On 08/06/2015 09:36 AM, Philipp Zabel wrote:
The DA9063 watchdog occasionally enters error condition and resets the system if the timeout is changed quickly after the timer was enabled. The method of disabling and waiting for > 150 µs before setting the new timeout is taken from the DA9052 driver.
Using sleep concerns me a bit, since during that time the watchdog is disabled. Bad enough that we have to do this to start with, but using usleep adds more risk to an already bad situation. How about udelay() instead ? Thanks, Guenter
Signed-off-by: Philipp Zabel <p.zabel@xxxxxxxxxxxxxx> --- drivers/watchdog/da9063_wdt.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/watchdog/da9063_wdt.c b/drivers/watchdog/da9063_wdt.c index b2e9201..87e14d8 100644 --- a/drivers/watchdog/da9063_wdt.c +++ b/drivers/watchdog/da9063_wdt.c @@ -67,6 +67,11 @@ static int _da9063_wdt_set_timeout(struct da9063_watchdog *wdt, mutex_lock(&wdt->mutex); ret = regmap_update_bits(da9063->regmap, DA9063_REG_CONTROL_D, + DA9063_TWDSCALE_MASK, DA9063_TWDSCALE_DISABLE); + + usleep_range(150, 300); + + ret = regmap_update_bits(da9063->regmap, DA9063_REG_CONTROL_D, DA9063_TWDSCALE_MASK, regval); wdt->defer_ping = false;
-- 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
