[PATCH v2] watchdog: da9063_wdt: don't trigger watchdog too fast

[PATCH v2] watchdog: da9063_wdt: don't trigger watchdog too fast

[Stefan Christ]

Triggering the watchdog faster than T_WDMIN=256ms leads to resets of the
DA9063 chip. The datasheet says that the watchdog must only be triggered
in the timeframe T_WDMIN to T_WDMAX. The T_WDMAX is configured in the
driver.

Signed-off-by: Stefan Christ <s.christ@phytec.de>
---
v2: use internal kernel API 'min_hw_heartbeat_ms' instead of driver specific
    implementation.

Hi,

this is a follow up patch to the patch

    [RFC 1/6] watchdog: da9063_wdt: don't trigger watchdog too fast

in my RFC patch series. Using the internal kernel watchdog API also makes my
second patch

    [RFC 2/6] watchdog: da9063_wdt: use delayed work to trigger

obsolete, because the kernel API uses a delayed worker already.

Kind regards,
    Stefan Christ
---
 drivers/watchdog/da9063_wdt.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/watchdog/da9063_wdt.c b/drivers/watchdog/da9063_wdt.c
index a100f64..5d6b4e5 100644
--- a/drivers/watchdog/da9063_wdt.c
+++ b/drivers/watchdog/da9063_wdt.c
@@ -34,6 +34,7 @@ static const unsigned int wdt_timeout[] = { 0, 2, 4, 8, 16, 32, 65, 131 };
 #define DA9063_WDT_MIN_TIMEOUT		wdt_timeout[DA9063_TWDSCALE_MIN]
 #define DA9063_WDT_MAX_TIMEOUT		wdt_timeout[DA9063_TWDSCALE_MAX]
 #define DA9063_WDG_TIMEOUT		wdt_timeout[3]
+#define DA9063_RESET_PROTECTION_MS	256
 
 struct da9063_watchdog {
 	struct da9063 *da9063;
@@ -171,6 +172,7 @@ static int da9063_wdt_probe(struct platform_device *pdev)
 	wdt->wdtdev.ops = &da9063_watchdog_ops;
 	wdt->wdtdev.min_timeout = DA9063_WDT_MIN_TIMEOUT;
 	wdt->wdtdev.max_timeout = DA9063_WDT_MAX_TIMEOUT;
+	wdt->wdtdev.min_hw_heartbeat_ms = DA9063_RESET_PROTECTION_MS;
 	wdt->wdtdev.timeout = DA9063_WDG_TIMEOUT;
 	wdt->wdtdev.parent = &pdev->dev;
 
-- 
1.9.1

Re: [PATCH v2] watchdog: da9063_wdt: don't trigger watchdog too fast

[Guenter Roeck]

On Wed, Jul 06, 2016 at 10:40:11AM +0200, Stefan Christ wrote:
> Triggering the watchdog faster than T_WDMIN=256ms leads to resets of the
> DA9063 chip. The datasheet says that the watchdog must only be triggered
> in the timeframe T_WDMIN to T_WDMAX. The T_WDMAX is configured in the
> driver.
> 
> Signed-off-by: Stefan Christ <s.christ@phytec.de>

