linux-kernel - Re: [PATCH v3 iwl-next] i40e: Use correct buffer size in i40e_dbg_command

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

Date:   Mon, 4 Dec 2023 13:51:11 +0100
From:   Alexander Lobakin <aleksander.lobakin@...el.com>
To:     Kunwu Chan <chentao@...inos.cn>
CC:     <jesse.brandeburg@...el.com>, <anthony.l.nguyen@...el.com>,
        <davem@...emloft.net>, <edumazet@...gle.com>, <kuba@...nel.org>,
        <pabeni@...hat.com>, <jeffrey.t.kirsher@...el.com>,
        <shannon.nelson@....com>, <kunwu.chan@...mail.com>,
        <intel-wired-lan@...ts.osuosl.org>, <netdev@...r.kernel.org>,
        <linux-kernel@...r.kernel.org>, Simon Horman <horms@...nel.org>
Subject: Re: [PATCH v3 iwl-next] i40e: Use correct buffer size in
 i40e_dbg_command_read

From: Kunwu Chan <chentao@...inos.cn>
Date: Mon, 4 Dec 2023 09:44:55 +0800

> The size of "i40e_dbg_command_buf" is 256, the size of "name"
> depends on "IFNAMSIZ", plus a null character and format size,
> the total size is more than 256.
> 
> Improve readability and maintainability by replacing a hardcoded string
> allocation and formatting by the use of the kasprintf() helper.
> 
> Fixes: 02e9c290814c ("i40e: debugfs interface")
> Signed-off-by: Kunwu Chan <chentao@...inos.cn>
> Suggested-by: Simon Horman <horms@...nel.org>
> Suggested-by: Alexander Lobakin <aleksander.lobakin@...el.com>
> ---
> v2
>    - Update the size calculation with IFNAMSIZ and sizeof(i40e_dbg_command_buf)
> v3
>    - Use kasprintf to improve readability and maintainability
> ---
>  drivers/net/ethernet/intel/i40e/i40e_debugfs.c | 15 +++++++--------
>  1 file changed, 7 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/net/ethernet/intel/i40e/i40e_debugfs.c b/drivers/net/ethernet/intel/i40e/i40e_debugfs.c
> index 88240571721a..a176de89de9c 100644
> --- a/drivers/net/ethernet/intel/i40e/i40e_debugfs.c
> +++ b/drivers/net/ethernet/intel/i40e/i40e_debugfs.c
> @@ -72,23 +72,22 @@ static ssize_t i40e_dbg_command_read(struct file *filp, char __user *buffer,
>  {
>  	struct i40e_pf *pf = filp->private_data;
>  	int bytes_not_copied;
> -	int buf_size = 256;
> -	char *buf;
> +	char *buf = NULL;

This is unneeded.

>  	int len;
>  
>  	/* don't allow partial reads */
>  	if (*ppos != 0)
>  		return 0;
> -	if (count < buf_size)
> -		return -ENOSPC;
>  
> -	buf = kzalloc(buf_size, GFP_KERNEL);
> +	buf = kasprintf(GFP_KERNEL, "%s: %s\n",
> +			pf->vsi[pf->lan_vsi]->netdev->name,
> +			i40e_dbg_command_buf);
>  	if (!buf)
>  		return -ENOSPC;
>  
> -	len = snprintf(buf, buf_size, "%s: %s\n",
> -		       pf->vsi[pf->lan_vsi]->netdev->name,
> -		       i40e_dbg_command_buf);
> +	len = strlen(buf);

strlen() doesn't include the terminating '\0', but you need to copy it
as well. Hence `strlen(buf) + 1`.

> +	if (count < len)
> +		return -ENOSPC;

Here you have memory leak in case the condition is true.

>  
>  	bytes_not_copied = copy_to_user(buffer, buf, len);
>  	kfree(buf);

Taking the above two into account, I'd suggest doing something like:

	if (!buf)
		return -ENOSPC;

	len = strlen(buf) + 1;

	if (count < len)
		bytes_not_copied = -ENOSPC;
	elseif (copy_to_user(buffer, buf, len))
		bytes_not_copied = -EFAULT;
	else
		bytes_not_copied = 0;

	kfree(buf);

	if (bytes_not_copied)
		return bytes_not_copied;

	*ppos = len;

	return len;
}

Thanks,
Olek