Gregory CLEMENT [Mon, 12 Jun 2017 15:34:53 +0000 (17:34 +0200)]
pinctrl: mvebu: remove the offset property for regmap
The offset property of the pinctrl node, when a regmap is used in the
device tree, was never used nor documented in the binding. Moreover, the
compatible string is enough to let the driver know which offset using.
So this patch removes the property and move the information at the driver
level.
Jerome Brunet [Thu, 8 Jun 2017 13:09:48 +0000 (15:09 +0200)]
pinctrl: meson-gxl: add tsin_a pins
Add Tsin A pins to bank DV and X.
We don't have a driver for the tsin yet but since the tsin A pinmux is
enabled by default at boot time, declaring this pinmux is required to
properly operate on GPIOX.
Without this change, GPIOX 8, 9, 10 and 11 can't be driven as GPIO output
as the tsin A seems to have priority.
Mika Westerberg [Tue, 6 Jun 2017 13:18:18 +0000 (16:18 +0300)]
pinctrl: intel: Make it possible to specify mode per pin in a group
On some SoCs not all pins in a group use the same mode when a certain
function is muxed out of them. This makes it possible to specify mode per
pin as an array instead in addition to single integer.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com> Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Mika Westerberg [Tue, 6 Jun 2017 13:18:17 +0000 (16:18 +0300)]
pinctrl: intel: Add support for variable size pad groups
The Intel GPIO hardware has a concept of pad groups, which means 1 to 32
pads occupying their own GPI_IS, GPI_IE, PAD_OWN and so on registers. The
existing hardware has the same amount of pads in each pad group (except the
last one) so it is possible to use community->gpp_size to calculate start
offset of each register.
With the next generation SoCs the pad group size is not always the same
anymore which means we cannot use community->gpp_size for register offset
calculations directly.
To support variable size pad groups we introduce struct intel_padgroup that
can be filled in by the client drivers according the hardware pad group
layout. The core driver will always use these when it performs calculations
for pad register offsets. The core driver will automatically populate pad
groups based on community->gpp_size if the driver does not provide any.
This makes sure the existing drivers still work as expected.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com> Signed-off-by: Chuah, Kim Tatt <kim.tatt.chuah@intel.com> Signed-off-by: Tan Jui Nee <jui.nee.tan@intel.com> Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Alexandre TORGUE [Wed, 31 May 2017 09:10:43 +0000 (11:10 +0200)]
pinctrl: stm32: remove useless check
There is no link between the number of elements of tab which contains all
pin desc (located in each pinctrl-stm32xxxx.c files) and the pin number
(defined in the tab).
Arnd Bergmann [Tue, 30 May 2017 09:11:28 +0000 (11:11 +0200)]
pinctrl: mcp23s08: improve I2C Kconfig dependency
With "SPI_MASTER=y && I2C=m", we can build mcp23s08 as a built-in driver,
which then results in a link failure:
drivers/pinctrl/built-in.o: In function `mcp23s08_probe_one.isra.0':
:(.text+0x7910): undefined reference to `__devm_regmap_init_i2c'
drivers/pinctrl/built-in.o: In function `mcp23s08_init':
:(.init.text+0x110): undefined reference to `i2c_register_driver'
drivers/pinctrl/built-in.o: In function `mcp23s08_exit':
:(.exit.text+0x3c): undefined reference to `i2c_del_driver'
To avoid the problem, this adds another dependency on I2C that enforces
mcp23s08 to be a loadable module whenever the I2C core is a module.
Fixes: 64ac43e6fa28 ("gpio: mcp23s08: move to pinctrl") Signed-off-by: Arnd Bergmann <arnd@arndb.de> Reviewed-by: Sebastian Reichel <sebastian.reichel@collabora.co.uk> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
David Wu [Fri, 26 May 2017 07:20:20 +0000 (15:20 +0800)]
pinctrl: rockchip: Add iomux-route switching support
On the some rockchip SOCS, some things like rk3399 specific uart2 can use
multiple pins. Somewhere between the pin io-cells and the uart it seems
to have some sort of switch to decide to which pin to actually route the
data.
+-------+ +--------+ /- GPIO4_B0 (pinmux 2)
| uart2 | -- | switch | --- GPIO4_C0 (pinmux 2)
+-------+ +--------+ \- GPIO4_C3 (pinmux 2)
(switch selects one of the 3 pins base on the GRF_SOC_CON7[BIT0, BIT1])
The routing switch is determined by one pin of a specific group to be set
to its special pinmux function. If the pinmux setting is wrong for that
pin the ip block won't work correctly anyway.
Paul Gortmaker [Mon, 22 May 2017 20:56:48 +0000 (16:56 -0400)]
pinctrl: bcm: clean up modular vs. non-modular distinctions
Fixups here tend to be more of a conglomerate of some of the other
repeated/systematic ones we've seen in the earlier pinctrl cleanups.
We remove module.h from code that isn't doing anything modular at
all; if they have __init sections, then replace it with init.h
One driver has a .remove that would be dispatched on module_exit,
and as that code is essentially orphaned, so we remove it. In case
anyone was previously doing the (pointless) unbind to get to that
function, we disable unbind for this one driver as well.
A couple bool drivers (hence non-modular) are converted over to
to builtin_platform_driver().
Since module_platform_driver() uses the same init level priority as
builtin_platform_driver() the init ordering remains unchanged with
this commit.
Also note that MODULE_DEVICE_TABLE is a no-op for non-modular code.
We also delete the MODULE_LICENSE tag etc. since all that information
was (or is now) contained at the top of the file in the comments.
Cc: Eric Anholt <eric@anholt.net> Cc: Florian Fainelli <f.fainelli@gmail.com> Cc: Jon Mason <jonmason@broadcom.com> Cc: Linus Walleij <linus.walleij@linaro.org> Cc: Ray Jui <rjui@broadcom.com> Cc: Scott Branden <sbranden@broadcom.com> Cc: Sherman Yin <syin@broadcom.com> Cc: bcm-kernel-feedback-list@broadcom.com Cc: linux-gpio@vger.kernel.org Cc: linux-rpi-kernel@lists.infradead.org Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com> Tested-by: Stefan Wahren <stefan.wahren@i2se.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Paul Gortmaker [Mon, 22 May 2017 20:56:47 +0000 (16:56 -0400)]
pinctrl: tegra: clean up modular vs. non-modular distinctions
None of the Kconfigs for any of these drivers are tristate,
meaning that they currently are not being built as a module by anyone.
Lets remove the modular code that is essentially orphaned, so that
when reading the drivers there is no doubt they are builtin-only. All
drivers get similar changes, so they are handled in batch.
We remove module.h from code that isn't doing anything modular at
all; if they have __init sections, then replace it with init.h.
A couple drivers have module_exit() code that is essentially orphaned,
and so we remove that.
Quite a few bool drivers (hence non-modular) are converted over to
to builtin_platform_driver().
Since module_platform_driver() uses the same init level priority as
builtin_platform_driver() the init ordering remains unchanged with
this commit.
Also note that MODULE_DEVICE_TABLE is a no-op for non-modular code.
We also delete the MODULE_LICENSE tag etc. since all that information
was (or is now) contained at the top of the file in the comments.
Gregory CLEMENT [Fri, 19 May 2017 16:04:43 +0000 (18:04 +0200)]
MAINTAINERS: extend mvebu SoC entry with pinctrl drivers
There was no entry for the mvebu pinctrl drivers. As they are tightly
linked to the SoCs and there is a lot of common code to support the
various pinctrl of each SoCs, then add a new entry for the mvebu
maintainers.
This moves irq property handling from spi/i2c specific code into
the generic mcp23s08_probe_one. This is possible because the
device properties are named equally.
As a side-effect this drops support for setting the properties via
pdata, which has no mainline users. If boardcode wants to enable
the chip as interrupt controller it can attach the device properties
instead.
Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.co.uk> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Instead of using custom caching, this switches to regmap based
caching. Before the conversion the debugfs file used uncached
values, so that it was easily possible to see power-loss related
problems. The new code will check and recover at this place.
The patch will also ensure, that irqs are not cleared by checking
register status in debugfs.
Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.co.uk> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
mcp23s08 support configuration of the pullups using the
pinconf framework. This removes the custom pullup configuration
from platform data, which has no upstream users.
Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.co.uk> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Paul Cercueil [Fri, 12 May 2017 16:53:07 +0000 (18:53 +0200)]
MIPS: jz4740: Remove custom GPIO code
All the drivers for the various hardware elements of the jz4740 SoC have
been modified to use the pinctrl framework for their pin configuration
needs.
As such, this platform code is now unused and can be deleted.
Signed-off-by: Paul Cercueil <paul@crapouillou.net> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Paul Cercueil [Fri, 12 May 2017 16:52:58 +0000 (18:52 +0200)]
MIPS: ingenic: Enable pinctrl for all ingenic SoCs
There is a pinctrl driver for each of the Ingenic SoCs supported by the
upstream Linux kernel. In order to switch away from the old GPIO
platform code, we now enable the pinctrl drivers by default for the
Ingenic SoCs.
Signed-off-by: Paul Cercueil <paul@crapouillou.net> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Paul Cercueil [Fri, 12 May 2017 16:53:06 +0000 (18:53 +0200)]
pwm: jz4740: Let the pinctrl driver configure the pins
Now that the JZ4740 and similar SoCs have a pinctrl driver, we rely on
the pins being properly configured before the driver probes.
One inherent problem of this new approach is that the pinctrl framework
does not allow us to configure each pin on demand, when the various PWM
channels are requested or released. For instance, the PWM channels can
be configured from sysfs, which would require all PWM pins to be configured
properly beforehand for the PWM function, eventually causing conflicts
with other platform or board drivers.
The proper solution here would be to modify the pwm-jz4740 driver to
handle only one PWM channel, and create an instance of this driver
for each one of the 8 PWM channels. Then, it could use the pinctrl
framework to dynamically configure the PWM pin it controls.
Until this can be done, the only jz4740 board supported upstream
(Qi lb60) can configure all of its connected PWM pins in PWM function
mode, since those are not used by other drivers nor by GPIOs on the
board.
Paul Cercueil [Fri, 12 May 2017 16:52:55 +0000 (18:52 +0200)]
dt/bindings: Document gpio-ingenic
This commit adds documentation for the devicetree bindings of the
gpio-ingenic driver, which handles GPIOs of the Ingenic SoCs
currently supported by the Linux kernel.
Signed-off-by: Paul Cercueil <paul@crapouillou.net> Acked-by: Rob Herring <robh@kernel.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Paul Cercueil [Fri, 12 May 2017 16:52:54 +0000 (18:52 +0200)]
dt/bindings: Document pinctrl-ingenic
This commit adds documentation for the devicetree bindings of the
pinctrl-ingenic driver, which handles pin configuration and pin
muxing of the Ingenic SoCs currently supported by the Linux kernel.
Signed-off-by: Paul Cercueil <paul@crapouillou.net> Acked-by: Rob Herring <robh@kernel.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
pinctrl: When claiming hog, skip maps not served by same device
When pinctrl device registers, it automatically claims hogs, that is,
maps that pinctrl device serves for itself.
It is possible that in addition to SoC's pinctrl device, other pinctrl
devices get registered. E.g. some gpio expander devies are registered
as pinctrl devices. For such devices, pinctrl maps could be defined
that set up SoC's pins (e.g. interrupt pin for gpio expander). Such
a map will have target device set to gpio expander.
Here is device tree snippet that causes this scenario:
Such a map will have target device set to gpio expander. However is not
a hog, it is a regular map that is claimed by core before gpio expander
device is probed.