Reviewed-by: Guenter Roeck <linux@roeck-us.net>

> ---
> v2: use internal kernel API 'min_hw_heartbeat_ms' instead of driver specific
>     implementation.
> 
> Hi,
> 
> this is a follow up patch to the patch
> 
>     [RFC 1/6] watchdog: da9063_wdt: don't trigger watchdog too fast
> 
> in my RFC patch series. Using the internal kernel watchdog API also makes my
> second patch
> 
>     [RFC 2/6] watchdog: da9063_wdt: use delayed work to trigger
> 
> obsolete, because the kernel API uses a delayed worker already.
> 
> Kind regards,
>     Stefan Christ
> ---
>  drivers/watchdog/da9063_wdt.c | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/drivers/watchdog/da9063_wdt.c b/drivers/watchdog/da9063_wdt.c
> index a100f64..5d6b4e5 100644
> --- a/drivers/watchdog/da9063_wdt.c
> +++ b/drivers/watchdog/da9063_wdt.c
> @@ -34,6 +34,7 @@ static const unsigned int wdt_timeout[] = { 0, 2, 4, 8, 16, 32, 65, 131 };
>  #define DA9063_WDT_MIN_TIMEOUT		wdt_timeout[DA9063_TWDSCALE_MIN]
>  #define DA9063_WDT_MAX_TIMEOUT		wdt_timeout[DA9063_TWDSCALE_MAX]
>  #define DA9063_WDG_TIMEOUT		wdt_timeout[3]
> +#define DA9063_RESET_PROTECTION_MS	256
>  
>  struct da9063_watchdog {
>  	struct da9063 *da9063;
> @@ -171,6 +172,7 @@ static int da9063_wdt_probe(struct platform_device *pdev)
>  	wdt->wdtdev.ops = &da9063_watchdog_ops;
>  	wdt->wdtdev.min_timeout = DA9063_WDT_MIN_TIMEOUT;
>  	wdt->wdtdev.max_timeout = DA9063_WDT_MAX_TIMEOUT;
> +	wdt->wdtdev.min_hw_heartbeat_ms = DA9063_RESET_PROTECTION_MS;
>  	wdt->wdtdev.timeout = DA9063_WDG_TIMEOUT;
>  	wdt->wdtdev.parent = &pdev->dev;
>  
> -- 
> 1.9.1
> 

Re: [PATCH v2] watchdog: da9063_wdt: don't trigger watchdog too fast

[Wim Van Sebroeck]

Hi Stefan,

> Triggering the watchdog faster than T_WDMIN=256ms leads to resets of the
> DA9063 chip. The datasheet says that the watchdog must only be triggered
> in the timeframe T_WDMIN to T_WDMAX. The T_WDMAX is configured in the
> driver.
> 
> Signed-off-by: Stefan Christ <s.christ@phytec.de>
> ---
> v2: use internal kernel API 'min_hw_heartbeat_ms' instead of driver specific
>     implementation.
> 
> Hi,
> 
> this is a follow up patch to the patch
> 
>     [RFC 1/6] watchdog: da9063_wdt: don't trigger watchdog too fast
> 
> in my RFC patch series. Using the internal kernel watchdog API also makes my
> second patch
> 
>     [RFC 2/6] watchdog: da9063_wdt: use delayed work to trigger
> 
> obsolete, because the kernel API uses a delayed worker already.
> 
> Kind regards,
>     Stefan Christ
> ---
>  drivers/watchdog/da9063_wdt.c | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/drivers/watchdog/da9063_wdt.c b/drivers/watchdog/da9063_wdt.c
> index a100f64..5d6b4e5 100644
> --- a/drivers/watchdog/da9063_wdt.c
> +++ b/drivers/watchdog/da9063_wdt.c
> @@ -34,6 +34,7 @@ static const unsigned int wdt_timeout[] = { 0, 2, 4, 8, 16, 32, 65, 131 };
>  #define DA9063_WDT_MIN_TIMEOUT		wdt_timeout[DA9063_TWDSCALE_MIN]
>  #define DA9063_WDT_MAX_TIMEOUT		wdt_timeout[DA9063_TWDSCALE_MAX]
>  #define DA9063_WDG_TIMEOUT		wdt_timeout[3]
> +#define DA9063_RESET_PROTECTION_MS	256
>  
>  struct da9063_watchdog {
>  	struct da9063 *da9063;
> @@ -171,6 +172,7 @@ static int da9063_wdt_probe(struct platform_device *pdev)
>  	wdt->wdtdev.ops = &da9063_watchdog_ops;
>  	wdt->wdtdev.min_timeout = DA9063_WDT_MIN_TIMEOUT;
>  	wdt->wdtdev.max_timeout = DA9063_WDT_MAX_TIMEOUT;
> +	wdt->wdtdev.min_hw_heartbeat_ms = DA9063_RESET_PROTECTION_MS;
>  	wdt->wdtdev.timeout = DA9063_WDG_TIMEOUT;
>  	wdt->wdtdev.parent = &pdev->dev;
>  
> -- 
> 1.9.1
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-watchdog" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

This patch has been added to linux-watchdog-next.

Kind regards,
Wim.