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: <2c51855b-cd74-5701-d9fc-ad75818e7c28@gmail.com>
Date:   Thu, 3 Nov 2022 23:01:07 +0100
From:   Philipp Hortmann <philipp.g.hortmann@...il.com>
To:     Larry Finger <Larry.Finger@...inger.net>,
        gregkh@...uxfoundation.org
Cc:     phil@...lpotter.co.uk, linux-staging@...ts.linux.dev,
        linux-kernel@...r.kernel.org, linux-wireless@...r.kernel.org,
        Randy Dunlap <rdunlap@...radead.org>
Subject: Re: [PATCH] staging: rtl8192e: Fix divide fault when calculating
 beacon age

On 11/3/22 21:05, Larry Finger wrote:
> When the configuration parameter CONFIG_HZ is less that 100, the compiler
> generates an error as follows:
> 
> ../drivers/staging/rtl8192e/rtllib_wx.c: In function 'rtl819x_translate_scan':
> ../drivers/staging/rtl8192e/rtllib_wx.c:220:57: warning: division by zero [-Wdiv-by-zero]
>    220 |                       (jiffies - network->last_scanned) / (HZ / 100));
>        |                                                         ^
> In file included from ../include/linux/skbuff.h:45,
>                   from ../include/linux/if_ether.h:19,
>                   from ../include/linux/etherdevice.h:20,
>                   from ../drivers/staging/rtl8192e/rtllib_wx.c:18:
> ../drivers/staging/rtl8192e/rtllib_wx.c: In function 'rtllib_wx_get_scan':
> ../drivers/staging/rtl8192e/rtllib_wx.c:261:70: warning: division by zero [-Wdiv-by-zero]
>    261 |                                    (jiffies - network->last_scanned) /
>        |
> 
> In fact, is HZ is not a multiple of 100, the calculation will be wrong, but it
> will compile correctly.
> 
> The fix is to get rid of the (HZ / 100) portion. To decrease any round-off
> errors, the compiler is forced to perform the 100 * jiffies-difference before
> dividing by HX. This patch is only compile tested.
> 
> Reported-by: Randy Dunlap <rdunlap@...radead.org>
> Signed-off-by: Larry Finger <Larry.Finger@...inger.net>
> ---
>   drivers/staging/rtl8192e/rtllib_wx.c | 6 +++---
>   1 file changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/staging/rtl8192e/rtllib_wx.c b/drivers/staging/rtl8192e/rtllib_wx.c
> index da2c41c9b92f..7013425102dd 100644
> --- a/drivers/staging/rtl8192e/rtllib_wx.c
> +++ b/drivers/staging/rtl8192e/rtllib_wx.c
> @@ -217,7 +217,7 @@ static inline char *rtl819x_translate_scan(struct rtllib_device *ieee,
>   	p = custom;
>   	p += scnprintf(p, MAX_CUSTOM_LEN - (p - custom),
>   		      " Last beacon: %lums ago",
> -		      (jiffies - network->last_scanned) / (HZ / 100));
> +		      (100 *(jiffies - network->last_scanned)) / HZ);
>   	iwe.u.data.length = p - custom;
>   	if (iwe.u.data.length)
>   		start = iwe_stream_add_point_rsl(info, start, stop,
> @@ -258,8 +258,8 @@ int rtllib_wx_get_scan(struct rtllib_device *ieee,
>   				   escape_essid(network->ssid,
>   						network->ssid_len),
>   				   network->bssid,
> -				   (jiffies - network->last_scanned) /
> -				   (HZ / 100));
> +				   (100 * (jiffies - network->last_scanned)) /
> +				   HZ);
>   	}
>   
>   	spin_unlock_irqrestore(&ieee->lock, flags);

Line length of the description is to long.

Tested-by: Philipp Hortmann <philipp.g.hortmann@...il.com>

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