Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Wed, 19 Oct 2022 07:28:45 -0400
From:   Brian Foster <bfoster@redhat.com>
To:     Ivan Babrou <ivan@cloudflare.com>
Cc:     linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
        kernel-team@cloudflare.com, Alexey Dobriyan <adobriyan@gmail.com>,
        Al Viro <viro@zeniv.linux.org.uk>,
        Theodore Ts'o <tytso@mit.edu>,
        David Laight <David.Laight@aculab.com>,
        Jonathan Corbet <corbet@lwn.net>,
        Andrew Morton <akpm@linux-foundation.org>,
        David Hildenbrand <david@redhat.com>,
        Johannes Weiner <hannes@cmpxchg.org>,
        Christoph Anton Mitterer <mail@christoph.anton.mitterer.name>,
        Mike Rapoport <rppt@kernel.org>,
        Paul Gortmaker <paul.gortmaker@windriver.com>,
        Kalesh Singh <kaleshsingh@google.com>
Subject: Re: [PATCH v3] proc: report open files as size in stat() for
 /proc/pid/fd
Message-ID: <Y0/fbSL0QDlTU6Yv@bfoster>
References: <20221018045844.37697-1-ivan@cloudflare.com>
 <Y07taqdJ/J3EyJoB@bfoster>
 <CABWYdi37Ts7KDshSvwMf34EKuUrz25duL7W8hOO8t1Xm53t2rA@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <CABWYdi37Ts7KDshSvwMf34EKuUrz25duL7W8hOO8t1Xm53t2rA@mail.gmail.com>
Precedence: bulk

On Tue, Oct 18, 2022 at 11:51:02AM -0700, Ivan Babrou wrote:
> On Tue, Oct 18, 2022 at 11:16 AM Brian Foster <bfoster@redhat.com> wrote:
> > > +static int proc_readfd_count(struct inode *inode)
> > > +{
> > > +     struct task_struct *p = get_proc_task(inode);
> > > +     struct fdtable *fdt;
> > > +     unsigned int open_fds = 0;
> > > +
> > > +     if (!p)
> > > +             return -ENOENT;
> >
> > Maybe this shouldn't happen, but do you mean to assign the error code to
> > stat->size in the caller? Otherwise this seems reasonable to me.
> 
> You are right. As unlikely as it is to happen, we shouldn't return
> negative size.
> 
> What's the idiomatic way to make this work? My two options are:
> 
> 1. Pass &stat->size into proc_readfd_count:
> 
>   if (S_ISDIR(inode->i_mode)) {
>     rv = proc_readfd_count(inode, &stat->size);
>     if (rv < 0)
>       goto out;
>   }
> 
> out:
>   return rv;
> 
> OR without a goto:
> 
>   if (S_ISDIR(inode->i_mode)) {
>     rv = proc_readfd_count(inode, &stat->size));
>     if (rv < 0)
>       return rv;
>   }
> 
>   return rv;
> 
> 2. Return negative count as error (as we don't expect negative amount
> of files open):
> 
>   if (S_ISDIR(inode->i_mode)) {
>     size = proc_readfd_count(inode);
>     if (size < 0)
>       return size;
>     stat->size = size;
>   }
> 

I suppose the latter is less of a change to the original patch..? Either
way seems reasonable to me. I have no strong preference FWIW.

Brian