public inbox for linux-nfs@vger.kernel.org
 help / color / mirror / Atom feed
From: Jeff Layton <jeff.layton@primarydata.com>
To: "J. Bruce Fields" <bfields@fieldses.org>
Cc: linux-fsdevel@vger.kernel.org, linux-nfs@vger.kernel.org,
	Christoph Hellwig <hch@infradead.org>,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2 07/17] nfsd: don't keep a pointer to the lease in nfs4_file
Date: Sat, 6 Sep 2014 08:33:03 -0400	[thread overview]
Message-ID: <20140906083303.6b79a775@tlielax.poochiereds.net> (raw)
In-Reply-To: <20140905214058.GA5443@fieldses.org>

On Fri, 5 Sep 2014 17:40:58 -0400
"J. Bruce Fields" <bfields@fieldses.org> wrote:

> On Thu, Sep 04, 2014 at 08:38:33AM -0400, Jeff Layton wrote:
> > Now that we don't need to pass in an actual lease pointer to
> > vfs_setlease on unlock, we can stop tracking a pointer to the lease in
> > the nfs4_file.
> > 
> > Switch all of the places that check the fi_lease to check fi_deleg_file
> > instead. We always set that at the same time so it will have the same
> > semantics.
> > 
> > Cc: J. Bruce Fields <bfields@fieldses.org>
> > Signed-off-by: Jeff Layton <jlayton@primarydata.com>
> > Reviewed-by: Christoph Hellwig <hch@lst.de>
> > ---
> >  fs/nfsd/nfs4state.c | 10 ++++------
> >  fs/nfsd/state.h     |  1 -
> >  2 files changed, 4 insertions(+), 7 deletions(-)
> > 
> > diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
> > index 0cd252916e1a..d0a6e8e022a2 100644
> > --- a/fs/nfsd/nfs4state.c
> > +++ b/fs/nfsd/nfs4state.c
> > @@ -680,11 +680,10 @@ static void nfs4_put_deleg_lease(struct nfs4_file *fp)
> >  {
> >  	lockdep_assert_held(&state_lock);
> >  
> > -	if (!fp->fi_lease)
> > +	if (!fp->fi_deleg_file)
> >  		return;
> >  	if (atomic_dec_and_test(&fp->fi_delegees)) {
> >  		vfs_setlease(fp->fi_deleg_file, F_UNLCK, NULL);
> > -		fp->fi_lease = NULL;
> >  		fput(fp->fi_deleg_file);
> >  		fp->fi_deleg_file = NULL;
> >  	}
> 
> Note the code in my for-3.18 branch is different here due to a couple
> patches I've already taken from you.
> 
> --b.
> 

Doh! Yes, I'll make sure to pull those into the topic branch where I'm
working on these and rebase on top of them. Thanks!

> > @@ -3061,8 +3060,8 @@ static void nfsd4_init_file(struct nfs4_file *fp, struct knfsd_fh *fh)
> >  	INIT_LIST_HEAD(&fp->fi_stateids);
> >  	INIT_LIST_HEAD(&fp->fi_delegations);
> >  	fh_copy_shallow(&fp->fi_fhandle, fh);
> > +	fp->fi_deleg_file = NULL;
> >  	fp->fi_had_conflict = false;
> > -	fp->fi_lease = NULL;
> >  	fp->fi_share_deny = 0;
> >  	memset(fp->fi_fds, 0, sizeof(fp->fi_fds));
> >  	memset(fp->fi_access, 0, sizeof(fp->fi_access));
> > @@ -3803,13 +3802,12 @@ static int nfs4_setlease(struct nfs4_delegation *dp)
> >  	if (fp->fi_had_conflict)
> >  		goto out_unlock;
> >  	/* Race breaker */
> > -	if (fp->fi_lease) {
> > +	if (fp->fi_deleg_file) {
> >  		status = 0;
> >  		atomic_inc(&fp->fi_delegees);
> >  		hash_delegation_locked(dp, fp);
> >  		goto out_unlock;
> >  	}
> > -	fp->fi_lease = fl;
> >  	fp->fi_deleg_file = filp;
> >  	atomic_set(&fp->fi_delegees, 1);
> >  	hash_delegation_locked(dp, fp);
> > @@ -3842,7 +3840,7 @@ nfs4_set_delegation(struct nfs4_client *clp, struct svc_fh *fh,
> >  	spin_lock(&state_lock);
> >  	spin_lock(&fp->fi_lock);
> >  	dp->dl_stid.sc_file = fp;
> > -	if (!fp->fi_lease) {
> > +	if (!fp->fi_deleg_file) {
> >  		spin_unlock(&fp->fi_lock);
> >  		spin_unlock(&state_lock);
> >  		status = nfs4_setlease(dp);
> > diff --git a/fs/nfsd/state.h b/fs/nfsd/state.h
> > index 4a89e00d7461..64f291a25a8c 100644
> > --- a/fs/nfsd/state.h
> > +++ b/fs/nfsd/state.h
> > @@ -477,7 +477,6 @@ struct nfs4_file {
> >  	atomic_t		fi_access[2];
> >  	u32			fi_share_deny;
> >  	struct file		*fi_deleg_file;
> > -	struct file_lock	*fi_lease;
> >  	atomic_t		fi_delegees;
> >  	struct knfsd_fh		fi_fhandle;
> >  	bool			fi_had_conflict;
> > -- 
> > 1.9.3
> > 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html


