linux-kernel - Re: [PATCH 2/7] watchdog: jz4740: Use devm

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:   Thu, 28 Dec 2017 21:22:18 +0100
From:   Paul Cercueil <paul@...pouillou.net>
To:     Guenter Roeck <linux@...ck-us.net>
Cc:     Ralf Baechle <ralf@...ux-mips.org>,
        Rob Herring <robh+dt@...nel.org>,
        Mark Rutland <mark.rutland@....com>,
        Wim Van Sebroeck <wim@...ana.be>, devicetree@...r.kernel.org,
        linux-mips@...ux-mips.org, linux-kernel@...r.kernel.org,
        linux-watchdog@...r.kernel.org
Subject: Re: [PATCH 2/7] watchdog: jz4740: Use devm_* functions



Le jeu. 28 déc. 2017 à 21:19, Guenter Roeck <linux@...ck-us.net> a 
écrit :
> On 12/28/2017 11:59 AM, Paul Cercueil wrote:
>> Hi Guenter,
>> 
>> Le jeu. 28 déc. 2017 à 18:48, Guenter Roeck <linux@...ck-us.net> a 
>> écrit :
>>> On 12/28/2017 08:29 AM, Paul Cercueil wrote:
>>>> - Use devm_clk_get instead of clk_get
>>>> - Use devm_watchdog_register_device instead of 
>>>> watchdog_register_device
>>>> 
>>>> Signed-off-by: Paul Cercueil <paul@...pouillou.net>
>>>> ---
>>>>   drivers/watchdog/jz4740_wdt.c | 27 ++++++++-------------------
>>>>   1 file changed, 8 insertions(+), 19 deletions(-)
>>>> 
>>>> diff --git a/drivers/watchdog/jz4740_wdt.c 
>>>> b/drivers/watchdog/jz4740_wdt.c
>>>> index 6955deb100ef..92d6ca8ceb49 100644
>>>> --- a/drivers/watchdog/jz4740_wdt.c
>>>> +++ b/drivers/watchdog/jz4740_wdt.c
>>>> @@ -178,40 +178,29 @@ static int jz4740_wdt_probe(struct 
>>>> platform_device *pdev)
>>>>         res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
>>>>       drvdata->base = devm_ioremap_resource(&pdev->dev, res);
>>>> -    if (IS_ERR(drvdata->base)) {
>>>> -        ret = PTR_ERR(drvdata->base);
>>>> -        goto err_out;
>>>> -    }
>>>> +    if (IS_ERR(drvdata->base))
>>>> +        return PTR_ERR(drvdata->base);
>>>>   -    drvdata->rtc_clk = clk_get(&pdev->dev, "rtc");
>>>> +    drvdata->rtc_clk = devm_clk_get(&pdev->dev, "rtc");
>>>>       if (IS_ERR(drvdata->rtc_clk)) {
>>>>           dev_err(&pdev->dev, "cannot find RTC clock\n");
>>>> -        ret = PTR_ERR(drvdata->rtc_clk);
>>>> -        goto err_out;
>>>> +        return PTR_ERR(drvdata->rtc_clk);
>>>>       }
>>>>   -    ret = watchdog_register_device(&drvdata->wdt);
>>>> +    ret = devm_watchdog_register_device(&pdev->dev, 
>>>> &drvdata->wdt);
>>>>       if (ret < 0)
>>>> -        goto err_disable_clk;
>>>> +        return ret;
>>>>         platform_set_drvdata(pdev, drvdata);
>>>> -    return 0;
>>>>   -err_disable_clk:
>>>> -    clk_put(drvdata->rtc_clk);
>>>> -err_out:
>>>> -    return ret;
>>>> +    return 0;
>>>>   }
>>>>     static int jz4740_wdt_remove(struct platform_device *pdev)
>>>>   {
>>>>       struct jz4740_wdt_drvdata *drvdata = 
>>>> platform_get_drvdata(pdev);
>>>>   -    jz4740_wdt_stop(&drvdata->wdt);
>>>> -    watchdog_unregister_device(&drvdata->wdt);
>>>> -    clk_put(drvdata->rtc_clk);
>>>> -
>>>> -    return 0;
>>>> +    return jz4740_wdt_stop(&drvdata->wdt);
>>> 
>>> If the watchdog is running, the module can not be unloaded. Even if 
>>> that wasn't
>>> the case, this defeats both WDIOF_MAGICCLOSE and 
>>> watchdog_set_nowayout().
>>> Are you sure this is what you want ? If so, please call
>>> watchdog_stop_on_unregister() before registration; this clarifies 
>>> that this
>>> is what you want, and you can drop the remove function.
>>> 
>>> Thanks,
>>> Guenter
>> 
>> This patch does not change the behaviour. We always used that driver 
>> built-in; what
>> should the behaviour be when unloading the module? Keep the watchdog 
>> hardware running
>> if configured for 'nowayout'?
>> 
> 
> One can still unload the driver. If you are fine with 
> bypassing/dfeating nowayout
> and WDIOF_MAGICCLOSE, may I ask why those are enabled in the first 
> place ?
> 

Who knows? That code is very old :)
I'm fine with removing the remove() function completely, if you think 
it makes more sense.

Thanks,
-Paul