[EXT4_TIND_BLOCK]), &lb); + if (retval) + goto err_out; + } + /* + * Build the last extent + */ + retval = finish_range(handle, tmp_inode, &lb); +err_out: + if (retval) + /* + * Failure case delete the extent information with the + * tmp_inode + */ + free_ext_block(handle, tmp_inode); + else { + retval = ext4_ext_swap_inode_data(handle, inode, tmp_inode); + if (retval) + /* + * if we fail to swap inode data free the extent + * details of the tmp inode + */ + free_ext_block(handle, tmp_inode); + } + + /* We mark the tmp_inode dirty via ext4_ext_tree_init. */ + retval = ext4_journal_ensure_credits(handle, 1, 0); + if (retval < 0) + goto out_stop; + /* + * Mark the tmp_inode as of size zero + */ + i_size_write(tmp_inode, 0); + + /* + * set the i_blocks count to zero + * so that the ext4_evict_inode() does the + * right job + * + * We don't need to take the i_lock because + * the inode is not visible to user space. + */ + tmp_inode->i_blocks = 0; + EXT4_I(tmp_inode)->i_csum_seed = tmp_csum_seed; + + /* Reset the extent details */ + ext4_ext_tree_init(handle, tmp_inode); +out_stop: + ext4_journal_stop(handle); +out_tmp_inode: + unlock_new_inode(tmp_inode); + iput(tmp_inode); +out_unlock: + ext4_writepages_up_write(inode->i_sb, alloc_ctx); + return retval; +} + +/* + * Migrate a simple extent-based inode to use the i_blocks[] array + */ +int ext4_ind_migrate(struct inode *inode) +{ + struct ext4_extent_header *eh; + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + struct ext4_super_block *es = sbi->s_es; + struct ext4_inode_info *ei = EXT4_I(inode); + struct ext4_extent *ex; + unsigned int i, len; + ext4_lblk_t start, end; + ext4_fsblk_t blk; + handle_t *handle; + int ret, ret2 = 0; + int alloc_ctx; + + if (!ext4_has_feature_extents(inode->i_sb) || + (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) + return -EINVAL; + + if (ext4_has_feature_bigalloc(inode->i_sb)) + return -EOPNOTSUPP; + + /* + * In order to get correct extent info, force all delayed allocation + * blocks to be allocated, otherwise delayed allocation blocks may not + * be reflected and bypass the checks on extent header. + */ + if (test_opt(inode->i_sb, DELALLOC)) + ext4_alloc_da_blocks(inode); + + alloc_ctx = ext4_writepages_down_write(inode->i_sb); + + handle = ext4_journal_start(inode, EXT4_HT_MIGRATE, 1); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + goto out_unlock; + } + + down_write(&EXT4_I(inode)->i_data_sem); + ret = ext4_ext_check_inode(inode); + if (ret) + goto errout; + + eh = ext_inode_hdr(inode); + ex = EXT_FIRST_EXTENT(eh); + if (ext4_blocks_count(es) > EXT4_MAX_BLOCK_FILE_PHYS || + eh->eh_depth != 0 || le16_to_cpu(eh->eh_entries) > 1) { + ret = -EOPNOTSUPP; + goto errout; + } + if (eh->eh_entries == 0) + blk = len = start = end = 0; + else { + len = le16_to_cpu(ex->ee_len); + blk = ext4_ext_pblock(ex); + start = le32_to_cpu(ex->ee_block); + end = start + len - 1; + if (end >= EXT4_NDIR_BLOCKS) { + ret = -EOPNOTSUPP; + goto errout; + } + } + + ext4_clear_inode_flag(inode, EXT4_INODE_EXTENTS); + memset(ei->i_data, 0, sizeof(ei->i_data)); + for (i = start; i <= end; i++) + ei->i_data[i] = cpu_to_le32(blk++); + ret2 = ext4_mark_inode_dirty(handle, inode); + if (unlikely(ret2 && !ret)) + ret = ret2; +errout: + up_write(&EXT4_I(inode)->i_data_sem); + ext4_journal_stop(handle); +out_unlock: + ext4_writepages_up_write(inode->i_sb, alloc_ctx); + return ret; +} diff --git a/fs/ext4l/mmp.c b/fs/ext4l/mmp.c new file mode 100644 index 00000000000..ab1ff51302f --- /dev/null +++ b/fs/ext4l/mmp.c @@ -0,0 +1,408 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/fs.h> +#include <linux/random.h> +#include <linux/buffer_head.h> +#include <linux/utsname.h> +#include <linux/kthread.h> + +#include "ext4.h" + +/* Checksumming functions */ +static __le32 ext4_mmp_csum(struct super_block *sb, struct mmp_struct *mmp) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + int offset = offsetof(struct mmp_struct, mmp_checksum); + __u32 csum; + + csum = ext4_chksum(sbi->s_csum_seed, (char *)mmp, offset); + + return cpu_to_le32(csum); +} + +static int ext4_mmp_csum_verify(struct super_block *sb, struct mmp_struct *mmp) +{ + if (!ext4_has_feature_metadata_csum(sb)) + return 1; + + return mmp->mmp_checksum == ext4_mmp_csum(sb, mmp); +} + +static void ext4_mmp_csum_set(struct super_block *sb, struct mmp_struct *mmp) +{ + if (!ext4_has_feature_metadata_csum(sb)) + return; + + mmp->mmp_checksum = ext4_mmp_csum(sb, mmp); +} + +/* + * Write the MMP block using REQ_SYNC to try to get the block on-disk + * faster. + */ +static int write_mmp_block_thawed(struct super_block *sb, + struct buffer_head *bh) +{ + struct mmp_struct *mmp = (struct mmp_struct *)(bh->b_data); + + ext4_mmp_csum_set(sb, mmp); + lock_buffer(bh); + bh->b_end_io = end_buffer_write_sync; + get_bh(bh); + submit_bh(REQ_OP_WRITE | REQ_SYNC | REQ_META | REQ_PRIO, bh); + wait_on_buffer(bh); + if (unlikely(!buffer_uptodate(bh))) + return -EIO; + return 0; +} + +static int write_mmp_block(struct super_block *sb, struct buffer_head *bh) +{ + int err; + + /* + * We protect against freezing so that we don't create dirty buffers + * on frozen filesystem. + */ + sb_start_write(sb); + err = write_mmp_block_thawed(sb, bh); + sb_end_write(sb); + return err; +} + +/* + * Read the MMP block. It _must_ be read from disk and hence we clear the + * uptodate flag on the buffer. + */ +static int read_mmp_block(struct super_block *sb, struct buffer_head **bh, + ext4_fsblk_t mmp_block) +{ + struct mmp_struct *mmp; + int ret; + + if (*bh) + clear_buffer_uptodate(*bh); + + /* This would be sb_bread(sb, mmp_block), except we need to be sure + * that the MD RAID device cache has been bypassed, and that the read + * is not blocked in the elevator. */ + if (!*bh) { + *bh = sb_getblk(sb, mmp_block); + if (!*bh) { + ret = -ENOMEM; + goto warn_exit; + } + } + + lock_buffer(*bh); + ret = ext4_read_bh(*bh, REQ_META | REQ_PRIO, NULL, false); + if (ret) + goto warn_exit; + + mmp = (struct mmp_struct *)((*bh)->b_data); + if (le32_to_cpu(mmp->mmp_magic) != EXT4_MMP_MAGIC) { + ret = -EFSCORRUPTED; + goto warn_exit; + } + if (!ext4_mmp_csum_verify(sb, mmp)) { + ret = -EFSBADCRC; + goto warn_exit; + } + return 0; +warn_exit: + brelse(*bh); + *bh = NULL; + ext4_warning(sb, "Error %d while reading MMP block %llu", + ret, mmp_block); + return ret; +} + +/* + * Dump as much information as possible to help the admin. + */ +void __dump_mmp_msg(struct super_block *sb, struct mmp_struct *mmp, + const char *function, unsigned int line, const char *msg) +{ + __ext4_warning(sb, function, line, "%s", msg); + __ext4_warning(sb, function, line, + "MMP failure info: last update time: %llu, last update node: %.*s, last update device: %.*s", + (unsigned long long)le64_to_cpu(mmp->mmp_time), + (int)sizeof(mmp->mmp_nodename), mmp->mmp_nodename, + (int)sizeof(mmp->mmp_bdevname), mmp->mmp_bdevname); +} + +/* + * kmmpd will update the MMP sequence every s_mmp_update_interval seconds + */ +static int kmmpd(void *data) +{ + struct super_block *sb = data; + struct ext4_super_block *es = EXT4_SB(sb)->s_es; + struct buffer_head *bh = EXT4_SB(sb)->s_mmp_bh; + struct mmp_struct *mmp; + ext4_fsblk_t mmp_block; + u32 seq = 0; + unsigned long failed_writes = 0; + int mmp_update_interval = le16_to_cpu(es->s_mmp_update_interval); + unsigned mmp_check_interval; + unsigned long last_update_time; + unsigned long diff; + int retval = 0; + + mmp_block = le64_to_cpu(es->s_mmp_block); + mmp = (struct mmp_struct *)(bh->b_data); + mmp->mmp_time = cpu_to_le64(ktime_get_real_seconds()); + /* + * Start with the higher mmp_check_interval and reduce it if + * the MMP block is being updated on time. + */ + mmp_check_interval = max(EXT4_MMP_CHECK_MULT * mmp_update_interval, + EXT4_MMP_MIN_CHECK_INTERVAL); + mmp->mmp_check_interval = cpu_to_le16(mmp_check_interval); + + memcpy(mmp->mmp_nodename, init_utsname()->nodename, + sizeof(mmp->mmp_nodename)); + + while (!kthread_should_stop() && !ext4_emergency_state(sb)) { + if (!ext4_has_feature_mmp(sb)) { + ext4_warning(sb, "kmmpd being stopped since MMP feature" + " has been disabled."); + goto wait_to_exit; + } + if (++seq > EXT4_MMP_SEQ_MAX) + seq = 1; + + mmp->mmp_seq = cpu_to_le32(seq); + mmp->mmp_time = cpu_to_le64(ktime_get_real_seconds()); + last_update_time = jiffies; + + retval = write_mmp_block(sb, bh); + /* + * Don't spew too many error messages. Print one every + * (s_mmp_update_interval * 60) seconds. + */ + if (retval) { + if ((failed_writes % 60) == 0) { + ext4_error_err(sb, -retval, + "Error writing to MMP block"); + } + failed_writes++; + } + + diff = jiffies - last_update_time; + if (diff < mmp_update_interval * HZ) + schedule_timeout_interruptible(mmp_update_interval * + HZ - diff); + + /* + * We need to make sure that more than mmp_check_interval + * seconds have not passed since writing. If that has happened + * we need to check if the MMP block is as we left it. + */ + diff = jiffies - last_update_time; + if (diff > mmp_check_interval * HZ) { + struct buffer_head *bh_check = NULL; + struct mmp_struct *mmp_check; + + retval = read_mmp_block(sb, &bh_check, mmp_block); + if (retval) { + ext4_error_err(sb, -retval, + "error reading MMP data: %d", + retval); + goto wait_to_exit; + } + + mmp_check = (struct mmp_struct *)(bh_check->b_data); + if (mmp->mmp_seq != mmp_check->mmp_seq || + memcmp(mmp->mmp_nodename, mmp_check->mmp_nodename, + sizeof(mmp->mmp_nodename))) { + dump_mmp_msg(sb, mmp_check, + "Error while updating MMP info. " + "The filesystem seems to have been" + " multiply mounted."); + ext4_error_err(sb, EBUSY, "abort"); + put_bh(bh_check); + retval = -EBUSY; + goto wait_to_exit; + } + put_bh(bh_check); + } + + /* + * Adjust the mmp_check_interval depending on how much time + * it took for the MMP block to be written. + */ + mmp_check_interval = clamp(EXT4_MMP_CHECK_MULT * diff / HZ, + EXT4_MMP_MIN_CHECK_INTERVAL, + EXT4_MMP_MAX_CHECK_INTERVAL); + mmp->mmp_check_interval = cpu_to_le16(mmp_check_interval); + } + + /* + * Unmount seems to be clean. + */ + mmp->mmp_seq = cpu_to_le32(EXT4_MMP_SEQ_CLEAN); + mmp->mmp_time = cpu_to_le64(ktime_get_real_seconds()); + + retval = write_mmp_block(sb, bh); + +wait_to_exit: + while (!kthread_should_stop()) { + set_current_state(TASK_INTERRUPTIBLE); + if (!kthread_should_stop()) + schedule(); + } + set_current_state(TASK_RUNNING); + return retval; +} + +void ext4_stop_mmpd(struct ext4_sb_info *sbi) +{ + if (sbi->s_mmp_tsk) { + kthread_stop(sbi->s_mmp_tsk); + brelse(sbi->s_mmp_bh); + sbi->s_mmp_tsk = NULL; + } +} + +/* + * Get a random new sequence number but make sure it is not greater than + * EXT4_MMP_SEQ_MAX. + */ +static unsigned int mmp_new_seq(void) +{ + return get_random_u32_below(EXT4_MMP_SEQ_MAX + 1); +} + +/* + * Protect the filesystem from being mounted more than once. + */ +int ext4_multi_mount_protect(struct super_block *sb, + ext4_fsblk_t mmp_block) +{ + struct ext4_super_block *es = EXT4_SB(sb)->s_es; + struct buffer_head *bh = NULL; + struct mmp_struct *mmp = NULL; + u32 seq; + unsigned int mmp_check_interval = le16_to_cpu(es->s_mmp_update_interval); + unsigned int wait_time = 0; + int retval; + + if (mmp_block < le32_to_cpu(es->s_first_data_block) || + mmp_block >= ext4_blocks_count(es)) { + ext4_warning(sb, "Invalid MMP block in superblock"); + retval = -EINVAL; + goto failed; + } + + retval = read_mmp_block(sb, &bh, mmp_block); + if (retval) + goto failed; + + mmp = (struct mmp_struct *)(bh->b_data); + + if (mmp_check_interval < EXT4_MMP_MIN_CHECK_INTERVAL) + mmp_check_interval = EXT4_MMP_MIN_CHECK_INTERVAL; + + /* + * If check_interval in MMP block is larger, use that instead of + * update_interval from the superblock. + */ + if (le16_to_cpu(mmp->mmp_check_interval) > mmp_check_interval) + mmp_check_interval = le16_to_cpu(mmp->mmp_check_interval); + + seq = le32_to_cpu(mmp->mmp_seq); + if (seq == EXT4_MMP_SEQ_CLEAN) + goto skip; + + if (seq == EXT4_MMP_SEQ_FSCK) { + dump_mmp_msg(sb, mmp, "fsck is running on the filesystem"); + retval = -EBUSY; + goto failed; + } + + wait_time = min(mmp_check_interval * 2 + 1, + mmp_check_interval + 60); + + /* Print MMP interval if more than 20 secs. */ + if (wait_time > EXT4_MMP_MIN_CHECK_INTERVAL * 4) + ext4_warning(sb, "MMP interval %u higher than expected, please" + " wait.\n", wait_time * 2); + + if (schedule_timeout_interruptible(HZ * wait_time) != 0) { + ext4_warning(sb, "MMP startup interrupted, failing mount\n"); + retval = -ETIMEDOUT; + goto failed; + } + + retval = read_mmp_block(sb, &bh, mmp_block); + if (retval) + goto failed; + mmp = (struct mmp_struct *)(bh->b_data); + if (seq != le32_to_cpu(mmp->mmp_seq)) { + dump_mmp_msg(sb, mmp, + "Device is already active on another node."); + retval = -EBUSY; + goto failed; + } + +skip: + /* + * write a new random sequence number. + */ + seq = mmp_new_seq(); + mmp->mmp_seq = cpu_to_le32(seq); + + /* + * On mount / remount we are protected against fs freezing (by s_umount + * semaphore) and grabbing freeze protection upsets lockdep + */ + retval = write_mmp_block_thawed(sb, bh); + if (retval) + goto failed; + + /* + * wait for MMP interval and check mmp_seq. + */ + if (schedule_timeout_interruptible(HZ * wait_time) != 0) { + ext4_warning(sb, "MMP startup interrupted, failing mount"); + retval = -ETIMEDOUT; + goto failed; + } + + retval = read_mmp_block(sb, &bh, mmp_block); + if (retval) + goto failed; + mmp = (struct mmp_struct *)(bh->b_data); + if (seq != le32_to_cpu(mmp->mmp_seq)) { + dump_mmp_msg(sb, mmp, + "Device is already active on another node."); + retval = -EBUSY; + goto failed; + } + + EXT4_SB(sb)->s_mmp_bh = bh; + + BUILD_BUG_ON(sizeof(mmp->mmp_bdevname) < BDEVNAME_SIZE); + snprintf(mmp->mmp_bdevname, sizeof(mmp->mmp_bdevname), + "%pg", bh->b_bdev); + + /* + * Start a kernel thread to update the MMP block periodically. + */ + EXT4_SB(sb)->s_mmp_tsk = kthread_run(kmmpd, sb, "kmmpd-%.*s", + (int)sizeof(mmp->mmp_bdevname), + mmp->mmp_bdevname); + if (IS_ERR(EXT4_SB(sb)->s_mmp_tsk)) { + EXT4_SB(sb)->s_mmp_tsk = NULL; + ext4_warning(sb, "Unable to create kmmpd thread for %s.", + sb->s_id); + retval = -ENOMEM; + goto failed; + } + + return 0; + +failed: + brelse(bh); + return retval; +} diff --git a/fs/ext4l/move_extent.c b/fs/ext4l/move_extent.c new file mode 100644 index 00000000000..4b091c21908 --- /dev/null +++ b/fs/ext4l/move_extent.c @@ -0,0 +1,706 @@ +// SPDX-License-Identifier: LGPL-2.1 +/* + * Copyright (c) 2008,2009 NEC Software Tohoku, Ltd. + * Written by Takashi Sato <t-sato@yk.jp.nec.com> + * Akira Fujita <a-fujita@rs.jp.nec.com> + */ + +#include <linux/fs.h> +#include <linux/quotaops.h> +#include <linux/slab.h> +#include <linux/sched/mm.h> +#include "ext4_jbd2.h" +#include "ext4.h" +#include "ext4_extents.h" + +/** + * get_ext_path() - Find an extent path for designated logical block number. + * @inode: inode to be searched + * @lblock: logical block number to find an extent path + * @path: pointer to an extent path + * + * ext4_find_extent wrapper. Return an extent path pointer on success, + * or an error pointer on failure. + */ +static inline struct ext4_ext_path * +get_ext_path(struct inode *inode, ext4_lblk_t lblock, + struct ext4_ext_path *path) +{ + path = ext4_find_extent(inode, lblock, path, EXT4_EX_NOCACHE); + if (IS_ERR(path)) + return path; + if (path[ext_depth(inode)].p_ext == NULL) { + ext4_free_ext_path(path); + return ERR_PTR(-ENODATA); + } + return path; +} + +/** + * ext4_double_down_write_data_sem() - write lock two inodes's i_data_sem + * @first: inode to be locked + * @second: inode to be locked + * + * Acquire write lock of i_data_sem of the two inodes + */ +void +ext4_double_down_write_data_sem(struct inode *first, struct inode *second) +{ + if (first < second) { + down_write(&EXT4_I(first)->i_data_sem); + down_write_nested(&EXT4_I(second)->i_data_sem, I_DATA_SEM_OTHER); + } else { + down_write(&EXT4_I(second)->i_data_sem); + down_write_nested(&EXT4_I(first)->i_data_sem, I_DATA_SEM_OTHER); + + } +} + +/** + * ext4_double_up_write_data_sem - Release two inodes' write lock of i_data_sem + * + * @orig_inode: original inode structure to be released its lock first + * @donor_inode: donor inode structure to be released its lock second + * Release write lock of i_data_sem of two inodes (orig and donor). + */ +void +ext4_double_up_write_data_sem(struct inode *orig_inode, + struct inode *donor_inode) +{ + up_write(&EXT4_I(orig_inode)->i_data_sem); + up_write(&EXT4_I(donor_inode)->i_data_sem); +} + +/** + * mext_check_coverage - Check that all extents in range has the same type + * + * @inode: inode in question + * @from: block offset of inode + * @count: block count to be checked + * @unwritten: extents expected to be unwritten + * @err: pointer to save error value + * + * Return 1 if all extents in range has expected type, and zero otherwise. + */ +static int +mext_check_coverage(struct inode *inode, ext4_lblk_t from, ext4_lblk_t count, + int unwritten, int *err) +{ + struct ext4_ext_path *path = NULL; + struct ext4_extent *ext; + int ret = 0; + ext4_lblk_t last = from + count; + while (from < last) { + path = get_ext_path(inode, from, path); + if (IS_ERR(path)) { + *err = PTR_ERR(path); + return ret; + } + ext = path[ext_depth(inode)].p_ext; + if (unwritten != ext4_ext_is_unwritten(ext)) + goto out; + from += ext4_ext_get_actual_len(ext); + } + ret = 1; +out: + ext4_free_ext_path(path); + return ret; +} + +/** + * mext_folio_double_lock - Grab and lock folio on both @inode1 and @inode2 + * + * @inode1: the inode structure + * @inode2: the inode structure + * @index1: folio index + * @index2: folio index + * @folio: result folio vector + * + * Grab two locked folio for inode's by inode order + */ +static int +mext_folio_double_lock(struct inode *inode1, struct inode *inode2, + pgoff_t index1, pgoff_t index2, struct folio *folio[2]) +{ + struct address_space *mapping[2]; + unsigned int flags; + + BUG_ON(!inode1 || !inode2); + if (inode1 < inode2) { + mapping[0] = inode1->i_mapping; + mapping[1] = inode2->i_mapping; + } else { + swap(index1, index2); + mapping[0] = inode2->i_mapping; + mapping[1] = inode1->i_mapping; + } + + flags = memalloc_nofs_save(); + folio[0] = __filemap_get_folio(mapping[0], index1, FGP_WRITEBEGIN, + mapping_gfp_mask(mapping[0])); + if (IS_ERR(folio[0])) { + memalloc_nofs_restore(flags); + return PTR_ERR(folio[0]); + } + + folio[1] = __filemap_get_folio(mapping[1], index2, FGP_WRITEBEGIN, + mapping_gfp_mask(mapping[1])); + memalloc_nofs_restore(flags); + if (IS_ERR(folio[1])) { + folio_unlock(folio[0]); + folio_put(folio[0]); + return PTR_ERR(folio[1]); + } + /* + * __filemap_get_folio() may not wait on folio's writeback if + * BDI not demand that. But it is reasonable to be very conservative + * here and explicitly wait on folio's writeback + */ + folio_wait_writeback(folio[0]); + folio_wait_writeback(folio[1]); + if (inode1 > inode2) + swap(folio[0], folio[1]); + + return 0; +} + +/* Force folio buffers uptodate w/o dropping folio's lock */ +static int mext_page_mkuptodate(struct folio *folio, size_t from, size_t to) +{ + struct inode *inode = folio->mapping->host; + sector_t block; + struct buffer_head *bh, *head; + unsigned int blocksize, block_start, block_end; + int nr = 0; + bool partial = false; + + BUG_ON(!folio_test_locked(folio)); + BUG_ON(folio_test_writeback(folio)); + + if (folio_test_uptodate(folio)) + return 0; + + blocksize = i_blocksize(inode); + head = folio_buffers(folio); + if (!head) + head = create_empty_buffers(folio, blocksize, 0); + + block = folio_pos(folio) >> inode->i_blkbits; + block_end = 0; + bh = head; + do { + block_start = block_end; + block_end = block_start + blocksize; + if (block_end <= from || block_start >= to) { + if (!buffer_uptodate(bh)) + partial = true; + continue; + } + if (buffer_uptodate(bh)) + continue; + if (!buffer_mapped(bh)) { + int err = ext4_get_block(inode, block, bh, 0); + if (err) + return err; + if (!buffer_mapped(bh)) { + folio_zero_range(folio, block_start, blocksize); + set_buffer_uptodate(bh); + continue; + } + } + lock_buffer(bh); + if (buffer_uptodate(bh)) { + unlock_buffer(bh); + continue; + } + ext4_read_bh_nowait(bh, 0, NULL, false); + nr++; + } while (block++, (bh = bh->b_this_page) != head); + + /* No io required */ + if (!nr) + goto out; + + bh = head; + do { + if (bh_offset(bh) + blocksize <= from) + continue; + if (bh_offset(bh) >= to) + break; + wait_on_buffer(bh); + if (buffer_uptodate(bh)) + continue; + return -EIO; + } while ((bh = bh->b_this_page) != head); +out: + if (!partial) + folio_mark_uptodate(folio); + return 0; +} + +/** + * move_extent_per_page - Move extent data per page + * + * @o_filp: file structure of original file + * @donor_inode: donor inode + * @orig_page_offset: page index on original file + * @donor_page_offset: page index on donor file + * @data_offset_in_page: block index where data swapping starts + * @block_len_in_page: the number of blocks to be swapped + * @unwritten: orig extent is unwritten or not + * @err: pointer to save return value + * + * Save the data in original inode blocks and replace original inode extents + * with donor inode extents by calling ext4_swap_extents(). + * Finally, write out the saved data in new original inode blocks. Return + * replaced block count. + */ +static int +move_extent_per_page(struct file *o_filp, struct inode *donor_inode, + pgoff_t orig_page_offset, pgoff_t donor_page_offset, + int data_offset_in_page, + int block_len_in_page, int unwritten, int *err) +{ + struct inode *orig_inode = file_inode(o_filp); + struct folio *folio[2] = {NULL, NULL}; + handle_t *handle; + ext4_lblk_t