linux-kernel - Re: [PATCH] rapidio: fix get device imbalance on error

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:   Fri, 21 Aug 2020 11:14:49 +0300
From:   Dan Carpenter <dan.carpenter@...cle.com>
To:     George Acosta <acostag.ubuntu@...il.com>
Cc:     Matt Porter <mporter@...nel.crashing.org>,
        Alexandre Bounine <alex.bou9@...il.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        John Hubbard <jhubbard@...dia.com>,
        "Gustavo A. R. Silva" <gustavoars@...nel.org>,
        Madhuparna Bhowmik <madhuparnabhowmik10@...il.com>,
        Kees Cook <keescook@...omium.org>, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] rapidio: fix get device imbalance on error

On Thu, Aug 20, 2020 at 10:44:57PM -0500, George Acosta wrote:
> Fix the imbalance in mport_cdev_open.
> Call put_device in error path to balance the
> refcount that increased by the get_device.
> 
> Fixes: e8de370188d0 ("rapidio: add mport char device driver")
> Signed-off-by: George Acosta <acostag.ubuntu@...il.com>
> ---
>  drivers/rapidio/devices/rio_mport_cdev.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/drivers/rapidio/devices/rio_mport_cdev.c b/drivers/rapidio/devices/rio_mport_cdev.c
> index c07ceec3c6d4..3b68e00eb40f 100644
> --- a/drivers/rapidio/devices/rio_mport_cdev.c
> +++ b/drivers/rapidio/devices/rio_mport_cdev.c
> @@ -1908,6 +1908,7 @@ static int mport_cdev_open(struct inode *inode, struct file *filp)
>  			  sizeof(struct rio_event) * MPORT_EVENT_DEPTH,
>  			  GFP_KERNEL);
>  	if (ret < 0) {
> +		put_device(&chdev->dev);
>  		dev_err(&chdev->dev, DRV_NAME ": kfifo_alloc failed\n");
>  		ret = -ENOMEM;
>  		goto err_fifo;
> -- 

If we ever hit this error path then the:

	list_add_tail(&priv->list, &chdev->file_list);

needs to be removed from the list as well or it lead to a use after
free.  Probably just move the list_add_tail() after the kfifo_alloc()
has succeeded. We need to remove it from the list in mport_cdev_release()
as well...

The error handling in this function is kind of rubbish.
1) Get rid of the out label and return directly (more readable).
1b) Use "return 0;" instead of "retur ret;" for the success path.
2) Name the labels after what they do not where they "come from".  In
other words do.

err_priv:
	kfree(priv);

3) Create a label to call put_device:

err_device:
	put_device(&chdev->dev);

Change all the error paths to use the goto instead of calling
put_device() before the goto.

regards,
dan carpenter