Commit f8245c26 authored by David Brownell's avatar David Brownell Committed by Linus Torvalds
Browse files

rtc: remove "RTC_ALM_SET mode" bugs 



This fixes a common glitch in how RTC drivers handle two "set alarm" modes,
by getting rid of the surprising/hidden one that was rarely implemented
correctly (and which could expose nonportable hardware-specific behavior).

The glitch comes from the /dev/rtcX logic implementing the legacy
RTC_ALM_SET (limited to 24 hours, needing RTC_AIE_ON) ioctl on top of the
RTC driver call providing access to the newer RTC_WKALM_SET (without those
limitations) by initializing the day/month/year fields to be invalid ...
that second mode.

Now, since few RTC drivers check those fields, and most hardware misbehaves
when faced with invalid date fields, many RTC drivers will set bogus alarm
times on those RTC_ALM_SET code paths.  (Several in-tree drivers have that
issue, and I also noticed it with code reviews on several new RTC drivers.)

This patch ensures that RTC drivers never see such invalid alarm fields, by
moving some logic out of rtc-omap into the RTC_ALM_SET code and adding an
explicit check (which will prevent the issue on other code paths).

Signed-off-by: default avatarDavid Brownell <dbrownell@users.sourceforge.net>
Cc: Scott Wood <scottwood@freescale.com>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: default avatarAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: default avatarLinus Torvalds <torvalds@linux-foundation.org>
parent 416ce32e

Documentation/rtc.txt

+2 −5
Original line number Diff line number Diff line
@@ -147,7 +147,7 @@ RTC class framework, but can't be supported by the older driver. 


    *	RTC_AIE_ON, RTC_AIE_OFF, RTC_ALM_SET, RTC_ALM_READ ... when the RTC

	is connected to an IRQ line, it can often issue an alarm IRQ up to

	24 hours in the future.

	24 hours in the future.  (Use RTC_WKALM_* by preference.)



    *	RTC_WKALM_SET, RTC_WKALM_RD ... RTCs that can issue alarms beyond

	the next 24 hours use a slightly more powerful API, which supports
@@ -175,10 +175,7 @@ driver returns ENOIOCTLCMD. Some common examples: 
	called with appropriate values.



    *	RTC_ALM_SET, RTC_ALM_READ, RTC_WKALM_SET, RTC_WKALM_RD: the

	set_alarm/read_alarm functions will be called.  To differentiate

	between the ALM and WKALM, check the larger fields of the rtc_wkalrm

	struct (like tm_year).  These will be set to -1 when using ALM and

	will be set to proper values when using WKALM.

	set_alarm/read_alarm functions will be called.



    *	RTC_IRQP_SET, RTC_IRQP_READ: the irq_set_freq function will be called

	to set the frequency while the framework will handle the read for you

drivers/rtc/interface.c

+4 −0
Original line number Diff line number Diff line
@@ -125,6 +125,10 @@ int rtc_set_alarm(struct rtc_device *rtc, struct rtc_wkalrm *alarm) 
{

	int err;



	err = rtc_valid_tm(&alarm->time);

	if (err != 0)

		return err;



	err = mutex_lock_interruptible(&rtc->ops_lock);

	if (err)

		return -EBUSY;

drivers/rtc/rtc-dev.c

+39 −3
Original line number Diff line number Diff line
@@ -277,12 +277,48 @@ static int rtc_dev_ioctl(struct inode *inode, struct file *file, 


		alarm.enabled = 0;

		alarm.pending = 0;

		alarm.time.tm_mday = -1;

		alarm.time.tm_mon = -1;

		alarm.time.tm_year = -1;

		alarm.time.tm_wday = -1;

		alarm.time.tm_yday = -1;

		alarm.time.tm_isdst = -1;



		/* RTC_ALM_SET alarms may be up to 24 hours in the future.

		 * Rather than expecting every RTC to implement "don't care"

		 * for day/month/year fields, just force the alarm to have

		 * the right values for those fields.

		 *

		 * RTC_WKALM_SET should be used instead.  Not only does it

		 * eliminate the need for a separate RTC_AIE_ON call, it

		 * doesn't have the "alarm 23:59:59 in the future" race.

		 *

		 * NOTE:  some legacy code may have used invalid fields as

		 * wildcards, exposing hardware "periodic alarm" capabilities.

		 * Not supported here.

		 */

		{

			unsigned long now, then;



			err = rtc_read_time(rtc, &tm);

			if (err < 0)

				return err;

			rtc_tm_to_time(&tm, &now);



			alarm.time.tm_mday = tm.tm_mday;

			alarm.time.tm_mon = tm.tm_mon;

			alarm.time.tm_year = tm.tm_year;

			err  = rtc_valid_tm(&alarm.time);

			if (err < 0)

				return err;

			rtc_tm_to_time(&alarm.time, &then);



			/* alarm may need to wrap into tomorrow */

			if (then < now) {

				rtc_time_to_tm(now + 24 * 60 * 60, &tm);

				alarm.time.tm_mday = tm.tm_mday;

				alarm.time.tm_mon = tm.tm_mon;

				alarm.time.tm_year = tm.tm_year;

			}

		}



		err = rtc_set_alarm(rtc, &alarm);

		break;

drivers/rtc/rtc-omap.c

+0 −28
Original line number Diff line number Diff line
@@ -289,34 +289,6 @@ static int omap_rtc_set_alarm(struct device *dev, struct rtc_wkalrm *alm) 
{

	u8 reg;



	/* Much userspace code uses RTC_ALM_SET, thus "don't care" for

	 * day/month/year specifies alarms up to 24 hours in the future.

	 * So we need to handle that ... but let's ignore the "don't care"

	 * values for hours/minutes/seconds.

	 */

	if (alm->time.tm_mday <= 0

			&& alm->time.tm_mon < 0

			&& alm->time.tm_year < 0) {

		struct rtc_time tm;

		unsigned long now, then;



		omap_rtc_read_time(dev, &tm);

		rtc_tm_to_time(&tm, &now);



		alm->time.tm_mday = tm.tm_mday;

		alm->time.tm_mon = tm.tm_mon;

		alm->time.tm_year = tm.tm_year;

		rtc_tm_to_time(&alm->time, &then);



		/* sometimes the alarm wraps into tomorrow */

		if (then < now) {

			rtc_time_to_tm(now + 24 * 60 * 60, &tm);

			alm->time.tm_mday = tm.tm_mday;

			alm->time.tm_mon = tm.tm_mon;

			alm->time.tm_year = tm.tm_year;

		}

	}



	if (tm2bcd(&alm->time) < 0)

		return -EINVAL;

CRA Git | Maintained and supported by SUSTech CRA and CCSE