Commit Graph

14 Commits

Author SHA1 Message Date
Linus Torvalds 2bd8040174 This is the bulk of GPIO changes for the v4.12 kernel cycle:
Core changes
 
 - Return NULL from gpiod_get_optional() when GPIOLIB is disabled.
   This was a much discussed change. It affects use cases where people
   write drivers that might or might not be using GPIO resources.
   I have decided that this is the lesser evil right now.
 
 - Make gpiod_count() behave consistently across different hardware
   descriptions.
 
 - Fix the syntax around open drain/open source to not infer active
   high/low semantics.
 
 New drivers
 
 - A new single-register fixed-direction framework driver for hardware
   that have lines controlled by a single register that just work in
   one direction (out or in), including IRQ support.
 
 - Support the Fintek F71889A GPIO SuperIO controller.
 
 - Support the National NI 169445 MMIO GPIO.
 
 - Support for the X-Gene derivative of the DWC GPIO controller
 
 - Support for the Rohm BD9571MWV-M PMIC GPIO controller.
 
 - Refactor the Gemini GPIO driver to a generic Faraday FTGPIO driver
   and replace both the Gemini and the Moxa ART custom drivers with
   this driver.
 
 Driver improvements
 
 - A whole slew of drivers have their spinlocks chaned to raw spinlocks
   as they provide irqchips, and thus we are progressing on realtime
   compliance.
 
 - Use devm_irq_alloc_descs() in a slew of drivers, getting managed
   resources.
 
 - Support for the embedded PWM controller inside the MVEBU driver.
 
 - Debounce, open source and open drain support for the Aspeed driver.
 
 - Misc smaller fixes like spelling and syntax and whatnot.
 -----BEGIN PGP SIGNATURE-----
 
 iQIcBAABAgAGBQJZCusBAAoJEEEQszewGV1zengP/i3YgmSXJdaf26t/NxG3klU3
 qx3akdQ3gEQ6BOlIY7Ew+b6qgMJynYNZa1trZgqTgLxPEKWjidyCgz+LHQ0OY+ir
 a7x6wr+gBj5WlgV+nBjVs4l6W8pVKCfnap/04OPEQDpxZOHs2LU5pqxxUZ9AxkKS
 urDFMDX55baFviQ+xAuHgamok87YoGP36A/e/fHIBepZmnochf0mCcPfIh0t8lRh
 s2x29PN5ZFRkl403RzjZfVCEMr9bMnSqmDquvPO++Kq0bL+3rOhuMErocd1Bg8ao
 LxBktkryujTaw699xK7Rq5SwcnOAPpaBY4NTmwsIJvAJuCh7qLy9JxQSBsSOT2bx
 61NWUt5T/Xsi0ECYZM4YvsNpUP6XrpSTyG3c8T3fY9vXYLNKZBv1ht6OODpLeuke
 DxULAWP+DdzUS8a3qfKQvIJzSTloU31a1MBG58DWNJ072EQfa2YNaVE75VQk/z5/
 0xZbSHdPY/0Xgx8ltpKu37bSO676JiVQZZ1HEAuti4h21+USYueYD2L8/Bx4k9e/
 4UaOcw3MaCDHP/sf5hg17kQBjhhS0lV9Zv6H9QbHZUocJTJlIU+vXtgkQlrfi3n8
 8j5m+ywVarmLtPqg1j2rqcw7LBCPe0qRXH3e5X/YmNMc3rH9bQz4cTo8ZSN9r8zS
 c17zGbbAqlGsBkpFAbQz
 =DGPb
 -----END PGP SIGNATURE-----

Merge tag 'gpio-v4.12-1' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpio

Pull GPIO updates from Linus Walleij:
 "This is the bulk of GPIO changes for the v4.12 kernel cycle.

  Core changes:

   - Return NULL from gpiod_get_optional() when GPIOLIB is disabled.
     This was a much discussed change. It affects use cases where people
     write drivers that might or might not be using GPIO resources. I
     have decided that this is the lesser evil right now.

