Power-on reset after the insertion of a battery does not always complete
successfully, leading to corrupted register content. The EXT_TEST bit
will stop the clock from running, but currently the driver will never
recover.
Safely handle the erroneous state by clearing EXT_TEST as part of the
usual set_time method.
Signed-off-by: Phil Elwell <phil@raspberrypi.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Link: https://lore.kernel.org/r/20211015111208.1757110-1-phil@raspberrypi.com
TQ-Systems' TQMa8Mx module (SoM) uses a pcf85063 as RTC. The default output
is 32768Hz. This is to provide the i.MX8M CKIL clock. Once the RTC driver
is probed, the clock is disabled and all i.MX8M functionality depending on
the 32 KHz clock will halt. In our case the whole system halts and a power
cycle is required.
Referencing the pcf85063 directly results in a deadlock. The kernel
will see, that i.MX8M system clock needs the RTC clock and do probe
deferral. But the i.MX8M I2C module never becomes usable without the
i.MX8M CKIL clock and thus the RTC's clock will not be probed. So
from the kernel's perspective this is a chicken-and-egg problem.
Technically everything is fine by not touching anything, since
the RTC clock correctly enables the clock on reset (i.e. on
battery backup power loss).
A workaround for this issue is describing the square wave pin
as fixed-clock, which is registered early and basically how
this pin is used on the i.MX8M.
This addresses the exact same issue as in commit f765e349c3 ("rtc:
m41t80: add support for fixed clock").
Signed-off-by: Alexander Stein <alexander.stein@ew.tq-group.com>
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
[Fixed return value 0 -> NULL]
Link: https://lore.kernel.org/r/20211013074954.997445-1-alexander.stein@ew.tq-group.com
The rtc device node is always NULL.
Since v5.12-rc1-dontuse/3c9ea42802a1fbf7ef29660ff8c6e526c58114f6 this
will lead to a NULL pointer dereference.
To fix this use the parent node which is the i2c client node as set by
devm_rtc_allocate_device().
Using the i2c client node seems to be what other similar drivers do
e.g. rtc-pcf8563.c.
Signed-off-by: Francois Gervais <fgervais@distech-controls.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Link: https://lore.kernel.org/r/20210310211026.27299-1-fgervais@distech-controls.com
Move the alarm callbacks in pcf85063_rtc_ops and use RTC_FEATURE_ALARM to
signal to the core whether alarms are available instead of having a
supplementary struct rtc_class_ops without alarm callbacks.
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Link: https://lore.kernel.org/r/20210110231752.1418816-9-alexandre.belloni@bootlin.com
rtc_register_device() is a managed interface but it doesn't use devres
by itself - instead it marks an rtc_device as "registered" and the devres
callback for devm_rtc_allocate_device() takes care of resource release.
This doesn't correspond with the design behind devres where managed
structures should not be aware of being managed. The correct solution
here is to register a separate devres callback for unregistering the
device.
While at it: rename rtc_register_device() to devm_rtc_register_device()
and add it to the list of managed interfaces in devres.rst. This way we
can avoid any potential confusion of driver developers who may expect
there to exist a corresponding unregister function.
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Link: https://lore.kernel.org/r/20201109163409.24301-8-brgl@bgdev.pl
rtc_nvmem_register() is a managed interface. It doesn't require any
release function to be called at driver detach. To avoid confusing
driver authors, let's rename it to devm_rtc_nvmem_register() and add it
to the list of managed interfaces in Documentation/.
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Link: https://lore.kernel.org/r/20201109163409.24301-6-brgl@bgdev.pl
Rationale:
Reduces attack surface on kernel devs opening the links for MITM
as HTTPS traffic is much harder to manipulate.
Deterministic algorithm:
For each file:
If not .svg:
For each line:
If doesn't contain `\bxmlns\b`:
For each link, `\bhttp://[^# \t\r\n]*(?:\w|/)`:
If both the HTTP and HTTPS versions
return 200 OK and serve the same content:
Replace HTTP with HTTPS.
Signed-off-by: Alexander A. Klimov <grandmaster@al2klimov.de>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Link: https://lore.kernel.org/r/20200706062727.18481-1-grandmaster@al2klimov.de
Remove RTC_VL_CLR handling because it is a disservice to userspace as it
removes the important information that the RTC data is invalid. This may
lead userspace to set an invalid system time later on.
Link: https://lore.kernel.org/r/20191214220259.621996-11-alexandre.belloni@bootlin.com
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Allow reading the oscillator status bit. Also allow clearing it even if
that makes little sense and can't be done in a race free way.
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
The PCF850363 has an offset correction with two modes:
With mode 0, the correction is triggered once every two hours and then
correction pulses are applied once per minute until the programmed
correction values have been implemented. This gives a step of 4.34 ppm.
With mode 1, the correction is triggered once every four minutes and then
correction pulses are applied once per second up to a maximum of 60 pulses.
When correction values greater than 60 pulses are used, additional
correction pulses are made in the 59 th second. This gives a step of 4.069
ppm.
Use the correction closest to the requested value.
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
The Micro Crystal RV8263 has the same IC as the pcf85063 but has an on
board crystal. This means that the CAP_SEL bit has to be cleared so the
correct capacitance is selected for the crystal.
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Add support for the alarms. The match on the weekday is not used as it it
not necessarily properly set.
The tested RTC shows a behaviour where setting an alarm on the second right
after an alarm that fired is not working, probably because of the circuit
that ensures an alarm only fires once. This is why uie_unsupported is set.
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
As stated in a comment pcf85063a and pcf85063tp don't have the same number
of registers. Especially, pcf85063tp doesn't have alarm support.
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Switch to regmap to simplify register accesses and remove the need for
pcf85063_stop_clock/pcf85063_start_clock.
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Only smbus reads and write are done in the driver, plain i2c functionality
is not required.
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Some of defines are not in use since
7b5768486a. Remove it to make the code
easier to read and understand.
Signed-off-by: Alexey Roslyakov <alexey.roslyakov@gmail.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Add support for specifying the xtal load capacitance in the DT node.
The pcf85063 supports xtal load capacitance of 7pF or 12.5pF.
If the rtc has the wrong configuration the time will
drift several hours/week.
The driver use the default value 7pF.
The DT may specify either 7000fF or 12500fF.
(The DT uses femto Farad to avoid decimal numbers).
Other values are warned and the driver uses the default value.
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Cc: Alexandre Belloni <alexandre.belloni@bootlin.com>
Cc: Urs Fässler <urs.fassler@bbv.ch>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Fix a bug that caused the Control_1 register to get zeroed whenever the
RTC time is set. The problem occurred between stopping and starting the
RTC clock, wherein the return value of a successful I2C write function
would get written to the register.
Also update variables of the start and stop functions to be more
consistent with the rest of the driver.
Signed-off-by: Alvin Šipraga <alvin@airtame.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Bit clear operation was missing ~
Signed-off-by: Michael McCormick <michael.mccormick@enatel.net>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
pcf85063_get_datetime and pcf85063_set_datetime are only used after casting
dev to an i2c_client. Remove that useless indirection.
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
The RTC core is always calling rtc_valid_tm after the read_time callback.
It is not necessary to call it before returning from the callback.
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Add a sanity check to see if chip is present. If we can not communicate
with the chip there is no point in registering a RTC device.
Signed-off-by: Mirza Krak <mirza.krak@hostmobility.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
The current rtc-pcf85063 driver only supports the PCF85063TP device.
Using the existing driver on a PCF85063A will result in the time being
set correctly into the RTC, but the RTC is held in the stopped state.
Therefore, the time will no longer advance and no error is indicated.
The PCF85063A device has a bigger memory map than the PCF85063TP.
The existing driver make use of an address rollover condition,
but the rollover point is different in the two devices.
Signed-off-by: Chris DeBruin <cdeb5783@gmail.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
The year range is not validated properly
As the driver has been mainlined in 2014, it is not an issue to stop
handling dates between 1970 and 2000 with the benefit of handling dates up
to 2100.
Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
No members of struct pcf85063 are used anymore, remove the whole structure.
Reviewed-by: Juergen Borleis <jbe@pengutronix.de>
Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
pcf85063_get_datetime() tries to handle a century bit but that bit is not
documented and the final value is never used anywhere else in the kernel.
Reviewed-by: Juergen Borleis <jbe@pengutronix.de>
Tested-by: Ulrich Ölmann <u.oelmann@pengutronix.de>
Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
When setting a new time/date the RTC's clock must be stopped first, in
order to write the time/date registers in an atomic manner.
So, this change stops the clock first and then writes the time/date
registers and the clock control register (to re-enable the clock) in one
turn.
Signed-off-by: Juergen Borleis <jbe@pengutronix.de>
Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
Check if the RTC signals an invalid time/date (due to a battery power loss
for example). In this case ignore the time/date until it is really set again.
Signed-off-by: Juergen Borleis <jbe@pengutronix.de>
Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
By using i2c_smbus_read_i2c_block_data() the code is now much simpler.
While at it: when reading the RTC's seconds register, all time/date registers
are frozen until the RTC's year register is read. So it is important to read
all time/date registers in one turn to not lose a second event. Make it more
clear why the read must happen in this way.
Signed-off-by: Juergen Borleis <jbe@pengutronix.de>
Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
Return an error when the date is invalid as the policy should be
implemented there.
Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
i2c_driver does not need to set an owner because i2c_register_driver()
will set it.
Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
Alexandre Belloni
Phil Elwell
Alexander Stein
Fabio Estevam
Francois Gervais
Bartosz Golaszewski
Gustavo A. R. Silva
Alexander A. Klimov
Michael McCormick
Alexey Roslyakov
Sam Ravnborg
Alvin Šipraga
Mirza Krak
Chris DeBruin
Alexandre Belloni
Juergen Borleis
Krzysztof Kozlowski
Søren Andersen