On Tue 18-03-25 13:22:56, Ojaswin Mujoo wrote:
> Presently we always BUG_ON if trying to start a transaction on a journal marked
> with JBD2_UNMOUNT, since this should never happen. However, while ltp running
> stress tests, it was observed that in case of some error handling paths, it is
> possible for update_super_work to start a transaction after the journal is
> destroyed eg:
>
> (umount)
> ext4_kill_sb
> kill_block_super
> generic_shutdown_super
> sync_filesystem /* commits all txns */
> evict_inodes
> /* might start a new txn */
> ext4_put_super
> flush_work(&sbi->s_sb_upd_work) /* flush the workqueue */
> jbd2_journal_destroy
> journal_kill_thread
> journal->j_flags |= JBD2_UNMOUNT;
> jbd2_journal_commit_transaction
> jbd2_journal_get_descriptor_buffer
> jbd2_journal_bmap
> ext4_journal_bmap
> ext4_map_blocks
> ...
> ext4_inode_error
> ext4_handle_error
> schedule_work(&sbi->s_sb_upd_work)
>
> /* work queue kicks in */
> update_super_work
> jbd2_journal_start
> start_this_handle
> BUG_ON(journal->j_flags &
> JBD2_UNMOUNT)
>
> Hence, introduce a new mount flag to indicate journal is destroying and only do
> a journaled (and deferred) update of sb if this flag is not set. Otherwise, just
> fallback to an un-journaled commit.
>
> Further, in the journal destroy path, we have the following sequence:
>
> 1. Set mount flag indicating journal is destroying
> 2. force a commit and wait for it
> 3. flush pending sb updates
>
> This sequence is important as it ensures that, after this point, there is no sb
> update that might be journaled so it is safe to update the sb outside the
> journal. (To avoid race discussed in 2d01ddc86606)
>
> Also, we don't need a similar check in ext4_grp_locked_error since it is only
> called from mballoc and AFAICT it would be always valid to schedule work here.
>
> Fixes: 2d01ddc86606 ("ext4: save error info to sb through journal if available")
> Reported-by: Mahesh Kumar <maheshkumar657g@xxxxxxxxx>
> Signed-off-by: Ojaswin Mujoo <ojaswin@xxxxxxxxxxxxx>
Looks good. Feel free to add:
Reviewed-by: Jan Kara <jack@xxxxxxx>
Honza
> ---
> fs/ext4/ext4.h | 3 ++-
> fs/ext4/ext4_jbd2.h | 15 +++++++++++++++
> fs/ext4/super.c | 16 ++++++++--------
> 3 files changed, 25 insertions(+), 9 deletions(-)
>
> diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
> index 2b7d781bfcad..0685bb68e64a 100644
> --- a/fs/ext4/ext4.h
> +++ b/fs/ext4/ext4.h
> @@ -1823,7 +1823,8 @@ static inline int ext4_valid_inum(struct super_block *sb, unsigned long ino)
> */
> enum {
> EXT4_MF_MNTDIR_SAMPLED,
> - EXT4_MF_FC_INELIGIBLE /* Fast commit ineligible */
> + EXT4_MF_FC_INELIGIBLE, /* Fast commit ineligible */
> + EXT4_MF_JOURNAL_DESTROY /* Journal is in process of destroying */
> };
>
> static inline void ext4_set_mount_flag(struct super_block *sb, int bit)
> diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
> index 9b3c9df02a39..3221714d9901 100644
> --- a/fs/ext4/ext4_jbd2.h
> +++ b/fs/ext4/ext4_jbd2.h
> @@ -437,6 +437,21 @@ static inline int ext4_journal_destroy(struct ext4_sb_info *sbi, journal_t *jour
> {
> 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;
>
> diff --git a/fs/ext4/super.c b/fs/ext4/super.c
> index 8ad664d47806..46f7c9922cda 100644
> --- a/fs/ext4/super.c
> +++ b/fs/ext4/super.c
> @@ -704,9 +704,13 @@ static void ext4_handle_error(struct super_block *sb, bool force_ro, int error,
> * 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.
> + * 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)
> + 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);
> @@ -1291,7 +1295,6 @@ static void ext4_put_super(struct super_block *sb)
> ext4_unregister_li_request(sb);
> ext4_quotas_off(sb, EXT4_MAXQUOTAS);
>
> - flush_work(&sbi->s_sb_upd_work);
> destroy_workqueue(sbi->rsv_conversion_wq);
> ext4_release_orphan_info(sb);
>
> @@ -1301,7 +1304,8 @@ static void ext4_put_super(struct super_block *sb)
> 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);
> @@ -4957,8 +4961,6 @@ static int ext4_load_and_init_journal(struct super_block *sb,
> return 0;
>
> out:
> - /* flush s_sb_upd_work before destroying the journal. */
> - flush_work(&sbi->s_sb_upd_work);
> ext4_journal_destroy(sbi, sbi->s_journal);
> return -EINVAL;
> }
> @@ -5648,8 +5650,6 @@ failed_mount8: __maybe_unused
> sbi->s_ea_block_cache = NULL;
>
> if (sbi->s_journal) {
> - /* flush s_sb_upd_work before journal destroy. */
> - flush_work(&sbi->s_sb_upd_work);
> ext4_journal_destroy(sbi, sbi->s_journal);
> }
> failed_mount3a:
> --
> 2.48.1
>
--
Jan Kara <jack@xxxxxxxx>
SUSE Labs, CR
