invalid default values in RTC chip
Hi,
My question relates to RTC driver linux/drivers/rtc/rtc-ds1307.c for device m41t00. Driver probe function will fail if some of the chip's registers contain invalid date time values i.e. if month register is 32 or minutes is 61. Is this correct behavior? Probe function's purpose is to check if the device is as was assumed (this time RTC and it is). The chip's values are incorrect but the chip works, even the m41t00 chip manual states that after initial powerup (RTC battery power applied) internal registers will contain random data.
There are two solutions first is driver patch and another is i2c-dev and i2cset tool to use from user space during bootup. Whitch one should be used?
PS. please add me to CC also.
JR
Quoted 11 lines · On Tue, 14 Oct 2008 12:00:21 +0300 J__ri Reitel <[email protected]>…
Hi,
My question relates to RTC driver
linux/drivers/rtc/rtc-ds1307.cfor device m41t00. Driver probe function will fail if some of the chip's registers contain invalid date time values i.e. if month register is 32 or minutes is 61. Is this correct behavior? Probe function's purpose is to check if the device is as was assumed (this time RTC and it is). The chip's values are incorrect but the chip works, even the m41t00 chip manual states that after initial powerup (RTC battery power applied) internal registers will contain random data.
Quoted 3 lines
There are two solutions first is driver patch and another is i2c-dev and i2cset tool to use from user space during bootup. Whitch one should be used?
(cc's added)
Quoted 11 lines · On Thu, 16 Oct 2008 13:58:46 -0700 Andrew Morton <[email protected]>…
Quoted 11 lines
My question relates to RTC driver
linux/drivers/rtc/rtc-ds1307.cfor device m41t00. Driver probe function will fail if some of the chip's registers contain invalid date time values i.e. if month register is 32 or minutes is 61. Is this correct behavior? Probe function's purpose is to check if the device is as was assumed (this time RTC and it is). The chip's values are incorrect but the chip works, even the m41t00 chip manual states that after initial powerup (RTC battery power applied) internal registers will contain random data.There are two solutions first is driver patch and another is i2c-dev and i2cset tool to use from user space during bootup. Whitch one should be used?
Given we transitioned to the new i2c model, which mandates a platform device to be declared, there's no more the need to check for the device in the probing function.
this should be fixed in the driver.
however, a driver may still return a failure when reading the time if there are inappropriate values in the registers.
the solution is to patch the driver for detection and then use a recent hwclock on bootup to write the time if required.
Signature · 6 lines
Best regards, Alessandro Zummo, Tower Technologies - Torino, Italy http://www.towertech.it
Quoted 7 lines · Alessandro Zummo
Quoted 4 lines · On Tue, 14 Oct 2008 12:00:21 +0300 Jüri Reitel <[email protected]>…
Quoted 4 lines
My question relates to RTC driver
linux/drivers/rtc/rtc-ds1307.cfor device m41t00. Driver probe function will fail if some of the chip's registers contain invalid date time values i.e. if month register is 32 or minutes is 61.
This stuff?
tmp = ds1307->regs[DS1307_REG_SECS];
tmp = BCD2BIN(tmp & 0x7f);
if (tmp > 60)
goto exit_bad;
tmp = BCD2BIN(ds1307->regs[DS1307_REG_MIN] & 0x7f);
if (tmp > 60)
goto exit_bad;
tmp = BCD2BIN(ds1307->regs[DS1307_REG_MDAY] & 0x3f);
if (tmp == 0 || tmp > 31)
goto exit_bad;
tmp = BCD2BIN(ds1307->regs[DS1307_REG_MONTH] & 0x1f);
if (tmp == 0 || tmp > 12)
goto exit_bad;Quoted 11 lines
Quoted 8 lines
Quoted 8 lines
Is this correct behavior? Probe function's purpose is
to check if the device is as was assumed (this time RTC and it is). The chip's values are incorrect but the chip works, even the m41t00 chip manual states that after initial powerup (RTC battery power applied) internal registers will contain random data.There are two solutions first is driver patch and another is i2c-dev and i2cset tool to use from user space during bootup. Whitch one should be used?
