The timer-of API does not provide a function to undo what has been done by
the timer_of_init() function.

Add a timer_of_exit() function.
Signed-off-by: NBenjamin Gaignard <benjamin.gaignard@linaro.org>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

Always accessing the compare register via the CM redirect region is
(relatively) slow. If the timer being updated is the current CPUs
then this can be shortcutted by writing to the CM VP local region.
Signed-off-by: NMatt Redfearn <matt.redfearn@mips.com>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

The function gic_next_event is always called with interrupts disabled, so
the local_irq_save / local_irq_restore are pointless - remove them.

[Daniel Lezcano: Fixed warning by removing unused variable 'flags']
Signed-off-by: NMatt Redfearn <matt.redfearn@mips.com>
Suggested-by: NDaniel Lezcano <daniel.lezcano@linaro.org>
Reported-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

Our ctags mangling script can't handle newlines inside of a
DEFINE_PER_CPU(), leading to an annoying message whenever tags are
built:

  ctags: Warning: drivers/clocksource/arm_arch_timer.c:302: null expansion of name pattern "\1"

This was dealt with elsewhere in commit:

  25528213 ("tags: Fix DEFINE_PER_CPU expansions")

... by ensuring each DEFINE_PER_CPU() was contained on a single line,
even where this would violate the usual code style (checkpatch warnings
and all).

Let's do the same for the arch timer driver, and get rid of the
distraction.
Signed-off-by: NMark Rutland <mark.rutland@arm.com>
Cc: Daniel Lezcano <daniel.lezcano@linaro.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

The ACPI GTDT code validates the CNTFRQ field of each MMIO timer
frame against the CNTFRQ system register of the current CPU, to
ensure that they are equal, which is mandated by the architecture.

However, reading the CNTFRQ field of a frame is not possible until
the RFRQ bit in the frame's CNTACRn register is set, and doing so
before that willl produce the following error:

  arch_timer: [Firmware Bug]: CNTFRQ mismatch: frame @ 0x00000000e0be0000: (0x00000000), CPU: (0x0ee6b280)
  arch_timer: Disabling MMIO timers due to CNTFRQ mismatch
  arch_timer: Failed to initialize memory-mapped timer.

The reason is that the CNTFRQ field is RES0 if access is not enabled.

So move the validation of CNTFRQ into the loop that iterates over the
timers to find the best frame, but defer it until after we have selected
the best frame, which should also have enabled the RFRQ bit.
Signed-off-by: NArd Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: NMark Rutland <mark.rutland@arm.com>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

We regularly run into build errors when a clocksource driver selects
CONFIG_TIMER_OF while CONFIG_GENERIC_CLOCKEVENTS is disabled:

In file included from drivers/clocksource/timer-of.c:25:0:
drivers/clocksource/timer-of.h:35:28: error: field 'clkevt' has incomplete type

At the moment, three drivers can show this behavior: ARMV7M_SYSTICK,
CLKSRC_ST_LPC and CLKSRC_NPS. We could add further dependencies as we did
many times, but I have looked a little bit more at what architectures
are left that don't use GENERIC_CLOCKEVENTS, and this shows that there
is a better solution.

On arch/frv and arch/ia64, we never select CONFIG_GENERIC_CLOCKEVENTS
and we also don't use ARCH_USES_GETTIMEOFFSET, which would
block the clocksource Kconfig menu. On m68k, some platforms use
CONFIG_GENERIC_CLOCKEVENTS, some use ARCH_USES_GETTIMEOFFSET, and some
use neither of them. The good news is that there is no configuration that
does not set CONFIG_GENERIC_CLOCKEVENTS but that wants to enable any of
the Kconfig symbols in the menu, so we can simply replace the dependency
with the stricter one. While in theory one could have a clocksource
driver without the clockevent infrastructure, this seems unlikely
to be relevant in the future any more.

We can probably drop some of the other dependencies as well now,
e.g. there should generally be no reason to depend on CONFIG_ARM
unless the driver uses architecture specific assembly.
Reported-by: NRandy Dunlap <rdunlap@infradead.org>
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Reviewed-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

pr_err() messages should end with a new-line to avoid other messages being
concatenated.
Signed-off-by: NArvind Yadav <arvind.yadav.cs@gmail.com>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

pr_err() messages should end with a new-line to avoid other messages being
concatenated.
Signed-off-by: NArvind Yadav <arvind.yadav.cs@gmail.com>
Reviewed-by: NAndreas Färber <afaerber@suse.de>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

pr_err() messages should end with a new-line to avoid other messages being
concatenated.
Signed-off-by: NArvind Yadav <arvind.yadav.cs@gmail.com>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

Use the existing of_device_get_match_data() helper instead of
open-coding its functionality.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

The in-driver channel configuration in sh_cmt_info.channels_mask is now
always set for all CMT devices instantiated from DT.

