Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Tue, 19 Jun 2018 10:42:48 +0300
From:   Dan Carpenter <dan.carpenter@oracle.com>
To:     Anton Vasilyev <vasilyev@ispras.ru>
Cc:     Andy Shevchenko <andy.shevchenko@gmail.com>,
        devel@driverdev.osuosl.org, ldv-project@linuxtesting.org,
        Johannes Thumshirn <jthumshirn@suse.de>,
        linux-kernel@vger.kernel.org, Sinan Kaya <okaya@codeaurora.org>,
        Hannes Reinecke <hare@suse.de>,
        Gaurav Pathak <gauravpathak129@gmail.com>
Subject: Re: your mail
Message-ID: <20180619074248.ssu5kza6wdwhzdqi@mwanda>
References: <CAHp75VetOuS+H4GU9UW-Ekczs22-Ju++jXWdvkMJGktbumDzwQ@mail.gmail.com>
 <20180613173128.32384-1-vasilyev@ispras.ru>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20180613173128.32384-1-vasilyev@ispras.ru>
User-Agent: NeoMutt/20170609 (1.8.3)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Thanks for this.  This is a lot of work.

On Wed, Jun 13, 2018 at 08:31:28PM +0300, Anton Vasilyev wrote:
> diff --git a/drivers/staging/rts5208/rtsx.c b/drivers/staging/rts5208/rtsx.c
> index 70e0b8623110..69e6abe14abf 100644
> --- a/drivers/staging/rts5208/rtsx.c
> +++ b/drivers/staging/rts5208/rtsx.c
> @@ -857,7 +857,7 @@ static int rtsx_probe(struct pci_dev *pci,
>  	dev->chip = kzalloc(sizeof(*dev->chip), GFP_KERNEL);
>  	if (!dev->chip) {
>  		err = -ENOMEM;
> -		goto errout;
> +		goto chip_alloc_fail;

The most recent successful allocation is scsi_host_alloc().  I was
really hoping this would say something like "goto err_free_host;" or
something.  The naming style here is a "come from" label which doesn't
say if it's going to free the scsi host or not...  It turns out we don't
free the the host, but we should:

err_put_host:
	scsi_host_put(host);

The kzalloc() has it's own error message built in, and all the other
error paths as well so the dev_err() is not super important to me...

Killing the threads seems actually really complicated so maybe we should
just have a separate error paths for that.  I'm not sure...

regards,
dan carpenter