linux-kernel - Re: [PATCH v2] watchdog: dw_wdt: fix overflow issue in dw_wdt_top_in

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 for Android: free password hash cracker in your pocket

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

Message-ID: <667541ac-48ab-f6ab-6b86-44aa96fe5bc7@roeck-us.net>
Date:   Mon, 14 Aug 2017 21:49:16 -0700
From:   Guenter Roeck <linux@...ck-us.net>
To:     Huibin Hong <huibin.hong@...k-chips.com>
Cc:     huangtao@...k-chips.com, Wim Van Sebroeck <wim@...ana.be>,
        linux-watchdog@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] watchdog: dw_wdt: fix overflow issue in
 dw_wdt_top_in_seconds

On 08/14/2017 07:45 PM, Huibin Hong wrote:
> If top is 15, (1 << (16 + top)) may be negative.
> 
> Signed-off-by: Huibin Hong <huibin.hong@...k-chips.com>
> ---
> 
> Changes in v2:
>      - Rebase mainline Linux 4.13-rc5
> 
>   drivers/watchdog/dw_wdt.c | 4 +++-
>   1 file changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/watchdog/dw_wdt.c b/drivers/watchdog/dw_wdt.c
> index 36be987..59ed958 100644
> --- a/drivers/watchdog/dw_wdt.c
> +++ b/drivers/watchdog/dw_wdt.c
> @@ -72,7 +72,9 @@ static inline int dw_wdt_top_in_seconds(struct dw_wdt *dw_wdt, unsigned top)
>   	 * There are 16 possible timeout values in 0..15 where the number of
>   	 * cycles is 2 ^ (16 + i) and the watchdog counts down.
>   	 */
> -	return (1U << (16 + top)) / dw_wdt->rate;
> +	unsigned int cycles = 1 << (16 + top);
> +
> +	return (cycles / clk_get_rate(dw_wdt.clk));

Sorry I didn't catch this earlier. Why not use dw_wdt->rate ?
It has been checked against 0, and presumably the rate does not change
on its own. If it can change, you'll have to submit a separate patch
dropping the use of dw_wdt->rate entirely and explaining the reason.
You'll also have to make sure that the driver somehow catches the
rate change and adjusts its settings automatically; otherwise, timeouts
will be off target, and the system can reset unexpectedly.

Thanks,
Guenter

>   }
>   
>   static int dw_wdt_get_top(struct dw_wdt *dw_wdt)
>