09 Aug, 2014
2 commits
-
Pull ARM SoC driver changes from Olof Johansson:
"A handful of driver-related changes. We've had a bunch of them going
in through other branches as well, so it's only a part of what we
really have this release.Larger pieces are:
- Removal of a now unused PWM driver for atmel
[ This includes AVR32 changes that have been appropriately acked ]
- Performance counter support for the arm CCN interconnect
- OMAP mailbox driver cleanups and consolidation
- PCI and SATA PHY drivers for SPEAr 13xx platforms
- Redefinition (with backwards compatibility!) of PCI DT bindings for
Tegra to better model regulators/power"Note: this merge also fixes up the semantic conflict with the new
calling convention for devm_phy_create(), see commit f0ed817638b5 ("phy:
core: Let node ptr of PHY point to PHY and not of PHY provider") that
came in through Greg's USB tree.Semantic merge patch by Stephen Rothwell through
the next tree.* tag 'drivers-for-3.17' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc: (38 commits)
bus: arm-ccn: Fix error handling at event allocation
mailbox/omap: add a parent structure for every IP instance
mailbox/omap: remove the private mailbox structure
mailbox/omap: consolidate OMAP mailbox driver
mailbox/omap: simplify the fifo assignment by using macros
mailbox/omap: remove omap_mbox_type_t from mailbox ops
mailbox/omap: remove OMAP1 mailbox driver
mailbox/omap: use devm_* interfaces
bus: ARM CCN: add PERF_EVENTS dependency
bus: ARM CCN PMU driver
PCI: spear: Remove spear13xx_pcie_remove()
PCI: spear: Fix Section mismatch compilation warning for probe()
ARM: tegra: Remove legacy PCIe power supply properties
PCI: tegra: Remove deprecated power supply properties
PCI: tegra: Implement accurate power supply scheme
ARM: SPEAr13xx: Update defconfigs
ARM: SPEAr13xx: Add pcie and miphy DT nodes
ARM: SPEAr13xx: Add bindings and dt node for misc block
ARM: SPEAr13xx: Fix static mapping table
phy: Add drivers for PCIe and SATA phy on SPEAr13xx
... -
Pull ARM SoC device-tree changes from Olof Johansson:
"Unlike the board branch, this keeps having large sets of changes for
every release, but that's quite expected and is so far working well.Most of this is plumbing for various device bindings and new
platforms, but there's also a bit of cleanup and code removal for
things that are moved from platform code to DT contents (some OMAP
clock code in particular).There's also a pinctrl driver for tegra here (appropriately acked),
that's introduced this way to make it more bisectable.I'm happy to say that there were no conflicts at all with this branch
this release, which means that changes are flowing through our tree as
expected instead of merged through driver maintainers (or at least not
done with conflicts).There are several new boards added, and a couple of SoCs. In no
particular order:- Rockchip RK3288 SoC support, including DTS for a dev board that
they have seeded with some community developers.
- Better support for Hardkernel Exynos4-based ODROID boards.
- CCF conversions (and dtsi contents) for several Renesas platforms.
- Gumstix Pepper (TI AM335x) board support
- TI eval board support for AM437x
- Allwinner A23 SoC, very similar to existing ones which mostly has
resulted in DT changes for support. Also includes support for an
Ippo tablet with the chipset.
- Allwinner A31 Hummingbird board support, not to be confused with
the SolidRun i.MX-based Hummingboard.
- Tegra30 Apalis board support"* tag 'dt-for-3.17' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc: (334 commits)
ARM: dts: Enable USB host0 (EHCI) on rk3288-evb
ARM: dts: add rk3288 ehci usb devices
ARM: dts: Turn on USB host vbus on rk3288-evb
ARM: tegra: apalis t30: fix device tree compatible node
ARM: tegra: paz00: Fix some indentation inconsistencies
ARM: zynq: DT: Clarify Xilinx Zynq platform
ARM: dts: rockchip: add watchdog node
ARM: dts: rockchip: remove pinctrl setting from radxarock uart2
ARM: dts: Add missing pinctrl for uart0/1 for exynos3250
ARM: dts: Remove duplicate 'interrput-parent' property for exynos3250
ARM: dts: Add TMU dt node to monitor the temperature for exynos3250
ARM: dts: Specify MAX77686 pmic interrupt for exynos5250-smdk5250
ARM: dts: cypress,cyapa trackpad is exynos5250-Snow only
ARM: dts: max77686 is exynos5250-snow only
ARM: zynq: DT: Remove DMA from board DTs
ARM: zynq: DT: Add CAN node
ARM: EXYNOS: Add exynos5260 PMU compatible string to DT match table
ARM: dts: Add PMU DT node for exynos5260 SoC
ARM: EXYNOS: Add support for Exynos5410 PMU
ARM: dts: Add PMU to exynos5410
...
