On 2024/5/6 22:18, Kemeng Shi wrote:
> We always set JBD2_UNMOUNT with j_state_lock held in journal_kill_thread.
> In kjournald2, we check JBD2_UNMOUNT flag two times under the same
> j_state_lock. Then the second check is unnecessary.
> Also see comment in struct journal_s about lock rule of j_flags.
>
> Signed-off-by: Kemeng Shi <shikemeng@xxxxxxxxxxxxxxx>
Looks good to me.
Reviewed-by: Zhang Yi <yi.zhang@xxxxxxxxxx>
> ---
> fs/jbd2/journal.c | 2 --
> 1 file changed, 2 deletions(-)
>
> diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
> index e8f592fbd6e1..ce9004f40ffb 100644
> --- a/fs/jbd2/journal.c
> +++ b/fs/jbd2/journal.c
> @@ -228,8 +228,6 @@ static int kjournald2(void *arg)
> if (transaction && time_after_eq(jiffies,
> transaction->t_expires))
> should_sleep = 0;
> - if (journal->j_flags & JBD2_UNMOUNT)
> - should_sleep = 0;
> if (should_sleep) {
> write_unlock(&journal->j_state_lock);
> schedule();
>
