linux-kernel - Re: [PATCH 1/2] filelock: add lease_dispose

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <697eb0d57f34882317e1f5cd73951f1e3b1e3175.camel@kernel.org>
Date: Wed, 03 Dec 2025 14:33:39 -0500
From: Jeff Layton <jlayton@...nel.org>
To: Chuck Lever <cel@...nel.org>, Alexander Viro <viro@...iv.linux.org.uk>, 
 Christian Brauner	 <brauner@...nel.org>, Jan Kara <jack@...e.cz>, Chuck
 Lever	 <chuck.lever@...cle.com>, Alexander Aring <alex.aring@...il.com>,
 "Matthew Wilcox (Oracle)" <willy@...radead.org>, Jonathan Corbet
 <corbet@....net>, NeilBrown <neil@...wn.name>, Olga Kornievskaia
 <okorniev@...hat.com>, Dai Ngo <Dai.Ngo@...cle.com>, Tom Talpey
 <tom@...pey.com>
Cc: linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org, 
	linux-doc@...r.kernel.org, linux-nfs@...r.kernel.org
Subject: Re: [PATCH 1/2] filelock: add lease_dispose_list() helper

On Wed, 2025-12-03 at 13:55 -0500, Chuck Lever wrote:
> 
> On Mon, Dec 1, 2025, at 10:08 AM, Jeff Layton wrote:
> > ...and call that from the lease handling code instead of
> > locks_dispose_list(). Remove the lease handling parts from
> > locks_dispose_list().
> 
> The actual change here isn't bothering me, but I'm having trouble
> understanding why it's needed. It doesn't appear to be a strict
> functional prerequisite for 2/2.
> 

It's not. We can table this patch for now if that's preferable, but I
do think it's a worthwhile cleanup.
 
> A little more context in the commit message would be helpful.
> Sample commit description:
> 
>   The lease-handling code paths always know they're disposing of leases,
>   yet locks_dispose_list() checks flags at runtime to determine whether
>   to call locks_free_lease() or locks_free_lock().
> 
>   Split out a dedicated lease_dispose_list() helper for lease code paths.
>   This makes the type handling explicit and prepares for the upcoming
>   lease_manager enhancements where lease-specific operations are being
>   consolidated.
> 

I may crib this if I end up resending it.

> But that reflects only my naive understanding of the patch. You
> might have something else in mind.
> 
> 

Nope, no ulterior motive here. It's just a nice to have cleanup that
helps to further separate the lock and lease handling code.

> > Signed-off-by: Jeff Layton <jlayton@...nel.org>
> > ---
> >  fs/locks.c | 29 +++++++++++++++++++----------
> >  1 file changed, 19 insertions(+), 10 deletions(-)
> > 
> > diff --git a/fs/locks.c b/fs/locks.c
> > index 
> > 7f4ccc7974bc8d3e82500ee692c6520b53f2280f..e974f8e180fe48682a271af4f143e6bc8e9c4d3b 
> > 100644
> > --- a/fs/locks.c
> > +++ b/fs/locks.c
> > @@ -369,10 +369,19 @@ locks_dispose_list(struct list_head *dispose)
> >  	while (!list_empty(dispose)) {
> >  		flc = list_first_entry(dispose, struct file_lock_core, flc_list);
> >  		list_del_init(&flc->flc_list);
> > -		if (flc->flc_flags & (FL_LEASE|FL_DELEG|FL_LAYOUT))
> > -			locks_free_lease(file_lease(flc));
> > -		else
> > -			locks_free_lock(file_lock(flc));
> > +		locks_free_lock(file_lock(flc));
> > +	}
> > +}
> > +
> > +static void
> > +lease_dispose_list(struct list_head *dispose)
> > +{
> > +	struct file_lock_core *flc;
> > +
> > +	while (!list_empty(dispose)) {
> > +		flc = list_first_entry(dispose, struct file_lock_core, flc_list);
> > +		list_del_init(&flc->flc_list);
> > +		locks_free_lease(file_lease(flc));
> >  	}
> >  }
> > 
> > @@ -1620,7 +1629,7 @@ int __break_lease(struct inode *inode, unsigned int flags)
> >  	spin_unlock(&ctx->flc_lock);
> >  	percpu_up_read(&file_rwsem);
> > 
> > -	locks_dispose_list(&dispose);
> > +	lease_dispose_list(&dispose);
> >  	error = wait_event_interruptible_timeout(new_fl->c.flc_wait,
> >  						 list_empty(&new_fl->c.flc_blocked_member),
> >  						 break_time);
> > @@ -1643,7 +1652,7 @@ int __break_lease(struct inode *inode, unsigned 
> > int flags)
> >  out:
> >  	spin_unlock(&ctx->flc_lock);
> >  	percpu_up_read(&file_rwsem);
> > -	locks_dispose_list(&dispose);
> > +	lease_dispose_list(&dispose);
> >  free_lock:
> >  	locks_free_lease(new_fl);
> >  	return error;
> > @@ -1726,7 +1735,7 @@ static int __fcntl_getlease(struct file *filp, 
> > unsigned int flavor)
> >  		spin_unlock(&ctx->flc_lock);
> >  		percpu_up_read(&file_rwsem);
> > 
> > -		locks_dispose_list(&dispose);
> > +		lease_dispose_list(&dispose);
> >  	}
> >  	return type;
> >  }
> > @@ -1895,7 +1904,7 @@ generic_add_lease(struct file *filp, int arg, 
> > struct file_lease **flp, void **pr
> >  out:
> >  	spin_unlock(&ctx->flc_lock);
> >  	percpu_up_read(&file_rwsem);
> > -	locks_dispose_list(&dispose);
> > +	lease_dispose_list(&dispose);
> >  	if (is_deleg)
> >  		inode_unlock(inode);
> >  	if (!error && !my_fl)
> > @@ -1931,7 +1940,7 @@ static int generic_delete_lease(struct file 
> > *filp, void *owner)
> >  		error = fl->fl_lmops->lm_change(victim, F_UNLCK, &dispose);
> >  	spin_unlock(&ctx->flc_lock);
> >  	percpu_up_read(&file_rwsem);
> > -	locks_dispose_list(&dispose);
> > +	lease_dispose_list(&dispose);
> >  	return error;
> >  }
> > 
> > @@ -2726,7 +2735,7 @@ locks_remove_lease(struct file *filp, struct 
> > file_lock_context *ctx)
> >  	spin_unlock(&ctx->flc_lock);
> >  	percpu_up_read(&file_rwsem);
> > 
> > -	locks_dispose_list(&dispose);
> > +	lease_dispose_list(&dispose);
> >  }
> > 
> >  /*
> > 
> > -- 
> > 2.52.0

-- 
Jeff Layton <jlayton@...nel.org>