08 Aug, 2014
1 commit
-
Pull pinctrl updates from Linus Walleij:
"This is the bulk pin control changes for the v3.17 merge development
cycle:- get rid of the .disable() callback from the driver callback vtable.
This callback was abused and counterintuitive since a pin or group
of pins can be said to always be in some setting, and never really
disabled. We now only enable a certain muxing, and move between
some certain muxings, we never "disable" a mux setting- some janitorial moving the MSM, Samsung and Nomadik and drivers to
their own subdirectories for a clearer view in the subsystem. This
will continue- kill off the use of the return value from gpiochip_remove(), this
will be done in parallel in the GPIO subsystem and hopefully not
trigger too many unchecked return value warnings before we get rid
of this altogether- a huge set of changes and improvements to the Allwinner sunxi
drivers especially for their latest A23 and A31 SoCs, and some
ground work for the new sun8i platform family- a large set of Rockchip driver improvements adding support for the
RK3288 SoC- advances in migration of older Freescale platforms to pin control,
especially i.MX1- Samsung and Exynos improvements
- support for the Qualcomm MSM8960 SoC
- use the gpiolib irqchip helpers for the ST SPEAr and Intel Baytrail
drivers- a bunch of nice janitorial work done with cppcheck"
* tag 'pinctrl-v3.17-1' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl: (61 commits)
pinctrl: baytrail: Convert to use gpiolib irqchip
pinctrl: sunxi: number gpio ranges starting from 0
pinctrl: sunxi: use gpiolib API to mark a GPIO used as an IRQ
pinctrl: rockchip: add drive-strength control for rk3288
pinctrl: rockchip: add separate type for rk3288
pinctrl: rockchip: set is_generic in pinconf_ops
pinctrl: msm: drop negativity check on unsigned value
pinctrl: remove all usage of gpio_remove ret val in driver/pinctl
pinctrl: qcom: Make muxing of gpio function explicit
pinctrl: nomadik: move all Nomadik drivers to subdir
pinctrl: samsung: Group all drivers in a sub-dir
sh-pfc: sh73a0: Introduce the use of devm_regulator_register
sh-pfc: Add renesas,pfc-r8a7791 to binding documentation
pinctrl: msm: move all qualcomm drivers to subdir
pinctrl: msm: Add msm8960 definitions
pinctrl: samsung: Allow pin value to be initialized using pinfunc
pinctrl: samsung: Allow grouping multiple pinmux/pinconf nodes
pinctrl: exynos: Consolidate irq_chips of GPIO and WKUP EINTs
pinctrl: samsung: Handle GPIO request and free using pinctrl helpers
pinctrl: samsung: Decouple direction setting from pinctrl
...
28 Jul, 2014
3 commits
-
Instead of open-coding irqchip handling in the driver we can take advantage
of the new irqchip helpers provided by the gpiolib core.While doing this we also make sure that we call gpiochip_irqchip_add()
after the gpiochip itself is registered as required.Signed-off-by: Mika Westerberg
Signed-off-by: Linus Walleij -
The pinctrl-sunxi driver originally used the pin number as the gpio
range offset. This resulted in large, bogus gpio numbers for the
new sun6i-a31-r pinctrl devices.This patch makes the driver number the gpios ranges starting from an
offset of 0, by subtracting the pin_base number from the pin number.
This also makes the system-wide gpio number match the pin number.Tested on sun8i with sysfs exported gpios.
