Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Message-ID: <ab91ce6f0c1b2e9549fc6e966db7514a988d0bf1.camel@themaw.net>
Subject: Re: [PATCH v6 3/7] kernfs: use VFS negative dentry caching
From:   Ian Kent <raven@themaw.net>
To:     Al Viro <viro@zeniv.linux.org.uk>
Cc:     Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        Tejun Heo <tj@kernel.org>, Eric Sandeen <sandeen@sandeen.net>,
        Fox Chen <foxhlchen@gmail.com>,
        Brice Goglin <brice.goglin@gmail.com>,
        Rick Lindsley <ricklind@linux.vnet.ibm.com>,
        David Howells <dhowells@redhat.com>,
        Miklos Szeredi <miklos@szeredi.hu>,
        Marcelo Tosatti <mtosatti@redhat.com>,
        "Eric W. Biederman" <ebiederm@xmission.com>,
        Carlos Maiolino <cmaiolino@redhat.com>,
        linux-fsdevel <linux-fsdevel@vger.kernel.org>,
        Kernel Mailing List <linux-kernel@vger.kernel.org>
Date:   Sat, 12 Jun 2021 09:08:05 +0800
In-Reply-To: <2ee74cbed729d66a38a5c7de9c4608d02fb89f26.camel@themaw.net>
References: <162322846765.361452.17051755721944717990.stgit@web.messagingengine.com>
         <162322862726.361452.10114120072438540655.stgit@web.messagingengine.com>
         <YMP6topegaTXGNgC@zeniv-ca.linux.org.uk>
         <2ee74cbed729d66a38a5c7de9c4608d02fb89f26.camel@themaw.net>
Content-Type: text/plain; charset="UTF-8"
User-Agent: Evolution 3.38.4 (3.38.4-1.fc33) 
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

On Sat, 2021-06-12 at 08:43 +0800, Ian Kent wrote:
> On Sat, 2021-06-12 at 00:07 +0000, Al Viro wrote:
> > On Wed, Jun 09, 2021 at 04:50:27PM +0800, Ian Kent wrote:
> > 
> > > +       if (d_really_is_negative(dentry)) {
> > > +               struct dentry *d_parent = dget_parent(dentry);
> > > +               struct kernfs_node *parent;
> > 
> > What the hell is dget_parent() for?  You don't do anything blocking
> > here, so why not simply grab dentry->d_lock - that'll stabilize
> > the value of ->d_parent just fine.  Just don't forget to drop the
> > lock before returning and that's it...
> 
> Thanks Al, I'll change it.

But if I change to take the read lock to ensure there's no operation
in progress for the revision check I would need the dget_parent(), yes?

> 
> > 
> > > +               /* If the kernfs parent node has changed discard
> > > and
> > > +                * proceed to ->lookup.
> > > +                */
> > > +               parent = kernfs_dentry_node(d_parent);
> > > +               if (parent) {
> > > +                       if (kernfs_dir_changed(parent, dentry)) {
> > > +                               dput(d_parent);
> > > +                               return 0;
> > > +                       }
> > > +               }
> > > +               dput(d_parent);
> > > +
> > > +               /* The kernfs node doesn't exist, leave the
> > > dentry
> > > +                * negative and return success.
> > > +                */
> > > +               return 1;
> > > +       }
>