projects/linux·commit 57e6319d

vrtc: change its year offset from 1960 to 1972

57e6319d · Feng Tang <[email protected]> · 2011-11-10 13:23:39 +00:00

Real world year equals the value in vrtc YEAR register plus an offset.
We used 1960 as the offset to make leap year consistent, but for a
device's first use, its YEAR register is 0 and the system year will
be parsed as 1960 which is not a valid UNIX time and will cause many
applications to fail mysteriously. So we use 1972 instead to fix this
issue.

Updated patch which adds a sanity check suggested by Mathias

This isn't a change in behaviour for systems, because 1972 is the one we
actually use. It's the old version in upstream which is out of sync with
all devices.

Signed-off-by: Feng Tang <[email protected]>
Signed-off-by: Alan Cox <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Commit
57e6319dd61d5ca10fe8dd57bcce8c0e2c480799
Parents
f2ee4421

Changes

2 files changed · +13 −10

FileStatusAddDelContent
arch/x86/platform/mrst/vrtc.cmodified+2−2View file
drivers/rtc/rtc-mrst.cmodified+11−8View file

arch/x86/platform/mrst/vrtc.c

View file
diff --git a/arch/x86/platform/mrst/vrtc.c b/arch/x86/platform/mrst/vrtc.c
index a8ac6f1..225bd0f 100644
--- a/arch/x86/platform/mrst/vrtc.c
+++ b/arch/x86/platform/mrst/vrtc.c
@@ -76,8 +76,8 @@ unsigned long vrtc_get_time(void)
spin_unlock_irqrestore(&rtc_lock, flags);
- /* vRTC YEAR reg contains the offset to 1960 */
- year += 1960;
+ /* vRTC YEAR reg contains the offset to 1972 */
+ year += 1972;
printk(KERN_INFO "vRTC: sec: %d min: %d hour: %d day: %d "
"mon: %d year: %d\n", sec, min, hour, mday, mon, year);

drivers/rtc/rtc-mrst.c

View file
diff --git a/drivers/rtc/rtc-mrst.c b/drivers/rtc/rtc-mrst.c
index d335448..bb21f44 100644
--- a/drivers/rtc/rtc-mrst.c
+++ b/drivers/rtc/rtc-mrst.c
@@ -76,12 +76,15 @@ static inline unsigned char vrtc_is_updating(void)
/*
* rtc_time's year contains the increment over 1900, but vRTC's YEAR
* register can't be programmed to value larger than 0x64, so vRTC
- * driver chose to use 1960 (1970 is UNIX time start point) as the base,
+ * driver chose to use 1972 (1970 is UNIX time start point) as the base,
* and does the translation at read/write time.
*
- * Why not just use 1970 as the offset? it's because using 1960 will
+ * Why not just use 1970 as the offset? it's because using 1972 will
* make it consistent in leap year setting for both vrtc and low-level
- * physical rtc devices.
+ * physical rtc devices. Then why not use 1960 as the offset? If we use
+ * 1960, for a device's first use, its YEAR register is 0 and the system
+ * year will be parsed as 1960 which is not a valid UNIX time and will
+ * cause many applications to fail mysteriously.
*/
static int mrst_read_time(struct device *dev, struct rtc_time *time)
{
@@ -99,10 +102,10 @@ static int mrst_read_time(struct device *dev, struct rtc_time *time)
time->tm_year = vrtc_cmos_read(RTC_YEAR);
spin_unlock_irqrestore(&rtc_lock, flags);
- /* Adjust for the 1960/1900 */
- time->tm_year += 60;
+ /* Adjust for the 1972/1900 */
+ time->tm_year += 72;
time->tm_mon--;
- return RTC_24H;
+ return rtc_valid_tm(time);
}
static int mrst_set_time(struct device *dev, struct rtc_time *time)
@@ -119,9 +122,9 @@ static int mrst_set_time(struct device *dev, struct rtc_time *time)
min = time->tm_min;
sec = time->tm_sec;
- if (yrs < 70 || yrs > 138)
+ if (yrs < 72 || yrs > 138)
return -EINVAL;
- yrs -= 60;
+ yrs -= 72;
spin_lock_irqsave(&rtc_lock, flags);