linux-kernel - Re: [PATCH v2 3/3] watchdog: sama5d4: Implement resume hook

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: <4016d976-cb6f-c303-eebe-c0bede771725@roeck-us.net>
Date:   Sun, 19 Feb 2017 09:05:15 -0800
From:   Guenter Roeck <linux@...ck-us.net>
To:     Alexandre Belloni <alexandre.belloni@...e-electrons.com>
Cc:     Wim Van Sebroeck <wim@...ana.be>,
        Nicolas Ferre <nicolas.ferre@...el.com>,
        linux-watchdog@...r.kernel.org,
        linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 3/3] watchdog: sama5d4: Implement resume hook

On 02/17/2017 07:22 AM, Alexandre Belloni wrote:
> On 17/02/2017 at 06:47:03 -0800, Guenter Roeck wrote:
>> On 02/16/2017 11:30 AM, Alexandre Belloni wrote:
>>> When resuming for the deepest state on sama5d2, it is necessary to restore
>>> MR as the registers are lost.
>>>
>>> Signed-off-by: Alexandre Belloni <alexandre.belloni@...e-electrons.com>
>>> ---
>>> Changes in v2:
>>>  - cache mr beofre suspending
>>>
>>>  drivers/watchdog/sama5d4_wdt.c | 32 ++++++++++++++++++++++++++++++++
>>>  1 file changed, 32 insertions(+)
>>>
>>> diff --git a/drivers/watchdog/sama5d4_wdt.c b/drivers/watchdog/sama5d4_wdt.c
>>> index 2a60251806d2..5ddeb4803dc3 100644
>>> --- a/drivers/watchdog/sama5d4_wdt.c
>>> +++ b/drivers/watchdog/sama5d4_wdt.c
>>> @@ -28,6 +28,7 @@
>>>  struct sama5d4_wdt {
>>>  	struct watchdog_device	wdd;
>>>  	void __iomem		*reg_base;
>>> +	u32			mr;
>>
>> Makes me wonder if we shouldn't just retain the original 'config'
>> (and maybe rename it to 'mr). After all, it _is_ used now.
>>
>>>  };
>>>
>>>  static int wdt_timeout = WDT_DEFAULT_TIMEOUT;
>>> @@ -248,11 +249,42 @@ static const struct of_device_id sama5d4_wdt_of_match[] = {
>>>  };
>>>  MODULE_DEVICE_TABLE(of, sama5d4_wdt_of_match);
>>>
>>> +#ifdef CONFIG_PM_SLEEP
>>> +static int sama5d4_wdt_suspend(struct device *dev)
>>> +{
>>> +	struct sama5d4_wdt *wdt = dev_get_drvdata(dev);
>>> +
>>> +	wdt->mr = wdt_read(wdt, AT91_WDT_MR);
>>> +
>>
>> Wouldn't you want to stop the watchdog here if it is running,
>> ie set AT91_WDT_WDDIS ?
>>
>
> Some existing customers want the watchdog to continue to run while the
> system is suspended.
>
>>> +	return 0;
>>> +}
>>> +
>>> +static int sama5d4_wdt_resume(struct device *dev)
>>> +{
>>> +	struct sama5d4_wdt *wdt = dev_get_drvdata(dev);
>>> +	u32 reg;
>>> +
>>> +	reg = wdt_read(wdt, AT91_WDT_MR);
>>> +	if (reg & AT91_WDT_WDDIS)
>>> +		wdt_write(wdt, AT91_WDT_MR, reg & ~AT91_WDT_WDDIS);
>>> +
>>> +	wdt_write(wdt, AT91_WDT_MR, wdt->mr & ~AT91_WDT_WDDIS);
>>
>> Is that necessary ? Why not just write wdt->mr unconditionally ?
>>
>
> Because you can change WDV and WDD *OR* WDDIS but not both at the same
> time.
>

Odd hardware ;-). Please add a comment explaining why the watchdog isn't
stopped on suspend, and why the AT91_WDT_MR has to be reprogrammed on resume.

Thanks,
Guenter