   - Make gpiod_count() behave consistently across different hardware
     descriptions.

   - Fix the syntax around open drain/open source to not infer active
     high/low semantics.

  New drivers:

   - A new single-register fixed-direction framework driver for hardware
     that have lines controlled by a single register that just work in
     one direction (out or in), including IRQ support.

   - Support the Fintek F71889A GPIO SuperIO controller.

   - Support the National NI 169445 MMIO GPIO.

   - Support for the X-Gene derivative of the DWC GPIO controller

   - Support for the Rohm BD9571MWV-M PMIC GPIO controller.

   - Refactor the Gemini GPIO driver to a generic Faraday FTGPIO driver
     and replace both the Gemini and the Moxa ART custom drivers with
     this driver.

  Driver improvements:

   - A whole slew of drivers have their spinlocks chaned to raw
     spinlocks as they provide irqchips, and thus we are progressing on
     realtime compliance.

   - Use devm_irq_alloc_descs() in a slew of drivers, getting managed
     resources.

   - Support for the embedded PWM controller inside the MVEBU driver.

   - Debounce, open source and open drain support for the Aspeed driver.

   - Misc smaller fixes like spelling and syntax and whatnot"

* tag 'gpio-v4.12-1' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpio: (77 commits)
  gpio: f7188x: Add a missing break
  gpio: omap: return error if requested debounce time is not possible
  gpio: Add ROHM BD9571MWV-M PMIC GPIO driver
  gpio: gpio-wcove: fix GPIO IRQ status mask
  gpio: DT bindings, move tca9554 from pcf857x to pca953x
  gpio: move tca9554 from pcf857x to pca953x
  gpio: arizona: Correct check whether the pin is an input
  gpio: Add XRA1403 DTS binding documentation
  dt-bindings: add exar to vendor prefixes list
  gpio: gpio-wcove: fix irq pending status bit width
  gpio: dwapb: use dwapb_read instead of readl_relaxed
  gpio: aspeed: Add open-source and open-drain support
  gpio: aspeed: Add debounce support
  gpio: aspeed: dt: Add optional clocks property
  gpio: aspeed: dt: Fix description alignment in bindings document
  gpio: mvebu: Add limited PWM support
  gpio: Use unsigned int for interrupt numbers
  gpio: f7188x: Add F71889A GPIO support.
  gpio: core: Decouple open drain/source flag with active low/high
  gpio: arizona: Correct handling for reading input GPIOs
  ...
2017-05-04 12:05:32 -07:00
Julia Cartwright 21d01c9c08 gpio: altera: make use of raw_spinlock variants
The altera gpio driver currently implements an irq_chip for handling
interrupts; due to how irq_chip handling is done, it's necessary for the
irq_chip methods to be invoked from hardirq context, even on a a real-time
kernel.  Because the spinlock_t type becomes a "sleeping" spinlock w/ RT
kernels, it is not suitable to be used with irq_chips.

A quick audit of the operations under the lock reveal that they do only
minimal, bounded work, and are therefore safe to do under a raw spinlock.

Signed-off-by: Julia Cartwright <julia@ni.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2017-03-16 21:52:06 +01:00
Phil Reid f759921cfb gpio: altera: Use handle_level_irq when configured as a level_high
When a threaded irq handler is chained attached to one of the gpio
pins when configure for level irq the altera_gpio_irq_leveL_high_handler
does not mask the interrupt while being handled by the chained irq.
This resulting in the threaded irq not getting enough cycles to complete
quickly enough before the irq was disabled as faulty. handle_level_irq
should be used in this situation instead of handle_simple_irq.

In gpiochip_irqchip_add set default handler to handle_bad_irq as
per Documentation/gpio/driver.txt. Then set the correct handler in
the set_type callback.