-- 
Jeff Layton <jlayton@primarydata.com>

  reply	other threads:[~2014-09-06 12:33 UTC|newest]

Thread overview: 49+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-09-04 12:38 [PATCH v2 00/17] locks: internal lease API overhaul Jeff Layton
2014-09-04 12:38 ` [PATCH v2 01/17] locks: consolidate "nolease" routines Jeff Layton
2014-09-04 12:41   ` Trond Myklebust
2014-09-04 12:49     ` Jeff Layton
2014-09-04 18:25       ` Trond Myklebust
2014-09-04 20:12         ` Christoph Hellwig
2014-09-05 11:48           ` Jeff Layton
2014-09-04 17:46   ` Christoph Hellwig
2014-09-04 12:38 ` [PATCH v2 02/17] security: make security_file_set_fowner, f_setown and __f_setown void return Jeff Layton
2014-09-04 17:47   ` Christoph Hellwig
2014-10-07 17:11   ` Dmitry Kasatkin
2014-10-07 17:17     ` Christoph Hellwig
2014-10-07 17:34       ` Dmitry Kasatkin
2014-10-07 18:02         ` Jeff Layton
2014-09-04 12:38 ` [PATCH v2 03/17] locks: close potential race in lease_get_mtime Jeff Layton
2014-09-04 12:38 ` [PATCH v2 04/17] nfsd: fix potential lease memory leak in nfs4_setlease Jeff Layton
2014-09-04 12:38 ` [PATCH v2 05/17] locks: generic_delete_lease doesn't need a file_lock at all Jeff Layton
2014-09-04 20:14   ` Christoph Hellwig
2014-09-05  0:29     ` Jeff Layton
2015-01-12 23:03   ` NeilBrown
2015-01-12 23:25     ` Jeff Layton
2015-01-13  2:14       ` NeilBrown
2014-09-04 12:38 ` [PATCH v2 06/17] locks: clean up vfs_setlease kerneldoc comments Jeff Layton
2014-09-04 12:38 ` [PATCH v2 07/17] nfsd: don't keep a pointer to the lease in nfs4_file Jeff Layton
2014-09-05 21:40   ` J. Bruce Fields
2014-09-06 12:33     ` Jeff Layton [this message]
2014-09-04 12:38 ` [PATCH v2 08/17] locks: plumb a "priv" pointer into the setlease routines Jeff Layton
2014-09-04 17:48   ` Christoph Hellwig
2014-09-04 12:38 ` [PATCH v2 09/17] locks: define a lm_setup handler for leases Jeff Layton
2014-09-04 17:49   ` Christoph Hellwig
2014-09-04 12:38 ` [PATCH v2 10/17] locks: move i_lock acquisition into generic_*_lease handlers Jeff Layton
2014-09-04 12:38 ` [PATCH v2 11/17] locks: move freeing of leases outside of i_lock Jeff Layton
2014-09-04 17:50   ` Christoph Hellwig
2014-09-05 14:03     ` Jeff Layton
2014-09-04 12:38 ` [PATCH v2 12/17] locks: update Documentation/filesystems with new setlease semantics Jeff Layton
2014-09-04 17:50   ` Christoph Hellwig
2014-09-05 14:02     ` Jeff Layton
2014-09-04 12:38 ` [PATCH v2 13/17] locks: remove i_have_this_lease check from __break_lease Jeff Layton
2014-09-04 17:51   ` Christoph Hellwig
2014-09-04 18:03     ` Jeff Layton
2014-09-04 12:38 ` [PATCH v2 14/17] locks: __break_lease cleanup in preparation of allowing direct removal of leases Jeff Layton
2014-09-04 18:07   ` Christoph Hellwig
2014-09-05 13:35     ` Jeff Layton
2014-09-04 12:38 ` [PATCH v2 15/17] locks: give lm_break a return value Jeff Layton
2014-09-04 18:08   ` Christoph Hellwig
2014-09-04 12:38 ` [PATCH v2 16/17] locks: set fl_owner for leases to filp instead of current->files Jeff Layton
2014-09-04 12:38 ` [PATCH v2 17/17] locks: clean up comments over fl_owner_t definition Jeff Layton
2014-09-04 17:53   ` Christoph Hellwig
2014-09-05 13:36     ` Jeff Layton

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20140906083303.6b79a775@tlielax.poochiereds.net \
    --to=jeff.layton@primarydata.com \
    --cc=bfields@fieldses.org \
    --cc=hch@infradead.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-nfs@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox