Instead of confusingly returning -EBUSY on failure to obtain an
interrupt, propagate the real error code. While at it, let the user know
why the interrupt could not be acquired.
Signed-off-by: NThierry Reding <treding@nvidia.com>
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

Fix a problem with commit 311ee9c1 ("rtc: cmos: allow using ACPI for
RTC alarm instead of HPET") defining `use_acpi_alarm' module parameter
even for non-ACPI platforms, which ignore it.  Wrap the definition into
#ifdef CONFIG_ACPI and use a static inline wrapper function, hardcoded
to return 0 and consequently optimized away for !ACPI, following the
existing pattern with HPET handling functions.
Signed-off-by: NMaciej W. Rozycki <macro@linux-mips.org>
Fixes: 311ee9c1 ("rtc: cmos: allow using ACPI for RTC alarm instead of HPET")
Cc: stable@vger.kernel.org # 4.18+
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

Fix a commit 311ee9c1 ("rtc: cmos: allow using ACPI for RTC alarm
instead of HPET") `rtc-cmos' regression causing a link error:

drivers/rtc/rtc-cmos.o: In function `cmos_platform_probe':
rtc-cmos.c:(.init.text+0x33c): undefined reference to `hpet_rtc_interrupt'
rtc-cmos.c:(.init.text+0x3f4): undefined reference to `hpet_rtc_interrupt'

with non-ACPI platforms using this driver.  The cause is the change of
the condition guarding the use of `hpet_rtc_interrupt'.

Previously it was a call to `is_hpet_enabled'.  That function is static
inline and has a hardcoded 0 result for non-ACPI platforms, which imply
!HPET_EMULATE_RTC.  Consequently the compiler optimized the whole block
away including the reference to `hpet_rtc_interrupt', which never made
it to the link stage.

Now the guarding condition is a call to `use_hpet_alarm', which is not
static inline and therefore the compiler may not be able to prove that
it actually always returns 0 for non-ACPI platforms.  Consequently the
build breaks with an unsatisfied reference, because `hpet_rtc_interrupt'
is nowhere defined at link time.

Fix the problem by marking `use_hpet_alarm' inline.  As the `inline'
keyword serves as an optimization hint rather than a requirement the
compiler is still free to choose whether inlining will be beneficial or
not for ACPI platforms.
Signed-off-by: NMaciej W. Rozycki <macro@linux-mips.org>
Fixes: 311ee9c1 ("rtc: cmos: allow using ACPI for RTC alarm instead of HPET")
Cc: stable@vger.kernel.org # 4.18+
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

Instead of lying to the core when the alarm is invalid, let it handle that
by returning the error.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>
Tested-by: Gregory CLEMENT <gregory.clement@bootlin.com>  (on Armada 375 DB)
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

Call the 64bit versions of rtc_time_to_tm now that the range is enforced by
the core.
Also remove the open coded rtc_tm_to_time64.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

rtc_time_to_tm always rturns a valid tm, there is no need to validate it.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>
Acked-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

Let the core handle offsetting and windowing the RTC range.
The RTC has a 24 bit counter for minutes plus a seconds counter.
Keep the epoch at the beginning of 2000.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>
Acked-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

Use rtc_add_group to add the sysfs group in a race free manner.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>
Reviewed-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

Instead of resetting the RTC to an bogus valid time, let userspace know
that the time is invalid when XSTP is set. Reset XSTP when setting the time
again.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

The name passed to devm_rtc_device_register is now unused. anyway, switch
to devm_rtc_allocate_device to avoid forward declaring rs5c348_driver.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

Since commit 8fb1ecb3 ("rtc: rtc-rs5c348: use devm_*() functions") the
kfree_exit label simply returns ret.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

Call the 64bit versions of rtc_time_to_tm and rtc_tm_to_time now that the
range is enforced by the core.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

The RTC is a 32bit seconds counter.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>
Tested-by: NGregory CLEMENT <gregory.clement@bootlin.com>
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

The IRQ is requested before the struct rtc is allocated and registered, but
this struct is used in the IRQ handler. This may lead to a NULL pointer
dereference.

Switch to devm_rtc_allocate_device/rtc_register_device to allocate the rtc
before requesting the IRQ.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

rtc_time64_to_tm has not been called rtc_time_to_tm64
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

Register frequency test using rtc_add_group to avoid a possible race
condition and simplify the code.