Thus when looking for hogs, it is not enough to check that map's target
device is set to pinctrl device being registered. Need also check that
map's control device is also set to the same.
Dong Aisheng [Fri, 19 May 2017 07:05:43 +0000 (15:05 +0800)]
pinctrl: imx: add soc specific mux_mode mask and shift property
MX7ULP MUX mode mask and shift bit is different from VF610.
Let's make it a platform specific property for the later easy of
adding MX7ULP support.
One trick in exist code that Vybrid hardcoded the config part
as 0xffff because its mux_config register BIT[15-0] are all configs
part. But it's not true in ULP, so use mux_mask instead to address
the difference.
Dong Aisheng [Fri, 19 May 2017 07:05:42 +0000 (15:05 +0800)]
pinctrl: imx: add generic pin config core support
The design is based on the exist architecture that the core will
provide a uniformed way to decode the generic pin config into platform
config register raw data according to the imx_cfg_params_decode maps
registered by platform.
Two useful macros, IMX_CFG_PARAMS_DECODE and IMX_CFG_PARAMS_DECODE_INVERT,
are created for platform to register decode map conveniently.
In order to cope with some special case, a platform specific fixup()
function is also available to use.
Note that rather than fully utilizing the generic pinconf support
provided by pinctrl core, IMX only adopts the device tree bindings
of generic pinconf. The config used in .pin_config_get[set] are raw
register data instead of generic one which makes us align the exist
using. And that's also why we cannot set pinconf_ops.is_generic.
If 'devm_kzalloc' fails, a NULL pointer will be dereferenced.
Return -ENOMEM instead, as done for the other memory allocation just a
few lines below.
BTW, change the 'devm_kzalloc' into a 'devm_kcalloc'.
Sean Wang [Mon, 1 May 2017 07:54:34 +0000 (15:54 +0800)]
pinctrl: mediatek: reuse pinctrl driver for mt7623
mt7623 pinctrl hardware can be compatible with mt2701 driver,
so the patch lets the pinctrl on mt7623 SoC reuse the driver
and deletes those redundant ones.
Signed-off-by: Sean Wang <sean.wang@mediatek.com> Acked-by: John Crispin <john@phrozen.org> Reviewed-by: Matthias Brugger <matthias.bgg@gmail.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
This adds the missing pins for the PWM controllers found in Amlogic
Meson8 SoCs. This includes the pins for PWM_A, PWM_B, PWM_C, PWM_D,
PWM_E and PWM_F controllers.
There is an additional PWM function with the name PWM_VS in the vendor
kernel sources which seems to be used for external video input. Thus
it's not part of this change as the IP block behind the pwm-meson driver
is not responsible for these pins.
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
pinctrl/amd: Update contact information for AMD pinctrl/amd
Updating the point of contact for AMD GPIO driver.
Signed-off-by: Shyam Sundar S K <Shyam-sundar.S-k@amd.com> Signed-off-by: Nehal Shah <Nehal-bakulchandra.Shah@amd.com> Cc: Ken Xue <Ken.Xue@amd.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
The Armada 37xx SoCs can handle interrupt through GPIO. However it can
only manage the edge ones.
The way the interrupt are managed is classical so we can use the generic
interrupt chip model.
The only unusual "feature" is that many interrupts are connected to the
parent interrupt controller. But we do not take advantage of this and use
the chained irq with all of them.
The pin controller on ZTE ZX platforms is kinda of hybrid. It consists
of a main controller and an auxiliary one. For example, on ZX296718 SoC,
the main controller is TOP_PMM and the auxiliary one is AON_IOCFG. Both
controllers work together to control pin multiplexing and configuration.
For most of pins, the pinmux function is controlled by main controller
only, and this type of pins are meant by term 'TOP pins'. For other
pins, the pinmux is controlled by both main and auxiliary controllers,
as the available multiplexing functions for the pin spread in both
controllers. This type of pins are called 'AON pins'. Though pinmux
implementation is quite different, pinconf is same for both types of
pins. Both are controlled by auxiliary controller, i.e. AON_IOCFG on
ZX296718.
