Re: [PATCH 07/10] fs: add a ->sync_lazytime method

From: Christoph Hellwig <hch@lst.de>
To: Jan Kara <jack@suse.cz>
Cc: Christoph Hellwig <hch@lst.de>,
	Christian Brauner <brauner@kernel.org>,
	Al Viro <viro@zeniv.linux.org.uk>,
	David Sterba <dsterba@suse.com>,
	Mike Marshall <hubcap@omnibond.com>,
	Martin Brandenburg <martin@omnibond.com>,
	Carlos Maiolino <cem@kernel.org>, Stefan Roesch <shr@fb.com>,
	Jeff Layton <jlayton@kernel.org>,
	linux-kernel@vger.kernel.org, linux-btrfs@vger.kernel.org,
	linux-fsdevel@vger.kernel.org, gfs2@lists.linux.dev,
	io-uring@vger.kernel.org, devel@lists.orangefs.org,
	linux-unionfs@vger.kernel.org, linux-mtd@lists.infradead.org,
	linux-xfs@vger.kernel.org, linux-nfs@vger.kernel.org
Subject: Re: [PATCH 07/10] fs: add a ->sync_lazytime method
Date: Thu, 18 Dec 2025 07:13:17 +0100	[thread overview]
Message-ID: <20251218061317.GA2775@lst.de> (raw)
In-Reply-To: <ghtgokkzdo7owrkfkpittqlc6xvjhr5w4eprbq5gcszqpmy7z3@7m3ecvlqfrzu>

On Wed, Dec 17, 2025 at 01:30:18PM +0100, Jan Kara wrote:
> >  	if (flags & I_DIRTY_INODE) {
> > +		bool was_dirty_time =
> > +			inode_state_read_once(inode) & I_DIRTY_TIME;
> > +
> >  		/*
> >  		 * Inode timestamp update will piggback on this dirtying.
> >  		 * We tell ->dirty_inode callback that timestamps need to
> >  		 * be updated by setting I_DIRTY_TIME in flags.
> >  		 */
> > -		if (inode_state_read_once(inode) & I_DIRTY_TIME) {
> > +		if (was_dirty_time) {
> >  			spin_lock(&inode->i_lock);
> >  			if (inode_state_read(inode) & I_DIRTY_TIME) {
> >  				inode_state_clear(inode, I_DIRTY_TIME);
> >  				flags |= I_DIRTY_TIME;
> > +				was_dirty_time = true;
> 
> This looks bogus. was_dirty_time is already true here. What I think you
> wanted here is to set it to false if locked I_DIRTY_TIME check failed.
> Otherwise the patch looks good.

Or better set it to false at initialization time and only set it to
true here to simply things a bit.  But otherwise: yes.