diff options
Diffstat (limited to 'fs/ext4/fast_commit.c')
-rw-r--r-- | fs/ext4/fast_commit.c | 210 |
1 files changed, 124 insertions, 86 deletions
diff --git a/fs/ext4/fast_commit.c b/fs/ext4/fast_commit.c index 2af962cbb835..ef05bfa87798 100644 --- a/fs/ext4/fast_commit.c +++ b/fs/ext4/fast_commit.c @@ -229,6 +229,12 @@ __releases(&EXT4_SB(inode->i_sb)->s_fc_lock) finish_wait(wq, &wait.wq_entry); } +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)); +} + /* * Inform Ext4's fast about start of an inode update * @@ -240,8 +246,7 @@ void ext4_fc_start_update(struct inode *inode) { struct ext4_inode_info *ei = EXT4_I(inode); - if (!test_opt2(inode->i_sb, JOURNAL_FAST_COMMIT) || - (EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY)) + if (ext4_fc_disabled(inode->i_sb)) return; restart: @@ -265,8 +270,7 @@ void ext4_fc_stop_update(struct inode *inode) { struct ext4_inode_info *ei = EXT4_I(inode); - if (!test_opt2(inode->i_sb, JOURNAL_FAST_COMMIT) || - (EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY)) + if (ext4_fc_disabled(inode->i_sb)) return; if (atomic_dec_and_test(&ei->i_fc_updates)) @@ -283,8 +287,7 @@ void ext4_fc_del(struct inode *inode) struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); struct ext4_fc_dentry_update *fc_dentry; - if (!test_opt2(inode->i_sb, JOURNAL_FAST_COMMIT) || - (EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY)) + if (ext4_fc_disabled(inode->i_sb)) return; restart: @@ -337,8 +340,7 @@ void ext4_fc_mark_ineligible(struct super_block *sb, int reason, handle_t *handl struct ext4_sb_info *sbi = EXT4_SB(sb); tid_t tid; - if (!test_opt2(sb, JOURNAL_FAST_COMMIT) || - (EXT4_SB(sb)->s_mount_state & EXT4_FC_REPLAY)) + if (ext4_fc_disabled(sb)) return; ext4_set_mount_flag(sb, EXT4_MF_FC_INELIGIBLE); @@ -493,10 +495,8 @@ void __ext4_fc_track_unlink(handle_t *handle, void ext4_fc_track_unlink(handle_t *handle, struct dentry *dentry) { struct inode *inode = d_inode(dentry); - struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); - if (!test_opt2(inode->i_sb, JOURNAL_FAST_COMMIT) || - (sbi->s_mount_state & EXT4_FC_REPLAY)) + if (ext4_fc_disabled(inode->i_sb)) return; if (ext4_test_mount_flag(inode->i_sb, EXT4_MF_FC_INELIGIBLE)) @@ -522,10 +522,8 @@ void __ext4_fc_track_link(handle_t *handle, void ext4_fc_track_link(handle_t *handle, struct dentry *dentry) { struct inode *inode = d_inode(dentry); - struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); - if (!test_opt2(inode->i_sb, JOURNAL_FAST_COMMIT) || - (sbi->s_mount_state & EXT4_FC_REPLAY)) + if (ext4_fc_disabled(inode->i_sb)) return; if (ext4_test_mount_flag(inode->i_sb, EXT4_MF_FC_INELIGIBLE)) @@ -551,10 +549,8 @@ void __ext4_fc_track_create(handle_t *handle, struct inode *inode, void ext4_fc_track_create(handle_t *handle, struct dentry *dentry) { struct inode *inode = d_inode(dentry); - struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); - if (!test_opt2(inode->i_sb, JOURNAL_FAST_COMMIT) || - (sbi->s_mount_state & EXT4_FC_REPLAY)) + if (ext4_fc_disabled(inode->i_sb)) return; if (ext4_test_mount_flag(inode->i_sb, EXT4_MF_FC_INELIGIBLE)) @@ -576,22 +572,20 @@ static int __track_inode(struct inode *inode, void *arg, bool update) void ext4_fc_track_inode(handle_t *handle, struct inode *inode) { - struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); 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 (!test_opt2(inode->i_sb, JOURNAL_FAST_COMMIT) || - (sbi->s_mount_state & EXT4_FC_REPLAY)) - return; - if (ext4_test_mount_flag(inode->i_sb, EXT4_MF_FC_INELIGIBLE)) return; @@ -634,15 +628,13 @@ static int __track_range(struct inode *inode, void *arg, bool update) void ext4_fc_track_range(handle_t *handle, struct inode *inode, ext4_lblk_t start, ext4_lblk_t end) { - struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); struct __track_range_args args; int ret; if (S_ISDIR(inode->i_mode)) return; - if (!test_opt2(inode->i_sb, JOURNAL_FAST_COMMIT) || - (sbi->s_mount_state & EXT4_FC_REPLAY)) + if (ext4_fc_disabled(inode->i_sb)) return; if (ext4_test_mount_flag(inode->i_sb, EXT4_MF_FC_INELIGIBLE)) @@ -710,10 +702,10 @@ static u8 *ext4_fc_reserve_space(struct super_block *sb, int len, u32 *crc) * After allocating len, we should have space at least for a 0 byte * padding. */ - if (len + sizeof(struct ext4_fc_tl) > bsize) + if (len + EXT4_FC_TAG_BASE_LEN > bsize) return NULL; - if (bsize - off - 1 > len + sizeof(struct ext4_fc_tl)) { + if (bsize - off - 1 > len + EXT4_FC_TAG_BASE_LEN) { /* * Only allocate from current buffer if we have enough space for * this request AND we have space to add a zero byte padding. @@ -730,10 +722,10 @@ static u8 *ext4_fc_reserve_space(struct super_block *sb, int len, u32 *crc) /* Need to add PAD tag */ tl = (struct ext4_fc_tl *)(sbi->s_fc_bh->b_data + off); tl->fc_tag = cpu_to_le16(EXT4_FC_TAG_PAD); - pad_len = bsize - off - 1 - sizeof(struct ext4_fc_tl); + pad_len = bsize - off - 1 - EXT4_FC_TAG_BASE_LEN; tl->fc_len = cpu_to_le16(pad_len); if (crc) - *crc = ext4_chksum(sbi, *crc, tl, sizeof(*tl)); + *crc = ext4_chksum(sbi, *crc, tl, EXT4_FC_TAG_BASE_LEN); if (pad_len > 0) ext4_fc_memzero(sb, tl + 1, pad_len, crc); ext4_fc_submit_bh(sb, false); @@ -775,7 +767,7 @@ static int ext4_fc_write_tail(struct super_block *sb, u32 crc) * 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, sizeof(tl) + sizeof(tail), &crc); + dst = ext4_fc_reserve_space(sb, EXT4_FC_TAG_BASE_LEN + sizeof(tail), &crc); if (!dst) return -ENOSPC; @@ -785,8 +777,8 @@ static int ext4_fc_write_tail(struct super_block *sb, u32 crc) tl.fc_len = cpu_to_le16(bsize - off - 1 + sizeof(struct ext4_fc_tail)); sbi->s_fc_bytes = round_up(sbi->s_fc_bytes, bsize); - ext4_fc_memcpy(sb, dst, &tl, sizeof(tl), &crc); - dst += sizeof(tl); + ext4_fc_memcpy(sb, dst, &tl, EXT4_FC_TAG_BASE_LEN, &crc); + dst += EXT4_FC_TAG_BASE_LEN; tail.fc_tid = cpu_to_le32(sbi->s_journal->j_running_transaction->t_tid); ext4_fc_memcpy(sb, dst, &tail.fc_tid, sizeof(tail.fc_tid), &crc); dst += sizeof(tail.fc_tid); @@ -808,15 +800,15 @@ static bool ext4_fc_add_tlv(struct super_block *sb, u16 tag, u16 len, u8 *val, struct ext4_fc_tl tl; u8 *dst; - dst = ext4_fc_reserve_space(sb, sizeof(tl) + len, crc); + 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); - ext4_fc_memcpy(sb, dst, &tl, sizeof(tl), crc); - ext4_fc_memcpy(sb, dst + sizeof(tl), val, len, crc); + ext4_fc_memcpy(sb, dst, &tl, EXT4_FC_TAG_BASE_LEN, crc); + ext4_fc_memcpy(sb, dst + EXT4_FC_TAG_BASE_LEN, val, len, crc); return true; } @@ -828,8 +820,8 @@ static bool ext4_fc_add_dentry_tlv(struct super_block *sb, u32 *crc, struct ext4_fc_dentry_info fcd; struct ext4_fc_tl tl; int dlen = fc_dentry->fcd_name.len; - u8 *dst = ext4_fc_reserve_space(sb, sizeof(tl) + sizeof(fcd) + dlen, - crc); + u8 *dst = ext4_fc_reserve_space(sb, + EXT4_FC_TAG_BASE_LEN + sizeof(fcd) + dlen, crc); if (!dst) return false; @@ -838,8 +830,8 @@ static bool ext4_fc_add_dentry_tlv(struct super_block *sb, u32 *crc, 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); - ext4_fc_memcpy(sb, dst, &tl, sizeof(tl), crc); - dst += sizeof(tl); + ext4_fc_memcpy(sb, dst, &tl, EXT4_FC_TAG_BASE_LEN, crc); + dst += EXT4_FC_TAG_BASE_LEN; ext4_fc_memcpy(sb, dst, &fcd, sizeof(fcd), crc); dst += sizeof(fcd); ext4_fc_memcpy(sb, dst, fc_dentry->fcd_name.name, dlen, crc); @@ -874,22 +866,25 @@ static int ext4_fc_write_inode(struct inode *inode, u32 *crc) 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, - sizeof(tl) + inode_len + sizeof(fc_inode.fc_ino), crc); + EXT4_FC_TAG_BASE_LEN + inode_len + sizeof(fc_inode.fc_ino), crc); if (!dst) - return -ECANCELED; + goto err; - if (!ext4_fc_memcpy(inode->i_sb, dst, &tl, sizeof(tl), crc)) - return -ECANCELED; - dst += sizeof(tl); + if (!ext4_fc_memcpy(inode->i_sb, dst, &tl, EXT4_FC_TAG_BASE_LEN, crc)) + goto err; + dst += EXT4_FC_TAG_BASE_LEN; if (!ext4_fc_memcpy(inode->i_sb, dst, &fc_inode, sizeof(fc_inode), crc)) - return -ECANCELED; + goto err; dst += sizeof(fc_inode); if (!ext4_fc_memcpy(inode->i_sb, dst, (u8 *)ext4_raw_inode(&iloc), inode_len, crc)) - return -ECANCELED; - - return 0; + goto err; + ret = 0; +err: + brelse(iloc.bh); + return ret; } /* @@ -1343,7 +1338,7 @@ struct dentry_info_args { }; static inline void tl_to_darg(struct dentry_info_args *darg, - struct ext4_fc_tl *tl, u8 *val) + struct ext4_fc_tl *tl, u8 *val) { struct ext4_fc_dentry_info fcd; @@ -1352,8 +1347,14 @@ static inline void tl_to_darg(struct dentry_info_args *darg, 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 = le16_to_cpu(tl->fc_len) - - sizeof(struct ext4_fc_dentry_info); + darg->dname_len = tl->fc_len - sizeof(struct ext4_fc_dentry_info); +} + +static inline void ext4_fc_get_tl(struct ext4_fc_tl *tl, u8 *val) +{ + memcpy(tl, val, EXT4_FC_TAG_BASE_LEN); + tl->fc_len = le16_to_cpu(tl->fc_len); + tl->fc_tag = le16_to_cpu(tl->fc_tag); } /* Unlink replay function */ @@ -1491,13 +1492,15 @@ static int ext4_fc_record_modified_inode(struct super_block *sb, int ino) if (state->fc_modified_inodes[i] == ino) return 0; if (state->fc_modified_inodes_used == state->fc_modified_inodes_size) { - state->fc_modified_inodes = krealloc( - state->fc_modified_inodes, + 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 (!state->fc_modified_inodes) + if (!fc_modified_inodes) return -ENOMEM; + state->fc_modified_inodes = fc_modified_inodes; state->fc_modified_inodes_size += EXT4_FC_REPLAY_REALLOC_INCREMENT; } @@ -1516,7 +1519,7 @@ static int ext4_fc_replay_inode(struct super_block *sb, struct ext4_fc_tl *tl, struct ext4_inode *raw_fc_inode; struct inode *inode = NULL; struct ext4_iloc iloc; - int inode_len, ino, ret, tag = le16_to_cpu(tl->fc_tag); + int inode_len, ino, ret, tag = tl->fc_tag; struct ext4_extent_header *eh; memcpy(&fc_inode, val, sizeof(fc_inode)); @@ -1541,7 +1544,7 @@ static int ext4_fc_replay_inode(struct super_block *sb, struct ext4_fc_tl *tl, if (ret) goto out; - inode_len = le16_to_cpu(tl->fc_len) - sizeof(struct ext4_fc_inode); + 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)); @@ -1682,15 +1685,18 @@ int ext4_fc_record_regions(struct super_block *sb, int ino, 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 = krealloc( - state->fc_regions, - state->fc_regions_size * - sizeof(struct ext4_fc_alloc_region), - GFP_KERNEL); - if (!state->fc_regions) - return -ENOMEM; + state->fc_regions = fc_regions; } region = &state->fc_regions[state->fc_regions_used++]; region->ino = ino; @@ -1770,8 +1776,7 @@ static int ext4_fc_replay_add_range(struct super_block *sb, ret = ext4_ext_insert_extent( NULL, inode, &path, &newex, 0); up_write((&EXT4_I(inode)->i_data_sem)); - ext4_ext_drop_refs(path); - kfree(path); + ext4_free_ext_path(path); if (ret) goto out; goto next; @@ -1926,8 +1931,7 @@ static void ext4_fc_set_bitmaps_and_counters(struct super_block *sb) for (j = 0; j < path->p_depth; j++) ext4_mb_mark_bb(inode->i_sb, path[j].p_block, 1, 1); - ext4_ext_drop_refs(path); - kfree(path); + ext4_free_ext_path(path); } cur += ret; ext4_mb_mark_bb(inode->i_sb, map.m_pblk, @@ -1972,6 +1976,34 @@ void ext4_fc_replay_cleanup(struct super_block *sb) kfree(sbi->s_fc_replay_state.fc_modified_inodes); } +static inline bool ext4_fc_tag_len_isvalid(struct ext4_fc_tl *tl, + u8 *val, u8 *end) +{ + if (val + tl->fc_len > end) + return false; + + /* Here only check ADD_RANGE/TAIL/HEAD which will read data when do + * journal rescan before do CRC check. Other tags length check will + * rely on CRC check. + */ + switch (tl->fc_tag) { + case EXT4_FC_TAG_ADD_RANGE: + return (sizeof(struct ext4_fc_add_range) == tl->fc_len); + case EXT4_FC_TAG_TAIL: + return (sizeof(struct ext4_fc_tail) <= tl->fc_len); + case EXT4_FC_TAG_HEAD: + return (sizeof(struct ext4_fc_head) == tl->fc_len); + 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: + default: + return true; + } +} + /* * Recovery Scan phase handler * @@ -2028,12 +2060,18 @@ static int ext4_fc_replay_scan(journal_t *journal, } state->fc_replay_expected_off++; - for (cur = start; cur < end; cur = cur + sizeof(tl) + le16_to_cpu(tl.fc_len)) { - memcpy(&tl, cur, sizeof(tl)); - val = cur + sizeof(tl); + 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 (!ext4_fc_tag_len_isvalid(&tl, val, end)) { + ret = state->fc_replay_num_tags ? + JBD2_FC_REPLAY_STOP : -ECANCELED; + goto out_err; + } ext4_debug("Scan phase, tag:%s, blk %lld\n", - tag2str(le16_to_cpu(tl.fc_tag)), bh->b_blocknr); - switch (le16_to_cpu(tl.fc_tag)) { + 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; @@ -2053,13 +2091,13 @@ static int ext4_fc_replay_scan(journal_t *journal, case EXT4_FC_TAG_PAD: state->fc_cur_tag++; state->fc_crc = ext4_chksum(sbi, state->fc_crc, cur, - sizeof(tl) + le16_to_cpu(tl.fc_len)); + 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(sbi, state->fc_crc, cur, - sizeof(tl) + + EXT4_FC_TAG_BASE_LEN + offsetof(struct ext4_fc_tail, fc_crc)); if (le32_to_cpu(tail.fc_tid) == expected_tid && @@ -2086,7 +2124,7 @@ static int ext4_fc_replay_scan(journal_t *journal, } state->fc_cur_tag++; state->fc_crc = ext4_chksum(sbi, state->fc_crc, cur, - sizeof(tl) + le16_to_cpu(tl.fc_len)); + EXT4_FC_TAG_BASE_LEN + tl.fc_len); break; default: ret = state->fc_replay_num_tags ? @@ -2141,19 +2179,20 @@ static int ext4_fc_replay(journal_t *journal, struct buffer_head *bh, start = (u8 *)bh->b_data; end = (__u8 *)bh->b_data + journal->j_blocksize - 1; - for (cur = start; cur < end; cur = cur + sizeof(tl) + le16_to_cpu(tl.fc_len)) { - memcpy(&tl, cur, sizeof(tl)); - val = cur + sizeof(tl); + 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(le16_to_cpu(tl.fc_tag))); + + ext4_debug("Replay phase, tag:%s\n", tag2str(tl.fc_tag)); state->fc_replay_num_tags--; - switch (le16_to_cpu(tl.fc_tag)) { + switch (tl.fc_tag) { case EXT4_FC_TAG_LINK: ret = ext4_fc_replay_link(sb, &tl, val); break; @@ -2174,19 +2213,18 @@ static int ext4_fc_replay(journal_t *journal, struct buffer_head *bh, break; case EXT4_FC_TAG_PAD: trace_ext4_fc_replay(sb, EXT4_FC_TAG_PAD, 0, - le16_to_cpu(tl.fc_len), 0); + tl.fc_len, 0); break; case EXT4_FC_TAG_TAIL: - trace_ext4_fc_replay(sb, EXT4_FC_TAG_TAIL, 0, - le16_to_cpu(tl.fc_len), 0); + 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, le16_to_cpu(tl.fc_tag), 0, - le16_to_cpu(tl.fc_len), 0); + trace_ext4_fc_replay(sb, tl.fc_tag, 0, tl.fc_len, 0); ret = -ECANCELED; break; } |