Re: [PATCH v3 4/6] watchdog: sp805: set WDOG_HW_RUNNING when appropriate

From: Ray Jui
Date: Thu May 24 2018 - 11:44:05 EST




On 5/24/2018 9:19 AM, Guenter Roeck wrote:
On Wed, May 23, 2018 at 05:15:22PM -0700, Ray Jui wrote:
If the watchdog hardware is already enabled during the boot process,
when the Linux watchdog driver loads, it should reset the watchdog and
tell the watchdog framework. As a result, ping can be generated from
the watchdog framework, until the userspace watchdog daemon takes over
control

Signed-off-by: Ray Jui <ray.jui@xxxxxxxxxxxx>
Reviewed-by: Vladimir Olovyannikov <vladimir.olovyannikov@xxxxxxxxxxxx>
Reviewed-by: Scott Branden <scott.branden@xxxxxxxxxxxx>

Reviewed-by: Guenter Roeck <linux@xxxxxxxxxxxx>

I have one question, though: Is it really correct that both
INT_ENABLE _and_ RESET_ENABLE have to be set to enable the watdog ?
What if only RESET_ENABLE is set ?

According to the SP805 TRM, INT_ENABLE needs to be set to high to enable the counter and the interrupt. Counter will be stopped if INT_ENABLE is cleared. So yes, INT_ENABLE needs to be set.

Thanks,

Ray


Thanks,
Guenter

---
drivers/watchdog/sp805_wdt.c | 19 +++++++++++++++++++
1 file changed, 19 insertions(+)

diff --git a/drivers/watchdog/sp805_wdt.c b/drivers/watchdog/sp805_wdt.c
index 1484609..d662a6f 100644
--- a/drivers/watchdog/sp805_wdt.c
+++ b/drivers/watchdog/sp805_wdt.c
@@ -42,6 +42,7 @@
/* control register masks */
#define INT_ENABLE (1 << 0)
#define RESET_ENABLE (1 << 1)
+ #define ENABLE_MASK (INT_ENABLE | RESET_ENABLE)
#define WDTINTCLR 0x00C
#define WDTRIS 0x010
#define WDTMIS 0x014
@@ -74,6 +75,15 @@ module_param(nowayout, bool, 0);
MODULE_PARM_DESC(nowayout,
"Set to 1 to keep watchdog running after device release");
+/* returns true if wdt is running; otherwise returns false */
+static bool wdt_is_running(struct watchdog_device *wdd)
+{
+ struct sp805_wdt *wdt = watchdog_get_drvdata(wdd);
+ u32 wdtcontrol = readl_relaxed(wdt->base + WDTCONTROL);
+
+ return (wdtcontrol & ENABLE_MASK) == ENABLE_MASK;
+}
+
/* This routine finds load value that will reset system in required timout */
static int wdt_setload(struct watchdog_device *wdd, unsigned int timeout)
{
@@ -239,6 +249,15 @@ sp805_wdt_probe(struct amba_device *adev, const struct amba_id *id)
watchdog_init_timeout(&wdt->wdd, 0, &adev->dev);
wdt_setload(&wdt->wdd, wdt->wdd.timeout);
+ /*
+ * If HW is already running, enable/reset the wdt and set the running
+ * bit to tell the wdt subsystem
+ */
+ if (wdt_is_running(&wdt->wdd)) {
+ wdt_enable(&wdt->wdd);
+ set_bit(WDOG_HW_RUNNING, &wdt->wdd.status);
+ }
+
ret = watchdog_register_device(&wdt->wdd);
if (ret) {
dev_err(&adev->dev, "watchdog_register_device() failed: %d\n",
--
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-watchdog" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html