| Message ID | 1359702386-21284-5-git-send-email-wenyou.yang@atmel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Hello, On Thursday 14 February 2013 00:15:22, Wim Van Sebroeck wrote: > Hi Wenyou, > > > Since the Watchdog Timer Mode Register can be only written only once, > > so the watchdog_info shall not support WDIOF_SETTIMEOUT > > and WDIOF_MAGICCLOSE options, remove them. > > Ik you keep using the timer, then you don't have to remove this. Why removing especially WDIOF_MAGICCLOSE (and the timer needed for that)? This is a feature removal. Without there isn't even a need for nowayout option as it is implicitly set anyway. Nevertheless I would prefere keeping the timer to allow a user to let the kernel triggering the timer upon MAGICCLOSE. Alexander
diff --git a/drivers/watchdog/at91sam9_wdt.c b/drivers/watchdog/at91sam9_wdt.c index ce7930b..c6d9f1f 100644 --- a/drivers/watchdog/at91sam9_wdt.c +++ b/drivers/watchdog/at91sam9_wdt.c @@ -152,8 +152,7 @@ static int at91wdt_enable(struct watchdog_device *wddev, unsigned int timeout) static const struct watchdog_info at91_wdt_info = { .identity = DRV_NAME, - .options = WDIOF_SETTIMEOUT | WDIOF_KEEPALIVEPING | - WDIOF_MAGICCLOSE, + .options = WDIOF_KEEPALIVEPING, }; static int at91wdt_start(struct watchdog_device *wddev)
Since the Watchdog Timer Mode Register can be only written only once, so the watchdog_info shall not support WDIOF_SETTIMEOUT and WDIOF_MAGICCLOSE options, remove them. Signed-off-by: Wenyou Yang <wenyou.yang@atmel.com> Cc: wim@iguana.be Cc: linux-watchdog@vger.kernel.org Cc: linux-kernel@vger.kernel.org --- drivers/watchdog/at91sam9_wdt.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)