This is a note to let you know that I've just added the patch titled ext4: fast commit may not fallback for ineligible commit to the 5.16-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary The filename of the patch is: ext4-fast-commit-may-not-fallback-for-ineligible-com.patch and it can be found in the queue-5.16 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable@xxxxxxxxxxxxxxx> know about it. commit 642f8c30b87603489a33051fd74a51c0a4447cf7 Author: Xin Yin <yinxin.x@xxxxxxxxxxxxx> Date: Mon Jan 17 17:36:54 2022 +0800 ext4: fast commit may not fallback for ineligible commit [ Upstream commit e85c81ba8859a4c839bcd69c5d83b32954133a5b ] For the follow scenario: 1. jbd start commit transaction n 2. task A get new handle for transaction n+1 3. task A do some ineligible actions and mark FC_INELIGIBLE 4. jbd complete transaction n and clean FC_INELIGIBLE 5. task A call fsync In this case fast commit will not fallback to full commit and transaction n+1 also not handled by jbd. Make ext4_fc_mark_ineligible() also record transaction tid for latest ineligible case, when call ext4_fc_cleanup() check current transaction tid, if small than latest ineligible tid do not clear the EXT4_MF_FC_INELIGIBLE. Reported-by: kernel test robot <lkp@xxxxxxxxx> Reported-by: Dan Carpenter <dan.carpenter@xxxxxxxxxx> Reported-by: Ritesh Harjani <riteshh@xxxxxxxxxxxxx> Suggested-by: Harshad Shirwadkar <harshadshirwadkar@xxxxxxxxx> Signed-off-by: Xin Yin <yinxin.x@xxxxxxxxxxxxx> Link: https://lore.kernel.org/r/20220117093655.35160-2-yinxin.x@xxxxxxxxxxxxx Signed-off-by: Theodore Ts'o <tytso@xxxxxxx> Cc: stable@xxxxxxxxxx Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx> diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 470fd3c2aef54..8b5015ea46199 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -1747,6 +1747,7 @@ struct ext4_sb_info { spinlock_t s_fc_lock; struct buffer_head *s_fc_bh; struct ext4_fc_stats s_fc_stats; + tid_t s_fc_ineligible_tid; #ifdef CONFIG_EXT4_DEBUG int s_fc_debug_max_replay; #endif @@ -2924,7 +2925,7 @@ void __ext4_fc_track_create(handle_t *handle, struct inode *inode, struct dentry *dentry); void ext4_fc_track_create(handle_t *handle, struct dentry *dentry); void ext4_fc_track_inode(handle_t *handle, struct inode *inode); -void ext4_fc_mark_ineligible(struct super_block *sb, int reason); +void ext4_fc_mark_ineligible(struct super_block *sb, int reason, handle_t *handle); void ext4_fc_start_update(struct inode *inode); void ext4_fc_stop_update(struct inode *inode); void ext4_fc_del(struct inode *inode); diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index d3a8d704d8b4f..d2667189be7e5 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -5342,7 +5342,7 @@ static int ext4_collapse_range(struct inode *inode, loff_t offset, loff_t len) ret = PTR_ERR(handle); goto out_mmap; } - ext4_fc_mark_ineligible(sb, EXT4_FC_REASON_FALLOC_RANGE); + ext4_fc_mark_ineligible(sb, EXT4_FC_REASON_FALLOC_RANGE, handle); down_write(&EXT4_I(inode)->i_data_sem); ext4_discard_preallocations(inode, 0); @@ -5482,7 +5482,7 @@ static int ext4_insert_range(struct inode *inode, loff_t offset, loff_t len) ret = PTR_ERR(handle); goto out_mmap; } - ext4_fc_mark_ineligible(sb, EXT4_FC_REASON_FALLOC_RANGE); + 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; diff --git a/fs/ext4/fast_commit.c b/fs/ext4/fast_commit.c index 1b935feec6f6b..0cdfc5003d91a 100644 --- a/fs/ext4/fast_commit.c +++ b/fs/ext4/fast_commit.c @@ -302,18 +302,32 @@ void ext4_fc_del(struct inode *inode) } /* - * Mark file system as fast commit ineligible. This means that next commit - * operation would result in a full jbd2 commit. + * 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) +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; if (!test_opt2(sb, JOURNAL_FAST_COMMIT) || (EXT4_SB(sb)->s_mount_state & EXT4_FC_REPLAY)) return; ext4_set_mount_flag(sb, EXT4_MF_FC_INELIGIBLE); + if (handle && !IS_ERR(handle)) + tid = handle->h_transaction->t_tid; + else { + read_lock(&sbi->s_journal->j_state_lock); + tid = sbi->s_journal->j_running_transaction ? + sbi->s_journal->j_running_transaction->t_tid : 0; + read_unlock(&sbi->s_journal->j_state_lock); + } + spin_lock(&sbi->s_fc_lock); + if (sbi->s_fc_ineligible_tid < tid) + sbi->s_fc_ineligible_tid = tid; + spin_unlock(&sbi->s_fc_lock); WARN_ON(reason >= EXT4_FC_REASON_MAX); sbi->s_fc_stats.fc_ineligible_reason_count[reason]++; } @@ -389,7 +403,7 @@ static int __track_dentry_update(struct inode *inode, void *arg, bool update) mutex_unlock(&ei->i_fc_lock); node = kmem_cache_alloc(ext4_fc_dentry_cachep, GFP_NOFS); if (!node) { - ext4_fc_mark_ineligible(inode->i_sb, EXT4_FC_REASON_NOMEM); + ext4_fc_mark_ineligible(inode->i_sb, EXT4_FC_REASON_NOMEM, NULL); mutex_lock(&ei->i_fc_lock); return -ENOMEM; } @@ -402,7 +416,7 @@ static int __track_dentry_update(struct inode *inode, void *arg, bool update) if (!node->fcd_name.name) { kmem_cache_free(ext4_fc_dentry_cachep, node); ext4_fc_mark_ineligible(inode->i_sb, - EXT4_FC_REASON_NOMEM); + EXT4_FC_REASON_NOMEM, NULL); mutex_lock(&ei->i_fc_lock); return -ENOMEM; } @@ -504,7 +518,7 @@ void ext4_fc_track_inode(handle_t *handle, struct inode *inode) if (ext4_should_journal_data(inode)) { ext4_fc_mark_ineligible(inode->i_sb, - EXT4_FC_REASON_INODE_JOURNAL_DATA); + EXT4_FC_REASON_INODE_JOURNAL_DATA, handle); return; } @@ -1182,7 +1196,7 @@ int ext4_fc_commit(journal_t *journal, tid_t commit_tid) * 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) +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); @@ -1230,7 +1244,10 @@ static void ext4_fc_cleanup(journal_t *journal, int full) &sbi->s_fc_q[FC_Q_MAIN]); ext4_clear_mount_flag(sb, EXT4_MF_FC_COMMITTING); - ext4_clear_mount_flag(sb, EXT4_MF_FC_INELIGIBLE); + if (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; diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 3bdfe010e17f9..2f5686dfa30d5 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -337,7 +337,7 @@ void ext4_evict_inode(struct inode *inode) return; no_delete: if (!list_empty(&EXT4_I(inode)->i_fc_list)) - ext4_fc_mark_ineligible(inode->i_sb, EXT4_FC_REASON_NOMEM); + ext4_fc_mark_ineligible(inode->i_sb, EXT4_FC_REASON_NOMEM, NULL); ext4_clear_inode(inode); /* We must guarantee clearing of inode... */ } @@ -5983,7 +5983,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) return PTR_ERR(handle); ext4_fc_mark_ineligible(inode->i_sb, - EXT4_FC_REASON_JOURNAL_FLAG_CHANGE); + EXT4_FC_REASON_JOURNAL_FLAG_CHANGE, handle); err = ext4_mark_inode_dirty(handle, inode); ext4_handle_sync(handle); ext4_journal_stop(handle); diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c index fd70bebb14370..f61b59045c6d3 100644 --- a/fs/ext4/ioctl.c +++ b/fs/ext4/ioctl.c @@ -169,7 +169,7 @@ static long swap_inode_boot_loader(struct super_block *sb, err = -EINVAL; goto err_out; } - ext4_fc_mark_ineligible(sb, EXT4_FC_REASON_SWAP_BOOT); + ext4_fc_mark_ineligible(sb, EXT4_FC_REASON_SWAP_BOOT, handle); /* Protect extent tree against block allocations via delalloc */ ext4_double_down_write_data_sem(inode, inode_bl); @@ -1075,7 +1075,7 @@ static long __ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) err = ext4_resize_fs(sb, n_blocks_count); if (EXT4_SB(sb)->s_journal) { - ext4_fc_mark_ineligible(sb, EXT4_FC_REASON_RESIZE); + ext4_fc_mark_ineligible(sb, EXT4_FC_REASON_RESIZE, NULL); jbd2_journal_lock_updates(EXT4_SB(sb)->s_journal); err2 = jbd2_journal_flush(EXT4_SB(sb)->s_journal, 0); jbd2_journal_unlock_updates(EXT4_SB(sb)->s_journal); diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 