orig_blk_offset, donor_blk_offset; + unsigned long blocksize = orig_inode->i_sb->s_blocksize; + unsigned int tmp_data_size, data_size, replaced_size; + int i, err2, jblocks, retries = 0; + int replaced_count = 0; + int from; + int blocks_per_page = PAGE_SIZE >> orig_inode->i_blkbits; + struct super_block *sb = orig_inode->i_sb; + struct buffer_head *bh = NULL; + + /* + * It needs twice the amount of ordinary journal buffers because + * inode and donor_inode may change each different metadata blocks. + */ +again: + *err = 0; + jblocks = ext4_meta_trans_blocks(orig_inode, block_len_in_page, + block_len_in_page) * 2; + handle = ext4_journal_start(orig_inode, EXT4_HT_MOVE_EXTENTS, jblocks); + if (IS_ERR(handle)) { + *err = PTR_ERR(handle); + return 0; + } + + orig_blk_offset = orig_page_offset * blocks_per_page + + data_offset_in_page; + + donor_blk_offset = donor_page_offset * blocks_per_page + + data_offset_in_page; + + /* Calculate data_size */ + if ((orig_blk_offset + block_len_in_page - 1) == + ((orig_inode->i_size - 1) >> orig_inode->i_blkbits)) { + /* Replace the last block */ + tmp_data_size = orig_inode->i_size & (blocksize - 1); + /* + * If data_size equal zero, it shows data_size is multiples of + * blocksize. So we set appropriate value. + */ + if (tmp_data_size == 0) + tmp_data_size = blocksize; + + data_size = tmp_data_size + + ((block_len_in_page - 1) << orig_inode->i_blkbits); + } else + data_size = block_len_in_page << orig_inode->i_blkbits; + + replaced_size = data_size; + + *err = mext_folio_double_lock(orig_inode, donor_inode, orig_page_offset, + donor_page_offset, folio); + if (unlikely(*err < 0)) + goto stop_journal; + /* + * If orig extent was unwritten it can become initialized + * at any time after i_data_sem was dropped, in order to + * serialize with delalloc we have recheck extent while we + * hold page's lock, if it is still the case data copy is not + * necessary, just swap data blocks between orig and donor. + */ + if (unwritten) { + ext4_double_down_write_data_sem(orig_inode, donor_inode); + /* If any of extents in range became initialized we have to + * fallback to data copying */ + unwritten = mext_check_coverage(orig_inode, orig_blk_offset, + block_len_in_page, 1, err); + if (*err) + goto drop_data_sem; + + unwritten &= mext_check_coverage(donor_inode, donor_blk_offset, + block_len_in_page, 1, err); + if (*err) + goto drop_data_sem; + + if (!unwritten) { + ext4_double_up_write_data_sem(orig_inode, donor_inode); + goto data_copy; + } + if (!filemap_release_folio(folio[0], 0) || + !filemap_release_folio(folio[1], 0)) { + *err = -EBUSY; + goto drop_data_sem; + } + replaced_count = ext4_swap_extents(handle, orig_inode, + donor_inode, orig_blk_offset, + donor_blk_offset, + block_len_in_page, 1, err); + drop_data_sem: + ext4_double_up_write_data_sem(orig_inode, donor_inode); + goto unlock_folios; + } +data_copy: + from = offset_in_folio(folio[0], + orig_blk_offset << orig_inode->i_blkbits); + *err = mext_page_mkuptodate(folio[0], from, from + replaced_size); + if (*err) + goto unlock_folios; + + /* At this point all buffers in range are uptodate, old mapping layout + * is no longer required, try to drop it now. */ + if (!filemap_release_folio(folio[0], 0) || + !filemap_release_folio(folio[1], 0)) { + *err = -EBUSY; + goto unlock_folios; + } + ext4_double_down_write_data_sem(orig_inode, donor_inode); + replaced_count = ext4_swap_extents(handle, orig_inode, donor_inode, + orig_blk_offset, donor_blk_offset, + block_len_in_page, 1, err); + ext4_double_up_write_data_sem(orig_inode, donor_inode); + if (*err) { + if (replaced_count) { + block_len_in_page = replaced_count; + replaced_size = + block_len_in_page << orig_inode->i_blkbits; + } else + goto unlock_folios; + } + /* Perform all necessary steps similar write_begin()/write_end() + * but keeping in mind that i_size will not change */ + bh = folio_buffers(folio[0]); + if (!bh) + bh = create_empty_buffers(folio[0], + 1 << orig_inode->i_blkbits, 0); + for (i = 0; i < from >> orig_inode->i_blkbits; i++) + bh = bh->b_this_page; + for (i = 0; i < block_len_in_page; i++) { + *err = ext4_get_block(orig_inode, orig_blk_offset + i, bh, 0); + if (*err < 0) + goto repair_branches; + bh = bh->b_this_page; + } + + block_commit_write(folio[0], from, from + replaced_size); + + /* Even in case of data=writeback it is reasonable to pin + * inode to transaction, to prevent unexpected data loss */ + *err = ext4_jbd2_inode_add_write(handle, orig_inode, + (loff_t)orig_page_offset << PAGE_SHIFT, replaced_size); + +unlock_folios: + folio_unlock(folio[0]); + folio_put(folio[0]); + folio_unlock(folio[1]); + folio_put(folio[1]); +stop_journal: + ext4_journal_stop(handle); + if (*err == -ENOSPC && + ext4_should_retry_alloc(sb, &retries)) + goto again; + /* Buffer was busy because probably is pinned to journal transaction, + * force transaction commit may help to free it. */ + if (*err == -EBUSY && retries++ < 4 && EXT4_SB(sb)->s_journal && + jbd2_journal_force_commit_nested(EXT4_SB(sb)->s_journal)) + goto again; + return replaced_count; + +repair_branches: + /* + * This should never ever happen! + * Extents are swapped already, but we are not able to copy data. + * Try to swap extents to it's original places + */ + ext4_double_down_write_data_sem(orig_inode, donor_inode); + replaced_count = ext4_swap_extents(handle, donor_inode, orig_inode, + orig_blk_offset, donor_blk_offset, + block_len_in_page, 0, &err2); + ext4_double_up_write_data_sem(orig_inode, donor_inode); + if (replaced_count != block_len_in_page) { + ext4_error_inode_block(orig_inode, (sector_t)(orig_blk_offset), + EIO, "Unable to copy data block," + " data will be lost."); + *err = -EIO; + } + replaced_count = 0; + goto unlock_folios; +} + +/** + * mext_check_arguments - Check whether move extent can be done + * + * @orig_inode: original inode + * @donor_inode: donor inode + * @orig_start: logical start offset in block for orig + * @donor_start: logical start offset in block for donor + * @len: the number of blocks to be moved + * + * Check the arguments of ext4_move_extents() whether the files can be + * exchanged with each other. + * Return 0 on success, or a negative error value on failure. + */ +static int +mext_check_arguments(struct inode *orig_inode, + struct inode *donor_inode, __u64 orig_start, + __u64 donor_start, __u64 *len) +{ + __u64 orig_eof, donor_eof; + unsigned int blkbits = orig_inode->i_blkbits; + unsigned int blocksize = 1 << blkbits; + + orig_eof = (i_size_read(orig_inode) + blocksize - 1) >> blkbits; + donor_eof = (i_size_read(donor_inode) + blocksize - 1) >> blkbits; + + + if (donor_inode->i_mode & (S_ISUID|S_ISGID)) { + ext4_debug("ext4 move extent: suid or sgid is set" + " to donor file [ino:orig %lu, donor %lu]\n", + orig_inode->i_ino, donor_inode->i_ino); + return -EINVAL; + } + + if (IS_IMMUTABLE(donor_inode) || IS_APPEND(donor_inode)) + return -EPERM; + + /* Ext4 move extent does not support swap files */ + if (IS_SWAPFILE(orig_inode) || IS_SWAPFILE(donor_inode)) { + ext4_debug("ext4 move extent: The argument files should not be swap files [ino:orig %lu, donor %lu]\n", + orig_inode->i_ino, donor_inode->i_ino); + return -ETXTBSY; + } + + if (ext4_is_quota_file(orig_inode) && ext4_is_quota_file(donor_inode)) { + ext4_debug("ext4 move extent: The argument files should not be quota files [ino:orig %lu, donor %lu]\n", + orig_inode->i_ino, donor_inode->i_ino); + return -EOPNOTSUPP; + } + + /* Ext4 move extent supports only extent based file */ + if (!(ext4_test_inode_flag(orig_inode, EXT4_INODE_EXTENTS))) { + ext4_debug("ext4 move extent: orig file is not extents " + "based file [ino:orig %lu]\n", orig_inode->i_ino); + return -EOPNOTSUPP; + } else if (!(ext4_test_inode_flag(donor_inode, EXT4_INODE_EXTENTS))) { + ext4_debug("ext4 move extent: donor file is not extents " + "based file [ino:donor %lu]\n", donor_inode->i_ino); + return -EOPNOTSUPP; + } + + if ((!orig_inode->i_size) || (!donor_inode->i_size)) { + ext4_debug("ext4 move extent: File size is 0 byte\n"); + return -EINVAL; + } + + /* Start offset should be same */ + if ((orig_start & ~(PAGE_MASK >> orig_inode->i_blkbits)) != + (donor_start & ~(PAGE_MASK >> orig_inode->i_blkbits))) { + ext4_debug("ext4 move extent: orig and donor's start " + "offsets are not aligned [ino:orig %lu, donor %lu]\n", + orig_inode->i_ino, donor_inode->i_ino); + return -EINVAL; + } + + if ((orig_start >= EXT_MAX_BLOCKS) || + (donor_start >= EXT_MAX_BLOCKS) || + (*len > EXT_MAX_BLOCKS) || + (donor_start + *len >= EXT_MAX_BLOCKS) || + (orig_start + *len >= EXT_MAX_BLOCKS)) { + ext4_debug("ext4 move extent: Can't handle over [%u] blocks " + "[ino:orig %lu, donor %lu]\n", EXT_MAX_BLOCKS, + orig_inode->i_ino, donor_inode->i_ino); + return -EINVAL; + } + if (orig_eof <= orig_start) + *len = 0; + else if (orig_eof < orig_start + *len - 1) + *len = orig_eof - orig_start; + if (donor_eof <= donor_start) + *len = 0; + else if (donor_eof < donor_start + *len - 1) + *len = donor_eof - donor_start; + if (!*len) { + ext4_debug("ext4 move extent: len should not be 0 " + "[ino:orig %lu, donor %lu]\n", orig_inode->i_ino, + donor_inode->i_ino); + return -EINVAL; + } + + return 0; +} + +/** + * ext4_move_extents - Exchange the specified range of a file + * + * @o_filp: file structure of the original file + * @d_filp: file structure of the donor file + * @orig_blk: start offset in block for orig + * @donor_blk: start offset in block for donor + * @len: the number of blocks to be moved + * @moved_len: moved block length + * + * This function returns 0 and moved block length is set in moved_len + * if succeed, otherwise returns error value. + * + */ +int +ext4_move_extents(struct file *o_filp, struct file *d_filp, __u64 orig_blk, + __u64 donor_blk, __u64 len, __u64 *moved_len) +{ + struct inode *orig_inode = file_inode(o_filp); + struct inode *donor_inode = file_inode(d_filp); + struct ext4_ext_path *path = NULL; + int blocks_per_page = PAGE_SIZE >> orig_inode->i_blkbits; + ext4_lblk_t o_end, o_start = orig_blk; + ext4_lblk_t d_start = donor_blk; + int ret; + + if (orig_inode->i_sb != donor_inode->i_sb) { + ext4_debug("ext4 move extent: The argument files " + "should be in same FS [ino:orig %lu, donor %lu]\n", + orig_inode->i_ino, donor_inode->i_ino); + return -EINVAL; + } + + /* orig and donor should be different inodes */ + if (orig_inode == donor_inode) { + ext4_debug("ext4 move extent: The argument files should not " + "be same inode [ino:orig %lu, donor %lu]\n", + orig_inode->i_ino, donor_inode->i_ino); + return -EINVAL; + } + + /* Regular file check */ + if (!S_ISREG(orig_inode->i_mode) || !S_ISREG(donor_inode->i_mode)) { + ext4_debug("ext4 move extent: The argument files should be " + "regular file [ino:orig %lu, donor %lu]\n", + orig_inode->i_ino, donor_inode->i_ino); + return -EINVAL; + } + + /* TODO: it's not obvious how to swap blocks for inodes with full + journaling enabled */ + if (ext4_should_journal_data(orig_inode) || + ext4_should_journal_data(donor_inode)) { + ext4_msg(orig_inode->i_sb, KERN_ERR, + "Online defrag not supported with data journaling"); + return -EOPNOTSUPP; + } + + if (IS_ENCRYPTED(orig_inode) || IS_ENCRYPTED(donor_inode)) { + ext4_msg(orig_inode->i_sb, KERN_ERR, + "Online defrag not supported for encrypted files"); + return -EOPNOTSUPP; + } + + /* Protect orig and donor inodes against a truncate */ + lock_two_nondirectories(orig_inode, donor_inode); + + /* Wait for all existing dio workers */ + inode_dio_wait(orig_inode); + inode_dio_wait(donor_inode); + + /* Protect extent tree against block allocations via delalloc */ + ext4_double_down_write_data_sem(orig_inode, donor_inode); + /* Check the filesystem environment whether move_extent can be done */ + ret = mext_check_arguments(orig_inode, donor_inode, orig_blk, + donor_blk, &len); + if (ret) + goto out; + o_end = o_start + len; + + *moved_len = 0; + while (o_start < o_end) { + struct ext4_extent *ex; + ext4_lblk_t cur_blk, next_blk; + pgoff_t orig_page_index, donor_page_index; + int offset_in_page; + int unwritten, cur_len; + + path = get_ext_path(orig_inode, o_start, path); + if (IS_ERR(path)) { + ret = PTR_ERR(path); + goto out; + } + ex = path[path->p_depth].p_ext; + cur_blk = le32_to_cpu(ex->ee_block); + cur_len = ext4_ext_get_actual_len(ex); + /* Check hole before the start pos */ + if (cur_blk + cur_len - 1 < o_start) { + next_blk = ext4_ext_next_allocated_block(path); + if (next_blk == EXT_MAX_BLOCKS) { + ret = -ENODATA; + goto out; + } + d_start += next_blk - o_start; + o_start = next_blk; + continue; + /* Check hole after the start pos */ + } else if (cur_blk > o_start) { + /* Skip hole */ + d_start += cur_blk - o_start; + o_start = cur_blk; + /* Extent inside requested range ?*/ + if (cur_blk >= o_end) + goto out; + } else { /* in_range(o_start, o_blk, o_len) */ + cur_len += cur_blk - o_start; + } + unwritten = ext4_ext_is_unwritten(ex); + if (o_end - o_start < cur_len) + cur_len = o_end - o_start; + + orig_page_index = o_start >> (PAGE_SHIFT - + orig_inode->i_blkbits); + donor_page_index = d_start >> (PAGE_SHIFT - + donor_inode->i_blkbits); + offset_in_page = o_start % blocks_per_page; + if (cur_len > blocks_per_page - offset_in_page) + cur_len = blocks_per_page - offset_in_page; + /* + * Up semaphore to avoid following problems: + * a. transaction deadlock among ext4_journal_start, + * ->write_begin via pagefault, and jbd2_journal_commit + * b. racing with ->read_folio, ->write_begin, and + * ext4_get_block in move_extent_per_page + */ + ext4_double_up_write_data_sem(orig_inode, donor_inode); + /* Swap original branches with new branches */ + *moved_len += move_extent_per_page(o_filp, donor_inode, + orig_page_index, donor_page_index, + offset_in_page, cur_len, + unwritten, &ret); + ext4_double_down_write_data_sem(orig_inode, donor_inode); + if (ret < 0) + break; + o_start += cur_len; + d_start += cur_len; + } + +out: + if (*moved_len) { + ext4_discard_preallocations(orig_inode); + ext4_discard_preallocations(donor_inode); + } + + ext4_free_ext_path(path); + ext4_double_up_write_data_sem(orig_inode, donor_inode); + unlock_two_nondirectories(orig_inode, donor_inode); + + return ret; +} diff --git a/fs/ext4l/orphan.c b/fs/ext4l/orphan.c new file mode 100644 index 00000000000..82d5e750145 --- /dev/null +++ b/fs/ext4l/orphan.c @@ -0,0 +1,659 @@ +/* + * Ext4 orphan inode handling + */ +#include <linux/fs.h> +#include <linux/quotaops.h> +#include <linux/buffer_head.h> + +#include "ext4.h" +#include "ext4_jbd2.h" + +static int ext4_orphan_file_add(handle_t *handle, struct inode *inode) +{ + int i, j, start; + struct ext4_orphan_info *oi = &EXT4_SB(inode->i_sb)->s_orphan_info; + int ret = 0; + bool found = false; + __le32 *bdata; + int inodes_per_ob = ext4_inodes_per_orphan_block(inode->i_sb); + int looped = 0; + + /* + * Find block with free orphan entry. Use CPU number for a naive hash + * for a search start in the orphan file + */ + start = raw_smp_processor_id()*13 % oi->of_blocks; + i = start; + do { + if (atomic_dec_if_positive(&oi->of_binfo[i].ob_free_entries) + >= 0) { + found = true; + break; + } + if (++i >= oi->of_blocks) + i = 0; + } while (i != start); + + if (!found) { + /* + * For now we don't grow or shrink orphan file. We just use + * whatever was allocated at mke2fs time. The additional + * credits we would have to reserve for each orphan inode + * operation just don't seem worth it. + */ + return -ENOSPC; + } + + ret = ext4_journal_get_write_access(handle, inode->i_sb, + oi->of_binfo[i].ob_bh, EXT4_JTR_ORPHAN_FILE); + if (ret) { + atomic_inc(&oi->of_binfo[i].ob_free_entries); + return ret; + } + + bdata = (__le32 *)(oi->of_binfo[i].ob_bh->b_data); + /* Find empty slot in a block */ + j = 0; + do { + if (looped) { + /* + * Did we walk through the block several times without + * finding free entry? It is theoretically possible + * if entries get constantly allocated and freed or + * if the block is corrupted. Avoid indefinite looping + * and bail. We'll use orphan list instead. + */ + if (looped > 3) { + atomic_inc(&oi->of_binfo[i].ob_free_entries); + return -ENOSPC; + } + cond_resched(); + } + while (bdata[j]) { + if (++j >= inodes_per_ob) { + j = 0; + looped++; + } + } + } while (cmpxchg(&bdata[j], (__le32)0, cpu_to_le32(inode->i_ino)) != + (__le32)0); + + EXT4_I(inode)->i_orphan_idx = i * inodes_per_ob + j; + ext4_set_inode_state(inode, EXT4_STATE_ORPHAN_FILE); + + return ext4_handle_dirty_metadata(handle, NULL, oi->of_binfo[i].ob_bh); +} + +/* + * ext4_orphan_add() links an unlinked or truncated inode into a list of + * such inodes, starting at the superblock, in case we crash before the + * file is closed/deleted, or in case the inode truncate spans multiple + * transactions and the last transaction is not recovered after a crash. + * + * At filesystem recovery time, we walk this list deleting unlinked + * inodes and truncating linked inodes in ext4_orphan_cleanup(). + * + * Orphan list manipulation functions must be called under i_rwsem unless + * we are just creating the inode or deleting it. + */ +int ext4_orphan_add(handle_t *handle, struct inode *inode) +{ + struct super_block *sb = inode->i_sb; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_iloc iloc; + int err = 0, rc; + bool dirty = false; + + if (!sbi->s_journal || is_bad_inode(inode)) + return 0; + + WARN_ON_ONCE(!(inode->i_state & (I_NEW | I_FREEING)) && + !inode_is_locked(inode)); + if (ext4_inode_orphan_tracked(inode)) + return 0; + + /* + * Orphan handling is only valid for files with data blocks + * being truncated, or files being unlinked. Note that we either + * hold i_rwsem, or the inode can not be referenced from outside, + * so i_nlink should not be bumped due to race + */ + ASSERT((S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) || + S_ISLNK(inode->i_mode)) || inode->i_nlink == 0); + + if (sbi->s_orphan_info.of_blocks) { + err = ext4_orphan_file_add(handle, inode); + /* + * Fallback to normal orphan list of orphan file is + * out of space + */ + if (err != -ENOSPC) + return err; + } + + BUFFER_TRACE(sbi->s_sbh, "get_write_access"); + err = ext4_journal_get_write_access(handle, sb, sbi->s_sbh, + EXT4_JTR_NONE); + if (err) + goto out; + + err = ext4_reserve_inode_write(handle, inode, &iloc); + if (err) + goto out; + + mutex_lock(&sbi->s_orphan_lock); + /* + * Due to previous errors inode may be already a part of on-disk + * orphan list. If so skip on-disk list modification. + */ + if (!NEXT_ORPHAN(inode) || NEXT_ORPHAN(inode) > + (le32_to_cpu(sbi->s_es->s_inodes_count))) { + /* Insert this inode at the head of the on-disk orphan list */ + NEXT_ORPHAN(inode) = le32_to_cpu(sbi->s_es->s_last_orphan); + lock_buffer(sbi->s_sbh); + sbi->s_es->s_last_orphan = cpu_to_le32(inode->i_ino); + ext4_superblock_csum_set(sb); + unlock_buffer(sbi->s_sbh); + dirty = true; + } + list_add(&EXT4_I(inode)->i_orphan, &sbi->s_orphan); + mutex_unlock(&sbi->s_orphan_lock); + + if (dirty) { + err = ext4_handle_dirty_metadata(handle, NULL, sbi->s_sbh); + rc = ext4_mark_iloc_dirty(handle, inode, &iloc); + if (!err) + err = rc; + if (err) { + /* + * We have to remove inode from in-memory list if + * addition to on disk orphan list failed. Stray orphan + * list entries can cause panics at unmount time. + */ + mutex_lock(&sbi->s_orphan_lock); + list_del_init(&EXT4_I(inode)->i_orphan); + mutex_unlock(&sbi->s_orphan_lock); + } + } else + brelse(iloc.bh); + + ext4_debug("superblock will point to %lu\n", inode->i_ino); + ext4_debug("orphan inode %lu will point to %d\n", + inode->i_ino, NEXT_ORPHAN(inode)); +out: + ext4_std_error(sb, err); + return err; +} + +static int ext4_orphan_file_del(handle_t *handle, struct inode *inode) +{ + struct ext4_orphan_info *oi = &EXT4_SB(inode->i_sb)->s_orphan_info; + __le32 *bdata; + int blk, off; + int inodes_per_ob = ext4_inodes_per_orphan_block(inode->i_sb); + int ret = 0; + + if (!handle) + goto out; + blk = EXT4_I(inode)->i_orphan_idx / inodes_per_ob; + off = EXT4_I(inode)->i_orphan_idx % inodes_per_ob; + if (WARN_ON_ONCE(blk >= oi->of_blocks)) + goto out; + + ret = ext4_journal_get_write_access(handle, inode->i_sb, + oi->of_binfo[blk].ob_bh, EXT4_JTR_ORPHAN_FILE); + if (ret) + goto out; + + bdata = (__le32 *)(oi->of_binfo[blk].ob_bh->b_data); + bdata[off] = 0; + atomic_inc(&oi->of_binfo[blk].ob_free_entries); + ret = ext4_handle_dirty_metadata(handle, NULL, oi->of_binfo[blk].ob_bh); +out: + ext4_clear_inode_state(inode, EXT4_STATE_ORPHAN_FILE); + INIT_LIST_HEAD(&EXT4_I(inode)->i_orphan); + + return ret; +} + +/* + * ext4_orphan_del() removes an unlinked or truncated inode from the list + * of such inodes stored on disk, because it is finally being cleaned up. + */ +int ext4_orphan_del(handle_t *handle, struct inode *inode) +{ + struct list_head *prev; + struct ext4_inode_info *ei = EXT4_I(inode); + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + __u32 ino_next; + struct ext4_iloc iloc; + int err = 0; + + if (!sbi->s_journal && !(sbi->s_mount_state & EXT4_ORPHAN_FS)) + return 0; + + WARN_ON_ONCE(!