Signed-off-by: Phil Reid <preid@electromag.com.au>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2017-03-06 14:35:22 +01:00
Paul Gortmaker 7b5409ee92 gpio: altera: fix implicit assumption module.h is present
The Kconfig for this file is:

drivers/gpio/Kconfig:config GPIO_ALTERA
drivers/gpio/Kconfig:   tristate "Altera GPIO"

...but however it does not include module.h -- it in turn gets it from
another header (gpio/driver.h) and we'd like to replace that with a
forward delcaration of "struct module;" but if we do, this file will
fail to compile.

So we fix this first to avoid putting build failures into the bisect
commit history.

Cc: Linus Walleij <linus.walleij@linaro.org>
Cc: Alexandre Courbot <gnurou@gmail.com>
Cc: linux-gpio@vger.kernel.org
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-09-15 14:00:47 +02:00
Phil Reid 73c13c8349 gpio: gpio-altera: Remove gpiochip on probe failure.
On failure to setup the irq altera_gpio_probe would return an error
but not go to cleanup. This resulted in kernel fault
"Unable to handle kernel paging request at virtual address xxxxxxxx"
later on in of_gpiochip_find_and_xlate.

Signed-off-by: Phil Reid <preid@electromag.com.au>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-01-27 10:50:22 +01:00
Linus Walleij 397d0773a9 gpio: altera: use gpiochip data pointer
This makes the driver use the data pointer added to the gpio_chip
to store a pointer to the state container instead of relying on
container_of().

Cc: Tien Hock Loh <thloh@altera.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-01-05 11:21:02 +01:00
Linus Walleij 58383c7842 gpio: change member .dev to .parent
The name .dev in a struct is normally reserved for a struct device
that is let us say a superclass to the thing described by the struct.
struct gpio_chip stands out by confusingly using a struct device *dev
to point to the parent device (such as a platform_device) that
represents the hardware. As we want to give gpio_chip:s real devices,
this is not working. We need to rename this member to parent.

This was done by two coccinelle scripts, I guess it is possible to
combine them into one, but I don't know such stuff. They look like
this:

@@
struct gpio_chip *var;
@@
-var->dev
+var->parent

and:

@@
struct gpio_chip var;
@@
-var.dev
+var.parent

and:

@@
struct bgpio_chip *var;
@@
-var->gc.dev
+var->gc.parent

Plus a few instances of bgpio that I couldn't figure out how
to teach Coccinelle to rewrite.

This patch hits all over the place, but I *strongly* prefer this
solution to any piecemal approaches that just exercise patch
mechanics all over the place. It mainly hits drivers/gpio and
drivers/pinctrl which is my own backyard anyway.

Cc: Haavard Skinnemoen <hskinnemoen@gmail.com>
Cc: Rafał Miłecki <zajec5@gmail.com>
Cc: Richard Purdie <rpurdie@rpsys.net>
Cc: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Cc: Alek Du <alek.du@intel.com>
Cc: Jaroslav Kysela <perex@perex.cz>
Cc: Takashi Iwai <tiwai@suse.com>
Acked-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Acked-by: Lee Jones <lee.jones@linaro.org>
Acked-by: Jiri Kosina <jkosina@suse.cz>
Acked-by: Hans-Christian Egtvedt <egtvedt@samfundet.no>
Acked-by: Jacek Anaszewski <j.anaszewski@samsung.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-11-19 09:24:35 +01:00
Linus Walleij 231d51b8a4 gpio: altera: use container_of() to get state container
The state container of the Altera GPIO driver is extracted from
the gpio_chip exploiting the fact that offsetof() the
struct gpio_chip inside the struct of_mm_gpio_chip are both 0, so
the container_of() is in practice a noop. However if a member
is added to struct altera_gpio_chip in front of
struct of_mm_gpio_chip, things will break. Using proper
container_of() avoids this problem.

Semantically this is a noop, the compiler will optimize it away,
but syntactically it makes me happier.