This patch also changes the GPIO bindings for R_PIO:
gpios = ;
Where B originally was the pinbank label (L or M) counted from A,
with this patch it becomes (L or M) counted from its pinbank base (L).Thus
gpios = ; /* PL11 */
becomes
gpios = ; /* PL11 */
IMO this is correct, as the binding shows the bank offset and pin offset
within the bank for the GPIO controller. But I'm worried it might be a
bit confusing.Signed-off-by: Chen-Yu Tsai
Acked-by: Maxime Ripard
Signed-off-by: Linus Walleij -
When an IRQ is started on a GPIO line, mark this GPIO as IRQ in
the gpiolib so we can keep track of the usage centrally.Signed-off-by: Chen-Yu Tsai
Acked-by: Maxime Ripard
Signed-off-by: Linus Walleij
23 Jul, 2014
4 commits
-
The rk3288 is the first Rockchip soc handling the drive strength on a per-pin
basis, while the older ones can set the drive-strength only for specific
pin-groups. Therefore limit setting the drive-strength to this soc for now.Signed-off-by: Heiko Stübner
Signed-off-by: Linus Walleij -
An upcoming pinctrl function of the rk3288 differs again from everything else,
so we'll need a separate type for it.Signed-off-by: Heiko Stübner
Signed-off-by: Linus Walleij -
The rockchip pinctrl driver implements the generic pinconfig, therefore
also state this, so that the default pinconf dump functions work.Signed-off-by: Heiko Stübner
Signed-off-by: Linus Walleij -
[linux-3.16-rc5/drivers/pinctrl/pinctrl-msm.c:145]: (style) Checking if
unsigned variable 'mux_bit' is less than zero.if (WARN_ON(g->mux_bit < 0))
return -EINVAL;Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=80491
Reported-by: David Binderman
Signed-off-by: Andrey Utkin
Signed-off-by: Linus Walleij
22 Jul, 2014
2 commits
-
Signed-off-by: abdoulaye berthe
Signed-off-by: Linus Walleij -
Instead of relying on pinmux->disable(), make the gpio function an
explicit function for all pins that supports it.Signed-off-by: Bjorn Andersson
Signed-off-by: Linus Walleij
20 Jul, 2014
1 commit
-
…nel/git/tegra/linux into next/dt
Merge "ARM: tegra: Add XUSB pad controller support" from Thierry Reding:
Adds device tree bindings and a driver for the XUSB pad controller found
on Tegra114 and later. This is a prerequisites for PCIe, SATA and XUSB
drivers which are all currently being reviewed or pending for merge.This is a separate branch in case it needs to be pulled into the pinctrl
tree to resolve conflicts.* tag 'tegra-for-3.17-xusb-padctl' of git://git.kernel.org/pub/scm/linux/kernel/git/tegra/linux:
pinctrl: Add NVIDIA Tegra XUSB pad controller support
of: Add NVIDIA Tegra XUSB pad controller bindingSigned-off-by: Olof Johansson <olof@lixom.net>
16 Jul, 2014
1 commit
-
st_gpio_irqmux_handler() reads the status register to find out
which banks inside the controller have pending IRQs.
For each banks having pending IRQs, it calls the corresponding handler.Problem is that current code restricts the number of possible banks inside the
controller to ST_GPIO_PINS_PER_BANK. This define represents the number of pins
inside a bank, so it shouldn't be used here.On STiH407, PIO_FRONT0 controller has 10 banks, so IRQs pending in the two
last banks (PIO18 & PIO19) aren't handled.This patch replace ST_GPIO_PINS_PER_BANK by the number of banks inside the
controller.Cc: Linus Walleij
Cc: #v3.15+
Acked-by: Srinivas Kandagatla
Signed-off-by: Maxime Coquelin
Signed-off-by: Linus Walleij
11 Jul, 2014
26 commits
-
We have a bunch of Nomadik family pin control drivers, so let's
move them into their own subdirectory.Signed-off-by: Linus Walleij
-
The XUSB pad controller found on NVIDIA Tegra SoCs provides several pads
that lanes can be assigned to in order to support a variety of interface
options: USB 2.0, USB 3.0, PCIe and SATA.In addition to the pin controller used to assign lanes to pads two PHYs
are exposed to allow the bricks for PCIe and SATA to be powered up and
down by PCIe and SATA drivers.Tested-by: Mikko Perttunen
Acked-by: Linus Walleij
Signed-off-by: Thierry Reding -
Group all pin control drivers of Samsung platform together in
a sub-directory for easy maintenance.Signed-off-by: Sachin Kamat
Signed-off-by: Linus Walleij -
This patch moves data allocated using regulator_register to
devm_regulator_register and does away the calls to regulator_unregister.