(inode->i_state & (I_NEW | I_FREEING)) && + !inode_is_locked(inode)); + if (ext4_test_inode_state(inode, EXT4_STATE_ORPHAN_FILE)) + return ext4_orphan_file_del(handle, inode); + + /* Do this quick check before taking global s_orphan_lock. */ + if (list_empty(&ei->i_orphan)) + return 0; + + if (handle) { + /* Grab inode buffer early before taking global s_orphan_lock */ + err = ext4_reserve_inode_write(handle, inode, &iloc); + } + + mutex_lock(&sbi->s_orphan_lock); + ext4_debug("remove inode %lu from orphan list\n", inode->i_ino); + + prev = ei->i_orphan.prev; + list_del_init(&ei->i_orphan); + + /* If we're on an error path, we may not have a valid + * transaction handle with which to update the orphan list on + * disk, but we still need to remove the inode from the linked + * list in memory. */ + if (!handle || err) { + mutex_unlock(&sbi->s_orphan_lock); + goto out_err; + } + + ino_next = NEXT_ORPHAN(inode); + if (prev == &sbi->s_orphan) { + ext4_debug("superblock will point to %u\n", ino_next); + BUFFER_TRACE(sbi->s_sbh, "get_write_access"); + err = ext4_journal_get_write_access(handle, inode->i_sb, + sbi->s_sbh, EXT4_JTR_NONE); + if (err) { + mutex_unlock(&sbi->s_orphan_lock); + goto out_brelse; + } + lock_buffer(sbi->s_sbh); + sbi->s_es->s_last_orphan = cpu_to_le32(ino_next); + ext4_superblock_csum_set(inode->i_sb); + unlock_buffer(sbi->s_sbh); + mutex_unlock(&sbi->s_orphan_lock); + err = ext4_handle_dirty_metadata(handle, NULL, sbi->s_sbh); + } else { + struct ext4_iloc iloc2; + struct inode *i_prev = + &list_entry(prev, struct ext4_inode_info, i_orphan)->vfs_inode; + + ext4_debug("orphan inode %lu will point to %u\n", + i_prev->i_ino, ino_next); + err = ext4_reserve_inode_write(handle, i_prev, &iloc2); + if (err) { + mutex_unlock(&sbi->s_orphan_lock); + goto out_brelse; + } + NEXT_ORPHAN(i_prev) = ino_next; + err = ext4_mark_iloc_dirty(handle, i_prev, &iloc2); + mutex_unlock(&sbi->s_orphan_lock); + } + if (err) + goto out_brelse; + NEXT_ORPHAN(inode) = 0; + err = ext4_mark_iloc_dirty(handle, inode, &iloc); +out_err: + ext4_std_error(inode->i_sb, err); + return err; + +out_brelse: + brelse(iloc.bh); + goto out_err; +} + +#ifdef CONFIG_QUOTA +static int ext4_quota_on_mount(struct super_block *sb, int type) +{ + return dquot_quota_on_mount(sb, + rcu_dereference_protected(EXT4_SB(sb)->s_qf_names[type], + lockdep_is_held(&sb->s_umount)), + EXT4_SB(sb)->s_jquota_fmt, type); +} +#endif + +static void ext4_process_orphan(struct inode *inode, + int *nr_truncates, int *nr_orphans) +{ + struct super_block *sb = inode->i_sb; + int ret; + + dquot_initialize(inode); + if (inode->i_nlink) { + if (test_opt(sb, DEBUG)) + ext4_msg(sb, KERN_DEBUG, + "%s: truncating inode %lu to %lld bytes", + __func__, inode->i_ino, inode->i_size); + ext4_debug("truncating inode %lu to %lld bytes\n", + inode->i_ino, inode->i_size); + inode_lock(inode); + truncate_inode_pages(inode->i_mapping, inode->i_size); + ret = ext4_truncate(inode); + if (ret) { + /* + * We need to clean up the in-core orphan list + * manually if ext4_truncate() failed to get a + * transaction handle. + */ + ext4_orphan_del(NULL, inode); + ext4_std_error(inode->i_sb, ret); + } + inode_unlock(inode); + (*nr_truncates)++; + } else { + if (test_opt(sb, DEBUG)) + ext4_msg(sb, KERN_DEBUG, + "%s: deleting unreferenced inode %lu", + __func__, inode->i_ino); + ext4_debug("deleting unreferenced inode %lu\n", + inode->i_ino); + (*nr_orphans)++; + } + iput(inode); /* The delete magic happens here! */ +} + +/* ext4_orphan_cleanup() walks a singly-linked list of inodes (starting at + * the superblock) which were deleted from all directories, but held open by + * a process at the time of a crash. We walk the list and try to delete these + * inodes at recovery time (only with a read-write filesystem). + * + * In order to keep the orphan inode chain consistent during traversal (in + * case of crash during recovery), we link each inode into the superblock + * orphan list_head and handle it the same way as an inode deletion during + * normal operation (which journals the operations for us). + * + * We only do an iget() and an iput() on each inode, which is very safe if we + * accidentally point at an in-use or already deleted inode. The worst that + * can happen in this case is that we get a "bit already cleared" message from + * ext4_free_inode(). The only reason we would point at a wrong inode is if + * e2fsck was run on this filesystem, and it must have already done the orphan + * inode cleanup for us, so we can safely abort without any further action. + */ +void ext4_orphan_cleanup(struct super_block *sb, struct ext4_super_block *es) +{ + unsigned int s_flags = sb->s_flags; + int nr_orphans = 0, nr_truncates = 0; + struct inode *inode; + int i, j; +#ifdef CONFIG_QUOTA + int quota_update = 0; +#endif + __le32 *bdata; + struct ext4_orphan_info *oi = &EXT4_SB(sb)->s_orphan_info; + int inodes_per_ob = ext4_inodes_per_orphan_block(sb); + + if (!es->s_last_orphan && !oi->of_blocks) { + ext4_debug("no orphan inodes to clean up\n"); + return; + } + + if (bdev_read_only(sb->s_bdev)) { + ext4_msg(sb, KERN_ERR, "write access " + "unavailable, skipping orphan cleanup"); + return; + } + + /* Check if feature set would not allow a r/w mount */ + if (!ext4_feature_set_ok(sb, 0)) { + ext4_msg(sb, KERN_INFO, "Skipping orphan cleanup due to " + "unknown ROCOMPAT features"); + return; + } + + if (EXT4_SB(sb)->s_mount_state & EXT4_ERROR_FS) { + /* don't clear list on RO mount w/ errors */ + if (es->s_last_orphan && !(s_flags & SB_RDONLY)) { + ext4_msg(sb, KERN_INFO, "Errors on filesystem, " + "clearing orphan list."); + es->s_last_orphan = 0; + } + ext4_debug("Skipping orphan recovery on fs with errors.\n"); + return; + } + + if (s_flags & SB_RDONLY) { + ext4_msg(sb, KERN_INFO, "orphan cleanup on readonly fs"); + sb->s_flags &= ~SB_RDONLY; + } +#ifdef CONFIG_QUOTA + /* + * Turn on quotas which were not enabled for read-only mounts if + * filesystem has quota feature, so that they are updated correctly. + */ + if (ext4_has_feature_quota(sb) && (s_flags & SB_RDONLY)) { + int ret = ext4_enable_quotas(sb); + + if (!ret) + quota_update = 1; + else + ext4_msg(sb, KERN_ERR, + "Cannot turn on quotas: error %d", ret); + } + + /* Turn on journaled quotas used for old sytle */ + for (i = 0; i < EXT4_MAXQUOTAS; i++) { + if (EXT4_SB(sb)->s_qf_names[i]) { + int ret = ext4_quota_on_mount(sb, i); + + if (!ret) + quota_update = 1; + else + ext4_msg(sb, KERN_ERR, + "Cannot turn on journaled " + "quota: type %d: error %d", i, ret); + } + } +#endif + + while (es->s_last_orphan) { + /* + * We may have encountered an error during cleanup; if + * so, skip the rest. + */ + if (EXT4_SB(sb)->s_mount_state & EXT4_ERROR_FS) { + ext4_debug("Skipping orphan recovery on fs with errors.\n"); + es->s_last_orphan = 0; + break; + } + + inode = ext4_orphan_get(sb, le32_to_cpu(es->s_last_orphan)); + if (IS_ERR(inode)) { + es->s_last_orphan = 0; + break; + } + + list_add(&EXT4_I(inode)->i_orphan, &EXT4_SB(sb)->s_orphan); + ext4_process_orphan(inode, &nr_truncates, &nr_orphans); + } + + for (i = 0; i < oi->of_blocks; i++) { + bdata = (__le32 *)(oi->of_binfo[i].ob_bh->b_data); + for (j = 0; j < inodes_per_ob; j++) { + if (!bdata[j]) + continue; + inode = ext4_orphan_get(sb, le32_to_cpu(bdata[j])); + if (IS_ERR(inode)) + continue; + ext4_set_inode_state(inode, EXT4_STATE_ORPHAN_FILE); + EXT4_I(inode)->i_orphan_idx = i * inodes_per_ob + j; + ext4_process_orphan(inode, &nr_truncates, &nr_orphans); + } + } + +#define PLURAL(x) (x), ((x) == 1) ? "" : "s" + + if (nr_orphans) + ext4_msg(sb, KERN_INFO, "%d orphan inode%s deleted", + PLURAL(nr_orphans)); + if (nr_truncates) + ext4_msg(sb, KERN_INFO, "%d truncate%s cleaned up", + PLURAL(nr_truncates)); +#ifdef CONFIG_QUOTA + /* Turn off quotas if they were enabled for orphan cleanup */ + if (quota_update) { + for (i = 0; i < EXT4_MAXQUOTAS; i++) { + if (sb_dqopt(sb)->files[i]) + dquot_quota_off(sb, i); + } + } +#endif + sb->s_flags = s_flags; /* Restore SB_RDONLY status */ +} + +void ext4_release_orphan_info(struct super_block *sb) +{ + int i; + struct ext4_orphan_info *oi = &EXT4_SB(sb)->s_orphan_info; + + if (!oi->of_blocks) + return; + for (i = 0; i < oi->of_blocks; i++) + brelse(oi->of_binfo[i].ob_bh); + kvfree(oi->of_binfo); +} + +static struct ext4_orphan_block_tail *ext4_orphan_block_tail( + struct super_block *sb, + struct buffer_head *bh) +{ + return (struct ext4_orphan_block_tail *)(bh->b_data + sb->s_blocksize - + sizeof(struct ext4_orphan_block_tail)); +} + +static int ext4_orphan_file_block_csum_verify(struct super_block *sb, + struct buffer_head *bh) +{ + __u32 calculated; + int inodes_per_ob = ext4_inodes_per_orphan_block(sb); + struct ext4_orphan_info *oi = &EXT4_SB(sb)->s_orphan_info; + struct ext4_orphan_block_tail *ot; + __le64 dsk_block_nr = cpu_to_le64(bh->b_blocknr); + + if (!ext4_has_feature_metadata_csum(sb)) + return 1; + + ot = ext4_orphan_block_tail(sb, bh); + calculated = ext4_chksum(oi->of_csum_seed, (__u8 *)&dsk_block_nr, + sizeof(dsk_block_nr)); + calculated = ext4_chksum(calculated, (__u8 *)bh->b_data, + inodes_per_ob * sizeof(__u32)); + return le32_to_cpu(ot->ob_checksum) == calculated; +} + +/* This gets called only when checksumming is enabled */ +void ext4_orphan_file_block_trigger(struct jbd2_buffer_trigger_type *triggers, + struct buffer_head *bh, + void *data, size_t size) +{ + struct super_block *sb = EXT4_TRIGGER(triggers)->sb; + __u32 csum; + int inodes_per_ob = ext4_inodes_per_orphan_block(sb); + struct ext4_orphan_info *oi = &EXT4_SB(sb)->s_orphan_info; + struct ext4_orphan_block_tail *ot; + __le64 dsk_block_nr = cpu_to_le64(bh->b_blocknr); + + csum = ext4_chksum(oi->of_csum_seed, (__u8 *)&dsk_block_nr, + sizeof(dsk_block_nr)); + csum = ext4_chksum(csum, (__u8 *)data, inodes_per_ob * sizeof(__u32)); + ot = ext4_orphan_block_tail(sb, bh); + ot->ob_checksum = cpu_to_le32(csum); +} + +int ext4_init_orphan_info(struct super_block *sb) +{ + struct ext4_orphan_info *oi = &EXT4_SB(sb)->s_orphan_info; + struct inode *inode; + int i, j; + int ret; + int free; + __le32 *bdata; + int inodes_per_ob = ext4_inodes_per_orphan_block(sb); + struct ext4_orphan_block_tail *ot; + ino_t orphan_ino = le32_to_cpu(EXT4_SB(sb)->s_es->s_orphan_file_inum); + + if (!ext4_has_feature_orphan_file(sb)) + return 0; + + inode = ext4_iget(sb, orphan_ino, EXT4_IGET_SPECIAL); + if (IS_ERR(inode)) { + ext4_msg(sb, KERN_ERR, "get orphan inode failed"); + return PTR_ERR(inode); + } + /* + * This is just an artificial limit to prevent corrupted fs from + * consuming absurd amounts of memory when pinning blocks of orphan + * file in memory. + */ + if (inode->i_size > 8 << 20) { + ext4_msg(sb, KERN_ERR, "orphan file too big: %llu", + (unsigned long long)inode->i_size); + ret = -EFSCORRUPTED; + goto out_put; + } + oi->of_blocks = inode->i_size >> sb->s_blocksize_bits; + oi->of_csum_seed = EXT4_I(inode)->i_csum_seed; + oi->of_binfo = kvmalloc_array(oi->of_blocks, + sizeof(struct ext4_orphan_block), + GFP_KERNEL); + if (!oi->of_binfo) { + ret = -ENOMEM; + goto out_put; + } + for (i = 0; i < oi->of_blocks; i++) { + oi->of_binfo[i].ob_bh = ext4_bread(NULL, inode, i, 0); + if (IS_ERR(oi->of_binfo[i].ob_bh)) { + ret = PTR_ERR(oi->of_binfo[i].ob_bh); + goto out_free; + } + if (!oi->of_binfo[i].ob_bh) { + ret = -EIO; + goto out_free; + } + ot = ext4_orphan_block_tail(sb, oi->of_binfo[i].ob_bh); + if (le32_to_cpu(ot->ob_magic) != EXT4_ORPHAN_BLOCK_MAGIC) { + ext4_error(sb, "orphan file block %d: bad magic", i); + ret = -EIO; + goto out_free; + } + if (!ext4_orphan_file_block_csum_verify(sb, + oi->of_binfo[i].ob_bh)) { + ext4_error(sb, "orphan file block %d: bad checksum", i); + ret = -EIO; + goto out_free; + } + bdata = (__le32 *)(oi->of_binfo[i].ob_bh->b_data); + free = 0; + for (j = 0; j < inodes_per_ob; j++) + if (bdata[j] == 0) + free++; + atomic_set(&oi->of_binfo[i].ob_free_entries, free); + } + iput(inode); + return 0; +out_free: + for (i--; i >= 0; i--) + brelse(oi->of_binfo[i].ob_bh); + kvfree(oi->of_binfo); +out_put: + iput(inode); + return ret; +} + +int ext4_orphan_file_empty(struct super_block *sb) +{ + struct ext4_orphan_info *oi = &EXT4_SB(sb)->s_orphan_info; + int i; + int inodes_per_ob = ext4_inodes_per_orphan_block(sb); + + if (!ext4_has_feature_orphan_file(sb)) + return 1; + for (i = 0; i < oi->of_blocks; i++) + if (atomic_read(&oi->of_binfo[i].ob_free_entries) != + inodes_per_ob) + return 0; + return 1; +} diff --git a/fs/ext4l/page-io.c b/fs/ext4l/page-io.c new file mode 100644 index 00000000000..39abfeec5f3 --- /dev/null +++ b/fs/ext4l/page-io.c @@ -0,0 +1,593 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * linux/fs/ext4/page-io.c + * + * This contains the new page_io functions for ext4 + * + * Written by Theodore Ts'o, 2010. + */ + +#include <linux/fs.h> +#include <linux/time.h> +#include <linux/highuid.h> +#include <linux/pagemap.h> +#include <linux/quotaops.h> +#include <linux/string.h> +#include <linux/buffer_head.h> +#include <linux/writeback.h> +#include <linux/pagevec.h> +#include <linux/mpage.h> +#include <linux/namei.h> +#include <linux/uio.h> +#include <linux/bio.h> +#include <linux/workqueue.h> +#include <linux/kernel.h> +#include <linux/slab.h> +#include <linux/mm.h> +#include <linux/sched/mm.h> + +#include "ext4_jbd2.h" +#include "xattr.h" +#include "acl.h" + +static struct kmem_cache *io_end_cachep; +static struct kmem_cache *io_end_vec_cachep; + +int __init ext4_init_pageio(void) +{ + io_end_cachep = KMEM_CACHE(ext4_io_end, SLAB_RECLAIM_ACCOUNT); + if (io_end_cachep == NULL) + return -ENOMEM; + + io_end_vec_cachep = KMEM_CACHE(ext4_io_end_vec, 0); + if (io_end_vec_cachep == NULL) { + kmem_cache_destroy(io_end_cachep); + return -ENOMEM; + } + return 0; +} + +void ext4_exit_pageio(void) +{ + kmem_cache_destroy(io_end_cachep); + kmem_cache_destroy(io_end_vec_cachep); +} + +struct ext4_io_end_vec *ext4_alloc_io_end_vec(ext4_io_end_t *io_end) +{ + struct ext4_io_end_vec *io_end_vec; + + io_end_vec = kmem_cache_zalloc(io_end_vec_cachep, GFP_NOFS); + if (!io_end_vec) + return ERR_PTR(-ENOMEM); + INIT_LIST_HEAD(&io_end_vec->list); + list_add_tail(&io_end_vec->list, &io_end->list_vec); + return io_end_vec; +} + +static void ext4_free_io_end_vec(ext4_io_end_t *io_end) +{ + struct ext4_io_end_vec *io_end_vec, *tmp; + + if (list_empty(&io_end->list_vec)) + return; + list_for_each_entry_safe(io_end_vec, tmp, &io_end->list_vec, list) { + list_del(&io_end_vec->list); + kmem_cache_free(io_end_vec_cachep, io_end_vec); + } +} + +struct ext4_io_end_vec *ext4_last_io_end_vec(ext4_io_end_t *io_end) +{ + BUG_ON(list_empty(&io_end->list_vec)); + return list_last_entry(&io_end->list_vec, struct ext4_io_end_vec, list); +} + +/* + * Print an buffer I/O error compatible with the fs/buffer.c. This + * provides compatibility with dmesg scrapers that look for a specific + * buffer I/O error message. We really need a unified error reporting + * structure to userspace ala Digital Unix's uerf system, but it's + * probably not going to happen in my lifetime, due to LKML politics... + */ +static void buffer_io_error(struct buffer_head *bh) +{ + printk_ratelimited(KERN_ERR "Buffer I/O error on device %pg, logical block %llu\n", + bh->b_bdev, + (unsigned long long)bh->b_blocknr); +} + +static void ext4_finish_bio(struct bio *bio) +{ + struct folio_iter fi; + + bio_for_each_folio_all(fi, bio) { + struct folio *folio = fi.folio; + struct folio *io_folio = NULL; + struct buffer_head *bh, *head; + size_t bio_start = fi.offset; + size_t bio_end = bio_start + fi.length; + unsigned under_io = 0; + unsigned long flags; + + if (fscrypt_is_bounce_folio(folio)) { + io_folio = folio; + folio = fscrypt_pagecache_folio(folio); + } + + if (bio->bi_status) { + int err = blk_status_to_errno(bio->bi_status); + mapping_set_error(folio->mapping, err); + } + bh = head = folio_buffers(folio); + /* + * We check all buffers in the folio under b_uptodate_lock + * to avoid races with other end io clearing async_write flags + */ + spin_lock_irqsave(&head->b_uptodate_lock, flags); + do { + if (bh_offset(bh) < bio_start || + bh_offset(bh) + bh->b_size > bio_end) { + if (buffer_async_write(bh)) + under_io++; + continue; + } + clear_buffer_async_write(bh); + if (bio->bi_status) { + set_buffer_write_io_error(bh); + buffer_io_error(bh); + } + } while ((bh = bh->b_this_page) != head); + spin_unlock_irqrestore(&head->b_uptodate_lock, flags); + if (!under_io) { + fscrypt_free_bounce_page(&io_folio->page); + folio_end_writeback(folio); + } + } +} + +static void ext4_release_io_end(ext4_io_end_t *io_end) +{ + struct bio *bio, *next_bio; + + BUG_ON(!list_empty(&io_end->list)); + BUG_ON(io_end->flag & EXT4_IO_END_UNWRITTEN); + WARN_ON(io_end->handle); + + for (bio = io_end->bio; bio; bio = next_bio) { + next_bio = bio->bi_private; + ext4_finish_bio(bio); + bio_put(bio); + } + ext4_free_io_end_vec(io_end); + kmem_cache_free(io_end_cachep, io_end); +} + +/* + * On successful IO, check a range of space and convert unwritten extents to + * written. On IO failure, check if journal abort is needed. Note that + * we are protected from truncate touching same part of extent tree by the + * fact that truncate code waits for all DIO to finish (thus exclusion from + * direct IO is achieved) and also waits for PageWriteback bits. Thus we + * cannot get to ext4_ext_truncate() before all IOs overlapping that range are + * completed (happens from ext4_free_ioend()). + */ +static int ext4_end_io_end(ext4_io_end_t *io_end) +{ + struct inode *inode = io_end->inode; + handle_t *handle = io_end->handle; + struct super_block *sb = inode->i_sb; + int ret = 0; + + ext4_debug("ext4_end_io_nolock: io_end 0x%p from inode %lu,list->next 0x%p," + "list->prev 0x%p\n", + io_end, inode->i_ino, io_end->list.next, io_end->list.prev); + + /* + * Do not convert the unwritten extents if data writeback fails, + * or stale data may be exposed. + */ + io_end->handle = NULL; /* Following call will use up the handle */ + if (unlikely(io_end->flag & EXT4_IO_END_FAILED)) { + ret = -EIO; + if (handle) + jbd2_journal_free_reserved(handle); + + if (test_opt(sb, DATA_ERR_ABORT)) + jbd2_journal_abort(EXT4_SB(sb)->s_journal, ret); + } else { + ret = ext4_convert_unwritten_io_end_vec(handle, io_end); + } + if (ret < 0 && !ext4_emergency_state(sb) && + io_end->flag & EXT4_IO_END_UNWRITTEN) { + ext4_msg(sb, KERN_EMERG, + "failed to convert unwritten extents to written " + "extents -- potential data loss! " + "(inode %lu, error %d)", inode->i_ino, ret); + } + + ext4_clear_io_unwritten_flag(io_end); + ext4_release_io_end(io_end); + return ret; +} + +static void dump_completed_IO(struct inode *inode, struct list_head *head) +{ +#ifdef EXT4FS_DEBUG + struct list_head *cur, *before, *after; + ext4_io_end_t *io_end, *io_end0, *io_end1; + + if (list_empty(head)) + return; + + ext4_debug("Dump inode %lu completed io list\n", inode->i_ino); + list_for_each_entry(io_end, head, list) { + cur = &io_end->list; + before = cur->prev; + io_end0 = container_of(before, ext4_io_end_t, list); + after = cur->next; + io_end1 = container_of(after, ext4_io_end_t, list); + + ext4_debug("io 0x%p from inode %lu,prev 0x%p,next 0x%p\n", + io_end, inode->i_ino, io_end0, io_end1); + } +#endif +} + +static bool ext4_io_end_defer_completion(ext4_io_end_t *io_end) +{ + if (io_end->flag & EXT4_IO_END_UNWRITTEN && + !list_empty(&io_end->list_vec)) + return true; + if (test_opt(io_end->inode->i_sb, DATA_ERR_ABORT) && + io_end->flag & EXT4_IO_END_FAILED && + !ext4_emergency_state(io_end->inode->i_sb)) + return true; + return false; +} + +/* Add the io_end to per-inode completed end_io list. */ +static void ext4_add_complete_io(ext4_io_end_t *io_end) +{ + struct ext4_inode_info *ei = EXT4_I(io_end->inode); + struct ext4_sb_info *sbi = EXT4_SB(io_end->inode->i_sb); + struct workqueue_struct *wq; + unsigned long flags; + + /* Only reserved conversions or pending IO errors will enter here. */ + WARN_ON(!(io_end->flag & EXT4_IO_END_DEFER_COMPLETION)); + WARN_ON(io_end->flag & EXT4_IO_END_UNWRITTEN && + !io_end->handle && sbi->s_journal); + WARN_ON(!io_end->bio); + + spin_lock_irqsave(&ei->i_completed_io_lock, flags); + wq = sbi->rsv_conversion_wq; + if (list_empty(&ei->i_rsv_conversion_list)) + queue_work(wq, &ei->i_rsv_conversion_work); + list_add_tail(&io_end->list, &ei->i_rsv_conversion_list); + spin_unlock_irqrestore(&ei->i_completed_io_lock, flags); +} + +static int ext4_do_flush_completed_IO(struct inode *inode, + struct list_head *head) +{ + ext4_io_end_t *io_end; + struct list_head unwritten; + unsigned long flags; + struct ext4_inode_info *ei = EXT4_I(inode); + int err, ret = 0; + + spin_lock_irqsave(&ei->i_completed_io_lock, flags); + dump_completed_IO(inode, head); + list_replace_init(head, &unwritten); + spin_unlock_irqrestore(&ei->i_completed_io_lock, flags); + + while (!list_empty(&unwritten)) { + io_end = list_entry(unwritten.next, ext4_io_end_t, list); + BUG_ON(!