netdev - Re: [PATCH net-next 3/4] net: hns3: fix strncpy() not using dest-buf length as length issue

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]

Message-ID: <ZGKOdijGtX03qV2p@corigine.com>
Date: Mon, 15 May 2023 21:57:41 +0200
From: Simon Horman <simon.horman@...igine.com>
To: Hao Lan <lanhao@...wei.com>
Cc: netdev@...r.kernel.org, yisen.zhuang@...wei.com, salil.mehta@...wei.com,
	davem@...emloft.net, edumazet@...gle.com, kuba@...nel.org,
	pabeni@...hat.com, richardcochran@...il.com,
	wangpeiyang1@...wei.com, shenjian15@...wei.com,
	chenhao418@...wei.com, wangjie125@...wei.com, yuanjilin@...rlc.com,
	cai.huoqing@...ux.dev, xiujianfeng@...wei.com
Subject: Re: [PATCH net-next 3/4] net: hns3: fix strncpy() not using dest-buf
 length as length issue

On Mon, May 15, 2023 at 09:46:42PM +0800, Hao Lan wrote:
> From: Hao Chen <chenhao418@...wei.com>
> 
> Now, strncpy() in hns3_dbg_fill_content() use src-length as copy-length,
> it may result in dest-buf overflow.
> 
> This patch is to fix intel compile warning for csky-linux-gcc (GCC) 12.1.0
> compiler.
> 
> The warning reports as below:
> 
> hclge_debugfs.c:92:25: warning: 'strncpy' specified bound depends on
> the length of the source argument [-Wstringop-truncation]
> 
> strncpy(pos, items[i].name, strlen(items[i].name));
> 
> hclge_debugfs.c:90:25: warning: 'strncpy' output truncated before
> terminating nul copying as many bytes from a string as its length
> [-Wstringop-truncation]
> 
> strncpy(pos, result[i], strlen(result[i]));
> 
> strncpy() use src-length as copy-length, it may result in
> dest-buf overflow.
> 
> So,this patch add some values check to avoid this issue.
> 
> Signed-off-by: Hao Chen <chenhao418@...wei.com>
> Reported-by: kernel test robot <lkp@...el.com>
> Closes: https://lore.kernel.org/lkml/202207170606.7WtHs9yS-lkp@intel.com/T/
> Signed-off-by: Hao Lan <lanhao@...wei.com>
> ---
>  .../ethernet/hisilicon/hns3/hns3_debugfs.c    | 31 ++++++++++++++-----
>  .../hisilicon/hns3/hns3pf/hclge_debugfs.c     | 29 ++++++++++++++---
>  2 files changed, 48 insertions(+), 12 deletions(-)
> 
> diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c b/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c
> index 4c3e90a1c4d0..cf415cb37685 100644
> --- a/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c
> +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c
> @@ -438,19 +438,36 @@ static void hns3_dbg_fill_content(char *content, u16 len,
>  				  const struct hns3_dbg_item *items,
>  				  const char **result, u16 size)
>  {
> +#define HNS3_DBG_LINE_END_LEN	2
>  	char *pos = content;
> +	u16 item_len;
>  	u16 i;
>  
> +	if (!len) {
> +		return;
> +	} else if (len <= HNS3_DBG_LINE_END_LEN) {
> +		*pos++ = '\0';
> +		return;
> +	}
> +
>  	memset(content, ' ', len);
> -	for (i = 0; i < size; i++) {
> -		if (result)
> -			strncpy(pos, result[i], strlen(result[i]));
> -		else
> -			strncpy(pos, items[i].name, strlen(items[i].name));
> +	len -= HNS3_DBG_LINE_END_LEN;
>  
> -		pos += strlen(items[i].name) + items[i].interval;
> +	for (i = 0; i < size; i++) {
> +		item_len = strlen(items[i].name) + items[i].interval;
> +		if (len < item_len)
> +			break;
> +
> +		if (result) {
> +			if (item_len < strlen(result[i]))
> +				break;
> +			memcpy(pos, result[i], strlen(result[i]));
> +		} else {
> +			memcpy(pos, items[i].name, strlen(items[i].name));

Hi,

The above memcpy() calls share the same property as the warning that
is being addressed: the length copied depends on the source not the
destination.

With the reworked code this seems safe. Which is good. But I wonder if,
given all the checking done, it makes sense to simply call strcpy() here.
Using strlen() as a length argument seems odd to me.

> +		}
> +		pos += item_len;
> +		len -= item_len;
>  	}
> -
>  	*pos++ = '\n';
>  	*pos++ = '\0';
>  }
> diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c
> index a0b46e7d863e..1354fd0461f7 100644
> --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c
> +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c
> @@ -88,16 +88,35 @@ static void hclge_dbg_fill_content(char *content, u16 len,
>  				   const struct hclge_dbg_item *items,
>  				   const char **result, u16 size)
>  {
> +#define HCLGE_DBG_LINE_END_LEN	2
>  	char *pos = content;
> +	u16 item_len;
>  	u16 i;
>  
> +	if (!len) {
> +		return;
> +	} else if (len <= HCLGE_DBG_LINE_END_LEN) {
> +		*pos++ = '\0';
> +		return;
> +	}
> +
>  	memset(content, ' ', len);
> +	len -= HCLGE_DBG_LINE_END_LEN;
> +
>  	for (i = 0; i < size; i++) {
> -		if (result)
> -			strncpy(pos, result[i], strlen(result[i]));
> -		else
> -			strncpy(pos, items[i].name, strlen(items[i].name));
> -		pos += strlen(items[i].name) + items[i].interval;
> +		item_len = strlen(items[i].name) + items[i].interval;
> +		if (len < item_len)
> +			break;
> +
> +		if (result) {
> +			if (item_len < strlen(result[i]))
> +				break;
> +			memcpy(pos, result[i], strlen(result[i]));
> +		} else {
> +			memcpy(pos, items[i].name, strlen(items[i].name));
> +		}
> +		pos += item_len;
> +		len -= item_len;
>  	}
>  	*pos++ = '\n';
>  	*pos++ = '\0';
> -- 
> 2.30.0
> 
>