The sh73a0_pinmux_soc_exit function is no longer needed and is removed.Signed-off-by: Himangi Saraogi
Acked-by: Julia Lawall
Acked-by: Laurent Pinchart
Signed-off-by: Linus Walleij -
We have four Qualcomm-related pin control drivers, and now there
are drivers coming in for the PMICs on these systems, so let's
create a qcom subdirectory to hold all the Qualcomm stuff.Acked-by: Ivan T. Ivanov
Acked-by: Bjorn Andersson
Signed-off-by: Linus Walleij -
Signed-off-by: Bjorn Andersson
Signed-off-by: Linus Walleij -
This patch extends the range of settings configurable via pinfunc API
to cover pin value as well. This allows configuration of default values
of pins, which is useful for pins that are not supposed to be used by
any dedicated driver, but need certain board-specific setting.Signed-off-by: Tomasz Figa
Acked-by: Kyungmin Park
Signed-off-by: Linus Walleij -
One of remaining limitations of current pinctrl-samsung driver was
the inability to parse multiple pinmux/pinconf group nodes grouped
inside a single device tree node. It made defining groups of pins for
single purpose, but with different parameters very inconvenient.This patch implements Tegra-like support for grouping multiple pinctrl
groups inside one device tree node, by completely changing the way
pin groups and functions are parsed from device tree. The code creating
pinctrl maps from DT nodes has been borrowed from pinctrl-tegra, while
the initial creation of groups and functions has been completely
rewritten with following assumptions:
- each group consists of just one pin and does not depend on data
from device tree,
- each function is represented by a device tree child node of the
pin controller, which in turn can contain multiple child nodes
for pins that need to have different configuration values.Device Tree bindings are fully backwards compatible. New functionality
can be used by defining a new pinctrl group consisting of several child
nodes, as on following example:sd4_bus8: sd4-bus-width8 {
part-1 {
samsung,pins = "gpk0-3", "gpk0-4",
"gpk0-5", "gpk0-6";
samsung,pin-function = ;
samsung,pin-pud = ;
samsung,pin-drv = ;
};
part-2 {
samsung,pins = "gpk1-3", "gpk1-4",
"gpk1-5", "gpk1-6";
samsung,pin-function = ;
samsung,pin-pud = ;
samsung,pin-drv = ;
};
};Tested on Exynos4210-Trats board and a custom Exynos4212-based one.
Signed-off-by: Tomasz Figa
Acked-by: Kyungmin Park
Reviewed-by: Stephen Warren
Cc: devicetree@vger.kernel.org
Cc: Rob Herring
Cc: Mark Rutland
Signed-off-by: Linus Walleij -
Handling of irq_chip operations for GPIO and WKUP external interrupts
is mostly the same, with the difference being offset of registers.