(io_end->flag & EXT4_IO_END_DEFER_COMPLETION)); + list_del_init(&io_end->list); + + err = ext4_end_io_end(io_end); + if (unlikely(!ret && err)) + ret = err; + } + return ret; +} + +/* + * Used to convert unwritten extents to written extents upon IO completion, + * or used to abort the journal upon IO errors. + */ +void ext4_end_io_rsv_work(struct work_struct *work) +{ + struct ext4_inode_info *ei = container_of(work, struct ext4_inode_info, + i_rsv_conversion_work); + ext4_do_flush_completed_IO(&ei->vfs_inode, &ei->i_rsv_conversion_list); +} + +ext4_io_end_t *ext4_init_io_end(struct inode *inode, gfp_t flags) +{ + ext4_io_end_t *io_end = kmem_cache_zalloc(io_end_cachep, flags); + + if (io_end) { + io_end->inode = inode; + INIT_LIST_HEAD(&io_end->list); + INIT_LIST_HEAD(&io_end->list_vec); + refcount_set(&io_end->count, 1); + } + return io_end; +} + +void ext4_put_io_end_defer(ext4_io_end_t *io_end) +{ + if (refcount_dec_and_test(&io_end->count)) { + if (ext4_io_end_defer_completion(io_end)) + return ext4_add_complete_io(io_end); + + ext4_release_io_end(io_end); + } +} + +int ext4_put_io_end(ext4_io_end_t *io_end) +{ + if (refcount_dec_and_test(&io_end->count)) { + if (ext4_io_end_defer_completion(io_end)) + return ext4_end_io_end(io_end); + + ext4_release_io_end(io_end); + } + return 0; +} + +ext4_io_end_t *ext4_get_io_end(ext4_io_end_t *io_end) +{ + refcount_inc(&io_end->count); + return io_end; +} + +/* BIO completion function for page writeback */ +static void ext4_end_bio(struct bio *bio) +{ + ext4_io_end_t *io_end = bio->bi_private; + sector_t bi_sector = bio->bi_iter.bi_sector; + + if (WARN_ONCE(!io_end, "io_end is NULL: %pg: sector %Lu len %u err %d\n", + bio->bi_bdev, + (long long) bio->bi_iter.bi_sector, + (unsigned) bio_sectors(bio), + bio->bi_status)) { + ext4_finish_bio(bio); + bio_put(bio); + return; + } + bio->bi_end_io = NULL; + + if (bio->bi_status) { + struct inode *inode = io_end->inode; + + ext4_warning(inode->i_sb, "I/O error %d writing to inode %lu " + "starting block %llu)", + bio->bi_status, inode->i_ino, + (unsigned long long) + bi_sector >> (inode->i_blkbits - 9)); + io_end->flag |= EXT4_IO_END_FAILED; + mapping_set_error(inode->i_mapping, + blk_status_to_errno(bio->bi_status)); + } + + if (ext4_io_end_defer_completion(io_end)) { + /* + * Link bio into list hanging from io_end. We have to do it + * atomically as bio completions can be racing against each + * other. + */ + bio->bi_private = xchg(&io_end->bio, bio); + ext4_put_io_end_defer(io_end); + } else { + /* + * Drop io_end reference early. Inode can get freed once + * we finish the bio. + */ + ext4_put_io_end_defer(io_end); + ext4_finish_bio(bio); + bio_put(bio); + } +} + +void ext4_io_submit(struct ext4_io_submit *io) +{ + struct bio *bio = io->io_bio; + + if (bio) { + if (io->io_wbc->sync_mode == WB_SYNC_ALL) + io->io_bio->bi_opf |= REQ_SYNC; + submit_bio(io->io_bio); + } + io->io_bio = NULL; +} + +void ext4_io_submit_init(struct ext4_io_submit *io, + struct writeback_control *wbc) +{ + io->io_wbc = wbc; + io->io_bio = NULL; + io->io_end = NULL; +} + +static void io_submit_init_bio(struct ext4_io_submit *io, + struct buffer_head *bh) +{ + struct bio *bio; + + /* + * bio_alloc will _always_ be able to allocate a bio if + * __GFP_DIRECT_RECLAIM is set, see comments for bio_alloc_bioset(). + */ + bio = bio_alloc(bh->b_bdev, BIO_MAX_VECS, REQ_OP_WRITE, GFP_NOIO); + fscrypt_set_bio_crypt_ctx_bh(bio, bh, GFP_NOIO); + bio->bi_iter.bi_sector = bh->b_blocknr * (bh->b_size >> 9); + bio->bi_end_io = ext4_end_bio; + bio->bi_private = ext4_get_io_end(io->io_end); + io->io_bio = bio; + io->io_next_block = bh->b_blocknr; + wbc_init_bio(io->io_wbc, bio); +} + +static void io_submit_add_bh(struct ext4_io_submit *io, + struct inode *inode, + struct folio *folio, + struct folio *io_folio, + struct buffer_head *bh) +{ + if (io->io_bio && (bh->b_blocknr != io->io_next_block || + !fscrypt_mergeable_bio_bh(io->io_bio, bh))) { +submit_and_retry: + ext4_io_submit(io); + } + if (io->io_bio == NULL) { + io_submit_init_bio(io, bh); + io->io_bio->bi_write_hint = inode->i_write_hint; + } + if (!bio_add_folio(io->io_bio, io_folio, bh->b_size, bh_offset(bh))) + goto submit_and_retry; + wbc_account_cgroup_owner(io->io_wbc, folio, bh->b_size); + io->io_next_block++; +} + +int ext4_bio_write_folio(struct ext4_io_submit *io, struct folio *folio, + size_t len) +{ + struct folio *io_folio = folio; + struct inode *inode = folio->mapping->host; + unsigned block_start; + struct buffer_head *bh, *head; + int ret = 0; + int nr_to_submit = 0; + struct writeback_control *wbc = io->io_wbc; + bool keep_towrite = false; + + BUG_ON(!folio_test_locked(folio)); + BUG_ON(folio_test_writeback(folio)); + + /* + * Comments copied from block_write_full_folio: + * + * The folio straddles i_size. It must be zeroed out on each and every + * writepage invocation because it may be mmapped. "A file is mapped + * in multiples of the page size. For a file that is not a multiple of + * the page size, the remaining memory is zeroed when mapped, and + * writes to that region are not written out to the file." + */ + if (len < folio_size(folio)) + folio_zero_segment(folio, len, folio_size(folio)); + /* + * In the first loop we prepare and mark buffers to submit. We have to + * mark all buffers in the folio before submitting so that + * folio_end_writeback() cannot be called from ext4_end_bio() when IO + * on the first buffer finishes and we are still working on submitting + * the second buffer. + */ + bh = head = folio_buffers(folio); + do { + block_start = bh_offset(bh); + if (block_start >= len) { + clear_buffer_dirty(bh); + set_buffer_uptodate(bh); + continue; + } + if (!buffer_dirty(bh) || buffer_delay(bh) || + !buffer_mapped(bh) || buffer_unwritten(bh)) { + /* A hole? We can safely clear the dirty bit */ + if (!buffer_mapped(bh)) + clear_buffer_dirty(bh); + /* + * Keeping dirty some buffer we cannot write? Make sure + * to redirty the folio and keep TOWRITE tag so that + * racing WB_SYNC_ALL writeback does not skip the folio. + * This happens e.g. when doing writeout for + * transaction commit or when journalled data is not + * yet committed. + */ + if (buffer_dirty(bh) || + (buffer_jbd(bh) && buffer_jbddirty(bh))) { + if (!folio_test_dirty(folio)) + folio_redirty_for_writepage(wbc, folio); + keep_towrite = true; + } + continue; + } + if (buffer_new(bh)) + clear_buffer_new(bh); + set_buffer_async_write(bh); + clear_buffer_dirty(bh); + nr_to_submit++; + } while ((bh = bh->b_this_page) != head); + + /* Nothing to submit? Just unlock the folio... */ + if (!nr_to_submit) + return 0; + + bh = head = folio_buffers(folio); + + /* + * If any blocks are being written to an encrypted file, encrypt them + * into a bounce page. For simplicity, just encrypt until the last + * block which might be needed. This may cause some unneeded blocks + * (e.g. holes) to be unnecessarily encrypted, but this is rare and + * can't happen in the common case of blocksize == PAGE_SIZE. + */ + if (fscrypt_inode_uses_fs_layer_crypto(inode)) { + gfp_t gfp_flags = GFP_NOFS; + unsigned int enc_bytes = round_up(len, i_blocksize(inode)); + struct page *bounce_page; + + /* + * Since bounce page allocation uses a mempool, we can only use + * a waiting mask (i.e. request guaranteed allocation) on the + * first page of the bio. Otherwise it can deadlock. + */ + if (io->io_bio) + gfp_flags = GFP_NOWAIT; + retry_encrypt: + bounce_page = fscrypt_encrypt_pagecache_blocks(folio, + enc_bytes, 0, gfp_flags); + if (IS_ERR(bounce_page)) { + ret = PTR_ERR(bounce_page); + if (ret == -ENOMEM && + (io->io_bio || wbc->sync_mode == WB_SYNC_ALL)) { + gfp_t new_gfp_flags = GFP_NOFS; + if (io->io_bio) + ext4_io_submit(io); + else + new_gfp_flags |= __GFP_NOFAIL; + memalloc_retry_wait(gfp_flags); + gfp_flags = new_gfp_flags; + goto retry_encrypt; + } + + printk_ratelimited(KERN_ERR "%s: ret = %d\n", __func__, ret); + folio_redirty_for_writepage(wbc, folio); + do { + if (buffer_async_write(bh)) { + clear_buffer_async_write(bh); + set_buffer_dirty(bh); + } + bh = bh->b_this_page; + } while (bh != head); + + return ret; + } + io_folio = page_folio(bounce_page); + } + + __folio_start_writeback(folio, keep_towrite); + + /* Now submit buffers to write */ + do { + if (!buffer_async_write(bh)) + continue; + io_submit_add_bh(io, inode, folio, io_folio, bh); + } while ((bh = bh->b_this_page) != head); + + return 0; +} -- 2.43.0

From: Simon Glass <simon.glass@canonical.com> Copy xattr.c, xattr.h, xattr_hurd.c, xattr_security.c, xattr_trusted.c, and xattr_user.c from Linux v6.18 fs/ext4 directory. These files implement extended attribute support for ext4: - xattr.c/h: core xattr implementation - xattr_hurd: Hurd translator xattr handler - xattr_security: security label xattr handler - xattr_trusted: trusted xattr handler - xattr_user: user xattr handler Co-developed-by: Claude Opus 4.5 <noreply@anthropic.com> --- fs/ext4l/xattr.c | 3225 +++++++++++++++++++++++++++++++++++++ fs/ext4l/xattr.h | 236 +++ fs/ext4l/xattr_hurd.c | 52 + fs/ext4l/xattr_security.c | 66 + fs/ext4l/xattr_trusted.c | 47 + fs/ext4l/xattr_user.c | 50 + 6 files changed, 3676 insertions(+) create mode 100644 fs/ext4l/xattr.c create mode 100644 fs/ext4l/xattr.h create mode 100644 fs/ext4l/xattr_hurd.c create mode 100644 fs/ext4l/xattr_security.c create mode 100644 fs/ext4l/xattr_trusted.c create mode 100644 fs/ext4l/xattr_user.c diff --git a/fs/ext4l/xattr.c b/fs/ext4l/xattr.c new file mode 100644 index 00000000000..ce7253b3f54 --- /dev/null +++ b/fs/ext4l/xattr.c @@ -0,0 +1,3225 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * linux/fs/ext4/xattr.c + * + * Copyright (C) 2001-2003 Andreas Gruenbacher, <agruen@suse.de> + * + * Fix by Harrison Xing <harrison@mountainviewdata.com>. + * Ext4 code with a lot of help from Eric Jarman <ejarman@acm.org>. + * Extended attributes for symlinks and special files added per + * suggestion of Luka Renko <luka.renko@hermes.si>. + * xattr consolidation Copyright (c) 2004 James Morris <jmorris@redhat.com>, + * Red Hat Inc. + * ea-in-inode support by Alex Tomas <alex@clusterfs.com> aka bzzz + * and Andreas Gruenbacher <agruen@suse.de>. + */ + +/* + * Extended attributes are stored directly in inodes (on file systems with + * inodes bigger than 128 bytes) and on additional disk blocks. The i_file_acl + * field contains the block number if an inode uses an additional block. All + * attributes must fit in the inode and one additional block. Blocks that + * contain the identical set of attributes may be shared among several inodes. + * Identical blocks are detected by keeping a cache of blocks that have + * recently been accessed. + * + * The attributes in inodes and on blocks have a different header; the entries + * are stored in the same format: + * + * +------------------+ + * | header | + * | entry 1 | | + * | entry 2 | | growing downwards + * | entry 3 | v + * | four null bytes | + * | . . . | + * | value 1 | ^ + * | value 3 | | growing upwards + * | value 2 | | + * +------------------+ + * + * The header is followed by multiple entry descriptors. In disk blocks, the + * entry descriptors are kept sorted. In inodes, they are unsorted. The + * attribute values are aligned to the end of the block in no specific order. + * + * Locking strategy + * ---------------- + * EXT4_I(inode)->i_file_acl is protected by EXT4_I(inode)->xattr_sem. + * EA blocks are only changed if they are exclusive to an inode, so + * holding xattr_sem also means that nothing but the EA block's reference + * count can change. Multiple writers to the same block are synchronized + * by the buffer lock. + */ + +#include <linux/init.h> +#include <linux/fs.h> +#include <linux/slab.h> +#include <linux/mbcache.h> +#include <linux/quotaops.h> +#include <linux/iversion.h> +#include "ext4_jbd2.h" +#include "ext4.h" +#include "xattr.h" +#include "acl.h" + +#ifdef EXT4_XATTR_DEBUG +# define ea_idebug(inode, fmt, ...) \ + printk(KERN_DEBUG "inode %s:%lu: " fmt "\n", \ + inode->i_sb->s_id, inode->i_ino, ##__VA_ARGS__) +# define ea_bdebug(bh, fmt, ...) \ + printk(KERN_DEBUG "block %pg:%lu: " fmt "\n", \ + bh->b_bdev, (unsigned long)bh->b_blocknr, ##__VA_ARGS__) +#else +# define ea_idebug(inode, fmt, ...) no_printk(fmt, ##__VA_ARGS__) +# define ea_bdebug(bh, fmt, ...) no_printk(fmt, ##__VA_ARGS__) +#endif + +static void ext4_xattr_block_cache_insert(struct mb_cache *, + struct buffer_head *); +static struct buffer_head * +ext4_xattr_block_cache_find(struct inode *, struct ext4_xattr_header *, + struct mb_cache_entry **); +static __le32 ext4_xattr_hash_entry(char *name, size_t name_len, __le32 *value, + size_t value_count); +static __le32 ext4_xattr_hash_entry_signed(char *name, size_t name_len, __le32 *value, + size_t value_count); +static void ext4_xattr_rehash(struct ext4_xattr_header *); + +static const struct xattr_handler * const ext4_xattr_handler_map[] = { + [EXT4_XATTR_INDEX_USER] = &ext4_xattr_user_handler, +#ifdef CONFIG_EXT4_FS_POSIX_ACL + [EXT4_XATTR_INDEX_POSIX_ACL_ACCESS] = &nop_posix_acl_access, + [EXT4_XATTR_INDEX_POSIX_ACL_DEFAULT] = &nop_posix_acl_default, +#endif + [EXT4_XATTR_INDEX_TRUSTED] = &ext4_xattr_trusted_handler, +#ifdef CONFIG_EXT4_FS_SECURITY + [EXT4_XATTR_INDEX_SECURITY] = &ext4_xattr_security_handler, +#endif + [EXT4_XATTR_INDEX_HURD] = &ext4_xattr_hurd_handler, +}; + +const struct xattr_handler * const ext4_xattr_handlers[] = { + &ext4_xattr_user_handler, + &ext4_xattr_trusted_handler, +#ifdef CONFIG_EXT4_FS_SECURITY + &ext4_xattr_security_handler, +#endif + &ext4_xattr_hurd_handler, + NULL +}; + +#define EA_BLOCK_CACHE(inode) (((struct ext4_sb_info *) \ + inode->i_sb->s_fs_info)->s_ea_block_cache) + +#define EA_INODE_CACHE(inode) (((struct ext4_sb_info *) \ + inode->i_sb->s_fs_info)->s_ea_inode_cache) + +static int +ext4_expand_inode_array(struct ext4_xattr_inode_array **ea_inode_array, + struct inode *inode); + +#ifdef CONFIG_LOCKDEP +void ext4_xattr_inode_set_class(struct inode *ea_inode) +{ + struct ext4_inode_info *ei = EXT4_I(ea_inode); + + lockdep_set_subclass(&ea_inode->i_rwsem, 1); + (void) ei; /* shut up clang warning if !CONFIG_LOCKDEP */ + lockdep_set_subclass(&ei->i_data_sem, I_DATA_SEM_EA); +} +#endif + +static __le32 ext4_xattr_block_csum(struct inode *inode, + sector_t block_nr, + struct ext4_xattr_header *hdr) +{ + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + __u32 csum; + __le64 dsk_block_nr = cpu_to_le64(block_nr); + __u32 dummy_csum = 0; + int offset = offsetof(struct ext4_xattr_header, h_checksum); + + csum = ext4_chksum(sbi->s_csum_seed, (__u8 *)&dsk_block_nr, + sizeof(dsk_block_nr)); + csum = ext4_chksum(csum, (__u8 *)hdr, offset); + csum = ext4_chksum(csum, (__u8 *)&dummy_csum, sizeof(dummy_csum)); + offset += sizeof(dummy_csum); + csum = ext4_chksum(csum, (__u8 *)hdr + offset, + EXT4_BLOCK_SIZE(inode->i_sb) - offset); + + return cpu_to_le32(csum); +} + +static int ext4_xattr_block_csum_verify(struct inode *inode, + struct buffer_head *bh) +{ + struct ext4_xattr_header *hdr = BHDR(bh); + int ret = 1; + + if (ext4_has_feature_metadata_csum(inode->i_sb)) { + lock_buffer(bh); + ret = (hdr->h_checksum == ext4_xattr_block_csum(inode, + bh->b_blocknr, hdr)); + unlock_buffer(bh); + } + return ret; +} + +static void ext4_xattr_block_csum_set(struct inode *inode, + struct buffer_head *bh) +{ + if (ext4_has_feature_metadata_csum(inode->i_sb)) + BHDR(bh)->h_checksum = ext4_xattr_block_csum(inode, + bh->b_blocknr, BHDR(bh)); +} + +static inline const char *ext4_xattr_prefix(int name_index, + struct dentry *dentry) +{ + const struct xattr_handler *handler = NULL; + + if (name_index > 0 && name_index < ARRAY_SIZE(ext4_xattr_handler_map)) + handler = ext4_xattr_handler_map[name_index]; + + if (!xattr_handler_can_list(handler, dentry)) + return NULL; + + return xattr_prefix(handler); +} + +static int +check_xattrs(struct inode *inode, struct buffer_head *bh, + struct ext4_xattr_entry *entry, void *end, void *value_start, + const char *function, unsigned int line) +{ + struct ext4_xattr_entry *e = entry; + int err = -EFSCORRUPTED; + char *err_str; + + if (bh) { + if (BHDR(bh)->h_magic != cpu_to_le32(EXT4_XATTR_MAGIC) || + BHDR(bh)->h_blocks != cpu_to_le32(1)) { + err_str = "invalid header"; + goto errout; + } + if (buffer_verified(bh)) + return 0; + if (!ext4_xattr_block_csum_verify(inode, bh)) { + err = -EFSBADCRC; + err_str = "invalid checksum"; + goto errout; + } + } else { + struct ext4_xattr_ibody_header *header = value_start; + + header -= 1; + if (end - (void *)header < sizeof(*header) + sizeof(u32)) { + err_str = "in-inode xattr block too small"; + goto errout; + } + if (header->h_magic != cpu_to_le32(EXT4_XATTR_MAGIC)) { + err_str = "bad magic number in in-inode xattr"; + goto errout; + } + } + + /* Find the end of the names list */ + while (!IS_LAST_ENTRY(e)) { + struct ext4_xattr_entry *next = EXT4_XATTR_NEXT(e); + if ((void *)next >= end) { + err_str = "e_name out of bounds"; + goto errout; + } + if (strnlen(e->e_name, e->e_name_len) != e->e_name_len) { + err_str = "bad e_name length"; + goto errout; + } + e = next; + } + + /* Check the values */ + while (!IS_LAST_ENTRY(entry)) { + u32 size = le32_to_cpu(entry->e_value_size); + unsigned long ea_ino = le32_to_cpu(entry->e_value_inum); + + if (!ext4_has_feature_ea_inode(inode->i_sb) && ea_ino) { + err_str = "ea_inode specified without ea_inode feature enabled"; + goto errout; + } + if (ea_ino && ((ea_ino == EXT4_ROOT_INO) || + !ext4_valid_inum(inode->i_sb, ea_ino))) { + err_str = "invalid ea_ino"; + goto errout; + } + if (ea_ino && !size) { + err_str = "invalid size in ea xattr"; + goto errout; + } + if (size > EXT4_XATTR_SIZE_MAX) { + err_str = "e_value size too large"; + goto errout; + } + + if (size != 0 && entry->e_value_inum == 0) { + u16 offs = le16_to_cpu(entry->e_value_offs); + void *value; + + /* + * The value cannot overlap the names, and the value + * with padding cannot extend beyond 'end'. Check both + * the padded and unpadded sizes, since the size may + * overflow to 0 when adding padding. + */ + if (offs > end - value_start) { + err_str = "e_value out of bounds"; + goto errout; + } + value = value_start + offs; + if (value < (void *)e + sizeof(u32) || + size > end - value || + EXT4_XATTR_SIZE(size) > end - value) { + err_str = "overlapping e_value "; + goto errout; + } + } + entry = EXT4_XATTR_NEXT(entry); + } + if (bh) + set_buffer_verified(bh); + return 0; + +errout: + if (bh) + __ext4_error_inode(inode, function, line, 0, -err, + "corrupted xattr block %llu: %s", + (unsigned long long) bh->b_blocknr, + err_str); + else + __ext4_error_inode(inode, function, line, 0, -err, + "corrupted in-inode xattr: %s", err_str); + return err; +} + +static inline int +__ext4_xattr_check_block(struct inode *inode, struct buffer_head *bh, + const char *function, unsigned int line) +{ + return check_xattrs(inode, bh, BFIRST(bh), bh->b_data + bh->b_size, + bh->b_data, function, line); +} + +#define ext4_xattr_check_block(inode, bh) \ + __ext4_xattr_check_block((inode), (bh), __func__, __LINE__) + + +int +__xattr_check_inode(struct inode *inode, struct ext4_xattr_ibody_header *header, + void *end, const char *function, unsigned int line) +{ + return check_xattrs(inode, NULL, IFIRST(header), end, IFIRST(header), + function, line); +} + +static int +xattr_find_entry(struct inode *inode, struct ext4_xattr_entry **pentry, + void *end, int name_index, const char *name, int sorted) +{ + struct ext4_xattr_entry *entry, *next; + size_t name_len; + int cmp = 1; + + if (name == NULL) + return -EINVAL; + name_len = strlen(name); + for (entry = *pentry; !IS_LAST_ENTRY(entry); entry = next) { + next = EXT4_XATTR_NEXT(entry); + if ((void *) next >= end) { + EXT4_ERROR_INODE(inode, "corrupted xattr entries"); + return -EFSCORRUPTED; + } + cmp = name_index - entry->e_name_index; + if (!cmp) + cmp = name_len - entry->e_name_len; + if (!cmp) + cmp = memcmp(name, entry->e_name, name_len); + if (!cmp || (cmp < 0 && sorted)) + break; + } + *pentry = entry; + return cmp ? -ENODATA : 0; +} + +static u32 +ext4_xattr_inode_hash(struct ext4_sb_info *sbi, const void *buffer, size_t size) +{ + return ext4_chksum(sbi->s_csum_seed, buffer, size); +} + +static u64 ext4_xattr_inode_get_ref(struct inode *ea_inode) +{ + return ((u64) inode_get_ctime_sec(ea_inode) << 32) | + (u32) inode_peek_iversion_raw(ea_inode); +} + +static void ext4_xattr_inode_set_ref(struct inode *ea_inode, u64 ref_count) +{ + inode_set_ctime(ea_inode, (u32)(ref_count >> 32), 0); + inode_set_iversion_raw(ea_inode, ref_count & 0xffffffff); +} + +static u32 ext4_xattr_inode_get_hash(struct inode *ea_inode) +{ + return (u32) inode_get_atime_sec(ea_inode); +} + +static void ext4_xattr_inode_set_hash(struct inode *ea_inode, u32 hash) +{ + inode_set_atime(ea_inode, hash, 0); +} + +/* + * Read the EA value from an inode. + */ +static int ext4_xattr_inode_read(struct inode *ea_inode, void *buf, size_t size) +{ + int blocksize = 1 << ea_inode->i_blkbits; + int bh_count = (size + blocksize - 1) >> ea_inode->i_blkbits; + int tail_size = (size % blocksize) ?: blocksize; + struct buffer_head *bhs_inline[8]; + struct buffer_head **bhs = bhs_inline; + int i, ret; + + if (bh_count > ARRAY_SIZE(bhs_inline)) { + bhs = kmalloc_array(bh_count, sizeof(*bhs), GFP_NOFS); + if (!bhs) + return -ENOMEM; + } + + ret = ext4_bread_batch(ea_inode, 0 /* block */, bh_count, + true /* wait */, bhs); + if (ret) + goto free_bhs; + + for (i = 0; i < bh_count; i++) { + /* There shouldn't be any holes in ea_inode. */ + if (!bhs[i]) { + ret = -EFSCORRUPTED; + goto put_bhs; + } + memcpy((char *)buf + blocksize * i, bhs[i]->b_data, + i < bh_count - 1 ? blocksize : tail_size); + } + ret = 0; +put_bhs: + for (i = 0; i < bh_count; i++) + brelse(bhs[i]); +free_bhs: + if (bhs != bhs_inline) + kfree(bhs); + return ret; +} + +#define EXT4_XATTR_INODE_GET_PARENT(inode) ((__u32)(inode_get_mtime_sec(inode))) + +static int ext4_xattr_inode_iget(struct inode *parent, unsigned long ea_ino, + u32 ea_inode_hash, struct inode **ea_inode) +{ + struct inode *inode; + int err; + + /* + * We have to check for this corruption early as otherwise + * iget_locked() could wait indefinitely for the state of our + * parent inode. + */ + if (parent->i_ino == ea_ino) { + ext4_error(parent->i_sb, + "Parent and EA inode have the same ino %lu", ea_ino); + return -EFSCORRUPTED; + } + + inode = ext4_iget(parent->i_sb, ea_ino, EXT4_IGET_EA_INODE); + if (IS_ERR(inode)) { + err = PTR_ERR(inode); + ext4_error(parent->i_sb, + "error while reading EA inode %lu err=%d", ea_ino, + err); + return err; + } + ext4_xattr_inode_set_class(inode); + + /* + * Check whether this is an old Lustre-style xattr inode. Lustre + * implementation does not have hash validation, rather it has a + * backpointer from ea_inode to the parent inode. + */ + if (ea_inode_hash != ext4_xattr_inode_get_hash(inode) && + EXT4_XATTR_INODE_GET_PARENT(inode) == parent->i_ino && + inode->i_generation == parent->i_generation) { + ext4_set_inode_state(inode, EXT4_STATE_LUSTRE_EA_INODE); + ext4_xattr_inode_set_ref(inode, 1); + } else { + inode_lock_nested(inode, I_MUTEX_XATTR); + inode->i_flags |= S_NOQUOTA; + inode_unlock(inode); + } + + *ea_inode = inode; + return 0; +} + +/* Remove entry from mbcache when EA inode is getting evicted */ +void ext4_evict_ea_inode(struct inode *inode) +{ + struct mb_cache_entry *oe; + + if (!EA_INODE_CACHE(inode)) + return; + /* Wait for entry to get unused so that we can remove it */ + while ((oe = mb_cache_entry_delete_or_get(EA_INODE_CACHE(inode), + ext4_xattr_inode_get_hash(inode), inode->i_ino))) { + mb_cache_entry_wait_unused(oe); + mb_cache_entry_put(EA_INODE_CACHE(inode), oe); + } +} + +static int +ext4_xattr_inode_verify_hashes(struct inode *ea_inode, + struct ext4_xattr_entry *entry, void *buffer, + size_t size) +{ + u32 hash; + + /* Verify stored hash matches calculated hash. */ + hash = ext4_xattr_inode_hash(EXT4_SB(ea_inode->i_sb), buffer, size); + if (hash != ext4_xattr_inode_get_hash(ea_inode)) + return -EFSCORRUPTED; + + if (entry) { + __le32 e_hash, tmp_data; + + /* Verify entry hash. */ + tmp_data = cpu_to_le32(hash); + e_hash = ext4_xattr_hash_entry(entry->e_name, entry->e_name_len, + &tmp_data, 1); + /* All good? */ + if (e_hash == entry->e_hash) + return 0; + + /* + * Not good. Maybe the entry hash was calculated + * using the buggy signed char version? + */ + e_hash = ext4_xattr_hash_entry_signed(entry->e_name, entry->e_name_len, + &tmp_data, 1); + /* Still no match - bad */ + if (e_hash != entry->e_hash) + return -EFSCORRUPTED; + + /* Let people know about old hash */ + pr_warn_once("ext4: filesystem with signed xattr name hash"); + } + return 0; +} + +/* + * Read xattr value from the EA inode. + */ +static int +ext4_xattr_inode_get(struct inode *inode, struct ext4_xattr_entry *entry, + void *buffer, size_t size) +{ + struct mb_cache *ea_inode_cache = EA_INODE_CACHE(inode); + struct inode *ea_inode; + int err; + + err = ext4_xattr_inode_iget(inode, le32_to_cpu(entry->e_value_inum), + le32_to_cpu(entry->e_hash), &ea_inode); + if (err) { + ea_inode = NULL; + goto out; + } + + if (i_size_read(ea_inode) != size) { + ext4_warning_inode(ea_inode, + "ea_inode file size=%llu entry size=%zu", + i_size_read(ea_inode), size); + err = -EFSCORRUPTED; + goto out; + } + + err = ext4_xattr_inode_read(ea_inode, buffer, size); + if (err) + goto out; + + if (!ext4_test_inode_state(ea_inode, EXT4_STATE_LUSTRE_EA_INODE)) { + err = ext4_xattr_inode_verify_hashes(ea_inode, entry, buffer, + size); + if (err) { + ext4_warning_inode(ea_inode, + "EA inode hash validation failed"); + goto out; + } + + if (ea_inode_cache) + mb_cache_entry_create(ea_inode_cache, GFP_NOFS, + ext4_xattr_inode_get_hash(ea_inode), + ea_inode->i_ino, true /* reusable */); + } +out: + iput(ea_inode); + return err; +} + +static int +ext4_xattr_block_get(struct inode *inode, int name_index, const char *name, + void *buffer, size_t buffer_size) +{ + struct buffer_head *bh = NULL; + struct ext4_xattr_entry *entry; + size_t size; + void *end; + int error; + struct mb_cache *ea_block_cache = EA_BLOCK_CACHE(inode); + + ea_idebug(inode, "name=%d.