Given we transitioned to the new i2c model, which mandates a platform device to be declared,
Actually "i2c_board_info", not a "platform_device" ...
Quoted 4 lines
there's no more the need
to check for the device in the probing function.this should be fixed in the driver.
Right. The code I snipped above dates from the old drivers/i2c/chips/ds1337.c code, as I recall, which was trying to defend against some random non-RTC chip sitting at the relevant address. As pretty much all legacy I2C drivers needed to do.
With "new style" I2C drivers, no guessing is needed and such defenses are superfluos. It's fair to notice that no chip is actually present ... but if there's a chip there, it should be assumed to match the declaration Linux was given.
Quoted 2 lines
however, a driver may still return a failure when reading the time if there are inappropriate values in the registers.
In this driver, ds1307_get_time() already has a
return rtc_valid_tm(t);
But in the just-merged alarm support, ds1307_read_alarm() doesn't do that. Actually that function should use a name like ds1337_read_alarm(), since the 1307 has no alarm!
Quoted 2 lines
the solution is to patch the driver for detection and then use a recent hwclock on bootup to write the time if required.
Exactly. Feel free to submit a patch removing those probe() checks, and fixing the code returning the alarm setting too.
- Dave
Quoted 75 lines · David Brownell
Quoted 6 lines · Alessandro Zummo
Quoted 2 lines · On Tue, 14 Oct 2008 12:00:21 +0300 Jüri Reitel <[email protected]>…
Quoted 2 lines
My question relates to RTC driver
linux/drivers/rtc/rtc-ds1307.cfor device m41t00. Driver probe function will fail if some of the chip's registers contain invalid date time values i.e. if month register is 32 or minutes is 61.This stuff?
tmp = ds1307->regs[DS1307_REG_SECS]; tmp = BCD2BIN(tmp & 0x7f); if (tmp > 60) goto exit_bad; tmp = BCD2BIN(ds1307->regs[DS1307_REG_MIN] & 0x7f); if (tmp > 60) goto exit_bad; tmp = BCD2BIN(ds1307->regs[DS1307_REG_MDAY] & 0x3f); if (tmp == 0 || tmp > 31) goto exit_bad; tmp = BCD2BIN(ds1307->regs[DS1307_REG_MONTH] & 0x1f); if (tmp == 0 || tmp > 12) goto exit_bad;Quoted 7 lines
Quoted 4 lines
Quoted 4 lines
Is this correct behavior? Probe function's purpose is to check if the device is as was assumed (this time RTC and it is). The chip's values are incorrect but the chip works, even the m41t00 chip manual states that after initial powerup (RTC battery power applied) internal registers will contain random data.
There are two solutions first is driver patch and another is i2c-dev and i2cset tool to use from user space during bootup. Whitch one should be used?
Given we transitioned to the new i2c model, which mandates a platform device to be declared,
Actually "
i2c_board_info", not a "platform_device" ...Quoted 5 lines
there's no more the need
to check for the device in the probing function.this should be fixed in the driver.
Right. The code I snipped above dates from the old
drivers/i2c/chips/ds1337.ccode, as I recall, which was trying to defend against some random non-RTC chip sitting at the relevant address. As pretty much all legacy I2C drivers needed to do.With "new style" I2C drivers, no guessing is needed and such defenses are superfluos. It's fair to notice that no chip is actually present ... but if there's a chip there, it should be assumed to match the declaration Linux was given.
Quoted 3 lines
however, a driver may still return a failure when reading the time if there are inappropriate values in the registers.
In this driver,
ds1307_get_time()already has areturn rtc_valid_tm(t);But in the just-merged alarm support,
ds1307_read_alarm()doesn't do that. Actually that function should use a name likeds1337_read_alarm(), since the 1307 has no alarm!
Quoted 9 lines
Quoted 3 lines
the solution is to patch the driver for detection and then use a recent hwclock on bootup to write the time if required.