This also moves the attribute to its proper location under the rtc device
instead of the i2c parent device.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

Use rtc_add_group to add the sysfs group in a race free manner.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

Instead of adding a binary sysfs attribute from the driver, use the core to
register an nvmem device.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

/proc is not the correct ABI to display debugging info. Remove
RTC_DS1685_PROC_REGS as the driver hasn't seen any real development since
it was included.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

Replace the license boilerplate by an SPDX identifier
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

The license text is specifying GPL v2 only but the MODULE_LICENSE is set to
GPL which means GNU Public License v2 or later. When MODULE_LICENSE and
boiler plate does not match, go for boiler plate license.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

The default word_size and stride of 1 are correct for the tx4939. Also fix
the nvmem folder name.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

The core header file is reserved for the core, stop including it.

Also reorder includes alphabetically.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

Move rtc_add_group and rtc_add_groups definition to rtc.h that is available
for all RTC drivers.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

A few entries are not placed correctly, reorder them.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

On m41txx you can enable open-drain OUT pin to check if offset is ok.
Enabling OUT pin with frequency_test_enable attribute, OUT pin will tick
512 times faster than 1s tick base.

Enable or Disable FT bit on CONTROL register if freq_test is 1 or 0.
Signed-off-by: NGiulio Benetti <giulio.benetti@micronovasrl.com>
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

m41txx chips can hold a calibration value to get correct clock bias.

Add offset handling (ranging between -63ppm and 126ppm) via sysfs.
Signed-off-by: NGiulio Benetti <giulio.benetti@micronovasrl.com>
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

Use rtc_add_group to add the common sysfs group to avoid a possible race
condition.

[Denis.Osterland@diehl.com: use to_i2c_client(dev->parent)]
Signed-off-by: NDenis Osterland <Denis.Osterland@diehl.com>
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

The move of atrim, dtrim usr sysfs properties from i2c device
to rtc device require to access them via dev->parent.
This patch also aligns timestamp0.

Stale mentions of irq_task are left in the kerneldoc after its removal.
Remove them.

There is still one indirect mention left but commit 3c8bb90e ("rtc: Fix
hrtimer deadlock") can probably be reverted now.
Reported-by: NLinus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

Ensure the non managed version of the un/registration functions is not used
anymore. No driver is using it anymore and they should not be necessary.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

Immutable branch between MFD and RTC due for the v4.20 merge window
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

The abx804 and abx805 chips have support for a simple watchdog
function that can trigger an external reset.
Signed-off-by: NJeremy Gebben <jgebben@sweptlaser.com>
Reviewed-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

This will allow additional data to be tracked, for future
improvements.
Signed-off-by: NJeremy Gebben <jgebben@sweptlaser.com>
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

The IRQ work is added before the struct rtc is allocated and registered,
but this struct is used in the IRQ handler. This may lead to a NULL pointer
dereference.

Switch to devm_rtc_allocate_device/rtc_register_device to allocate the rtc
before calling menelaus_add_irq_work.

Also, this solves a possible leak as the RTC is never released.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: NLee Jones <lee.jones@linaro.org>

Switch to devm_rtc_allocate_device to simplify the erro and driver removal
paths.
Acked-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

The IRQ is requested before the struct rtc is allocated and registered, but
this struct is used in the IRQ handler. This may lead to a NULL pointer
dereference.

Switch to devm_rtc_allocate_device/rtc_register_device to allocate the rtc
before requesting the IRQ.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

The IRQ is requested before the struct rtc is allocated and registered, but
this struct is used in the IRQ handler. This may lead to a NULL pointer
dereference.

Switch to devm_rtc_allocate_device/rtc_register_device to allocate the rtc
before requesting the IRQ.
Acked-by: NEddie Huang <eddie.huang@mediatek.com>
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

The old rtc driver is getting in the way of some compat_ioctl
simplification. Looking up the loongson64 git history, it seems
that everyone uses the more modern but compatible RTC_CMOS driver
anyway, so let's remove the special case for loongson64.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

Fixes gcc '-Wunused-but-set-variable' warning:

drivers/rtc/rtc-mrst.c: In function 'mrst_procfs':
drivers/rtc/rtc-mrst.c:264:29: warning:
 variable 'valid' set but not used [-Wunused-but-set-variable]
  unsigned char rtc_control, valid;
Signed-off-by: NYueHaibing <yuehaibing@huawei.com>
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>

The Epson rx8803 is mostly similar to the Microcrystal RV8803 but the size
of the offset register is 4 bits vs 6 bits but it has a configurable
temperature compensation.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@bootlin.com>