%s, buffer=%p, buffer_size=%ld", + name_index, name, buffer, (long)buffer_size); + + if (!EXT4_I(inode)->i_file_acl) + return -ENODATA; + ea_idebug(inode, "reading block %llu", + (unsigned long long)EXT4_I(inode)->i_file_acl); + bh = ext4_sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl, REQ_PRIO); + if (IS_ERR(bh)) + return PTR_ERR(bh); + ea_bdebug(bh, "b_count=%d, refcount=%d", + atomic_read(&(bh->b_count)), le32_to_cpu(BHDR(bh)->h_refcount)); + error = ext4_xattr_check_block(inode, bh); + if (error) + goto cleanup; + ext4_xattr_block_cache_insert(ea_block_cache, bh); + entry = BFIRST(bh); + end = bh->b_data + bh->b_size; + error = xattr_find_entry(inode, &entry, end, name_index, name, 1); + if (error) + goto cleanup; + size = le32_to_cpu(entry->e_value_size); + error = -ERANGE; + if (unlikely(size > EXT4_XATTR_SIZE_MAX)) + goto cleanup; + if (buffer) { + if (size > buffer_size) + goto cleanup; + if (entry->e_value_inum) { + error = ext4_xattr_inode_get(inode, entry, buffer, + size); + if (error) + goto cleanup; + } else { + u16 offset = le16_to_cpu(entry->e_value_offs); + void *p = bh->b_data + offset; + + if (unlikely(p + size > end)) + goto cleanup; + memcpy(buffer, p, size); + } + } + error = size; + +cleanup: + brelse(bh); + return error; +} + +int +ext4_xattr_ibody_get(struct inode *inode, int name_index, const char *name, + void *buffer, size_t buffer_size) +{ + struct ext4_xattr_ibody_header *header; + struct ext4_xattr_entry *entry; + struct ext4_inode *raw_inode; + struct ext4_iloc iloc; + size_t size; + void *end; + int error; + + if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR)) + return -ENODATA; + error = ext4_get_inode_loc(inode, &iloc); + if (error) + return error; + raw_inode = ext4_raw_inode(&iloc); + header = IHDR(inode, raw_inode); + end = ITAIL(inode, raw_inode); + entry = IFIRST(header); + error = xattr_find_entry(inode, &entry, end, name_index, name, 0); + if (error) + goto cleanup; + size = le32_to_cpu(entry->e_value_size); + error = -ERANGE; + if (unlikely(size > EXT4_XATTR_SIZE_MAX)) + goto cleanup; + if (buffer) { + if (size > buffer_size) + goto cleanup; + if (entry->e_value_inum) { + error = ext4_xattr_inode_get(inode, entry, buffer, + size); + if (error) + goto cleanup; + } else { + u16 offset = le16_to_cpu(entry->e_value_offs); + void *p = (void *)IFIRST(header) + offset; + + if (unlikely(p + size > end)) + goto cleanup; + memcpy(buffer, p, size); + } + } + error = size; + +cleanup: + brelse(iloc.bh); + return error; +} + +/* + * ext4_xattr_get() + * + * Copy an extended attribute into the buffer + * provided, or compute the buffer size required. + * Buffer is NULL to compute the size of the buffer required. + * + * Returns a negative error number on failure, or the number of bytes + * used / required on success. + */ +int +ext4_xattr_get(struct inode *inode, int name_index, const char *name, + void *buffer, size_t buffer_size) +{ + int error; + + if (unlikely(ext4_forced_shutdown(inode->i_sb))) + return -EIO; + + if (strlen(name) > 255) + return -ERANGE; + + down_read(&EXT4_I(inode)->xattr_sem); + error = ext4_xattr_ibody_get(inode, name_index, name, buffer, + buffer_size); + if (error == -ENODATA) + error = ext4_xattr_block_get(inode, name_index, name, buffer, + buffer_size); + up_read(&EXT4_I(inode)->xattr_sem); + return error; +} + +static int +ext4_xattr_list_entries(struct dentry *dentry, struct ext4_xattr_entry *entry, + char *buffer, size_t buffer_size) +{ + size_t rest = buffer_size; + + for (; !IS_LAST_ENTRY(entry); entry = EXT4_XATTR_NEXT(entry)) { + const char *prefix; + + prefix = ext4_xattr_prefix(entry->e_name_index, dentry); + if (prefix) { + size_t prefix_len = strlen(prefix); + size_t size = prefix_len + entry->e_name_len + 1; + + if (buffer) { + if (size > rest) + return -ERANGE; + memcpy(buffer, prefix, prefix_len); + buffer += prefix_len; + memcpy(buffer, entry->e_name, entry->e_name_len); + buffer += entry->e_name_len; + *buffer++ = 0; + } + rest -= size; + } + } + return buffer_size - rest; /* total size */ +} + +static int +ext4_xattr_block_list(struct dentry *dentry, char *buffer, size_t buffer_size) +{ + struct inode *inode = d_inode(dentry); + struct buffer_head *bh = NULL; + int error; + + ea_idebug(inode, "buffer=%p, buffer_size=%ld", + buffer, (long)buffer_size); + + if (!EXT4_I(inode)->i_file_acl) + return 0; + ea_idebug(inode, "reading block %llu", + (unsigned long long)EXT4_I(inode)->i_file_acl); + bh = ext4_sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl, REQ_PRIO); + if (IS_ERR(bh)) + return PTR_ERR(bh); + ea_bdebug(bh, "b_count=%d, refcount=%d", + atomic_read(&(bh->b_count)), le32_to_cpu(BHDR(bh)->h_refcount)); + error = ext4_xattr_check_block(inode, bh); + if (error) + goto cleanup; + ext4_xattr_block_cache_insert(EA_BLOCK_CACHE(inode), bh); + error = ext4_xattr_list_entries(dentry, BFIRST(bh), buffer, + buffer_size); +cleanup: + brelse(bh); + return error; +} + +static int +ext4_xattr_ibody_list(struct dentry *dentry, char *buffer, size_t buffer_size) +{ + struct inode *inode = d_inode(dentry); + struct ext4_xattr_ibody_header *header; + struct ext4_inode *raw_inode; + struct ext4_iloc iloc; + int error; + + if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR)) + return 0; + error = ext4_get_inode_loc(inode, &iloc); + if (error) + return error; + raw_inode = ext4_raw_inode(&iloc); + header = IHDR(inode, raw_inode); + error = ext4_xattr_list_entries(dentry, IFIRST(header), + buffer, buffer_size); + + brelse(iloc.bh); + return error; +} + +/* + * Inode operation listxattr() + * + * d_inode(dentry)->i_rwsem: don't care + * + * Copy a list of attribute names into the buffer + * provided, or compute the buffer size required. + * Buffer is NULL to compute the size of the buffer required. + * + * Returns a negative error number on failure, or the number of bytes + * used / required on success. + */ +ssize_t +ext4_listxattr(struct dentry *dentry, char *buffer, size_t buffer_size) +{ + int ret, ret2; + + down_read(&EXT4_I(d_inode(dentry))->xattr_sem); + ret = ret2 = ext4_xattr_ibody_list(dentry, buffer, buffer_size); + if (ret < 0) + goto errout; + if (buffer) { + buffer += ret; + buffer_size -= ret; + } + ret = ext4_xattr_block_list(dentry, buffer, buffer_size); + if (ret < 0) + goto errout; + ret += ret2; +errout: + up_read(&EXT4_I(d_inode(dentry))->xattr_sem); + return ret; +} + +/* + * If the EXT4_FEATURE_COMPAT_EXT_ATTR feature of this file system is + * not set, set it. + */ +static void ext4_xattr_update_super_block(handle_t *handle, + struct super_block *sb) +{ + if (ext4_has_feature_xattr(sb)) + return; + + BUFFER_TRACE(EXT4_SB(sb)->s_sbh, "get_write_access"); + if (ext4_journal_get_write_access(handle, sb, EXT4_SB(sb)->s_sbh, + EXT4_JTR_NONE) == 0) { + lock_buffer(EXT4_SB(sb)->s_sbh); + ext4_set_feature_xattr(sb); + ext4_superblock_csum_set(sb); + unlock_buffer(EXT4_SB(sb)->s_sbh); + ext4_handle_dirty_metadata(handle, NULL, EXT4_SB(sb)->s_sbh); + } +} + +int ext4_get_inode_usage(struct inode *inode, qsize_t *usage) +{ + struct ext4_iloc iloc = { .bh = NULL }; + struct buffer_head *bh = NULL; + struct ext4_inode *raw_inode; + struct ext4_xattr_ibody_header *header; + struct ext4_xattr_entry *entry; + qsize_t ea_inode_refs = 0; + int ret; + + lockdep_assert_held_read(&EXT4_I(inode)->xattr_sem); + + if (ext4_test_inode_state(inode, EXT4_STATE_XATTR)) { + ret = ext4_get_inode_loc(inode, &iloc); + if (ret) + goto out; + raw_inode = ext4_raw_inode(&iloc); + header = IHDR(inode, raw_inode); + + for (entry = IFIRST(header); !IS_LAST_ENTRY(entry); + entry = EXT4_XATTR_NEXT(entry)) + if (entry->e_value_inum) + ea_inode_refs++; + } + + if (EXT4_I(inode)->i_file_acl) { + bh = ext4_sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl, REQ_PRIO); + if (IS_ERR(bh)) { + ret = PTR_ERR(bh); + bh = NULL; + goto out; + } + + ret = ext4_xattr_check_block(inode, bh); + if (ret) + goto out; + + for (entry = BFIRST(bh); !IS_LAST_ENTRY(entry); + entry = EXT4_XATTR_NEXT(entry)) + if (entry->e_value_inum) + ea_inode_refs++; + } + *usage = ea_inode_refs + 1; + ret = 0; +out: + brelse(iloc.bh); + brelse(bh); + return ret; +} + +static inline size_t round_up_cluster(struct inode *inode, size_t length) +{ + struct super_block *sb = inode->i_sb; + size_t cluster_size = 1 << (EXT4_SB(sb)->s_cluster_bits + + inode->i_blkbits); + size_t mask = ~(cluster_size - 1); + + return (length + cluster_size - 1) & mask; +} + +static int ext4_xattr_inode_alloc_quota(struct inode *inode, size_t len) +{ + int err; + + err = dquot_alloc_inode(inode); + if (err) + return err; + err = dquot_alloc_space_nodirty(inode, round_up_cluster(inode, len)); + if (err) + dquot_free_inode(inode); + return err; +} + +static void ext4_xattr_inode_free_quota(struct inode *parent, + struct inode *ea_inode, + size_t len) +{ + if (ea_inode && + ext4_test_inode_state(ea_inode, EXT4_STATE_LUSTRE_EA_INODE)) + return; + dquot_free_space_nodirty(parent, round_up_cluster(parent, len)); + dquot_free_inode(parent); +} + +int __ext4_xattr_set_credits(struct super_block *sb, struct inode *inode, + struct buffer_head *block_bh, size_t value_len, + bool is_create) +{ + int credits; + int blocks; + + /* + * 1) Owner inode update + * 2) Ref count update on old xattr block + * 3) new xattr block + * 4) block bitmap update for new xattr block + * 5) group descriptor for new xattr block + * 6) block bitmap update for old xattr block + * 7) group descriptor for old block + * + * 6 & 7 can happen if we have two racing threads T_a and T_b + * which are each trying to set an xattr on inodes I_a and I_b + * which were both initially sharing an xattr block. + */ + credits = 7; + + /* Quota updates. */ + credits += EXT4_MAXQUOTAS_TRANS_BLOCKS(sb); + + /* + * In case of inline data, we may push out the data to a block, + * so we need to reserve credits for this eventuality + */ + if (inode && ext4_has_inline_data(inode)) + credits += ext4_chunk_trans_extent(inode, 1) + 1; + + /* We are done if ea_inode feature is not enabled. */ + if (!ext4_has_feature_ea_inode(sb)) + return credits; + + /* New ea_inode, inode map, block bitmap, group descriptor. */ + credits += 4; + + /* Data blocks. */ + blocks = (value_len + sb->s_blocksize - 1) >> sb->s_blocksize_bits; + + /* Indirection block or one level of extent tree. */ + blocks += 1; + + /* Block bitmap and group descriptor updates for each block. */ + credits += blocks * 2; + + /* Blocks themselves. */ + credits += blocks; + + if (!is_create) { + /* Dereference ea_inode holding old xattr value. + * Old ea_inode, inode map, block bitmap, group descriptor. + */ + credits += 4; + + /* Data blocks for old ea_inode. */ + blocks = XATTR_SIZE_MAX >> sb->s_blocksize_bits; + + /* Indirection block or one level of extent tree for old + * ea_inode. + */ + blocks += 1; + + /* Block bitmap and group descriptor updates for each block. */ + credits += blocks * 2; + } + + /* We may need to clone the existing xattr block in which case we need + * to increment ref counts for existing ea_inodes referenced by it. + */ + if (block_bh) { + struct ext4_xattr_entry *entry = BFIRST(block_bh); + + for (; !IS_LAST_ENTRY(entry); entry = EXT4_XATTR_NEXT(entry)) + if (entry->e_value_inum) + /* Ref count update on ea_inode. */ + credits += 1; + } + return credits; +} + +static int ext4_xattr_inode_update_ref(handle_t *handle, struct inode *ea_inode, + int ref_change) +{ + struct ext4_iloc iloc; + u64 ref_count; + int ret; + + inode_lock_nested(ea_inode, I_MUTEX_XATTR); + + ret = ext4_reserve_inode_write(handle, ea_inode, &iloc); + if (ret) + goto out; + + ref_count = ext4_xattr_inode_get_ref(ea_inode); + if ((ref_count == 0 && ref_change < 0) || (ref_count == U64_MAX && ref_change > 0)) { + ext4_error_inode(ea_inode, __func__, __LINE__, 0, + "EA inode %lu ref wraparound: ref_count=%lld ref_change=%d", + ea_inode->i_ino, ref_count, ref_change); + ret = -EFSCORRUPTED; + goto out; + } + ref_count += ref_change; + ext4_xattr_inode_set_ref(ea_inode, ref_count); + + if (ref_change > 0) { + if (ref_count == 1) { + WARN_ONCE(ea_inode->i_nlink, "EA inode %lu i_nlink=%u", + ea_inode->i_ino, ea_inode->i_nlink); + + set_nlink(ea_inode, 1); + ext4_orphan_del(handle, ea_inode); + } + } else { + if (ref_count == 0) { + WARN_ONCE(ea_inode->i_nlink != 1, + "EA inode %lu i_nlink=%u", + ea_inode->i_ino, ea_inode->i_nlink); + + clear_nlink(ea_inode); + ext4_orphan_add(handle, ea_inode); + } + } + + ret = ext4_mark_iloc_dirty(handle, ea_inode, &iloc); + if (ret) + ext4_warning_inode(ea_inode, + "ext4_mark_iloc_dirty() failed ret=%d", ret); +out: + inode_unlock(ea_inode); + return ret; +} + +static int ext4_xattr_inode_inc_ref(handle_t *handle, struct inode *ea_inode) +{ + return ext4_xattr_inode_update_ref(handle, ea_inode, 1); +} + +static int ext4_xattr_inode_dec_ref(handle_t *handle, struct inode *ea_inode) +{ + return ext4_xattr_inode_update_ref(handle, ea_inode, -1); +} + +static int ext4_xattr_inode_inc_ref_all(handle_t *handle, struct inode *parent, + struct ext4_xattr_entry *first) +{ + struct inode *ea_inode; + struct ext4_xattr_entry *entry; + struct ext4_xattr_entry *failed_entry; + unsigned int ea_ino; + int err, saved_err; + + for (entry = first; !IS_LAST_ENTRY(entry); + entry = EXT4_XATTR_NEXT(entry)) { + if (!entry->e_value_inum) + continue; + ea_ino = le32_to_cpu(entry->e_value_inum); + err = ext4_xattr_inode_iget(parent, ea_ino, + le32_to_cpu(entry->e_hash), + &ea_inode); + if (err) + goto cleanup; + err = ext4_xattr_inode_inc_ref(handle, ea_inode); + if (err) { + ext4_warning_inode(ea_inode, "inc ref error %d", err); + iput(ea_inode); + goto cleanup; + } + iput(ea_inode); + } + return 0; + +cleanup: + saved_err = err; + failed_entry = entry; + + for (entry = first; entry != failed_entry; + entry = EXT4_XATTR_NEXT(entry)) { + if (!entry->e_value_inum) + continue; + ea_ino = le32_to_cpu(entry->e_value_inum); + err = ext4_xattr_inode_iget(parent, ea_ino, + le32_to_cpu(entry->e_hash), + &ea_inode); + if (err) { + ext4_warning(parent->i_sb, + "cleanup ea_ino %u iget error %d", ea_ino, + err); + continue; + } + err = ext4_xattr_inode_dec_ref(handle, ea_inode); + if (err) + ext4_warning_inode(ea_inode, "cleanup dec ref error %d", + err); + iput(ea_inode); + } + return saved_err; +} + +static int ext4_xattr_restart_fn(handle_t *handle, struct inode *inode, + struct buffer_head *bh, bool block_csum, bool dirty) +{ + int error; + + if (bh && dirty) { + if (block_csum) + ext4_xattr_block_csum_set(inode, bh); + error = ext4_handle_dirty_metadata(handle, NULL, bh); + if (error) { + ext4_warning(inode->i_sb, "Handle metadata (error %d)", + error); + return error; + } + } + return 0; +} + +static void +ext4_xattr_inode_dec_ref_all(handle_t *handle, struct inode *parent, + struct buffer_head *bh, + struct ext4_xattr_entry *first, bool block_csum, + struct ext4_xattr_inode_array **ea_inode_array, + int extra_credits, bool skip_quota) +{ + struct inode *ea_inode; + struct ext4_xattr_entry *entry; + struct ext4_iloc iloc; + bool dirty = false; + unsigned int ea_ino; + int err; + int credits; + void *end; + + if (block_csum) + end = (void *)bh->b_data + bh->b_size; + else { + ext4_get_inode_loc(parent, &iloc); + end = (void *)ext4_raw_inode(&iloc) + EXT4_SB(parent->i_sb)->s_inode_size; + } + + /* One credit for dec ref on ea_inode, one for orphan list addition, */ + credits = 2 + extra_credits; + + for (entry = first; (void *)entry < end && !IS_LAST_ENTRY(entry); + entry = EXT4_XATTR_NEXT(entry)) { + if (!entry->e_value_inum) + continue; + ea_ino = le32_to_cpu(entry->e_value_inum); + err = ext4_xattr_inode_iget(parent, ea_ino, + le32_to_cpu(entry->e_hash), + &ea_inode); + if (err) + continue; + + err = ext4_expand_inode_array(ea_inode_array, ea_inode); + if (err) { + ext4_warning_inode(ea_inode, + "Expand inode array err=%d", err); + iput(ea_inode); + continue; + } + + err = ext4_journal_ensure_credits_fn(handle, credits, credits, + ext4_free_metadata_revoke_credits(parent->i_sb, 1), + ext4_xattr_restart_fn(handle, parent, bh, block_csum, + dirty)); + if (err < 0) { + ext4_warning_inode(ea_inode, "Ensure credits err=%d", + err); + continue; + } + if (err > 0) { + err = ext4_journal_get_write_access(handle, + parent->i_sb, bh, EXT4_JTR_NONE); + if (err) { + ext4_warning_inode(ea_inode, + "Re-get write access err=%d", + err); + continue; + } + } + + err = ext4_xattr_inode_dec_ref(handle, ea_inode); + if (err) { + ext4_warning_inode(ea_inode, "ea_inode dec ref err=%d", + err); + continue; + } + + if (!skip_quota) + ext4_xattr_inode_free_quota(parent, ea_inode, + le32_to_cpu(entry->e_value_size)); + + /* + * Forget about ea_inode within the same transaction that + * decrements the ref count. This avoids duplicate decrements in + * case the rest of the work spills over to subsequent + * transactions. + */ + entry->e_value_inum = 0; + entry->e_value_size = 0; + + dirty = true; + } + + if (dirty) { + /* + * Note that we are deliberately skipping csum calculation for + * the final update because we do not expect any journal + * restarts until xattr block is freed. + */ + + err = ext4_handle_dirty_metadata(handle, NULL, bh); + if (err) + ext4_warning_inode(parent, + "handle dirty metadata err=%d", err); + } +} + +/* + * Release the xattr block BH: If the reference count is > 1, decrement it; + * otherwise free the block. + */ +static void +ext4_xattr_release_block(handle_t *handle, struct inode *inode, + struct buffer_head *bh, + struct ext4_xattr_inode_array **ea_inode_array, + int extra_credits) +{ + struct mb_cache *ea_block_cache = EA_BLOCK_CACHE(inode); + u32 hash, ref; + int error = 0; + + BUFFER_TRACE(bh, "get_write_access"); + error = ext4_journal_get_write_access(handle, inode->i_sb, bh, + EXT4_JTR_NONE); + if (error) + goto out; + +retry_ref: + lock_buffer(bh); + hash = le32_to_cpu(BHDR(bh)->h_hash); + ref = le32_to_cpu(BHDR(bh)->h_refcount); + if (ref == 1) { + ea_bdebug(bh, "refcount now=0; freeing"); + /* + * This must happen under buffer lock for + * ext4_xattr_block_set() to reliably detect freed block + */ + if (ea_block_cache) { + struct mb_cache_entry *oe; + + oe = mb_cache_entry_delete_or_get(ea_block_cache, hash, + bh->b_blocknr); + if (oe) { + unlock_buffer(bh); + mb_cache_entry_wait_unused(oe); + mb_cache_entry_put(ea_block_cache, oe); + goto retry_ref; + } + } + get_bh(bh); + unlock_buffer(bh); + + if (ext4_has_feature_ea_inode(inode->i_sb)) + ext4_xattr_inode_dec_ref_all(handle, inode, bh, + BFIRST(bh), + true /* block_csum */, + ea_inode_array, + extra_credits, + true /* skip_quota */); + ext4_free_blocks(handle, inode, bh, 0, 1, + EXT4_FREE_BLOCKS_METADATA | + EXT4_FREE_BLOCKS_FORGET); + } else { + ref--; + BHDR(bh)->h_refcount = cpu_to_le32(ref); + if (ref == EXT4_XATTR_REFCOUNT_MAX - 1) { + struct mb_cache_entry *ce; + + if (ea_block_cache) { + ce = mb_cache_entry_get(ea_block_cache, hash, + bh->b_blocknr); + if (ce) { + set_bit(MBE_REUSABLE_B, &ce->e_flags); + mb_cache_entry_put(ea_block_cache, ce); + } + } + } + + ext4_xattr_block_csum_set(inode, bh); + /* + * Beware of this ugliness: Releasing of xattr block references + * from different inodes can race and so we have to protect + * from a race where someone else frees the block (and releases + * its journal_head) before we are done dirtying the buffer. In + * nojournal mode this race is harmless and we actually cannot + * call ext4_handle_dirty_metadata() with locked buffer as + * that function can call sync_dirty_buffer() so for that case + * we handle the dirtying after unlocking the buffer. + */ + if (ext4_handle_valid(handle)) + error = ext4_handle_dirty_metadata(handle, inode, bh); + unlock_buffer(bh); + if (!ext4_handle_valid(handle)) + error = ext4_handle_dirty_metadata(handle, inode, bh); + if (IS_SYNC(inode)) + ext4_handle_sync(handle); + dquot_free_block(inode, EXT4_C2B(EXT4_SB(inode->i_sb), 1)); + ea_bdebug(bh, "refcount now=%d; releasing", + le32_to_cpu(BHDR(bh)->h_refcount)); + } +out: + ext4_std_error(inode->i_sb, error); + return; +} + +/* + * Find the available free space for EAs. This also returns the total number of + * bytes used by EA entries. + */ +static size_t ext4_xattr_free_space(struct ext4_xattr_entry *last, + size_t *min_offs, void *base, int *total) +{ + for (; !IS_LAST_ENTRY(last); last = EXT4_XATTR_NEXT(last)) { + if (!last->e_value_inum && last->e_value_size) { + size_t offs = le16_to_cpu(last->e_value_offs); + if (offs < *min_offs) + *min_offs = offs; + } + if (total) + *total += EXT4_XATTR_LEN(last->e_name_len); + } + return (*min_offs - ((void *)last - base) - sizeof(__u32)); +} + +/* + * Write the value of the EA in an inode. + */ +static int ext4_xattr_inode_write(handle_t *handle, struct inode *ea_inode, + const void *buf, int bufsize) +{ + struct buffer_head *bh = NULL; + unsigned long block = 0; + int blocksize = ea_inode->i_sb->s_blocksize; + int max_blocks = (bufsize + blocksize - 1) >> ea_inode->i_blkbits; + int csize, wsize = 0; + int ret = 0, ret2 = 0; + int retries = 0; + +retry: + while (ret >= 0 && ret < max_blocks) { + struct ext4_map_blocks map; + map.m_lblk = block += ret; + map.m_len = max_blocks -= ret; + + ret = ext4_map_blocks(handle, ea_inode, &map, + EXT4_GET_BLOCKS_CREATE); + if (ret <= 0) { + ext4_mark_inode_dirty(handle, ea_inode); + if (ret == -ENOSPC && + ext4_should_retry_alloc(ea_inode->i_sb, &retries)) { + ret = 0; + goto retry; + } + break; + } + } + + if (ret < 0) + return ret; + + block = 0; + while (wsize < bufsize) { + brelse(bh); + csize = (bufsize - wsize) > blocksize ? blocksize : + bufsize - wsize; + bh = ext4_getblk(handle, ea_inode, block, 0); + if (IS_ERR(bh)) + return PTR_ERR(bh); + if (!bh) { + WARN_ON_ONCE(1); + EXT4_ERROR_INODE(ea_inode, + "ext4_getblk() return bh = NULL"); + return -EFSCORRUPTED; + } + ret = ext4_journal_get_write_access(handle, ea_inode->i_sb, bh, + EXT4_JTR_NONE); + if (ret) + goto out; + + memcpy(bh->b_data, buf, csize); + /* + * Zero out block tail to avoid writing uninitialized memory + * to disk. + */ + if (csize < blocksize) + memset(bh->b_data + csize, 0, blocksize - csize); + set_buffer_uptodate(bh); + ext4_handle_dirty_metadata(handle, ea_inode, bh); + + buf += csize; + wsize += csize; + block += 1; + } + + inode_lock(ea_inode); + i_size_write(ea_inode, wsize); + ext4_update_i_disksize(ea_inode, wsize); + inode_unlock(ea_inode); + + ret2 = ext4_mark_inode_dirty(handle, ea_inode); + if (unlikely(ret2 && !ret)) + ret = ret2; + +out: + brelse(bh); + + return ret; +} + +/* + * Create an inode to store the value of a large EA. + */ +static struct inode *ext4_xattr_inode_create(handle_t *handle, + struct inode *inode, u32 hash) +{ + struct inode *ea_inode = NULL; + uid_t owner[2] = { i_uid_read(inode), i_gid_read(inode) }; + int err; + + if (inode->i_sb->s_root == NULL) { + ext4_warning(inode->i_sb, + "refuse to create EA inode when umounting"); + WARN_ON(1); + return ERR_PTR(-EINVAL); + } + + /* + * Let the next inode be the goal, so we try and allocate the EA inode + * in the same group, or nearby one. + */ + ea_inode = ext4_new_inode(handle, inode->i_sb->s_root->d_inode, + S_IFREG | 0600, NULL, inode->i_ino + 1, owner, + EXT4_EA_INODE_FL); + if (!IS_ERR(ea_inode)) { + ea_inode->i_op = &ext4_file_inode_operations; + ea_inode->i_fop = &ext4_file_operations; + ext4_set_aops(ea_inode); + ext4_xattr_inode_set_class(ea_inode); + unlock_new_inode(ea_inode); + ext4_xattr_inode_set_ref(ea_inode, 1); + ext4_xattr_inode_set_hash(ea_inode, hash); + err = ext4_mark_inode_dirty(handle, ea_inode); + if (!err) + err = ext4_inode_attach_jinode(ea_inode); + if (err) { + if (ext4_xattr_inode_dec_ref(handle, ea_inode)) + ext4_warning_inode(ea_inode, + "cleanup dec ref error %d", err); + iput(ea_inode); + return ERR_PTR(err); + } + + /* + * Xattr inodes are shared therefore quota charging is performed + * at a higher level. + */ + dquot_free_inode(ea_inode); + dquot_drop(ea_inode); + inode_lock(ea_inode); + ea_inode->i_flags |= S_NOQUOTA; + inode_unlock(ea_inode); + } + + return ea_inode; +} + +static struct inode * +ext4_xattr_inode_cache_find(struct inode *inode, const void *value, + size_t value_len, u32 hash) +{ + struct inode *ea_inode; + struct mb_cache_entry *ce; + struct mb_cache *ea_inode_cache = EA_INODE_CACHE(inode); + void *ea_data; + + if (!ea_inode_cache) + return NULL; + + ce = mb_cache_entry_find_first(ea_inode_cache, hash); + if (!ce) + return NULL; + + WARN_ON_ONCE(ext4_handle_valid(journal_current_handle()) && + !(current->flags & PF_MEMALLOC_NOFS)); + + ea_data = kvmalloc(value_len, GFP_NOFS); + if (!ea_data) { + mb_cache_entry_put(ea_inode_cache, ce); + return NULL; + } + + while (ce) { + ea_inode = ext4_iget(inode->i_sb, ce->e_value, + EXT4_IGET_EA_INODE); + if (IS_ERR(ea_inode)) + goto next_entry; + ext4_xattr_inode_set_class(ea_inode); + if (i_size_read(ea_inode) == value_len && + !ext4_xattr_inode_read(ea_inode, ea_data, value_len) && + !ext4_xattr_inode_verify_hashes(ea_inode, NULL, ea_data, + value_len) && + !memcmp(value, ea_data, value_len)) { + mb_cache_entry_touch(ea_inode_cache, ce); + mb_cache_entry_put(ea_inode_cache, ce); + kvfree(ea_data); + return ea_inode; + } + iput(ea_inode); + next_entry: + ce = mb_cache_entry_find_next(ea_inode_cache, ce); + } + kvfree(ea_data); + return NULL; +} + +/* + * Add value of the EA in an inode. + */ +static struct inode *ext4_xattr_inode_lookup_create(handle_t *handle, + struct inode *inode, const void *value, size_t value_len) +{ + struct inode *ea_inode; + u32 hash; + int err; + + /* Account inode & space to quota even if sharing... */ + err = ext4_xattr_inode_alloc_quota(inode, value_len); + if (err) + return ERR_PTR(err); + + hash = ext4_xattr_inode_hash(EXT4_SB(inode->i_sb), value, value_len); + ea_inode = ext4_xattr_inode_cache_find(inode, value, value_len, hash); + if (ea_inode) { + err = ext4_xattr_inode_inc_ref(handle, ea_inode); + if (err) + goto out_err; + return ea_inode; + } + + /* Create an inode for the EA value */ + ea_inode = ext4_xattr_inode_create(handle, inode, hash); + if (IS_ERR(ea_inode)) { + ext4_xattr_inode_free_quota(inode, NULL, value_len); + return ea_inode; + } + + err = ext4_xattr_inode_write(handle, ea_inode, value, value_len); + if (err) { + if (ext4_xattr_inode_dec_ref(handle, ea_inode)) + ext4_warning_inode(ea_inode, "cleanup dec ref error %d", err); + goto out_err; + } + + if (EA_INODE_CACHE(inode)) + mb_cache_entry_create(EA_INODE_CACHE(inode), GFP_NOFS, hash, + ea_inode->i_ino, true /* reusable */); + return ea_inode; +out_err: + iput(ea_inode); + ext4_xattr_inode_free_quota(inode, NULL, value_len); + return ERR_PTR(err); +} + +/* + * Reserve min(block_size/8, 1024) bytes for xattr entries/names if ea_inode + * feature is enabled. + */ +#define EXT4_XATTR_BLOCK_RESERVE(inode) min(i_blocksize(inode)/8, 1024U) + +static int ext4_xattr_set_entry(struct ext4_xattr_info *i, + struct ext4_xattr_search *s, + handle_t *handle, struct inode *inode, + struct inode *new_ea_inode, + bool is_block) +{ + struct ext4_xattr_entry *last, *next; + struct ext4_xattr_entry *here = s->here; + size_t min_offs = s->end - s->base, name_len = strlen(i->name); + int in_inode = i->in_inode; + struct inode *old_ea_inode = NULL; + size_t old_size, new_size; + int ret; + + /* Space used by old and new values. */ + old_size = (!s->not_found && !here->e_value_inum) ? + EXT4_XATTR_SIZE(le32_to_cpu(here->e_value_size)) : 0; + new_size = (i->value && !in_inode) ? EXT4_XATTR_SIZE(i->value_len) : 0; + + /* + * Optimization for the simple case when old and new values have the + * same padded sizes. Not applicable if external inodes are involved. + */ + if (new_size && new_size == old_size) { + size_t offs = le16_to_cpu(here->e_value_offs); + void *val = s->base + offs; + + here->e_value_size = cpu_to_le32(i->value_len); + if (i->value == EXT4_ZERO_XATTR_VALUE) { + memset(val, 0, new_size); + } else { + memcpy(val, i->value, i->value_len); + /* Clear padding bytes. */ + memset(val + i->value_len, 0, new_size - i->value_len); + } + goto update_hash; + } + + /* Compute min_offs and last. */ + last = s->first; + for (; !IS_LAST_ENTRY(last); last = next) { + next = EXT4_XATTR_NEXT(last); + if ((void *)next >= s->end) { + EXT4_ERROR_INODE(inode, "corrupted xattr entries"); + ret = -EFSCORRUPTED; + goto out; + } + if (!last->e_value_inum && last->e_value_size) { + size_t offs = le16_to_cpu(last->e_value_offs); + if (offs < min_offs) + min_offs = offs; + } + } + + /* Check whether we have enough space. */ + if (i->value) { + size_t free; + + free = min_offs - ((void *)last - s->base) - sizeof(__u32); + if (!s->not_found) + free += EXT4_XATTR_LEN(name_len) + old_size; + + if (free < EXT4_XATTR_LEN(name_len) + new_size) { + ret = -ENOSPC; + goto out; + } + + /* + * If storing the value in an external inode is an option, + * reserve space for xattr entries/names in the external + * attribute block so that a long value does not occupy the + * whole space and prevent further entries being added. + */ + if (ext4_has_feature_ea_inode(inode->i_sb) && + new_size && is_block && + (min_offs + old_size - new_size) < + EXT4_XATTR_BLOCK_RESERVE(inode)) { + ret = -ENOSPC; + goto out; + } + } + + /* + * Getting access to old and new ea inodes is subject to failures. + * Finish that work before doing any modifications to the xattr data. + */ + if (!s->not_found && here->e_value_inum) { + ret = ext4_xattr_inode_iget(inode, + le32_to_cpu(here->e_value_inum), + le32_to_cpu(here->e_hash), + &old_ea_inode); + if (ret) { + old_ea_inode = NULL; + goto out; + } + + /* We are ready to release ref count on the old_ea_inode. */ + ret = ext4_xattr_inode_dec_ref(handle, old_ea_inode); + if (ret) + goto out; + + ext4_xattr_inode_free_quota(inode, old_ea_inode, + le32_to_cpu(here->e_value_size)); + } + + /* No failures allowed past this point. */ + + if (!s->not_found && here->e_value_size && !here->e_value_inum) { + /* Remove the old value. */ + void *first_val = s->base + min_offs; + size_t offs = le16_to_cpu(here->e_value_offs); + void *val = s->base + offs; + + memmove(first_val + old_size, first_val, val - first_val); + memset(first_val, 0, old_size); + min_offs += old_size; + + /* Adjust all value offsets. */ + last = s->first; + while (!IS_LAST_ENTRY(last)) { + size_t o = le16_to_cpu(last->e_value_offs); + + if (!last->e_value_inum && + last->e_value_size && o < offs) + last->e_value_offs = cpu_to_le16(o + old_size); + last = EXT4_XATTR_NEXT(last); + } + } + + if (!i->value) { + /* Remove old name. */ + size_t size = EXT4_XATTR_LEN(name_len); + + last = ENTRY((void *)last - size); + memmove(here, (void *)here + size, + (void *)last - (void *)here + sizeof(__u32)); + memset(last, 0, size); + + /* + * Update i_inline_off - moved ibody region might contain + * system.data attribute. Handling a failure here won't + * cause other complications for setting an xattr. + */ + if (!is_block && ext4_has_inline_data(inode)) { + ret = ext4_find_inline_data_nolock(inode); + if (ret) { + ext4_warning_inode(inode, + "unable to update i_inline_off"); + goto out; + } + } + } else if (s->not_found) { + /* Insert new name. */ + size_t size = EXT4_XATTR_LEN(name_len); + size_t rest = (void *)last - (void *)here + sizeof(__u32); + + memmove((void *)here + size, here, rest); + memset(here, 0, size); + here->e_name_index = i->name_index; + here->e_name_len = name_len; + memcpy(here->e_name, i->name, name_len); + } else { + /* This is an update, reset value info. */ + here->e_value_inum = 0; + here->e_value_offs = 0; + here->e_value_size = 0; + } + + if (i->value) { + /* Insert new value. */ + if (in_inode) { + here->e_value_inum = cpu_to_le32(new_ea_inode->i_ino); + } else if (i->value_len) { + void *val = s->base + min_offs - new_size; + + here->e_value_offs = cpu_to_le16(min_offs - new_size); + if (i->value == EXT4_ZERO_XATTR_VALUE) { + memset(val, 0, new_size); + } else { + memcpy(val, i->value, i->value_len); + /* Clear padding bytes. */ + memset(val + i->value_len, 0, + new_size - i->value_len); + } + } + here->e_value_size = cpu_to_le32(i->value_len); + } + +update_hash: + if (i->value) { + __le32 hash = 0; + + /* Entry hash calculation. */ + if (in_inode) { + __le32 crc32c_hash; + + /* + * Feed crc32c hash instead of the raw value for entry + * hash calculation. This is to avoid walking + * potentially long value buffer again. + */ + crc32c_hash = cpu_to_le32( + ext4_xattr_inode_get_hash(new_ea_inode)); + hash = ext4_xattr_hash_entry(here->e_name, + here->e_name_len, + &crc32c_hash, 1); + } else if (is_block) { + __le32 *value = s->base + le16_to_cpu( + here->e_value_offs); + + hash = ext4_xattr_hash_entry(here->e_name, + here->e_name_len, value, + new_size >> 2); + } + here->e_hash = hash; + } + + if (is_block) + ext4_xattr_rehash((struct ext4_xattr_header *)s->base); + + ret = 0; +out: + iput(old_ea_inode); + return ret; +} + +struct ext4_xattr_block_find { + struct ext4_xattr_search s; + struct buffer_head *bh; +}; + +static int +ext4_xattr_block_find(struct inode *inode, struct ext4_xattr_info *i, + struct ext4_xattr_block_find *bs) +{ + struct super_block *sb = inode->i_sb; + int error; + + ea_idebug(inode, "name=%d.%s, value=%p, value_len=%ld", + i->name_index, i->name, i->value, (long)i->value_len); + + if (EXT4_I(inode)->i_file_acl) { + /* The inode already has an extended attribute block. */ + bs->bh = ext4_sb_bread(sb, EXT4_I(inode)->i_file_acl, REQ_PRIO); + if (IS_ERR(bs->bh)) { + error = PTR_ERR(bs->bh); + bs->bh = NULL; + return error; + } + ea_bdebug(bs->bh, "b_count=%d, refcount=%d", + atomic_read(&(bs->bh->b_count)), + le32_to_cpu(BHDR(bs->bh)->h_refcount)); + error = ext4_xattr_check_block(inode, bs->bh); + if (error) + return error; + /* Find the named attribute. */ + bs->s.base = BHDR(bs->bh); + bs->s.first = BFIRST(bs->bh); + bs->s.end = bs->bh->b_data + bs->bh->b_size; + bs->s.here = bs->s.first; + error = xattr_find_entry(inode, &bs->s.here, bs->s.end, + i->name_index, i->name, 1); + if (error && error != -ENODATA) + return error; + bs->s.not_found = error; + } + return 0; +} + +static int +ext4_xattr_block_set(handle_t *handle, struct inode *inode, + struct ext4_xattr_info *i, + struct ext4_xattr_block_find *bs) +{ + struct super_block *sb = inode->i_sb; + struct buffer_head *new_bh = NULL; + struct ext4_xattr_search s_copy = bs->s; + struct ext4_xattr_search *s = &s_copy; + struct mb_cache_entry *ce = NULL; + int error = 0; + struct mb_cache *ea_block_cache = EA_BLOCK_CACHE(inode); + struct inode *ea_inode = NULL, *tmp_inode; + size_t old_ea_inode_quota = 0; + unsigned int ea_ino; + +#define header(x) ((struct ext4_xattr_header *)(x)) + + /* If we need EA inode, prepare it before locking the buffer */ + if (i->value && i->in_inode) { + WARN_ON_ONCE(!i->value_len); + + ea_inode = ext4_xattr_inode_lookup_create(handle, inode, + i->value, i->value_len); + if (IS_ERR(ea_inode)) { + error = PTR_ERR(ea_inode); + ea_inode = NULL; + goto cleanup; + } + } + + if (s->base) { + int offset = (char *)s->here - bs->bh->b_data; + + BUFFER_TRACE(bs->bh, "get_write_access"); + error = ext4_journal_get_write_access(handle, sb, bs->bh, + EXT4_JTR_NONE); + if (error) + goto cleanup; + + lock_buffer(bs->bh); + + if (header(s->base)->h_refcount == cpu_to_le32(1)) { + __u32 hash = le32_to_cpu(BHDR(bs->bh)->h_hash); + + /* + * This must happen under buffer lock for + * ext4_xattr_block_set() to reliably detect modified + * block + */ + if (ea_block_cache) { + struct mb_cache_entry *oe; + + oe = mb_cache_entry_delete_or_get(ea_block_cache, + hash, bs->bh->b_blocknr); + if (oe) { + /* + * Xattr block is getting reused. Leave + * it alone. + */ + mb_cache_entry_put(ea_block_cache, oe); + goto clone_block; + } + } + ea_bdebug(bs->bh, "modifying in-place"); + error = ext4_xattr_set_entry(i, s, handle, inode, + ea_inode, true /* is_block */); + ext4_xattr_block_csum_set(inode, bs->bh); + unlock_buffer(bs->bh); + if (error == -EFSCORRUPTED) + goto bad_block; + if (!error) + error = ext4_handle_dirty_metadata(handle, + inode, + bs->bh); + if (error) + goto cleanup; + goto inserted; + } +clone_block: + unlock_buffer(bs->bh); + ea_bdebug(bs->bh, "cloning"); + s->base = kmemdup(BHDR(bs->bh), bs->bh->b_size, GFP_NOFS); + error = -ENOMEM; + if (s->base == NULL) + goto cleanup; + s->first = ENTRY(header(s->base)+1); + header(s->base)->h_refcount = cpu_to_le32(1); + s->here = ENTRY(s->base + offset); + s->end = s->base + bs->bh->b_size; + + /* + * If existing entry points to an xattr inode, we need + * to prevent ext4_xattr_set_entry() from decrementing + * ref count on it because the reference belongs to the + * original block. In this case, make the entry look + * like it has an empty value. + */ + if (!s->not_found && s->here->e_value_inum) { + ea_ino = le32_to_cpu(s->here->e_value_inum); + error = ext4_xattr_inode_iget(inode, ea_ino, + le32_to_cpu(s->here->e_hash), + &tmp_inode); + if (error) + goto cleanup; + + if (!ext4_test_inode_state(tmp_inode, + EXT4_STATE_LUSTRE_EA_INODE)) { + /* + * Defer quota free call for previous + * inode until success is guaranteed. + */ + old_ea_inode_quota = le32_to_cpu( + s->here->e_value_size); + } + iput(tmp_inode); + + s->here->e_value_inum = 0; + s->here->e_value_size = 0; + } + } else { + /* Allocate a buffer where we construct the new block. */ + s->base = kzalloc(sb->s_blocksize, GFP_NOFS); + error = -ENOMEM; + if (s->base == NULL) + goto cleanup; + header(s->base)->h_magic = cpu_to_le32(EXT4_XATTR_MAGIC); + header(s->base)->h_blocks = cpu_to_le32(1); + header(s->base)->h_refcount = cpu_to_le32(1); + s->first = ENTRY(header(s->base)+1); + s->here = ENTRY(header(s->base)+1); + s->end = s->base + sb->s_blocksize; + } + + error = ext4_xattr_set_entry(i, s, handle, inode, ea_inode, + true /* is_block */); + if (error == -EFSCORRUPTED) + goto bad_block; + if (error) + goto cleanup; + +inserted: + if (!IS_LAST_ENTRY(s->first)) { + new_bh = ext4_xattr_block_cache_find(inode, header(s->base), &ce); + if (IS_ERR(new_bh)) { + error = PTR_ERR(new_bh); + new_bh = NULL; + goto cleanup; + } + + if (new_bh) { + /* We found an identical block in the cache. */ + if (new_bh == bs->bh) + ea_bdebug(new_bh, "keeping"); + else { + u32 ref; + +#ifdef EXT4_XATTR_DEBUG + WARN_ON_ONCE(dquot_initialize_needed(inode)); +#endif + /* The old block is released after updating + the inode. */ + error = dquot_alloc_block(inode, + EXT4_C2B(EXT4_SB(sb), 1)); + if (error) + goto cleanup; + BUFFER_TRACE(new_bh, "get_write_access"); + error = ext4_journal_get_write_access( + handle, sb, new_bh, + EXT4_JTR_NONE); + if (error) + goto cleanup_dquot; + lock_buffer(new_bh); + /* + * We have to be careful about races with + * adding references to xattr block. Once we + * hold buffer lock xattr block's state is + * stable so we can check the additional + * reference fits. + */ + ref = le32_to_cpu(BHDR(new_bh)->h_refcount) + 1; + if (ref > EXT4_XATTR_REFCOUNT_MAX) { + /* + * Undo everything and check mbcache + * again. + */ + unlock_buffer(new_bh); + dquot_free_block(inode, + EXT4_C2B(EXT4_SB(sb), + 1)); + brelse(new_bh); + mb_cache_entry_put(ea_block_cache, ce); + ce = NULL; + new_bh = NULL; + goto inserted; + } + BHDR(new_bh)->h_refcount = cpu_to_le32(ref); + if (ref == EXT4_XATTR_REFCOUNT_MAX) + clear_bit(MBE_REUSABLE_B, &ce->e_flags); + ea_bdebug(new_bh, "reusing; refcount now=%d", + ref); + ext4_xattr_block_csum_set(inode, new_bh); + unlock_buffer(new_bh); + error = ext4_handle_dirty_metadata(handle, + inode, + new_bh); + if (error) + goto cleanup_dquot; + } + mb_cache_entry_touch(ea_block_cache, ce); + mb_cache_entry_put(ea_block_cache, ce); + ce = NULL; + } else if (bs->bh && s->base == bs->bh->b_data) { + /* We were modifying this block in-place. */ + ea_bdebug(bs->bh, "keeping this block"); + ext4_xattr_block_cache_insert(ea_block_cache, bs->bh); + new_bh = bs->bh; + get_bh(new_bh); + } else { + /* We need to allocate a new block */ + ext4_fsblk_t goal, block; + +#ifdef EXT4_XATTR_DEBUG + WARN_ON_ONCE(dquot_initialize_needed(inode)); +#endif + goal = ext4_group_first_block_no(sb, + EXT4_I(inode)->i_block_group); + block = ext4_new_meta_blocks(handle, inode, goal, 0, + NULL, &error); + if (error) + goto cleanup; + + ea_idebug(inode, "creating block %llu", + (unsigned long long)block); + + new_bh = sb_getblk(sb, block); + if (unlikely(!new_bh)) { + error = -ENOMEM; +getblk_failed: + ext4_free_blocks(handle, inode, NULL, block, 1, + EXT4_FREE_BLOCKS_METADATA); + goto cleanup; + } + error = ext4_xattr_inode_inc_ref_all(handle, inode, + ENTRY(header(s->base)+1)); + if (error) + goto getblk_failed; + if (ea_inode) { + /* Drop the extra ref on ea_inode. */ + error = ext4_xattr_inode_dec_ref(handle, + ea_inode); + if (error) + ext4_warning_inode(ea_inode, + "dec ref error=%d", + error); + iput(ea_inode); + ea_inode = NULL; + } + + lock_buffer(new_bh); + error = ext4_journal_get_create_access(handle, sb, + new_bh, EXT4_JTR_NONE); + if (error) { + unlock_buffer(new_bh); + error = -EIO; + goto getblk_failed; + } + memcpy(new_bh->b_data, s->base, new_bh->b_size); + ext4_xattr_block_csum_set(inode, new_bh); + set_buffer_uptodate(new_bh); + unlock_buffer(new_bh); + ext4_xattr_block_cache_insert(ea_block_cache, new_bh); + error = ext4_handle_dirty_metadata(handle, inode, + new_bh); + if (error) + goto cleanup; + } + } + + if (old_ea_inode_quota) + ext4_xattr_inode_free_quota(inode, NULL, old_ea_inode_quota); + + /* Update the inode. */ + EXT4_I(inode)->i_file_acl = new_bh ? new_bh->b_blocknr : 0; + + /* Drop the previous xattr block. */ + if (bs->bh && bs->bh != new_bh) { + struct ext4_xattr_inode_array *ea_inode_array = NULL; + + ext4_xattr_release_block(handle, inode, bs->bh, + &ea_inode_array, + 0 /* extra_credits */); + ext4_xattr_inode_array_free(ea_inode_array); + } + error = 0; + +cleanup: + if (ea_inode) { + if (error) { + int error2; + + error2 = ext4_xattr_inode_dec_ref(handle, ea_inode); + if (error2) + ext4_warning_inode(ea_inode, "dec ref error=%d", + error2); + ext4_xattr_inode_free_quota(inode, ea_inode, + i_size_read(ea_inode)); + } + iput(ea_inode); + } + if (ce) + mb_cache_entry_put(ea_block_cache, ce); + brelse(new_bh); + if (!