[PATCH v2] watchdog: intel-mid_wdt: Keep watchdog running

[PATCH v2] watchdog: intel-mid_wdt: Keep watchdog running

[Andy Shevchenko]

Firmware followed by bootloader leaves watchdog running.

Keep it running in the driver.

User will not need any additional options to reboot in case of panic
during boot.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
---
 drivers/watchdog/intel-mid_wdt.c | 12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

diff --git a/drivers/watchdog/intel-mid_wdt.c b/drivers/watchdog/intel-mid_wdt.c
index 45e4d02221b5..f2c66689f42c 100644
--- a/drivers/watchdog/intel-mid_wdt.c
+++ b/drivers/watchdog/intel-mid_wdt.c
@@ -147,8 +147,16 @@ static int mid_wdt_probe(struct platform_device *pdev)
 		return ret;
 	}
 
-	/* Make sure the watchdog is not running */
-	wdt_stop(wdt_dev);
+	/*
+	 * Make sure the watchdog is serviced.
+	 *
+	 * The firmware followed by U-Boot leaves the watchdog running
+	 * with the default threshold 60 seconds. Our default timeout is
+	 * 90 seconds, but internal worker divides it by two, which is
+	 * 45 seconds and should be enough (less by 15 seconds than
+	 * threshold).
+	 */
+	set_bit(WDOG_HW_RUNNING, &wdt_dev->status);
 
 	ret = devm_watchdog_register_device(&pdev->dev, wdt_dev);
 	if (ret) {
-- 
2.11.0

Re: [PATCH v2] watchdog: intel-mid_wdt: Keep watchdog running

[Andy Shevchenko]

On Fri, 2017-03-10 at 18:51 +0200, Andy Shevchenko wrote:
> Firmware followed by bootloader leaves watchdog running.
> 
> Keep it running in the driver.
> 
> User will not need any additional options to reboot in case of panic
> during boot.

> -	/* Make sure the watchdog is not running */
> -	wdt_stop(wdt_dev);
> +	/*
> +	 * Make sure the watchdog is serviced.
> +	 *
> +	 * The firmware followed by U-Boot leaves the watchdog
> running
> +	 * with the default threshold 60 seconds. Our default timeout
> is
> +	 * 90 seconds, but internal worker divides it by two, which
> is
> +	 * 45 seconds and should be enough (less by 15 seconds than
> +	 * threshold).
> +	 */
> +	set_bit(WDOG_HW_RUNNING, &wdt_dev->status);

Giving one more thought here perhaps more robust will be to start
watchdog unconditionally here?

In that case it will be no matter how it's configured in firmware and/or
boot loader.

>  
>  	ret = devm_watchdog_register_device(&pdev->dev, wdt_dev);
>  	if (ret) {

-- 
Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Intel Finland Oy

Re: [PATCH v2] watchdog: intel-mid_wdt: Keep watchdog running

[Guenter Roeck]

On Fri, Mar 10, 2017 at 07:05:49PM +0200, Andy Shevchenko wrote:
> On Fri, 2017-03-10 at 18:51 +0200, Andy Shevchenko wrote:
> > Firmware followed by bootloader leaves watchdog running.
> > 
> > Keep it running in the driver.
> > 
> > User will not need any additional options to reboot in case of panic
> > during boot.
> 
> > -	/* Make sure the watchdog is not running */
> > -	wdt_stop(wdt_dev);
> > +	/*
> > +	 * Make sure the watchdog is serviced.
> > +	 *
> > +	 * The firmware followed by U-Boot leaves the watchdog
> > running
> > +	 * with the default threshold 60 seconds. Our default timeout
> > is
> > +	 * 90 seconds, but internal worker divides it by two, which
> > is
> > +	 * 45 seconds and should be enough (less by 15 seconds than
> > +	 * threshold).
> > +	 */
> > +	set_bit(WDOG_HW_RUNNING, &wdt_dev->status);
> 
> Giving one more thought here perhaps more robust will be to start
> watchdog unconditionally here?
> 
Yes, that would be another option, and it sounds like a good idea.
Let me know if you want to do that.

Thanks,
Guenter

> In that case it will be no matter how it's configured in firmware and/or
> boot loader.
> 
> >  
> >  	ret = devm_watchdog_register_device(&pdev->dev, wdt_dev);
> >  	if (ret) {
> 
> -- 
> Andy Shevchenko <andriy.shevchenko@linux.intel.com>
> Intel Finland Oy

Re: [PATCH v2] watchdog: intel-mid_wdt: Keep watchdog running

[Andy Shevchenko]

On Fri, 2017-03-10 at 09:16 -0800, Guenter Roeck wrote:
> On Fri, Mar 10, 2017 at 07:05:49PM +0200, Andy Shevchenko wrote:
> > On Fri, 2017-03-10 at 18:51 +0200, Andy Shevchenko wrote:

> > > +	/*
> > > +	 * Make sure the watchdog is serviced.
> > > +	 *
> > > +	 * The firmware followed by U-Boot leaves the watchdog
> > > running
> > > +	 * with the default threshold 60 seconds. Our default
> > > timeout
> > > is
> > > +	 * 90 seconds, but internal worker divides it by two,
> > > which
> > > is
> > > +	 * 45 seconds and should be enough (less by 15 seconds
> > > than
> > > +	 * threshold).
> > > +	 */
> > > +	set_bit(WDOG_HW_RUNNING, &wdt_dev->status);
> > 
> > Giving one more thought here perhaps more robust will be to start
> > watchdog unconditionally here?
> 
> Yes, that would be another option, and it sounds like a good idea.
> Let me know if you want to do that.

Yes, I prefer it over the rest of solutions.

-- 
Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Intel Finland Oy

Re: [PATCH v2] watchdog: intel-mid_wdt: Keep watchdog running

[Guenter Roeck]

On Fri, Mar 10, 2017 at 07:26:57PM +0200, Andy Shevchenko wrote:
> On Fri, 2017-03-10 at 09:16 -0800, Guenter Roeck wrote:
> > On Fri, Mar 10, 2017 at 07:05:49PM +0200, Andy Shevchenko wrote:
> > > On Fri, 2017-03-10 at 18:51 +0200, Andy Shevchenko wrote:
> 
> > > > +	/*
> > > > +	 * Make sure the watchdog is serviced.
> > > > +	 *
> > > > +	 * The firmware followed by U-Boot leaves the watchdog
> > > > running
> > > > +	 * with the default threshold 60 seconds. Our default
> > > > timeout
> > > > is
> > > > +	 * 90 seconds, but internal worker divides it by two,
> > > > which
> > > > is
> > > > +	 * 45 seconds and should be enough (less by 15 seconds
> > > > than
> > > > +	 * threshold).
> > > > +	 */
> > > > +	set_bit(WDOG_HW_RUNNING, &wdt_dev->status);
> > > 
> > > Giving one more thought here perhaps more robust will be to start
> > > watchdog unconditionally here?
> > 
> > Yes, that would be another option, and it sounds like a good idea.
> > Let me know if you want to do that.
> 
> Yes, I prefer it over the rest of solutions.
> 
Ok, I'll wait for v3 of the patch.

Thanks,
Guenter