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: <2ce8e183f03855af6c16b2a555473cca3fbbfef6.camel@gmail.com>
Subject: Re: [PATCH V3] scsi: ufs-debugfs: Add error counters
From:   Bean Huo <huobean@gmail.com>
To:     Adrian Hunter <adrian.hunter@intel.com>,
        "Martin K . Petersen" <martin.petersen@oracle.com>,
        "James E . J . Bottomley" <jejb@linux.ibm.com>
Cc:     linux-scsi@vger.kernel.org, linux-kernel@vger.kernel.org,
        Alim Akhtar <alim.akhtar@samsung.com>,
        Avri Altman <avri.altman@wdc.com>,
        Can Guo <cang@codeaurora.org>,
        Stanley Chu <stanley.chu@mediatek.com>
Date:   Fri, 18 Dec 2020 19:57:55 +0100
In-Reply-To: <17957d71-d45b-d5f6-8ef2-453402a23268@intel.com>
References: <20201218122027.27472-1-adrian.hunter@intel.com>
         <de305f4d6034950908e8e889c4af5442431e7d15.camel@gmail.com>
         <17957d71-d45b-d5f6-8ef2-453402a23268@intel.com>
Content-Type: text/plain; charset="UTF-8"
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Precedence: bulk

On Fri, 2020-12-18 at 20:34 +0200, Adrian Hunter wrote:
> It is OK to pass NULL or error codes as parent dentry to
> debugfs_create_...
> functions.
> 
> If ufs_debugfs_root is NULL (which won't happen) then the directory
> will be
> created in debugfs root i.e. /sys/kernel/debug, using the device
> name.
> 
> If ufs_debugfs_root is an error code, then debugfs_create_dir will
> return an
> error code, and following debugfs_create_file() will return an error
> code.
> 
> > 
> > > +    debugfs_create_file("stats", 0400, hba->debugfs_root, hba,
> > > &ufs_debugfs_stats_fops);
> > 
> >        if (!debugfs_create_file("stats", 0400, hba->debugfs_root,
> > hba,
> >                &ufs_debugfs_stats_fops)) {
> >                debugfs_remove(hba->debugfs_root);
> >                return -ENOMEM;
> 
> Being without debugfs files is not a problem, so there is no reason
> to
> return an error.  It is relatively rare in the kernel that code
> checks the
> return value of debugfs_create_file().  We really don't want to fail
> probing
> just because of debugfs.
> 
> However, because debugfs' only real resource is a small amount of
> memory, it
> is extremely unlikely it will fail in that sense.  Although you can
> force it
> to fail by adding the kernel command line parameter debugfs=off

Adrian
Sounds your choice is correct. thanks.

Reviewed-by: Bean Huo <beanhuo@micron.com>