Hence the "renesas,channels-mask" property is no longer checked, and its
handling can be removed, cfr. commit 4e18111f ("devicetree:
bindings: Remove deprecated properties").
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

Document in the driver that "renesas,cmt-48-gen2" is deprecated, but
still supported for backward compatibility with old DTBs, cfr. commit
4e18111f ("devicetree: bindings: Remove deprecated
properties").
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

Remove driver matching support for the unused "renesas,cmt-32" and
"renesas,cmt-32-fast" compatible values, cfr. commit 203bb347
("devicetree: bindings: Remove unused 32-bit CMT bindings").

As this removes the last user of SH_CMT_32BIT_FAST, all support for this
variant is removed from the driver.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

Add support for the new R-Car Gen2 CMT0 and CMT1 bindings. Support
for the old DT binding is still kept around, however devices using
such binding will be treated as a low-feature CMT0 device. If users
want to make use of CMT1-specific features then they need to update
their DTBs. No special CMT1-specific features are however implemented
by his patch, only DT bindings are redone as groundwork for future
feature patches.
Signed-off-by: NMagnus Damm <damm+renesas@opensource.se>
Acked-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

Always use 0x3f as channel mask for the SH_CMT_48BIT type of devices.
Once this patch is applied the "renesas,channels-mask" property will
be ignored by the driver for older devices matching SH_CMT_48BIT. In
the future when all CMT types store channel mask in the driver then
we will be able to deprecate and remove "renesas,channels-mask" from DTS.
Signed-off-by: NMagnus Damm <damm+renesas@opensource.se>
Acked-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

Make this const as it is only used during a copy operation and add
__initconst as this usage is during the initialization phase.
Signed-off-by: NBhumika Goyal <bhumirks@gmail.com>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: julia.lawall@lip6.fr
Cc: daniel.lezcano@linaro.org
Link: http://lkml.kernel.org/r/1504030631-10812-1-git-send-email-bhumirks@gmail.com

gcc-4.6 and older fail to inline integrator_clocksource_init, so they
end up showing a harmless warning:

WARNING: vmlinux.o(.text+0x4aa94c): Section mismatch in reference from the function integrator_clocksource_init() to the function .init.text:clocksource_mmio_init()
The function integrator_clocksource_init() references
the function __init clocksource_mmio_init().
This is often because integrator_clocksource_init lacks a __init
annotation or the annotation of clocksource_mmio_init is wrong.

Add the missing __init annotation that makes it build cleanly with all
compilers.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Thierry Reding <treding@nvidia.com>
Cc: Linus Walleij <linus.walleij@linaro.org>
Cc: Daniel Lezcano <daniel.lezcano@linaro.org>
Link: http://lkml.kernel.org/r/20170915194310.1170514-1-arnd@arndb.de

Now that we have a custom printf format specifier, convert users of
full_name to use %pOF instead. This is preparation to remove storing
of the full path string for each node.
Signed-off-by: NRob Herring <robh@kernel.org>
Cc: Daniel Lezcano <daniel.lezcano@linaro.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Marc Gonzalez <marc_gonzalez@sigmadesigns.com>
Cc: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Cc: Alexandre Torgue <alexandre.torgue@st.com>
Cc: linux-arm-kernel@lists.infradead.org
Acked-by: NMarc Gonzalez <marc_gonzalez@sigmadesigns.com>
Acked-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

Switch from calling functions exported by the GIC interrupt controller
to using new accessors provided by asm/mips-gic.h. This will allow the
counter-handling functionality to be removed from the interrupt
controller driver, where it doesn't really belong, and also allow for
inlining of the accesses to the GIC.
Signed-off-by: NPaul Burton <paul.burton@imgtec.com>
Acked-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Daniel Lezcano <daniel.lezcano@linaro.org>
Cc: Jason Cooper <jason@lakedaemon.net>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/17021/Signed-off-by: NRalf Baechle <ralf@linux-mips.org>

The bcm2835_timer_init() function emits an error message in case of a memory
allocation failure. This is pointless as the mm core does that already.

Remove this message.

This issue was detected by using the Coccinelle software.
Signed-off-by: NMarkus Elfring <elfring@users.sourceforge.net>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

IMX Timer/PWM Module (TPM) supports both timer and pwm function while
this patch only adds the timer support. PWM would be added later.

The TPM counter, compare and capture registers are clocked by an
asynchronous clock that can remain enabled in low power modes.

NOTE: We observed in a very small probability, the bus fabric
contention between GPU and A7 may results a few cycles delay
of writing CNT registers which may cause the min_delta event got
missed, so we need add a ETIME check here in case it happened.

Cc: Daniel Lezcano <daniel.lezcano@linaro.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Shawn Guo <shawnguo@kernel.org>
Cc: Anson Huang <Anson.Huang@nxp.com>
Cc: Bai Ping <ping.bai@nxp.com>
Signed-off-by: NDong Aisheng <aisheng.dong@nxp.com>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

In v4.13, CLKSRC_PISTACHIO can select TIMER_OF on architectures without
GENERIC_CLOCKEVENTS, resulting in a struct clock_event_device missing
some required features and build breakage compiling timer_of.c. One of
the symbols selecting TIMER_OF is CLKSRC_PISTACHIO, so add the
dependency on GENERIC_CLOCKEVENTS.

Thanks to kbuild test robot for finding this error
(https://lkml.org/lkml/2017/7/16/249)
Signed-off-by: NMatt Redfearn <matt.redfearn@imgtec.com>
Suggested-by: NIan Abbott <abbotti@mev.co.uk>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

The current code checks the return value of the of_io_request_and_map()
function as it was returning a NULL pointer in case of error.

However, it returns an error code encoded in the pointer return value, not a
NULL value. Fix this by checking the returned pointer against IS_ERR() and
return the error with PTR_ERR().
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

Propagate the return values of platform_get_irq and devm_request_irq on
failure.

Cc: Frans Klaver <fransklaver@gmail.com>
Signed-off-by: NGustavo A. R. Silva <garsilva@embeddedor.com>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

The loop to find the best memory frame in arch_timer_mem_acpi_init()
initializes the loop counter with itself ('i = i'), which is suspicious
in the first place and pointed out by clang. The loop condition is
'i < timer_count' and a prior for loop exits when 'i' reaches
'timer_count', therefore the second loop is never executed.

Initialize the loop counter with 0 to iterate over all timers, which
supposedly was the intention before the typo monster attacked.

Fixes: c2743a36 ("clocksource: arm_arch_timer: add GTDT support for memory-mapped timer")
Signed-off-by: NMatthias Kaehlcke <mka@chromium.org>
Reported-by: NArd Biesheuvel <ard.biesheuvel@linaro.org>
Acked-by: NMark Rutland <mark.rutland@arm.com>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

Use the new static_branch_enable_cpuslocked() function to switch
the workaround static key on the CPU hotplug path.
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-arm-kernel@lists.infradead.org
Link: http://lkml.kernel.org/r/20170801080257.5056-5-marc.zyngier@arm.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

of_irq_get_byname() may return a negative error number as well as 0 on
failure, while timer_irq_init() only checks for 0, blithely continuing with
the call to request_[percpu_]irq() -- those functions expect *unsigned int*,
so would probably fail anyway when a large IRQ number resulting from a
conversion of a negative error number is passed to them... This, however,
is incorrect behavior -- error number is not IRQ number.

Filter out the negative error numbers, complain, and return them to the
timer_irq_init()'s callers...

Fixes: dc11bae7 ("clocksource/drivers: Add timer-of common init routine")
Signed-off-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Daniel Lezcano <daniel.lezcano@linaro.org>
Link: http://lkml.kernel.org/r/20170717180114.678825147@cogentembedded.com

'clk' is a valid pointer at this point. So calling PTR_ERR on it is
pointess.

Return the error code from 'clk_prepare_enable()' if it fails instead.
Signed-off-by: NChristophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

In case of error at init time, rollback iomapping.
Signed-off-by: NArvind Yadav <arvind.yadav.cs@gmail.com>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

Now that AVR32 is gone, we can use the proper IO accessors that are
correctly handling endianness.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@free-electrons.com>
Acked-by: NNicolas Ferre <nicolas.ferre@microchip.com>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

Previously a framework to factor out the drivers init function has been
merged.