Exactly. Feel free to submit a patch removing those
probe()checks, and fixing the code returning the alarm setting too.
I included patch that just removes the RTC time registers check, when i compiled the module and booted up with invalid values in RTC then system automatically set system time and RTC time to 1.1.2000 (this is nice because i dont have to use date and hwclock manually)
i checked the rtc_valid_tm function and saw that it also checks for the year value among others but in ds1307_read_alarm (now this function is renamed to ds1337_read_alarm) year is set to -1 with some other fields so rtc_valid_tm function can not be used in ds1337_read_alarm function.
Quoted 4 lines
- Dave
This patch removes rtc date time values check in probe function
drivers/rtc/rtc-ds1307.c +0 -0
@@ -298,7 +298,7 @@ static int ds1307_set_time(struct device *dev, struct rtc_time *t)return 0 ;
}
-static int ds1307_read_alarm(struct device *dev, struct rtc_wkalrm *t) +static int ds1337_read_alarm(struct device *dev, struct rtc_wkalrm *t)
{
struct i2c_client *client = to_i2c_client(dev);
struct ds1307 *ds1307 = i2c_get_clientdata(client);
@@ -353,7 +353,7 @@ static int ds1307_read_alarm(struct device *dev, struct rtc_wkalrm *t)
return 0;
}-static int ds1307_set_alarm(struct device *dev, struct rtc_wkalrm *t) +static int ds1337_set_alarm(struct device *dev, struct rtc_wkalrm *t)
{
struct i2c_client *client = to_i2c_client(dev);
struct ds1307 *ds1307 = i2c_get_clientdata(client);
@@ -475,8 +475,8 @@ static int ds1307_ioctl(struct device *dev, unsigned int cmd, unsigned long arg)
static const struct rtc_class_ops ds13xx_rtc_ops = {
.read_time = ds1307_get_time,
.set_time = ds1307_set_time,
- .read_alarm = ds1307_read_alarm,
- .set_alarm = ds1307_set_alarm,
+ .read_alarm = ds1337_read_alarm,
+ .set_alarm = ds1337_set_alarm,
.ioctl = ds1307_ioctl,
};@@ -707,22 +707,6 @@ read_rtc:
break;
}- tmp = ds1307->regs[DS1307_REG_SECS]; - tmp = bcd2bin(tmp & 0x7f); - if (tmp > 60)
- goto exit_bad;
- tmp = bcd2bin(ds1307->regs[DS1307_REG_MIN] & 0x7f);
- if (tmp > 60)
- goto exit_bad;
-
- tmp = bcd2bin(ds1307->regs[DS1307_REG_MDAY] & 0x3f);
- if (tmp == 0 || tmp > 31)
- goto exit_bad;
-
- tmp = bcd2bin(ds1307->regs[DS1307_REG_MONTH] & 0x1f);
- if (tmp == 0 || tmp > 12)
- goto exit_bad;
-
tmp = ds1307->regs[DS1307_REG_HOUR];
switch (ds1307->type) {
case ds_1340:@@ -779,13 +763,6 @@ read_rtc:
return 0;
-exit_bad:
- dev_dbg(&client->dev, "%s: %02x %02x %02x %02x %02x %02x %02x\n",
- "bogus register",
- ds1307->regs[0], ds1307->regs[1],
- ds1307->regs[2], ds1307->regs[3],
- ds1307->regs[4], ds1307->regs[5],
- ds1307->regs[6]);
exit_irq:
if (ds1307->rtc)
rtc_device_unregister(ds1307->rtc);Quoted 5 lines · Jüri Reitel
I included patch that just removes the RTC time registers check, when i compiled the module and booted up with invalid values in RTC then system automatically set system time and RTC time to 1.1.2000 (this is nice because i dont have to use date and hwclock manually)
I wouldn't rely on the initial time setting being valid; it's not guaranteed by specs for all the chips, as I recall.