However currently the driver has all the code duplicated for both EINT
types, which is undesirable, because changes in irq_chip operations have
to be done to both instances of the same code.This patch fixes this by creating exynos_irq_chip struct that has normal
irq_chip struct embedded and contain differences between particular EINT
types, which are three register offsets. One instance of code is removed
and the new structure is used instead to fetch necessary data instead of
samsung_pin_ctrl struct used previously.While at it, the patch removes Exynos-specific fields from
aforementioned structure to improve layering of the driver.Signed-off-by: Tomasz Figa
Signed-off-by: Linus Walleij -
This patch adds .request() and .free() operations to gpio_chip of
pinctrl-samsung driver, which call pinctrl request and free helpers to
request and free pinctrl pin along with GPIO pin.Signed-off-by: Tomasz Figa
Signed-off-by: Linus Walleij -
This patch makes the pinctrl-samsung driver configure GPIO direction on
its own, without using the pinctrl_gpio_direction_*() "helpers". The
rationale behind this change is as follows:
- pinctrl-samsung does not need translation from GPIO namespace to
pinctrl namespace to handle GPIO operations - GPIO chip and offset
therein are enough to calculate necessary offsets and bit masks in
constant time,
- the pinctrl_gpio_direction_*() functions do not do anything useful
other than translating the pin into pinctrl namespace and calling the
.gpio_set_direction() from pinmux_ops of the controller,
- the undesirable side effect of using those helpers is losing the
ability to change GPIO direction in atomic context, because they
explicitly use a mutex for synchronization,Results of this patch are:
- fixed warnings about scheduling while atomic in code that needs to
set GPIO direction in atomic context (e.g. interrupt handler),
- reduced overhead of bitbanging drivers that use gpio_direction_*(),
e.g. i2c-gpio.Signed-off-by: Tomasz Figa
Signed-off-by: Linus Walleij -
Add CAN0/1 data/clock pin groups to R8A7791 PFC driver.
Signed-off-by: Sergei Shtylyov
Acked-by: Laurent Pinchart
Signed-off-by: Linus Walleij -
For Baytrail, you should never set a GPIO set to direct_irq
to output mode. When direct_irq_en is set for a GPIO, it is
tied directly to an APIC internally, and making the pad output
does not make any sense. Assert a WARN() in the event this happens.Signed-off-by: Eric Ernst
Acked-by: Mika Westerberg
Signed-off-by: Linus Walleij -
Remove variable that are never used
This was found using a static code analysis program called cppcheck.
Signed-off-by: Rickard Strandqvist
Acked-by: Patrice Chotard
Signed-off-by: Linus Walleij -
%d in format string used, but the type is unsigned int
This was found using a static code analysis program called cppcheck
Signed-off-by: Rickard Strandqvist
Signed-off-by: Linus Walleij -
%d in format string used, but the type is unsigned int
This was found using a static code analysis program called cppcheck
Signed-off-by: Rickard Strandqvist
Signed-off-by: Linus Walleij -
Remove variable that are never used
This was found using a static code analysis program called cppcheck.
Signed-off-by: Rickard Strandqvist
Signed-off-by: Linus Walleij -
Remove variable that are never used
This was found using a static code analysis program called cppcheck.
Signed-off-by: Rickard Strandqvist
Signed-off-by: Linus Walleij -
Remove checking if a unsigned is less than zero
This was found using a static code analysis program called cppcheck.
Signed-off-by: Rickard Strandqvist
Acked-by: Srinivas Kandagatla
Acked-by: Maxime Coquelin
Signed-off-by: Linus Walleij -
Remove checking if a unsigned is less than zero
This was found using a static code analysis program called cppcheck.
Signed-off-by: Rickard Strandqvist
Signed-off-by: Linus Walleij -
Remove checking if a unsigned is less than zero
This was found using a static code analysis program called cppcheck.
Signed-off-by: Rickard Strandqvist
Signed-off-by: Linus Walleij -
Remove checking if a unsigned is less than zero
This was found using a static code analysis program called cppcheck.
Signed-off-by: Rickard Strandqvist
Signed-off-by: Linus Walleij -
no .irq_set_wake API is available for pinctrl-st driver.
Add the IRQCHIP_SKIP_SET_WAKE flag to inform irq handler
not to call this API.Signed-off-by: David Paris
Signed-off-by: Linus Walleij -
bcm281xx_pinctrl_probe is local to this file. Make it static.
Signed-off-by: Sachin Kamat
Signed-off-by: Linus Walleij -
This patch adds the BUS_HOLD (Keeper) bias option for pins.
Signed-off-by: Andy Gross
Acked-by: Bjorn Andersson
Signed-off-by: Linus Walleij -
The A23 has a R_PIO pin controller, similar to the one found on the A31 SoC.
Add support for the pins controlled by the R_PIO controller.Signed-off-by: Chen-Yu Tsai
Acked-by: Maxime Ripard
Signed-off-by: Linus Walleij