Use this common framework in this driver, we get:

Before:

text    data     bss     dec   hex filename
1787     384      12    2183   887 drivers/clocksource/sun4i_timer.o

After:

text    data     bss     dec   hex filename
1407     512       0    1919   77f drivers/clocksource/sun4i_timer.o
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>
Tested-by: NChen-Yu Tsai <wens@csie.org>

A typo in the code checks the return value of iomap against !NULL
and, thus, fails everytime the mapping succeed.

Fix this by inverting the condition in the check.
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

If none of the flags are set, 'ret' is uninitialized as pointed out
by gcc:

drivers/clocksource/timer-of.c: In function 'timer_of_init':
drivers/clocksource/timer-of.c:160:9: error: 'ret' may be used uninitialized in this function [-Werror=maybe-uninitialized]

Since calling the function without any of the flags is an error,
set the return value to -EINVAL for that case.

[ tglx: Get rid of the silly backwards goto while at it ]

Fixes: dc11bae7 ("clocksource/drivers: Add timer-of common init routine")
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Daniel Lezcano <daniel.lezcano@linaro.org>
Link: http://lkml.kernel.org/r/20170621215005.3870011-1-arnd@arndb.de

The Actions Semi S900 SoC provides four 32-bit timers, TIMER0/1/2/3,
but no 2Hz timers.

An S900 datasheet can be found in 96Boards documentation:
https://github.com/96boards/documentation/blob/master/ConsumerEdition/Bubblegum-96/HardwareDocs/SoC_bubblegum96.pdfAcked-by: NDaniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: NAndreas Färber <afaerber@suse.de>

The Actions Semi S500 SoC provides four timers, 2Hz0/1 and 32-bit TIMER0/1.

Use TIMER0 as clocksource and TIMER1 as clockevents.

Based on LeMaker linux-actions tree.

An S500 datasheet can be found on the LeMaker Guitar pages:
http://www.lemaker.org/product-guitar-download-29.htmlAcked-by: NDaniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: NAndreas Färber <afaerber@suse.de>

The sched_clock() and delay timer callbacks can just call
each other and we can save an #ifdef.
Suggested-by: NDaniel Lezcano <daniel.lezcano@linaro.org>
Cc: Andrew Jeffery <andrew@aj.id.au>
Cc: Joel Stanley <joel@jms.id.au>
Cc: Jonas Jensen <jonas.jensen@gmail.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

This timer is often used on the ARM architecture, so as with so
many siblings, we can implement delay timers, removing the need
for the system to calibrate jiffys at boot, and potentially
handling CPU frequency scaling on targets.

We cannot just protect the Kconfig with a "depends on ARM" because
it is already known that different architectures are using Faraday
IP blocks, so it is better to make things open-ended and use

Result on boot dmesg:

Switching to timer-based delay loop, resolution 40n
Calibrating delay loop (skipped), value calculated using
  timer frequency.. 50.00 BogoMIPS (lpj=250000)

This is accurately the timer frequency, 250MHz on the APB
bus.

Cc: Andrew Jeffery <andrew@aj.id.au>
Cc: Joel Stanley <joel@jms.id.au>
Cc: Jonas Jensen <jonas.jensen@gmail.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>
Tested-by: NJonas Jensen <jonas.jensen@gmail.com>
Tested-by: NAndrew Jeffery <andrew@aj.id.au>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

The different drivers are all using the same pattern when initializing.

 1. Get the base address
 2. Get the irq number
 3. Get the clock
 4. Prepare and enable the clock
 5. Get the rate
 6. Request an interrupt

Instead of repeating again and again these steps in all the drivers, let's
provide a common init routine to give the opportunity to factor all of them
out.

We can expect a significant kernel size improvement when the common routine
will be used in all the drivers.
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

On sama5d2, power to the core may be cut while entering suspend mode. It is
necessary to save and restore the TCB registers.
Signed-off-by: NAlexandre Belloni <alexandre.belloni@free-electrons.com>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>

The sched_clock() call should be really fast so we want to
avoid an extra if() clause on the read path if possible.

Implement two sched_clock_read() functions, one if the timer
counts up and one if it counts down. Incidentally this also
mirrors how clocksource_mmio_init() works and make things
simple and easy to understand.
Suggested-by: NDaniel Lezcano <daniel.lezcano@linaro.org>
Cc: Andrew Jeffery <andrew@aj.id.au>
Cc: Joel Stanley <joel@jms.id.au>
Cc: Jonas Jensen <jonas.jensen@gmail.com>
Signed-off-by: NLinus Walleij <linus.walleij@linaro.org>
Signed-off-by: NDaniel Lezcano <daniel.lezcano@linaro.org>