Received-SPF: pass (google.com: domain of linux-ext4-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Thu, 13 Jan 2022 18:08:42 +0530
From:   Ritesh Harjani <riteshh@linux.ibm.com>
To:     Jan Kara <jack@suse.cz>
Cc:     linux-ext4@vger.kernel.org, linux-fsdevel@vger.kernel.org,
        linux-kernel@vger.kernel.org, Jan Kara <jack@suse.com>,
        Andreas Dilger <adilger.kernel@dilger.ca>, tytso@mit.edu,
        Eric Whitney <enwlinux@gmail.com>
Subject: Re: [PATCH 6/6] jbd2: No need to use t_handle_lock in
 jbd2_journal_wait_updates
Message-ID: <20220113123842.3rpfcyecylt5n3wo@riteshh-domain>
References: <cover.1642044249.git.riteshh@linux.ibm.com>
 <e7e0f8c54306591a3a9c8fead1e0e54358052ab6.1642044249.git.riteshh@linux.ibm.com>
 <20220113112749.d5tfszcksvxvshnn@quack3.lan>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20220113112749.d5tfszcksvxvshnn@quack3.lan>
Precedence: bulk

On 22/01/13 12:27PM, Jan Kara wrote:
> On Thu 13-01-22 08:56:29, Ritesh Harjani wrote:
> > Since jbd2_journal_wait_updates() uses waitq based on t_updates atomic_t
> > variable. So from code review it looks like we don't need to use
> > t_handle_lock spinlock for checking t_updates value.
> > Hence this patch gets rid of the spinlock protection in
> > jbd2_journal_wait_updates()
> >
> > Signed-off-by: Ritesh Harjani <riteshh@linux.ibm.com>
>
> This patch looks good. Feel free to add:
>
> Reviewed-by: Jan Kara <jack@suse.cz>
>
> Actually looking at it, t_handle_lock seems to be very much unused. I agree

I too had this thought in mind. Thanks for taking a deeper look into it :)

>
> we don't need it when waiting for outstanding handles but the only
> remaining uses are:
>
> 1) jbd2_journal_extend() where it is not needed either - we use
> atomic_add_return() to manipulate t_outstanding_credits and hold
> j_state_lock for reading which provides us enough exclusion.
>
> 2) update_t_max_wait() - this is the only valid use of t_handle_lock but we
> can just switch it to cmpxchg loop with a bit of care. Something like:
>
> 	unsigned long old;
>
> 	ts = jbd2_time_diff(ts, transaction->t_start);
> 	old = transaction->t_max_wait;
> 	while (old < ts)
> 		old = cmpxchg(&transaction->t_max_wait, old, ts);
>
> So perhaps you can add two more patches to remove other t_handle_lock uses
> and drop it completely.

Thanks for providing the details Jan :)
Agree with jbd2_journal_extend(). I did looked a bit around t_max_wait and
I agree that something like above could work. I will spend some more time around
that code and will submit those changes together in v2.

-ritesh

>
> 								Honza
>
> > ---
> >  include/linux/jbd2.h | 4 ----
> >  1 file changed, 4 deletions(-)
> >
> > diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
> > index 34b051aa9009..9bef47622b9d 100644
> > --- a/include/linux/jbd2.h
> > +++ b/include/linux/jbd2.h
> > @@ -1768,22 +1768,18 @@ static inline void jbd2_journal_wait_updates(journal_t *journal)
> >  	if (!commit_transaction)
> >  		return;
> >
> > -	spin_lock(&commit_transaction->t_handle_lock);
> >  	while (atomic_read(&commit_transaction->t_updates)) {
> >  		DEFINE_WAIT(wait);
> >
> >  		prepare_to_wait(&journal->j_wait_updates, &wait,
> >  					TASK_UNINTERRUPTIBLE);
> >  		if (atomic_read(&commit_transaction->t_updates)) {
> > -			spin_unlock(&commit_transaction->t_handle_lock);
> >  			write_unlock(&journal->j_state_lock);
> >  			schedule();
> >  			write_lock(&journal->j_state_lock);
> > -			spin_lock(&commit_transaction->t_handle_lock);
> >  		}
> >  		finish_wait(&journal->j_wait_updates, &wait);
> >  	}
> > -	spin_unlock(&commit_transaction->t_handle_lock);
> >  }
> >
> >  /*
> > --
> > 2.31.1
> >
> --
> Jan Kara <jack@suse.com>
> SUSE Labs, CR