The patch adds the ZTE ZX core pinctrl driver to support this hybrid
pin controller as well as ZX296718 SoC specific pin data.
Linus Torvalds [Mon, 22 May 2017 01:26:54 +0000 (18:26 -0700)]
x86: fix 32-bit case of __get_user_asm_u64()
The code to fetch a 64-bit value from user space was entirely buggered,
and has been since the code was merged in early 2016 in commit b2f680380ddf ("x86/mm/32: Add support for 64-bit __get_user() on 32-bit
kernels").
Happily the buggered routine is almost certainly entirely unused, since
the normal way to access user space memory is just with the non-inlined
"get_user()", and the inlined version didn't even historically exist.
The normal "get_user()" case is handled by external hand-written asm in
arch/x86/lib/getuser.S that doesn't have either of these issues.
There were two independent bugs in __get_user_asm_u64():
- it still did the STAC/CLAC user space access marking, even though
that is now done by the wrapper macros, see commit 11f1a4b9755f
("x86: reorganize SMAP handling in user space accesses").
This didn't result in a semantic error, it just means that the
inlined optimized version was hugely less efficient than the
allegedly slower standard version, since the CLAC/STAC overhead is
quite high on modern Intel CPU's.
- the double register %eax/%edx was marked as an output, but the %eax
part of it was touched early in the asm, and could thus clobber other
inputs to the asm that gcc didn't expect it to touch.
In particular, that meant that the generated code could look like
this:
mov (%eax),%eax
mov 0x4(%eax),%edx
where the load of %edx obviously was _supposed_ to be from the 32-bit
word that followed the source of %eax, but because %eax was
overwritten by the first instruction, the source of %edx was
basically random garbage.
The fixes are trivial: remove the extraneous STAC/CLAC entries, and mark
the 64-bit output as early-clobber to let gcc know that no inputs should
alias with the output register.
Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: Benjamin LaHaise <bcrl@kvack.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: stable@kernel.org # v4.8+ Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Linus Torvalds [Sun, 21 May 2017 22:25:46 +0000 (15:25 -0700)]
Clean up x86 unsafe_get/put_user() type handling
Al noticed that unsafe_put_user() had type problems, and fixed them in
commit a7cc722fff0b ("fix unsafe_put_user()"), which made me look more
at those functions.
It turns out that unsafe_get_user() had a type issue too: it limited the
largest size of the type it could handle to "unsigned long". Which is
fine with the current users, but doesn't match our existing normal
get_user() semantics, which can also handle "u64" even when that does
not fit in a long.
While at it, also clean up the type cast in unsafe_put_user(). We
actually want to just make it an assignment to the expected type of the
pointer, because we actually do want warnings from types that don't
convert silently. And it makes the code more readable by not having
that one very long and complex line.
[ This patch might become stable material if we ever end up back-porting
any new users of the unsafe uaccess code, but as things stand now this
doesn't matter for any current existing uses. ]
Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Linus Torvalds [Sun, 21 May 2017 19:06:44 +0000 (12:06 -0700)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Pull misc uaccess fixes from Al Viro:
"Fix for unsafe_put_user() (no callers currently in mainline, but
anyone starting to use it will step into that) + alpha osf_wait4()
infoleak fix"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
osf_wait4(): fix infoleak
fix unsafe_put_user()
Linus Torvalds [Sun, 21 May 2017 18:52:00 +0000 (11:52 -0700)]
Merge branch 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull scheduler fix from Thomas Gleixner:
"A single scheduler fix:
Prevent idle task from ever being preempted. That makes sure that
synchronize_rcu_tasks() which is ignoring idle task does not pretend
that no task is stuck in preempted state. If that happens and idle was
preempted on a ftrace trampoline the machine crashes due to
inconsistent state"
* 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched/core: Call __schedule() from do_idle() without enabling preemption