Quoted 4 lines
i checked the
rtc_valid_tmfunction and saw that it also checks for the year value among others but inds1307_read_alarm(now this function is renamed tods1337_read_alarm) year is set to -1 with some other fields sortc_valid_tmfunction can not be used inds1337_read_alarmfunction.
Right. I forget what alarm issues I was thinking of, but that wasn't it ... other code in the RTC framework handles morphing that time into a valid date.
Quoted 2 lines
This patch removes rtc date time values check in probe function
Thanks. This looks OK ... got a Signed-off-By for us?
Quoted 72 lines
drivers/rtc/rtc-ds1307.c +0
-0@@ -298,7 +298,7 @@ static int ds1307_set_time(struct device *dev, struct rtc_time *t) return 0; }drivers/rtc/rtc-ds1307.c +1
-1-static int ds1307_read_alarm (struct device *dev,struct rtc_wkalrm *t) +static int ds1337_read_alarm (struct device *dev,struct rtc_wkalrm *t){
structi2c_client*client =to_i2c_client(dev); struct ds1307 *ds1307 =i2c_get_clientdata(client);drivers/rtc/rtc-ds1307.c +0
-0@@ -353,7 +353,7 @@ static int ds1307_read_alarm(struct device *dev, struct rtc_wkalrm *t) return 0; }drivers/rtc/rtc-ds1307.c +1
-1-static int ds1307_set_alarm (struct device *dev,struct rtc_wkalrm *t) +static int ds1337_set_alarm (struct device *dev,struct rtc_wkalrm *t){
structi2c_client*client =to_i2c_client(dev); struct ds1307 *ds1307 =i2c_get_clientdata(client);drivers/rtc/rtc-ds1307.c +2
-2@@ -475,8 +475,8 @@ static int ds1307_ioctl(struct device *dev, unsigned int cmd, unsigned long arg) static const struct rtc_class_ops ds13xx_rtc_ops = { .read_time = ds1307_get_time, .set_time = ds1307_set_time, - .read_alarm = ds1307_read_alarm, - .set_alarm = ds1307_set_alarm, + .read_alarm = ds1337_read_alarm, + .set_alarm = ds1337_set_alarm, .ioctl = ds1307_ioctl, };drivers/rtc/rtc-ds1307.c +0
-0@@ -707,22 +707,6 @@ read_rtc: break; }
- tmp = ds1307->regs[DS1307_REG_SECS];
- tmp =bcd2bin(tmp & 0x7f);
- if (tmp > 60)
- gotoexit_bad;
- tmp =bcd2bin(ds1307->regs[DS1307_REG_MIN] & 0x7f);
- if (tmp > 60)
- gotoexit_bad;drivers/rtc/rtc-ds1307.c +0
-9- - tmp = bcd2bin(ds1307->regs[DS1307_REG_MDAY] & 0x3f); - if (tmp == 0 || tmp > 31) - goto exit_bad; - - tmp = bcd2bin(ds1307->regs[DS1307_REG_MONTH] & 0x1f); - if (tmp == 0 || tmp > 12) - goto exit_bad; -tmp = ds1307->regs[DS1307_REG_HOUR];
switch (ds1307->type) {
caseds_1340:drivers/rtc/rtc-ds1307.c +0
-0@@ -779,13 +763,6 @@ read_rtc:
return 0;
-exit_bad:
- dev_dbg(&client->dev, "%s: %02x %02x %02x %02x %02x %02x %02x\n",
- "bogus register",
- ds1307->regs[0], ds1307->regs[1],
- ds1307->regs[2], ds1307->regs[3],
- ds1307->regs[4], ds1307->regs[5],
- ds1307->regs[6]);
exit_irq:
if (ds1307->rtc)
rtc_device_unregister(ds1307->rtc);
Quoted 87 lines · David Brownell
Quoted 2 lines · Jüri Reitel
I included patch that just removes the RTC time registers check, when i compiled the module and booted up with invalid values in RTC then system automatically set system time and RTC time to 1.1.2000 (this is nice because i dont have to use date and hwclock manually)
I wouldn't rely on the initial time setting being valid; it's not guaranteed by specs for all the chips, as I recall.