(bs->bh && s->base == bs->bh->b_data)) + kfree(s->base); + + return error; + +cleanup_dquot: + dquot_free_block(inode, EXT4_C2B(EXT4_SB(sb), 1)); + goto cleanup; + +bad_block: + EXT4_ERROR_INODE(inode, "bad block %llu", + EXT4_I(inode)->i_file_acl); + goto cleanup; + +#undef header +} + +int ext4_xattr_ibody_find(struct inode *inode, struct ext4_xattr_info *i, + struct ext4_xattr_ibody_find *is) +{ + struct ext4_xattr_ibody_header *header; + struct ext4_inode *raw_inode; + int error; + + if (!EXT4_INODE_HAS_XATTR_SPACE(inode)) + return 0; + + raw_inode = ext4_raw_inode(&is->iloc); + header = IHDR(inode, raw_inode); + is->s.base = is->s.first = IFIRST(header); + is->s.here = is->s.first; + is->s.end = ITAIL(inode, raw_inode); + if (ext4_test_inode_state(inode, EXT4_STATE_XATTR)) { + /* Find the named attribute. */ + error = xattr_find_entry(inode, &is->s.here, is->s.end, + i->name_index, i->name, 0); + if (error && error != -ENODATA) + return error; + is->s.not_found = error; + } + return 0; +} + +int ext4_xattr_ibody_set(handle_t *handle, struct inode *inode, + struct ext4_xattr_info *i, + struct ext4_xattr_ibody_find *is) +{ + struct ext4_xattr_ibody_header *header; + struct ext4_xattr_search *s = &is->s; + struct inode *ea_inode = NULL; + int error; + + if (!EXT4_INODE_HAS_XATTR_SPACE(inode)) + return -ENOSPC; + + /* If we need EA inode, prepare it before locking the buffer */ + if (i->value && i->in_inode) { + WARN_ON_ONCE(!i->value_len); + + ea_inode = ext4_xattr_inode_lookup_create(handle, inode, + i->value, i->value_len); + if (IS_ERR(ea_inode)) + return PTR_ERR(ea_inode); + } + error = ext4_xattr_set_entry(i, s, handle, inode, ea_inode, + false /* is_block */); + if (error) { + if (ea_inode) { + int error2; + + error2 = ext4_xattr_inode_dec_ref(handle, ea_inode); + if (error2) + ext4_warning_inode(ea_inode, "dec ref error=%d", + error2); + + ext4_xattr_inode_free_quota(inode, ea_inode, + i_size_read(ea_inode)); + iput(ea_inode); + } + return error; + } + header = IHDR(inode, ext4_raw_inode(&is->iloc)); + if (!IS_LAST_ENTRY(s->first)) { + header->h_magic = cpu_to_le32(EXT4_XATTR_MAGIC); + ext4_set_inode_state(inode, EXT4_STATE_XATTR); + } else { + header->h_magic = cpu_to_le32(0); + ext4_clear_inode_state(inode, EXT4_STATE_XATTR); + } + iput(ea_inode); + return 0; +} + +static int ext4_xattr_value_same(struct ext4_xattr_search *s, + struct ext4_xattr_info *i) +{ + void *value; + + /* When e_value_inum is set the value is stored externally. */ + if (s->here->e_value_inum) + return 0; + if (le32_to_cpu(s->here->e_value_size) != i->value_len) + return 0; + value = ((void *)s->base) + le16_to_cpu(s->here->e_value_offs); + return !memcmp(value, i->value, i->value_len); +} + +static struct buffer_head *ext4_xattr_get_block(struct inode *inode) +{ + struct buffer_head *bh; + int error; + + if (!EXT4_I(inode)->i_file_acl) + return NULL; + bh = ext4_sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl, REQ_PRIO); + if (IS_ERR(bh)) + return bh; + error = ext4_xattr_check_block(inode, bh); + if (error) { + brelse(bh); + return ERR_PTR(error); + } + return bh; +} + +/* + * ext4_xattr_set_handle() + * + * Create, replace or remove an extended attribute for this inode. Value + * is NULL to remove an existing extended attribute, and non-NULL to + * either replace an existing extended attribute, or create a new extended + * attribute. The flags XATTR_REPLACE and XATTR_CREATE + * specify that an extended attribute must exist and must not exist + * previous to the call, respectively. + * + * Returns 0, or a negative error number on failure. + */ +int +ext4_xattr_set_handle(handle_t *handle, struct inode *inode, int name_index, + const char *name, const void *value, size_t value_len, + int flags) +{ + struct ext4_xattr_info i = { + .name_index = name_index, + .name = name, + .value = value, + .value_len = value_len, + .in_inode = 0, + }; + struct ext4_xattr_ibody_find is = { + .s = { .not_found = -ENODATA, }, + }; + struct ext4_xattr_block_find bs = { + .s = { .not_found = -ENODATA, }, + }; + int no_expand; + int error; + + if (!name) + return -EINVAL; + if (strlen(name) > 255) + return -ERANGE; + + ext4_write_lock_xattr(inode, &no_expand); + + /* Check journal credits under write lock. */ + if (ext4_handle_valid(handle)) { + struct buffer_head *bh; + int credits; + + bh = ext4_xattr_get_block(inode); + if (IS_ERR(bh)) { + error = PTR_ERR(bh); + goto cleanup; + } + + credits = __ext4_xattr_set_credits(inode->i_sb, inode, bh, + value_len, + flags & XATTR_CREATE); + brelse(bh); + + if (jbd2_handle_buffer_credits(handle) < credits) { + error = -ENOSPC; + goto cleanup; + } + WARN_ON_ONCE(!(current->flags & PF_MEMALLOC_NOFS)); + } + + error = ext4_reserve_inode_write(handle, inode, &is.iloc); + if (error) + goto cleanup; + + if (ext4_test_inode_state(inode, EXT4_STATE_NEW)) { + struct ext4_inode *raw_inode = ext4_raw_inode(&is.iloc); + memset(raw_inode, 0, EXT4_SB(inode->i_sb)->s_inode_size); + ext4_clear_inode_state(inode, EXT4_STATE_NEW); + } + + error = ext4_xattr_ibody_find(inode, &i, &is); + if (error) + goto cleanup; + if (is.s.not_found) + error = ext4_xattr_block_find(inode, &i, &bs); + if (error) + goto cleanup; + if (is.s.not_found && bs.s.not_found) { + error = -ENODATA; + if (flags & XATTR_REPLACE) + goto cleanup; + error = 0; + if (!value) + goto cleanup; + } else { + error = -EEXIST; + if (flags & XATTR_CREATE) + goto cleanup; + } + + if (!value) { + if (!is.s.not_found) + error = ext4_xattr_ibody_set(handle, inode, &i, &is); + else if (!bs.s.not_found) + error = ext4_xattr_block_set(handle, inode, &i, &bs); + } else { + error = 0; + /* Xattr value did not change? Save us some work and bail out */ + if (!is.s.not_found && ext4_xattr_value_same(&is.s, &i)) + goto cleanup; + if (!bs.s.not_found && ext4_xattr_value_same(&bs.s, &i)) + goto cleanup; + + if (ext4_has_feature_ea_inode(inode->i_sb) && + (EXT4_XATTR_SIZE(i.value_len) > + EXT4_XATTR_MIN_LARGE_EA_SIZE(inode->i_sb->s_blocksize))) + i.in_inode = 1; +retry_inode: + error = ext4_xattr_ibody_set(handle, inode, &i, &is); + if (!error && !bs.s.not_found) { + i.value = NULL; + error = ext4_xattr_block_set(handle, inode, &i, &bs); + } else if (error == -ENOSPC) { + if (EXT4_I(inode)->i_file_acl && !bs.s.base) { + brelse(bs.bh); + bs.bh = NULL; + error = ext4_xattr_block_find(inode, &i, &bs); + if (error) + goto cleanup; + } + error = ext4_xattr_block_set(handle, inode, &i, &bs); + if (!error && !is.s.not_found) { + i.value = NULL; + error = ext4_xattr_ibody_set(handle, inode, &i, + &is); + } else if (error == -ENOSPC) { + /* + * Xattr does not fit in the block, store at + * external inode if possible. + */ + if (ext4_has_feature_ea_inode(inode->i_sb) && + i.value_len && !i.in_inode) { + i.in_inode = 1; + goto retry_inode; + } + } + } + } + if (!error) { + ext4_xattr_update_super_block(handle, inode->i_sb); + inode_set_ctime_current(inode); + inode_inc_iversion(inode); + if (!value) + no_expand = 0; + error = ext4_mark_iloc_dirty(handle, inode, &is.iloc); + /* + * The bh is consumed by ext4_mark_iloc_dirty, even with + * error != 0. + */ + is.iloc.bh = NULL; + if (IS_SYNC(inode)) + ext4_handle_sync(handle); + } + ext4_fc_mark_ineligible(inode->i_sb, EXT4_FC_REASON_XATTR, handle); + +cleanup: + brelse(is.iloc.bh); + brelse(bs.bh); + ext4_write_unlock_xattr(inode, &no_expand); + return error; +} + +int ext4_xattr_set_credits(struct inode *inode, size_t value_len, + bool is_create, int *credits) +{ + struct buffer_head *bh; + int err; + + *credits = 0; + + if (!EXT4_SB(inode->i_sb)->s_journal) + return 0; + + down_read(&EXT4_I(inode)->xattr_sem); + + bh = ext4_xattr_get_block(inode); + if (IS_ERR(bh)) { + err = PTR_ERR(bh); + } else { + *credits = __ext4_xattr_set_credits(inode->i_sb, inode, bh, + value_len, is_create); + brelse(bh); + err = 0; + } + + up_read(&EXT4_I(inode)->xattr_sem); + return err; +} + +/* + * ext4_xattr_set() + * + * Like ext4_xattr_set_handle, but start from an inode. This extended + * attribute modification is a filesystem transaction by itself. + * + * Returns 0, or a negative error number on failure. + */ +int +ext4_xattr_set(struct inode *inode, int name_index, const char *name, + const void *value, size_t value_len, int flags) +{ + handle_t *handle; + struct super_block *sb = inode->i_sb; + int error, retries = 0; + int credits; + + error = dquot_initialize(inode); + if (error) + return error; + +retry: + error = ext4_xattr_set_credits(inode, value_len, flags & XATTR_CREATE, + &credits); + if (error) + return error; + + handle = ext4_journal_start(inode, EXT4_HT_XATTR, credits); + if (IS_ERR(handle)) { + error = PTR_ERR(handle); + } else { + int error2; + + error = ext4_xattr_set_handle(handle, inode, name_index, name, + value, value_len, flags); + ext4_fc_mark_ineligible(inode->i_sb, EXT4_FC_REASON_XATTR, + handle); + error2 = ext4_journal_stop(handle); + if (error == -ENOSPC && + ext4_should_retry_alloc(sb, &retries)) + goto retry; + if (error == 0) + error = error2; + } + + return error; +} + +/* + * Shift the EA entries in the inode to create space for the increased + * i_extra_isize. + */ +static void ext4_xattr_shift_entries(struct ext4_xattr_entry *entry, + int value_offs_shift, void *to, + void *from, size_t n) +{ + struct ext4_xattr_entry *last = entry; + int new_offs; + + /* We always shift xattr headers further thus offsets get lower */ + BUG_ON(value_offs_shift > 0); + + /* Adjust the value offsets of the entries */ + for (; !IS_LAST_ENTRY(last); last = EXT4_XATTR_NEXT(last)) { + if (!last->e_value_inum && last->e_value_size) { + new_offs = le16_to_cpu(last->e_value_offs) + + value_offs_shift; + last->e_value_offs = cpu_to_le16(new_offs); + } + } + /* Shift the entries by n bytes */ + memmove(to, from, n); +} + +/* + * Move xattr pointed to by 'entry' from inode into external xattr block + */ +static int ext4_xattr_move_to_block(handle_t *handle, struct inode *inode, + struct ext4_inode *raw_inode, + struct ext4_xattr_entry *entry) +{ + struct ext4_xattr_ibody_find *is = NULL; + struct ext4_xattr_block_find *bs = NULL; + char *buffer = NULL, *b_entry_name = NULL; + size_t value_size = le32_to_cpu(entry->e_value_size); + struct ext4_xattr_info i = { + .value = NULL, + .value_len = 0, + .name_index = entry->e_name_index, + .in_inode = !!entry->e_value_inum, + }; + struct ext4_xattr_ibody_header *header = IHDR(inode, raw_inode); + int needs_kvfree = 0; + int error; + + is = kzalloc(sizeof(struct ext4_xattr_ibody_find), GFP_NOFS); + bs = kzalloc(sizeof(struct ext4_xattr_block_find), GFP_NOFS); + b_entry_name = kmalloc(entry->e_name_len + 1, GFP_NOFS); + if (!is || !bs || !b_entry_name) { + error = -ENOMEM; + goto out; + } + + is->s.not_found = -ENODATA; + bs->s.not_found = -ENODATA; + is->iloc.bh = NULL; + bs->bh = NULL; + + /* Save the entry name and the entry value */ + if (entry->e_value_inum) { + buffer = kvmalloc(value_size, GFP_NOFS); + if (!buffer) { + error = -ENOMEM; + goto out; + } + needs_kvfree = 1; + error = ext4_xattr_inode_get(inode, entry, buffer, value_size); + if (error) + goto out; + } else { + size_t value_offs = le16_to_cpu(entry->e_value_offs); + buffer = (void *)IFIRST(header) + value_offs; + } + + memcpy(b_entry_name, entry->e_name, entry->e_name_len); + b_entry_name[entry->e_name_len] = '\0'; + i.name = b_entry_name; + + error = ext4_get_inode_loc(inode, &is->iloc); + if (error) + goto out; + + error = ext4_xattr_ibody_find(inode, &i, is); + if (error) + goto out; + + i.value = buffer; + i.value_len = value_size; + error = ext4_xattr_block_find(inode, &i, bs); + if (error) + goto out; + + /* Move ea entry from the inode into the block */ + error = ext4_xattr_block_set(handle, inode, &i, bs); + if (error) + goto out; + + /* Remove the chosen entry from the inode */ + i.value = NULL; + i.value_len = 0; + error = ext4_xattr_ibody_set(handle, inode, &i, is); + +out: + kfree(b_entry_name); + if (needs_kvfree && buffer) + kvfree(buffer); + if (is) + brelse(is->iloc.bh); + if (bs) + brelse(bs->bh); + kfree(is); + kfree(bs); + + return error; +} + +static int ext4_xattr_make_inode_space(handle_t *handle, struct inode *inode, + struct ext4_inode *raw_inode, + int isize_diff, size_t ifree, + size_t bfree, int *total_ino) +{ + struct ext4_xattr_ibody_header *header = IHDR(inode, raw_inode); + struct ext4_xattr_entry *small_entry; + struct ext4_xattr_entry *entry; + struct ext4_xattr_entry *last; + unsigned int entry_size; /* EA entry size */ + unsigned int total_size; /* EA entry size + value size */ + unsigned int min_total_size; + int error; + + while (isize_diff > ifree) { + entry = NULL; + small_entry = NULL; + min_total_size = ~0U; + last = IFIRST(header); + /* Find the entry best suited to be pushed into EA block */ + for (; !IS_LAST_ENTRY(last); last = EXT4_XATTR_NEXT(last)) { + /* never move system.data out of the inode */ + if ((last->e_name_len == 4) && + (last->e_name_index == EXT4_XATTR_INDEX_SYSTEM) && + !memcmp(last->e_name, "data", 4)) + continue; + total_size = EXT4_XATTR_LEN(last->e_name_len); + if (!last->e_value_inum) + total_size += EXT4_XATTR_SIZE( + le32_to_cpu(last->e_value_size)); + if (total_size <= bfree && + total_size < min_total_size) { + if (total_size + ifree < isize_diff) { + small_entry = last; + } else { + entry = last; + min_total_size = total_size; + } + } + } + + if (entry == NULL) { + if (small_entry == NULL) + return -ENOSPC; + entry = small_entry; + } + + entry_size = EXT4_XATTR_LEN(entry->e_name_len); + total_size = entry_size; + if (!entry->e_value_inum) + total_size += EXT4_XATTR_SIZE( + le32_to_cpu(entry->e_value_size)); + error = ext4_xattr_move_to_block(handle, inode, raw_inode, + entry); + if (error) + return error; + + *total_ino -= entry_size; + ifree += total_size; + bfree -= total_size; + } + + return 0; +} + +/* + * Expand an inode by new_extra_isize bytes when EAs are present. + * Returns 0 on success or negative error number on failure. + */ +int ext4_expand_extra_isize_ea(struct inode *inode, int new_extra_isize, + struct ext4_inode *raw_inode, handle_t *handle) +{ + struct ext4_xattr_ibody_header *header; + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + static unsigned int mnt_count; + size_t min_offs; + size_t ifree, bfree; + int total_ino; + void *base, *end; + int error = 0, tried_min_extra_isize = 0; + int s_min_extra_isize = le16_to_cpu(sbi->s_es->s_min_extra_isize); + int isize_diff; /* How much do we need to grow i_extra_isize */ + +retry: + isize_diff = new_extra_isize - EXT4_I(inode)->i_extra_isize; + if (EXT4_I(inode)->i_extra_isize >= new_extra_isize) + return 0; + + header = IHDR(inode, raw_inode); + + /* + * Check if enough free space is available in the inode to shift the + * entries ahead by new_extra_isize. + */ + + base = IFIRST(header); + end = ITAIL(inode, raw_inode); + min_offs = end - base; + total_ino = sizeof(struct ext4_xattr_ibody_header) + sizeof(u32); + + ifree = ext4_xattr_free_space(base, &min_offs, base, &total_ino); + if (ifree >= isize_diff) + goto shift; + + /* + * Enough free space isn't available in the inode, check if + * EA block can hold new_extra_isize bytes. + */ + if (EXT4_I(inode)->i_file_acl) { + struct buffer_head *bh; + + bh = ext4_sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl, REQ_PRIO); + if (IS_ERR(bh)) { + error = PTR_ERR(bh); + goto cleanup; + } + error = ext4_xattr_check_block(inode, bh); + if (error) { + brelse(bh); + goto cleanup; + } + base = BHDR(bh); + end = bh->b_data + bh->b_size; + min_offs = end - base; + bfree = ext4_xattr_free_space(BFIRST(bh), &min_offs, base, + NULL); + brelse(bh); + if (bfree + ifree < isize_diff) { + if (!tried_min_extra_isize && s_min_extra_isize) { + tried_min_extra_isize++; + new_extra_isize = s_min_extra_isize; + goto retry; + } + error = -ENOSPC; + goto cleanup; + } + } else { + bfree = inode->i_sb->s_blocksize; + } + + error = ext4_xattr_make_inode_space(handle, inode, raw_inode, + isize_diff, ifree, bfree, + &total_ino); + if (error) { + if (error == -ENOSPC && !tried_min_extra_isize && + s_min_extra_isize) { + tried_min_extra_isize++; + new_extra_isize = s_min_extra_isize; + goto retry; + } + goto cleanup; + } +shift: + /* Adjust the offsets and shift the remaining entries ahead */ + ext4_xattr_shift_entries(IFIRST(header), EXT4_I(inode)->i_extra_isize + - new_extra_isize, (void *)raw_inode + + EXT4_GOOD_OLD_INODE_SIZE + new_extra_isize, + (void *)header, total_ino); + EXT4_I(inode)->i_extra_isize = new_extra_isize; + + if (ext4_has_inline_data(inode)) + error = ext4_find_inline_data_nolock(inode); + +cleanup: + if (error && (mnt_count != le16_to_cpu(sbi->s_es->s_mnt_count))) { + ext4_warning(inode->i_sb, "Unable to expand inode %lu. Delete some EAs or run e2fsck.", + inode->i_ino); + mnt_count = le16_to_cpu(sbi->s_es->s_mnt_count); + } + return error; +} + +#define EIA_INCR 16 /* must be 2^n */ +#define EIA_MASK (EIA_INCR - 1) + +/* Add the large xattr @inode into @ea_inode_array for deferred iput(). + * If @ea_inode_array is new or full it will be grown and the old + * contents copied over. + */ +static int +ext4_expand_inode_array(struct ext4_xattr_inode_array **ea_inode_array, + struct inode *inode) +{ + if (*ea_inode_array == NULL) { + /* + * Start with 15 inodes, so it fits into a power-of-two size. + */ + (*ea_inode_array) = kmalloc( + struct_size(*ea_inode_array, inodes, EIA_MASK), + GFP_NOFS); + if (*ea_inode_array == NULL) + return -ENOMEM; + (*ea_inode_array)->count = 0; + } else if (((*ea_inode_array)->count & EIA_MASK) == EIA_MASK) { + /* expand the array once all 15 + n * 16 slots are full */ + struct ext4_xattr_inode_array *new_array = NULL; + + new_array = kmalloc( + struct_size(*ea_inode_array, inodes, + (*ea_inode_array)->count + EIA_INCR), + GFP_NOFS); + if (new_array == NULL) + return -ENOMEM; + memcpy(new_array, *ea_inode_array, + struct_size(*ea_inode_array, inodes, + (*ea_inode_array)->count)); + kfree(*ea_inode_array); + *ea_inode_array = new_array; + } + (*ea_inode_array)->count++; + (*ea_inode_array)->inodes[(*ea_inode_array)->count - 1] = inode; + return 0; +} + +/* + * ext4_xattr_delete_inode() + * + * Free extended attribute resources associated with this inode. Traverse + * all entries and decrement reference on any xattr inodes associated with this + * inode. This is called immediately before an inode is freed. We have exclusive + * access to the inode. If an orphan inode is deleted it will also release its + * references on xattr block and xattr inodes. + */ +int ext4_xattr_delete_inode(handle_t *handle, struct inode *inode, + struct ext4_xattr_inode_array **ea_inode_array, + int extra_credits) +{ + struct buffer_head *bh = NULL; + struct ext4_xattr_ibody_header *header; + struct ext4_iloc iloc = { .bh = NULL }; + struct ext4_xattr_entry *entry; + struct inode *ea_inode; + int error; + + error = ext4_journal_ensure_credits(handle, extra_credits, + ext4_free_metadata_revoke_credits(inode->i_sb, 1)); + if (error < 0) { + EXT4_ERROR_INODE(inode, "ensure credits (error %d)", error); + goto cleanup; + } + + if (ext4_has_feature_ea_inode(inode->i_sb) && + ext4_test_inode_state(inode, EXT4_STATE_XATTR)) { + + error = ext4_get_inode_loc(inode, &iloc); + if (error) { + EXT4_ERROR_INODE(inode, "inode loc (error %d)", error); + goto cleanup; + } + + error = ext4_journal_get_write_access(handle, inode->i_sb, + iloc.bh, EXT4_JTR_NONE); + if (error) { + EXT4_ERROR_INODE(inode, "write access (error %d)", + error); + goto cleanup; + } + + header = IHDR(inode, ext4_raw_inode(&iloc)); + if (header->h_magic == cpu_to_le32(EXT4_XATTR_MAGIC)) + ext4_xattr_inode_dec_ref_all(handle, inode, iloc.bh, + IFIRST(header), + false /* block_csum */, + ea_inode_array, + extra_credits, + false /* skip_quota */); + } + + if (EXT4_I(inode)->i_file_acl) { + bh = ext4_sb_bread(inode->i_sb, EXT4_I(inode)->i_file_acl, REQ_PRIO); + if (IS_ERR(bh)) { + error = PTR_ERR(bh); + if (error == -EIO) { + EXT4_ERROR_INODE_ERR(inode, EIO, + "block %llu read error", + EXT4_I(inode)->i_file_acl); + } + bh = NULL; + goto cleanup; + } + error = ext4_xattr_check_block(inode, bh); + if (error) + goto cleanup; + + if (ext4_has_feature_ea_inode(inode->i_sb)) { + for (entry = BFIRST(bh); !IS_LAST_ENTRY(entry); + entry = EXT4_XATTR_NEXT(entry)) { + if (!entry->e_value_inum) + continue; + error = ext4_xattr_inode_iget(inode, + le32_to_cpu(entry->e_value_inum), + le32_to_cpu(entry->e_hash), + &ea_inode); + if (error) + continue; + ext4_xattr_inode_free_quota(inode, ea_inode, + le32_to_cpu(entry->e_value_size)); + iput(ea_inode); + } + + } + + ext4_xattr_release_block(handle, inode, bh, ea_inode_array, + extra_credits); + /* + * Update i_file_acl value in the same transaction that releases + * block. + */ + EXT4_I(inode)->i_file_acl = 0; + error = ext4_mark_inode_dirty(handle, inode); + if (error) { + EXT4_ERROR_INODE(inode, "mark inode dirty (error %d)", + error); + goto cleanup; + } + ext4_fc_mark_ineligible(inode->i_sb, EXT4_FC_REASON_XATTR, handle); + } + error = 0; +cleanup: + brelse(iloc.bh); + brelse(bh); + return error; +} + +void ext4_xattr_inode_array_free(struct ext4_xattr_inode_array *ea_inode_array) +{ + int idx; + + if (ea_inode_array == NULL) + return; + + for (idx = 0; idx < ea_inode_array->count; ++idx) + iput(ea_inode_array->inodes[idx]); + kfree(ea_inode_array); +} + +/* + * ext4_xattr_block_cache_insert() + * + * Create a new entry in the extended attribute block cache, and insert + * it unless such an entry is already in the cache. + */ +static void +ext4_xattr_block_cache_insert(struct mb_cache *ea_block_cache, + struct buffer_head *bh) +{ + struct ext4_xattr_header *header = BHDR(bh); + __u32 hash = le32_to_cpu(header->h_hash); + int reusable = le32_to_cpu(header->h_refcount) < + EXT4_XATTR_REFCOUNT_MAX; + int error; + + if (!ea_block_cache) + return; + error = mb_cache_entry_create(ea_block_cache, GFP_NOFS, hash, + bh->b_blocknr, reusable); + if (error) { + if (error == -EBUSY) + ea_bdebug(bh, "already in cache"); + } else + ea_bdebug(bh, "inserting [%x]", (int)hash); +} + +/* + * ext4_xattr_cmp() + * + * Compare two extended attribute blocks for equality. + * + * Returns 0 if the blocks are equal, 1 if they differ. + */ +static int +ext4_xattr_cmp(struct ext4_xattr_header *header1, + struct ext4_xattr_header *header2) +{ + struct ext4_xattr_entry *entry1, *entry2; + + entry1 = ENTRY(header1+1); + entry2 = ENTRY(header2+1); + while (!IS_LAST_ENTRY(entry1)) { + if (IS_LAST_ENTRY(entry2)) + return 1; + if (entry1->e_hash != entry2->e_hash || + entry1->e_name_index != entry2->e_name_index || + entry1->e_name_len != entry2->e_name_len || + entry1->e_value_size != entry2->e_value_size || + entry1->e_value_inum != entry2->e_value_inum || + memcmp(entry1->e_name, entry2->e_name, entry1->e_name_len)) + return 1; + if (!entry1->e_value_inum && + memcmp((char *)header1 + le16_to_cpu(entry1->e_value_offs), + (char *)header2 + le16_to_cpu(entry2->e_value_offs), + le32_to_cpu(entry1->e_value_size))) + return 1; + + entry1 = EXT4_XATTR_NEXT(entry1); + entry2 = EXT4_XATTR_NEXT(entry2); + } + if (!IS_LAST_ENTRY(entry2)) + return 1; + return 0; +} + +/* + * ext4_xattr_block_cache_find() + * + * Find an identical extended attribute block. + * + * Returns a pointer to the block found, or NULL if such a block was not + * found, or an error pointer if an error occurred while reading ea block. + */ +static struct buffer_head * +ext4_xattr_block_cache_find(struct inode *inode, + struct ext4_xattr_header *header, + struct mb_cache_entry **pce) +{ + __u32 hash = le32_to_cpu(header->h_hash); + struct mb_cache_entry *ce; + struct mb_cache *ea_block_cache = EA_BLOCK_CACHE(inode); + + if (!ea_block_cache) + return NULL; + if (!header->h_hash) + return NULL; /* never share */ + ea_idebug(inode, "looking for cached blocks [%x]", (int)hash); + ce = mb_cache_entry_find_first(ea_block_cache, hash); + while (ce) { + struct buffer_head *bh; + + bh = ext4_sb_bread(inode->i_sb, ce->e_value, REQ_PRIO); + if (IS_ERR(bh)) { + if (PTR_ERR(bh) != -ENOMEM) + EXT4_ERROR_INODE(inode, "block %lu read error", + (unsigned long)ce->e_value); + mb_cache_entry_put(ea_block_cache, ce); + return bh; + } else if (ext4_xattr_cmp(header, BHDR(bh)) == 0) { + *pce = ce; + return bh; + } + brelse(bh); + ce = mb_cache_entry_find_next(ea_block_cache, ce); + } + return NULL; +} + +#define NAME_HASH_SHIFT 5 +#define VALUE_HASH_SHIFT 16 + +/* + * ext4_xattr_hash_entry() + * + * Compute the hash of an extended attribute. + */ +static __le32 ext4_xattr_hash_entry(char *name, size_t name_len, __le32 *value, + size_t value_count) +{ + __u32 hash = 0; + + while (name_len--) { + hash = (hash << NAME_HASH_SHIFT) ^ + (hash >> (8*sizeof(hash) - NAME_HASH_SHIFT)) ^ + (unsigned char)*name++; + } + while (value_count--) { + hash = (hash << VALUE_HASH_SHIFT) ^ + (hash >> (8*sizeof(hash) - VALUE_HASH_SHIFT)) ^ + le32_to_cpu(*value++); + } + return cpu_to_le32(hash); +} + +/* + * ext4_xattr_hash_entry_signed() + * + * Compute the hash of an extended attribute incorrectly. + */ +static __le32 ext4_xattr_hash_entry_signed(char *name, size_t name_len, __le32 *value, size_t value_count) +{ + __u32 hash = 0; + + while (name_len--) { + hash = (hash << NAME_HASH_SHIFT) ^ + (hash >> (8*sizeof(hash) - NAME_HASH_SHIFT)) ^ + (signed char)*name++; + } + while (value_count--) { + hash = (hash << VALUE_HASH_SHIFT) ^ + (hash >> (8*sizeof(hash) - VALUE_HASH_SHIFT)) ^ + le32_to_cpu(*value++); + } + return cpu_to_le32(hash); +} + +#undef NAME_HASH_SHIFT +#undef VALUE_HASH_SHIFT + +#define BLOCK_HASH_SHIFT 16 + +/* + * ext4_xattr_rehash() + * + * Re-compute the extended attribute hash value after an entry has changed. + */ +static void ext4_xattr_rehash(struct ext4_xattr_header *header) +{ + struct ext4_xattr_entry *here; + __u32 hash = 0; + + here = ENTRY(header+1); + while (!IS_LAST_ENTRY(here)) { + if (!here->e_hash) { + /* Block is not shared if an entry's hash value == 0 */ + hash = 0; + break; + } + hash = (hash << BLOCK_HASH_SHIFT) ^ + (hash >> (8*sizeof(hash) - BLOCK_HASH_SHIFT)) ^ + le32_to_cpu(here->e_hash); + here = EXT4_XATTR_NEXT(here); + } + header->h_hash = cpu_to_le32(hash); +} + +#undef BLOCK_HASH_SHIFT + +#define HASH_BUCKET_BITS 10 + +struct mb_cache * +ext4_xattr_create_cache(void) +{ + return mb_cache_create(HASH_BUCKET_BITS); +} + +void ext4_xattr_destroy_cache(struct mb_cache *cache) +{ + if (cache) + mb_cache_destroy(cache); +} + diff --git a/fs/ext4l/xattr.h b/fs/ext4l/xattr.h new file mode 100644 index 00000000000..1fedf44d4fb --- /dev/null +++ b/fs/ext4l/xattr.h @@ -0,0 +1,236 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + File: fs/ext4/xattr.h + + On-disk format of extended attributes for the ext4 filesystem. + + (C) 2001 Andreas Gruenbacher, <a.gruenbacher@computer.org> +*/ + +#include <linux/xattr.h> + +/* Magic value in attribute blocks */ +#define EXT4_XATTR_MAGIC 0xEA020000 + +/* Maximum number of references to one attribute block */ +#define EXT4_XATTR_REFCOUNT_MAX 1024 + +/* Name indexes */ +#define EXT4_XATTR_INDEX_USER 1 +#define EXT4_XATTR_INDEX_POSIX_ACL_ACCESS 2 +#define EXT4_XATTR_INDEX_POSIX_ACL_DEFAULT 3 +#define EXT4_XATTR_INDEX_TRUSTED 4 +#define EXT4_XATTR_INDEX_LUSTRE 5 +#define EXT4_XATTR_INDEX_SECURITY 6 +#define EXT4_XATTR_INDEX_SYSTEM 7 +#define EXT4_XATTR_INDEX_RICHACL 8 +#define EXT4_XATTR_INDEX_ENCRYPTION 9 +#define EXT4_XATTR_INDEX_HURD 10 /* Reserved for Hurd */ + +struct ext4_xattr_header { + __le32 h_magic; /* magic number for identification */ + __le32 h_refcount; /* reference count */ + __le32 h_blocks; /* number of disk blocks used */ + __le32 h_hash; /* hash value of all attributes */ + __le32 h_checksum; /* crc32c(uuid+blknum+xattrblock) */ + __u32 h_reserved[3]; /* zero right now */ +}; + +struct ext4_xattr_ibody_header { + __le32 h_magic; /* magic number for identification */ +}; + +struct ext4_xattr_entry { + __u8 e_name_len; /* length of name */ + __u8 e_name_index; /* attribute name index */ + __le16 e_value_offs; /* offset in disk block of value */ + __le32 e_value_inum; /* inode in which the value is stored */ + __le32 e_value_size; /* size of attribute value */ + __le32 e_hash; /* hash value of name and value */ + char e_name[]; /* attribute name */ +}; + +#define EXT4_XATTR_PAD_BITS 2 +#define EXT4_XATTR_PAD (1<<EXT4_XATTR_PAD_BITS) +#define EXT4_XATTR_ROUND (EXT4_XATTR_PAD-1) +#define EXT4_XATTR_LEN(name_len) \ + (((name_len) + EXT4_XATTR_ROUND + \ + sizeof(struct ext4_xattr_entry)) & ~EXT4_XATTR_ROUND) +#define EXT4_XATTR_NEXT(entry) \ + ((struct ext4_xattr_entry *)( \ + (char *)(entry) + EXT4_XATTR_LEN((entry)->e_name_len))) +#define EXT4_XATTR_SIZE(size) \ + (((size) + EXT4_XATTR_ROUND) & ~EXT4_XATTR_ROUND) + +#define IHDR(inode, raw_inode) \ + ((struct ext4_xattr_ibody_header *) \ + ((void *)raw_inode + \ + EXT4_GOOD_OLD_INODE_SIZE + \ + EXT4_I(inode)->i_extra_isize)) +#define ITAIL(inode, raw_inode) \ + ((void *)(raw_inode) + \ + EXT4_SB((inode)->i_sb)->s_inode_size) +#define IFIRST(hdr) ((struct ext4_xattr_entry *)((hdr)+1)) + +/* + * XATTR_SIZE_MAX is currently 64k, but for the purposes of checking + * for file system consistency errors, we use a somewhat bigger value. + * This allows XATTR_SIZE_MAX to grow in the future, but by using this + * instead of INT_MAX for certain consistency checks, we don't need to + * worry about arithmetic overflows. (Actually XATTR_SIZE_MAX is + * defined in include/uapi/linux/limits.h, so changing it is going + * not going to be trivial....) + */ +#define EXT4_XATTR_SIZE_MAX (1 << 24) + +/* + * The minimum size of EA value when you start storing it in an external inode + * size of block - size of header - size of 1 entry - 4 null bytes + */ +#define EXT4_XATTR_MIN_LARGE_EA_SIZE(b) \ + ((b) - EXT4_XATTR_LEN(3) - sizeof(struct ext4_xattr_header) - 4) + +#define BHDR(bh) ((struct ext4_xattr_header *)((bh)->b_data)) +#define ENTRY(ptr) ((struct ext4_xattr_entry *)(ptr)) +#define BFIRST(bh) ENTRY(BHDR(bh)+1) +#define IS_LAST_ENTRY(entry) (*(__u32 *)(entry) == 0) + +#define EXT4_ZERO_XATTR_VALUE ((void *)-1) + +/* + * If we want to add an xattr to the inode, we should make sure that + * i_extra_isize is not 0 and that the inode size is not less than + * EXT4_GOOD_OLD_INODE_SIZE + extra_isize + pad. + * EXT4_GOOD_OLD_INODE_SIZE extra_isize header entry pad data + * |--------------------------|------------|------|---------|---|-------| + */ +#define EXT4_INODE_HAS_XATTR_SPACE(inode) \ + ((EXT4_I(inode)->i_extra_isize != 0) && \ + (EXT4_GOOD_OLD_INODE_SIZE + EXT4_I(inode)->i_extra_isize + \ + sizeof(struct ext4_xattr_ibody_header) + EXT4_XATTR_PAD <= \ + EXT4_INODE_SIZE((inode)->i_sb))) + +struct ext4_xattr_info { + const char *name; + const void *value; + size_t value_len; + int name_index; + int in_inode; +}; + +struct ext4_xattr_search { + struct ext4_xattr_entry *first; + void *base; + void *end; + struct ext4_xattr_entry *here; + int not_found; +}; + +struct ext4_xattr_ibody_find { + struct ext4_xattr_search s; + struct ext4_iloc iloc; +}; + +struct ext4_xattr_inode_array { + unsigned int count; + struct inode *inodes[] __counted_by(count); +}; + +extern const struct xattr_handler ext4_xattr_user_handler; +extern const struct xattr_handler ext4_xattr_trusted_handler; +extern const struct xattr_handler ext4_xattr_security_handler; +extern const struct xattr_handler ext4_xattr_hurd_handler; + +#define EXT4_XATTR_NAME_ENCRYPTION_CONTEXT "c" + +/* + * The EXT4_STATE_NO_EXPAND is overloaded and used for two purposes. + * The first is to signal that there the inline xattrs and data are + * taking up so much space that we might as well not keep trying to + * expand it. The second is that xattr_sem is taken for writing, so + * we shouldn't try to recurse into the inode expansion. For this + * second case, we need to make sure that we take save and restore the + * NO_EXPAND state flag appropriately. + */ +static inline void ext4_write_lock_xattr(struct inode *inode, int *save) +{ + down_write(&EXT4_I(inode)->xattr_sem); + *save = ext4_test_inode_state(inode, EXT4_STATE_NO_EXPAND); + ext4_set_inode_state(inode, EXT4_STATE_NO_EXPAND); +} + +static inline int ext4_write_trylock_xattr(struct inode *inode, int *save) +{ + if (down_write_trylock(&EXT4_I(inode)->xattr_sem) == 0) + return 0; + *save = ext4_test_inode_state(inode, EXT4_STATE_NO_EXPAND); + ext4_set_inode_state(inode, EXT4_STATE_NO_EXPAND); + return 1; +} + +static inline void ext4_write_unlock_xattr(struct inode *inode, int *save) +{ + if (*save == 0) + ext4_clear_inode_state(inode, EXT4_STATE_NO_EXPAND); + up_write(&EXT4_I(inode)->xattr_sem); +} + +extern ssize_t ext4_listxattr(struct dentry *, char *, size_t); + +extern int ext4_xattr_get(struct inode *, int, const char *, void *, size_t); +extern int ext4_xattr_set(struct inode *, int, const char *, const void *, size_t, int); +extern int ext4_xattr_set_handle(handle_t *, struct inode *, int, const char *, const void *, size_t, int); +extern int ext4_xattr_set_credits(struct inode *inode, size_t value_len, + bool is_create, int *credits); +extern int __ext4_xattr_set_credits(struct super_block *sb, struct inode *inode, + struct buffer_head *block_bh, size_t value_len, + bool is_create); + +extern int ext4_xattr_delete_inode(handle_t *handle, struct inode *inode, + struct ext4_xattr_inode_array **array, + int extra_credits); +extern void ext4_xattr_inode_array_free(struct ext4_xattr_inode_array *array); + +extern int ext4_expand_extra_isize_ea(struct inode *inode, int new_extra_isize, + struct ext4_inode *raw_inode, handle_t *handle); +extern void ext4_evict_ea_inode(struct inode *inode); + +extern const struct xattr_handler * const ext4_xattr_handlers[]; + +extern int ext4_xattr_ibody_find(struct inode *inode, struct ext4_xattr_info *i, + struct ext4_xattr_ibody_find *is); +extern int ext4_xattr_ibody_get(struct inode *inode, int name_index, + const char *name, + void *buffer, size_t buffer_size); +extern int ext4_xattr_ibody_set(handle_t *handle, struct inode *inode, + struct ext4_xattr_info *i, + struct ext4_xattr_ibody_find *is); + +extern struct mb_cache *ext4_xattr_create_cache(void); +extern void ext4_xattr_destroy_cache(struct mb_cache *); + +extern int +__xattr_check_inode(struct inode *inode, struct ext4_xattr_ibody_header *header, + void *end, const char *function, unsigned int line); + +#define xattr_check_inode(inode, header, end) \ + __xattr_check_inode((inode), (header), (end), __func__, __LINE__) + +#ifdef CONFIG_EXT4_FS_SECURITY +extern int ext4_init_security(handle_t *handle, struct inode *inode, + struct inode *dir, const struct qstr *qstr); +#else +static inline int ext4_init_security(handle_t *handle, struct inode *inode, + struct inode *dir, const struct qstr *qstr) +{ + return 0; +} +#endif + +#ifdef CONFIG_LOCKDEP +extern void ext4_xattr_inode_set_class(struct inode *ea_inode); +#else +static inline void ext4_xattr_inode_set_class(struct inode *ea_inode) { } +#endif + +extern int ext4_get_inode_usage(struct inode *inode, qsize_t *usage); diff --git a/fs/ext4l/xattr_hurd.c b/fs/ext4l/xattr_hurd.c new file mode 100644 index 00000000000..8a5842e4cd9 --- /dev/null +++ b/fs/ext4l/xattr_hurd.c @@ -0,0 +1,52 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * linux/fs/ext4/xattr_hurd.c + * Handler for extended gnu attributes for the Hurd. + * + * Copyright (C) 2001 by Andreas Gruenbacher, <a.gruenbacher@computer.org> + * Copyright (C) 2020 by Jan (janneke) Nieuwenhuizen, <janneke@gnu.org> + */ + +#include <linux/init.h> +#include <linux/string.h> +#include "ext4.h" +#include "xattr.h" + +static bool +ext4_xattr_hurd_list(struct dentry *dentry) +{ + return test_opt(dentry->d_sb, XATTR_USER); +} + +static int +ext4_xattr_hurd_get(const struct xattr_handler *handler, + struct dentry *unused, struct inode *inode, + const char *name, void *buffer, size_t size) +{ + if (!test_opt(inode->i_sb, XATTR_USER)) + return -EOPNOTSUPP; + + return ext4_xattr_get(inode, EXT4_XATTR_INDEX_HURD, + name, buffer, size); +} + +static int +ext4_xattr_hurd_set(const struct xattr_handler *handler, + struct mnt_idmap *idmap, + struct dentry *unused, struct inode *inode, + const char *name, const void *value, + size_t size, int flags) +{ + if (!test_opt(inode->i_sb, XATTR_USER)) + return -EOPNOTSUPP; + + return ext4_xattr_set(inode, EXT4_XATTR_INDEX_HURD, + name, value, size, flags); +} + +const struct xattr_handler ext4_xattr_hurd_handler = { + .prefix = XATTR_HURD_PREFIX, + .list = ext4_xattr_hurd_list, + .get = ext4_xattr_hurd_get, + .set = ext4_xattr_hurd_set, +}; diff --git a/fs/ext4l/xattr_security.c b/fs/ext4l/xattr_security.c new file mode 100644 index 00000000000..776cf11d24c --- /dev/null +++ b/fs/ext4l/xattr_security.c @@ -0,0 +1,66 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * linux/fs/ext4/xattr_security.c + * Handler for storing security labels as extended attributes. + */ + +#include <linux/string.h> +#include <linux/fs.h> +#include <linux/security.h> +#include <linux/slab.h> +#include "ext4_jbd2.h" +#include "ext4.h" +#include "xattr.h" + +static int +ext4_xattr_security_get(const struct xattr_handler *handler, + struct dentry *unused, struct inode *inode, + const char *name, void *buffer, size_t size) +{ + return ext4_xattr_get(inode, EXT4_XATTR_INDEX_SECURITY, + name, buffer, size); +} + +static int +ext4_xattr_security_set(const struct xattr_handler *handler, + struct mnt_idmap *idmap, + struct dentry *unused, struct inode *inode, + const char *name, const void *value, + size_t size, int flags) +{ + return ext4_xattr_set(inode, EXT4_XATTR_INDEX_SECURITY, + name, value, size, flags); +} + +static int +ext4_initxattrs(struct inode *inode, const struct xattr *xattr_array, + void *fs_info) +{ + const struct xattr *xattr; + handle_t *handle = fs_info; + int err = 0; + + for (xattr = xattr_array; xattr->name != NULL; xattr++) { + err = ext4_xattr_set_handle(handle, inode, + EXT4_XATTR_INDEX_SECURITY, + xattr->name, xattr->value, + xattr->value_len, XATTR_CREATE); + if (err < 0) + break; + } + return err; +} + +int +ext4_init_security(handle_t *handle, struct inode *inode, struct inode *dir, + const struct qstr *qstr) +{ + return security_inode_init_security(inode, dir, qstr, + &ext4_initxattrs, handle); +} + +const struct xattr_handler ext4_xattr_security_handler = { + .prefix = XATTR_SECURITY_PREFIX, + .get = ext4_xattr_security_get, + .set = ext4_xattr_security_set, +}; diff --git a/fs/ext4l/xattr_trusted.c b/fs/ext4l/xattr_trusted.c new file mode 100644 index 00000000000..9811eb0ab27 --- /dev/null +++ b/fs/ext4l/xattr_trusted.c @@ -0,0 +1,47 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * linux/fs/ext4/xattr_trusted.c + * Handler for trusted extended attributes. + * + * Copyright (C) 2003 by Andreas Gruenbacher, <a.gruenbacher@computer.org> + */ + +#include <linux/string.h> +#include <linux/capability.h> +#include <linux/fs.h> +#include "ext4_jbd2.h" +#include "ext4.h" +#include "xattr.h" + +static bool +ext4_xattr_trusted_list(struct dentry *dentry) +{ + return capable(CAP_SYS_ADMIN); +} + +static int +ext4_xattr_trusted_get(const struct xattr_handler *handler, + struct dentry *unused, struct inode *inode, + const char *name, void *buffer, size_t size) +{ + return ext4_xattr_get(inode, EXT4_XATTR_INDEX_TRUSTED, + name, buffer, size); +} + +static int +ext4_xattr_trusted_set(const struct xattr_handler *handler, + struct mnt_idmap *idmap, + struct dentry *unused, struct inode *inode, + const char *name, const void *value, + size_t size, int flags) +{ + return ext4_xattr_set(inode, EXT4_XATTR_INDEX_TRUSTED, + name, value, size, flags); +} + +const struct xattr_handler ext4_xattr_trusted_handler = { + .prefix = XATTR_TRUSTED_PREFIX, + .list = ext4_xattr_trusted_list, + .get = ext4_xattr_trusted_get, + .set = ext4_xattr_trusted_set, +}; diff --git a/fs/ext4l/xattr_user.c b/fs/ext4l/xattr_user.c new file mode 100644 index 00000000000..4b70bf4e762 --- /dev/null +++ b/fs/ext4l/xattr_user.c @@ -0,0 +1,50 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * linux/fs/ext4/xattr_user.c + * Handler for extended user attributes. + * + * Copyright (C) 2001 by Andreas Gruenbacher, <a.gruenbacher@computer.org> + */ + +#include <linux/string.h> +#include <linux/fs.h> +#include "ext4_jbd2.h" +#include "ext4.h" +#include "xattr.h" + +static bool +ext4_xattr_user_list(struct dentry *dentry) +{ + return test_opt(dentry->d_sb, XATTR_USER); +} + +static int +ext4_xattr_user_get(const struct xattr_handler *handler, + struct dentry *unused, struct inode *inode, + const char *name, void *buffer, size_t size) +{ + if (!test_opt(inode->i_sb, XATTR_USER)) + return -EOPNOTSUPP; + return ext4_xattr_get(inode, EXT4_XATTR_INDEX_USER, + name, buffer, size); +} + +static int +ext4_xattr_user_set(const struct xattr_handler *handler, + struct mnt_idmap *idmap, + struct dentry *unused, struct inode *inode, + const char *name, const void *value, + size_t size, int flags) +{ + if (!test_opt(inode->i_sb, XATTR_USER)) + return -EOPNOTSUPP; + return ext4_xattr_set(inode, EXT4_XATTR_INDEX_USER, + name, value, size, flags); +} + +const struct xattr_handler ext4_xattr_user_handler = { + .prefix = XATTR_USER_PREFIX, + .list = ext4_xattr_user_list, + .get = ext4_xattr_user_get, + .set = ext4_xattr_user_set, +}; -- 2.43.0