Cc: Tien Hock Loh <thloh@altera.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-02 04:19:32 -07:00
Thomas Gleixner bd0b9ac405 genirq: Remove irq argument from irq flow handlers
Most interrupt flow handlers do not use the irq argument. Those few
which use it can retrieve the irq number from the irq descriptor.

Remove the argument.

Search and replace was done with coccinelle and some extra helper
scripts around it. Thanks to Julia for her help!

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Julia Lawall <Julia.Lawall@lip6.fr>
Cc: Jiang Liu <jiang.liu@linux.intel.com>
2015-09-16 15:47:51 +02:00
Masahiro Yamada 41ec66c922 gpio: altera: use of_mm_gpiochip_remove() to fix memory leak
This driver calls of_mm_gpiochip_add() to add a memory mapped gpio
chip.  So, of_mm_gpiochip_remove() should be used when removing it.

The direct call of gpiochip_remove() misses unmapping the register
and freeing the label.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Alexandre Courbot <acourbot@nvidia.com>
Acked-by: Tien Hock Loh <thloh@altera.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-07-16 00:11:03 +02:00
Masahiro Yamada 1c8b5d688d gpio: altera: fix return value of altera_gpio_remove()
The remove callback never succeeds, which seems odd.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Alexandre Courbot <acourbot@nvidia.com>
Acked-by: Tien Hock Loh <thloh@altera.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-07-16 00:11:02 +02:00
Daniel Lockyer 38e003f4b5 gpio: Fix checkpatch.pl issues
This patch fixes some issues given by checkpatch. Fixes include
bracket placement, spacing and indenting.

Signed-off-by: Daniel Lockyer <thisisdaniellockyer@gmail.com>
Reviewed-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-06-16 11:00:06 +02:00
kbuild test robot c4b40493de altera_gpio_probe() can be static
Signed-off-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-03-27 09:28:40 +01:00
Tien Hock Loh c5abbba932 drivers/gpio: Altera soft IP GPIO driver
Adds a new driver for Altera soft GPIO IP. The driver is able to do
read/write and allows GPIO to be a interrupt controller.

Tested on Altera GHRD on interrupt handling and IO.

v10:
- Updated conflicting device tree parameters
- Removed unused headers
- Used macro instead of magic numbers for ngpio
- Code readability cleanup using ?: and temporal variables
- Removed leftover garbage and unnecessary function calls
- Checked bgpio_init but unusable because Altera GPIO may not
  be a multiple of 8 bits

v9:
- Removed duplicated initialization on set_type using temporals
  to improve code readability in calling generic_handle_irq
- Using ?: ternary to reduce code size

v8:
- Using for_each_set_bit
- Added const for struct definition
- Removed naggy pr_err
- Sort alpha header
- Remove unused macros
- Use fixed width data types instead of unsigned long
- Whitespace issue fixes
- Removed _relaxed function for better compatibility across different
  CPU
- Changed irq_create_mapping to platform_get_irq updated implementation
  to use gpiochip_irqchip_add
- Reserve interrupt-cells number 2 in device tree binding for future
  use
- Remove confusing sections on devicetree bindings
- Added tristate Kconfig help text

v7:
- Used dev_warn instead of pr_warn
- Clean up unnecesarry if else indentation

v6:
- Added irq_startup and irq_shutdown
- Changed bitwise clamping style
- Cleanup bitwise operation to improve readability change naming of
  mapped irqs from virq to mapped_irq

v5:
- Dispose irq_domain mapping correctly
- Update optional binding description in binding docs

v4:
- Added vendor prefix to devicetree binding for IP specific properties
  using MMIO GPIO helper library instead of manually map PIO to memory
- altera_gpio_chip inline struct documentation to kerneldoc
- Using dev_ print to print a better failure message

v2, v3:
- Do not reference NO_IRQ
- Updated irq_set_type to only allow the hardware configured irq type

Signed-off-by: Tien Hock Loh <thloh@altera.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-03-07 20:30:49 +01:00