[PATCH] consider alrm->enable in pcf50633_rtc_set_alarm

[Werner Almesberger]

Hi Balaji,

Mickey mentioned to me that he had trouble with the RTC wakeup interrupt.
I had a quick look at the problem and it seems that alrm->enable doesn't
get propagated when setting the alarm time with RTC_WKALM_SET.

Does something like my patch below look right ? We also don't handle
alrm->pending, but I'm not sure if we have to.

I tested this only very lightly since my current andy-tracking crashes
in soc_suspend. If nobody else beats me to it, I'll have a look at it
tomorrow.

- Werner

---------------------------------- cut here -----------------------------------

According to Documentation/rtc.txt, RTC_WKALM_SET sets the alarm time
and enables/disables the alarm. We implement RTC_WKALM_SET through
pcf50633_rtc_set_alarm. The enabling/disabling part was missing.

Signed-off-by: Werner Almesberger <werner at openmoko.org>
Reported-by: Michael 'Mickey' Lauer <mickey at openmoko.org>

---

diff --git a/drivers/rtc/rtc-pcf50633.c b/drivers/rtc/rtc-pcf50633.c
index f4dd87e..0d6b006 100644
--- a/drivers/rtc/rtc-pcf50633.c
+++ b/drivers/rtc/rtc-pcf50633.c
@@ -245,8 +245,9 @@ static int pcf50633_rtc_set_alarm(struct device *dev, struct rtc_wkalrm *alrm)
 	ret = pcf50633_write_block(rtc->pcf, PCF50633_REG_RTCSCA,
 				PCF50633_TI_EXTENT, &pcf_tm.time[0]);
 
-	if (!alarm_masked)
+	if (!alarm_masked || alrm->enabled)
 		pcf50633_irq_unmask(rtc->pcf, PCF50633_IRQ_ALARM);
+	rtc->alarm_enabled = alrm->enabled;
 
 	return ret;
 }