Quoted 2 lines
i checked the
rtc_valid_tmfunction and saw that it also checks for the year value among others but inds1307_read_alarm(now this function is renamed tods1337_read_alarm) year is set to -1 with some other fields sortc_valid_tmfunction can not be used inds1337_read_alarmfunction.Right. I forget what alarm issues I was thinking of, but that wasn't it ... other code in the RTC framework handles morphing that time into a valid date.
Quoted 2 lines
This patch removes rtc date time values check in probe function
Thanks. This looks OK ... got a Signed-off-By for us?
Quoted 61 lines
drivers/rtc/rtc-ds1307.c +1
-1@@ -298,7 +298,7 @@ static int ds1307_set_time(struct device *dev, struct rtc_time *t)return 0 ; } -static int ds1307_read_alarm (struct device *dev,struct rtc_wkalrm *t) +static int ds1337_read_alarm (struct device *dev,struct rtc_wkalrm *t) {struct i2c_client *client =to_i2c_client (dev);struct ds1307 *ds1307 = i2c_get_clientdata(client);drivers/rtc/rtc-ds1307.c +1
-1@@ -353,7 +353,7 @@ static int ds1307_read_alarm(struct device *dev, struct rtc_wkalrm *t)return 0 ; } -static int ds1307_set_alarm (struct device *dev,struct rtc_wkalrm *t) +static int ds1337_set_alarm (struct device *dev,struct rtc_wkalrm *t) {struct i2c_client *client =to_i2c_client (dev);struct ds1307 *ds1307 = i2c_get_clientdata(client);drivers/rtc/rtc-ds1307.c +2
-18@@ -475,8 +475,8 @@ static int ds1307_ioctl(struct device *dev, unsigned int cmd, unsigned long arg)static const struct rtc_class_ops ds13xx_rtc_ops = { .read_time = ds1307_get_time, .set_time = ds1307_set_time, - .read_alarm = ds1307_read_alarm, - .set_alarm = ds1307_set_alarm, + .read_alarm = ds1337_read_alarm, + .set_alarm = ds1337_set_alarm, .ioctl = ds1307_ioctl, };@@ -707,22 +707,6 @@ read_rtc:break ; } - tmp = ds1307->regs[DS1307_REG_SECS ]; - tmp =bcd2bin (tmp &0x7f ); -if (tmp >60 ) - goto exit_bad; - tmp =bcd2bin (ds1307->regs[DS1307_REG_MIN ] &0x7f ); -if (tmp >60 ) - goto exit_bad; - - tmp =bcd2bin (ds1307->regs[DS1307_REG_MDAY ] &0x3f ); -if (tmp ==0 || tmp >31 ) - goto exit_bad; - - tmp =bcd2bin (ds1307->regs[DS1307_REG_MONTH ] &0x1f ); -if (tmp ==0 || tmp >12 ) - goto exit_bad; - tmp = ds1307->regs[DS1307_REG_HOUR ];switch (ds1307->type) {case ds_1340:@@ -779,13 +763,6 @@ read_rtc:return 0 ;
Quoted 17 lines
drivers/rtc/rtc-ds1307.c read_rtc: · line 768
Quoted 12 lines
drivers/rtc/rtc-ds1307.c +0
-7-exit_bad: -dev_dbg (&client->dev,"%s: %02x %02x %02x %02x %02x %02x %02x\n" , -"bogus register" , - ds1307->regs[0 ], ds1307->regs[1 ], - ds1307->regs[2 ], ds1307->regs[3 ], - ds1307->regs[4 ], ds1307->regs[5 ], - ds1307->regs[6 ]); exit_irq:if (ds1307->rtc)rtc_device_unregister (ds1307->rtc);
Thank you pointing to not rely on random values on RTC chip, i had luck when reinserted battery from/to RTC in random initial values (that gave me 1.1.2000).
JR