52c9bd154122a..47b9f87dbc6f7 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -3889,7 +3889,7 @@ static int ext4_rename(struct user_namespace *mnt_userns, struct inode *old_dir, * dirents in directories. */ ext4_fc_mark_ineligible(old.inode->i_sb, - EXT4_FC_REASON_RENAME_DIR); + EXT4_FC_REASON_RENAME_DIR, handle); } else { if (new.inode) ext4_fc_track_unlink(handle, new.dentry); @@ -4049,7 +4049,7 @@ static int ext4_cross_rename(struct inode *old_dir, struct dentry *old_dentry, if (unlikely(retval)) goto end_rename; ext4_fc_mark_ineligible(new.inode->i_sb, - EXT4_FC_REASON_CROSS_RENAME); + EXT4_FC_REASON_CROSS_RENAME, handle); if (old.dir_bh) { retval = ext4_rename_dir_finish(handle, &old, new.dir->i_ino); if (retval) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index d304b72593d76..888b2db92924d 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -4627,6 +4627,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) sbi->s_fc_bytes = 0; ext4_clear_mount_flag(sb, EXT4_MF_FC_INELIGIBLE); ext4_clear_mount_flag(sb, EXT4_MF_FC_COMMITTING); + sbi->s_fc_ineligible_tid = 0; spin_lock_init(&sbi->s_fc_lock); memset(&sbi->s_fc_stats, 0, sizeof(sbi->s_fc_stats)); sbi->s_fc_replay_state.fc_regions = NULL; diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c index 1e0fc1ed845bf..0423253490986 100644 --- a/fs/ext4/xattr.c +++ b/fs/ext4/xattr.c @@ -2408,7 +2408,7 @@ ext4_xattr_set_handle(handle_t *handle, struct inode *inode, int name_index, if (IS_SYNC(inode)) ext4_handle_sync(handle); } - ext4_fc_mark_ineligible(inode->i_sb, EXT4_FC_REASON_XATTR); + ext4_fc_mark_ineligible(inode->i_sb, EXT4_FC_REASON_XATTR, handle); cleanup: brelse(is.iloc.bh); @@ -2486,7 +2486,7 @@ ext4_xattr_set(struct inode *inode, int name_index, const char *name, if (error == 0) error = error2; } - ext4_fc_mark_ineligible(inode->i_sb, EXT4_FC_REASON_XATTR); + ext4_fc_mark_ineligible(inode->i_sb, EXT4_FC_REASON_XATTR, NULL); return error; } @@ -2920,7 +2920,7 @@ int ext4_xattr_delete_inode(handle_t *handle, struct inode *inode, error); goto cleanup; } - ext4_fc_mark_ineligible(inode->i_sb, EXT4_FC_REASON_XATTR); + ext4_fc_mark_ineligible(inode->i_sb, EXT4_FC_REASON_XATTR, handle); } error = 0; cleanup: diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c index 3cc4ab2ba7f4f..d188fa913a075 100644 --- a/fs/jbd2/commit.c +++ b/fs/jbd2/commit.c @@ -1170,7 +1170,7 @@ void jbd2_journal_commit_transaction(journal_t *journal) if (journal->j_commit_callback) journal->j_commit_callback(journal, commit_transaction); if (journal->j_fc_cleanup_callback) - journal->j_fc_cleanup_callback(journal, 1); + journal->j_fc_cleanup_callback(journal, 1, commit_transaction->t_tid); trace_jbd2_end_commit(journal, commit_transaction); jbd_debug(1, "JBD2: commit %d complete, head %d\n", diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index bd9ac98916043..1f8493ef181d6 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -769,7 +769,7 @@ EXPORT_SYMBOL(jbd2_fc_begin_commit); static int __jbd2_fc_end_commit(journal_t *journal, tid_t tid, bool fallback) { if (journal->j_fc_cleanup_callback) - journal->j_fc_cleanup_callback(journal, 0); + journal->j_fc_cleanup_callback(journal, 0, tid); write_lock(&journal->j_state_lock); journal->j_flags &= ~JBD2_FAST_COMMIT_ONGOING; if (fallback) diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h index fd933c45281af..d63b8106796e2 100644 --- a/include/linux/jbd2.h +++ b/include/linux/jbd2.h @@ -1295,7 +1295,7 @@ struct journal_s * Clean-up after fast commit or full commit. JBD2 calls this function * after every commit operation. */ - void (*j_fc_cleanup_callback)(struct journal_s *journal, int); + void (*j_fc_cleanup_callback)(struct journal_s *journal, int full, tid_t tid); /** * @j_fc_replay_callback: