2
0
mirror of https://github.com/edk2-porting/linux-next.git synced 2024-12-25 13:43:55 +08:00
linux-next/drivers/gpio/gpiolib.c

4125 lines
110 KiB
C
Raw Normal View History

#include <linux/bitmap.h>
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/interrupt.h>
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
#include <linux/irq.h>
#include <linux/spinlock.h>
#include <linux/list.h>
gpio: sysfs interface This adds a simple sysfs interface for GPIOs. /sys/class/gpio /export ... asks the kernel to export a GPIO to userspace /unexport ... to return a GPIO to the kernel /gpioN ... for each exported GPIO #N /value ... always readable, writes fail for input GPIOs /direction ... r/w as: in, out (default low); write high, low /gpiochipN ... for each gpiochip; #N is its first GPIO /base ... (r/o) same as N /label ... (r/o) descriptive, not necessarily unique /ngpio ... (r/o) number of GPIOs; numbered N .. N+(ngpio - 1) GPIOs claimed by kernel code may be exported by its owner using a new gpio_export() call, which should be most useful for driver debugging. Such exports may optionally be done without a "direction" attribute. Userspace may ask to take over a GPIO by writing to a sysfs control file, helping to cope with incomplete board support or other "one-off" requirements that don't merit full kernel support: echo 23 > /sys/class/gpio/export ... will gpio_request(23, "sysfs") and gpio_export(23); use /sys/class/gpio/gpio-23/direction to (re)configure it, when that GPIO can be used as both input and output. echo 23 > /sys/class/gpio/unexport ... will gpio_free(23), when it was exported as above The extra D-space footprint is a few hundred bytes, except for the sysfs resources associated with each exported GPIO. The additional I-space footprint is about two thirds of the current size of gpiolib (!). Since no /dev node creation is involved, no "udev" support is needed. Related changes: * This adds a device pointer to "struct gpio_chip". When GPIO providers initialize that, sysfs gpio class devices become children of that device instead of being "virtual" devices. * The (few) gpio_chip providers which have such a device node have been updated. * Some gpio_chip drivers also needed to update their module "owner" field ... for which missing kerneldoc was added. * Some gpio_chips don't support input GPIOs. Those GPIOs are now flagged appropriately when the chip is registered. Based on previous patches, and discussion both on and off LKML. A Documentation/ABI/testing/sysfs-gpio update is ready to submit once this merges to mainline. [akpm@linux-foundation.org: a few maintenance build fixes] Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Guennadi Liakhovetski <g.liakhovetski@pengutronix.de> Cc: Greg KH <greg@kroah.com> Cc: Kay Sievers <kay.sievers@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 16:46:07 +08:00
#include <linux/device.h>
#include <linux/err.h>
#include <linux/debugfs.h>
#include <linux/seq_file.h>
#include <linux/gpio.h>
#include <linux/of_gpio.h>
#include <linux/idr.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 16:04:11 +08:00
#include <linux/slab.h>
ACPI / driver core: Store an ACPI device pointer in struct acpi_dev_node Modify struct acpi_dev_node to contain a pointer to struct acpi_device associated with the given device object (that is, its ACPI companion device) instead of an ACPI handle corresponding to it. Introduce two new macros for manipulating that pointer in a CONFIG_ACPI-safe way, ACPI_COMPANION() and ACPI_COMPANION_SET(), and rework the ACPI_HANDLE() macro to take the above changes into account. Drop the ACPI_HANDLE_SET() macro entirely and rework its users to use ACPI_COMPANION_SET() instead. For some of them who used to pass the result of acpi_get_child() directly to ACPI_HANDLE_SET() introduce a helper routine acpi_preset_companion() doing an equivalent thing. The main motivation for doing this is that there are things represented by struct acpi_device objects that don't have valid ACPI handles (so called fixed ACPI hardware features, such as power and sleep buttons) and we would like to create platform device objects for them and "glue" them to their ACPI companions in the usual way (which currently is impossible due to the lack of valid ACPI handles). However, there are more reasons why it may be useful. First, struct acpi_device pointers allow of much better type checking than void pointers which are ACPI handles, so it should be more difficult to write buggy code using modified struct acpi_dev_node and the new macros. Second, the change should help to reduce (over time) the number of places in which the result of ACPI_HANDLE() is passed to acpi_bus_get_device() in order to obtain a pointer to the struct acpi_device associated with the given "physical" device, because now that pointer is returned by ACPI_COMPANION() directly. Finally, the change should make it easier to write generic code that will build both for CONFIG_ACPI set and unset without adding explicit compiler directives to it. Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Tested-by: Mika Westerberg <mika.westerberg@linux.intel.com> # on Haswell Reviewed-by: Mika Westerberg <mika.westerberg@linux.intel.com> Reviewed-by: Aaron Lu <aaron.lu@intel.com> # for ATA and SDIO part
2013-11-12 05:41:56 +08:00
#include <linux/acpi.h>
#include <linux/gpio/driver.h>
#include <linux/gpio/machine.h>
#include <linux/pinctrl/consumer.h>
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
#include <linux/cdev.h>
#include <linux/fs.h>
#include <linux/uaccess.h>
#include <linux/compat.h>
gpio: userspace ABI for reading/writing GPIO lines This adds a userspace ABI for reading and writing GPIO lines. The mechanism returns an anonymous file handle to a request to read/write n offsets from a gpiochip. This file handle in turn accepts two ioctl()s: one that reads and one that writes values to the selected lines. - Handles can be requested as input/output, active low, open drain, open source, however when you issue a request for n lines with GPIO_GET_LINEHANDLE_IOCTL, they must all have the same flags, i.e. all inputs or all outputs, all open drain etc. If a granular control of the flags for each line is desired, they need to be requested individually, not in a batch. - The GPIOHANDLE_GET_LINE_VALUES_IOCTL read ioctl() can be issued also to output lines to verify that the hardware is in the expected state. - It reads and writes up to GPIOHANDLES_MAX lines at once, utilizing the .set_multiple() call in the driver if possible, making the call efficient if several lines can be written with a single register update. The limitation of GPIOHANDLES_MAX to 64 lines is done under the assumption that we may expect hardware that can issue a transaction updating 64 bits at an instant but unlikely anything larger than that. ChangeLog v2->v3: - Use gpiod_get_value_cansleep() so we support also slowpath GPIO drivers. - Fix up the UAPI docs kerneldoc. - Allocate the anonymous fd last, so that the release function don't get called until that point of something fails. After this point, skip the errorpath. ChangeLog v1->v2: - Handle ioctl_compat() properly based on a similar patch to the other ioctl() handling code. - Use _IOWR() as we pass pointers both in and out of the ioctl() - Use kmalloc() and kfree() for the linehandled, do not try to be fancy with devm_* it doesn't work the way I thought. - Fix const-correctness on the linehandle name field. Acked-by: Michael Welling <mwelling@ieee.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-04-26 16:35:29 +08:00
#include <linux/anon_inodes.h>
gpio: GPIO_GET_LINE{HANDLE,EVENT}_IOCTL: Fix file descriptor leak When allocating a new line handle or event a file is allocated that it is associated to. The file is attached to a file descriptor of the current process and the file descriptor is returned to userspace using copy_to_user(). If this copy operation fails the line handle or event allocation is aborted, all acquired resources are freed and an error is returned. But the file struct is not freed and left attached to the userspace application and even though the file descriptor number was not copied it is trivial to guess. If a userspace application performs a IOCTL on such a left over file descriptor it will trigger a use-after-free and if the file descriptor is closed (latest when the application exits) a double-free is triggered. anon_inode_getfd() performs 3 tasks, allocate a file struct, allocate a file descriptor for the current process and install the file struct in the file descriptor. As soon as the file struct is installed in the file descriptor it is accessible by userspace (even if the IOCTL itself hasn't completed yet), this means uninstalling the fd on the error path is not an option, since userspace might already got a reference to the file. Instead anon_inode_getfd() needs to be broken into its individual steps. The allocation of the file struct and file descriptor is done first, then the copy_to_user() is executed and only if it succeeds the file is installed. Since the file struct is reference counted it can not be just freed, but its reference needs to be dropped, which will also call the release() callback, which will free the state attached to the file. So in this case the normal error cleanup path should not be taken. Cc: stable@vger.kernel.org Fixes: d932cd49182f ("gpio: free handles in fringe cases") Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-10-24 19:59:15 +08:00
#include <linux/file.h>
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
#include <linux/kfifo.h>
#include <linux/poll.h>
#include <linux/timekeeping.h>
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
#include <uapi/linux/gpio.h>
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
#include "gpiolib.h"
#define CREATE_TRACE_POINTS
#include <trace/events/gpio.h>
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/* Implementation infrastructure for GPIO interfaces.
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
*
* The GPIO programming interface allows for inlining speed-critical
* get/set operations for common cases, so that access to SOC-integrated
* GPIOs can sometimes cost only an instruction or two per bit.
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
*/
/* When debugging, extend minimal trust to callers and platform code.
* Also emit diagnostic messages that may help initial bringup, when
* board setup or driver bugs are most common.
*
* Otherwise, minimize overhead in what may be bitbanging codepaths.
*/
#ifdef DEBUG
#define extra_checks 1
#else
#define extra_checks 0
#endif
/* Device and char device-related information */
static DEFINE_IDA(gpio_ida);
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
static dev_t gpio_devt;
#define GPIO_DEV_MAX 256 /* 256 GPIO chip devices supported */
static struct bus_type gpio_bus_type = {
.name = "gpio",
};
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/* gpio_lock prevents conflicts during gpio_desc[] table updates.
* While any GPIO is requested, its gpio_chip is not removable;
* each GPIO's "requested" flag serves as a lock and refcount.
*/
DEFINE_SPINLOCK(gpio_lock);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
static DEFINE_MUTEX(gpio_lookup_lock);
static LIST_HEAD(gpio_lookup_list);
LIST_HEAD(gpio_devices);
static void gpiochip_free_hogs(struct gpio_chip *chip);
static int gpiochip_add_irqchip(struct gpio_chip *gpiochip,
struct lock_class_key *key);
static void gpiochip_irqchip_remove(struct gpio_chip *gpiochip);
static int gpiochip_irqchip_init_valid_mask(struct gpio_chip *gpiochip);
static void gpiochip_irqchip_free_valid_mask(struct gpio_chip *gpiochip);
static bool gpiolib_initialized;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
static inline void desc_set_label(struct gpio_desc *d, const char *label)
{
d->label = label;
}
/**
* gpio_to_desc - Convert a GPIO number to its descriptor
* @gpio: global GPIO number
*
* Returns:
* The GPIO descriptor associated with the given GPIO, or %NULL if no GPIO
* with the given number exists in the system.
*/
struct gpio_desc *gpio_to_desc(unsigned gpio)
{
struct gpio_device *gdev;
unsigned long flags;
spin_lock_irqsave(&gpio_lock, flags);
list_for_each_entry(gdev, &gpio_devices, list) {
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
if (gdev->base <= gpio &&
gdev->base + gdev->ngpio > gpio) {
spin_unlock_irqrestore(&gpio_lock, flags);
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
return &gdev->descs[gpio - gdev->base];
}
}
spin_unlock_irqrestore(&gpio_lock, flags);
if (!gpio_is_valid(gpio))
WARN(1, "invalid GPIO %d\n", gpio);
return NULL;
}
EXPORT_SYMBOL_GPL(gpio_to_desc);
/**
* gpiochip_get_desc - get the GPIO descriptor corresponding to the given
* hardware number for this chip
* @chip: GPIO chip
* @hwnum: hardware number of the GPIO for this chip
*
* Returns:
* A pointer to the GPIO descriptor or %ERR_PTR(-EINVAL) if no GPIO exists
* in the given chip for the specified hardware number.
*/
struct gpio_desc *gpiochip_get_desc(struct gpio_chip *chip,
u16 hwnum)
{
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
struct gpio_device *gdev = chip->gpiodev;
if (hwnum >= gdev->ngpio)
return ERR_PTR(-EINVAL);
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
return &gdev->descs[hwnum];
}
/**
* desc_to_gpio - convert a GPIO descriptor to the integer namespace
* @desc: GPIO descriptor
*
* This should disappear in the future but is needed since we still
* use GPIO numbers for error messages and sysfs nodes.
*
* Returns:
* The global GPIO number for the GPIO specified by its descriptor.
*/
int desc_to_gpio(const struct gpio_desc *desc)
{
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
return desc->gdev->base + (desc - &desc->gdev->descs[0]);
}
EXPORT_SYMBOL_GPL(desc_to_gpio);
/**
* gpiod_to_chip - Return the GPIO chip to which a GPIO descriptor belongs
* @desc: descriptor to return the chip of
*/
struct gpio_chip *gpiod_to_chip(const struct gpio_desc *desc)
{
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
if (!desc || !desc->gdev || !desc->gdev->chip)
return NULL;
return desc->gdev->chip;
}
EXPORT_SYMBOL_GPL(gpiod_to_chip);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/* dynamic allocation of GPIOs, e.g. on a hotplugged device */
static int gpiochip_find_base(int ngpio)
{
struct gpio_device *gdev;
int base = ARCH_NR_GPIOS - ngpio;
list_for_each_entry_reverse(gdev, &gpio_devices, list) {
/* found a free space? */
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
if (gdev->base + gdev->ngpio <= base)
break;
else
/* nope, check the space right before the chip */
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
base = gdev->base - ngpio;
}
if (gpio_is_valid(base)) {
pr_debug("%s: found new base at %d\n", __func__, base);
return base;
} else {
pr_err("%s: cannot find free range\n", __func__);
return -ENOSPC;
}
}
/**
* gpiod_get_direction - return the current direction of a GPIO
* @desc: GPIO to get the direction of
*
* Return GPIOF_DIR_IN or GPIOF_DIR_OUT, or an error code in case of error.
*
* This function may sleep if gpiod_cansleep() is true.
*/
int gpiod_get_direction(struct gpio_desc *desc)
{
struct gpio_chip *chip;
unsigned offset;
int status = -EINVAL;
chip = gpiod_to_chip(desc);
offset = gpio_chip_hwgpio(desc);
if (!chip->get_direction)
return status;
status = chip->get_direction(chip, offset);
if (status > 0) {
/* GPIOF_DIR_IN, or other positive */
status = 1;
clear_bit(FLAG_IS_OUT, &desc->flags);
}
if (status == 0) {
/* GPIOF_DIR_OUT */
set_bit(FLAG_IS_OUT, &desc->flags);
}
return status;
}
EXPORT_SYMBOL_GPL(gpiod_get_direction);
/*
* Add a new chip to the global chips list, keeping the list of chips sorted
* by range(means [base, base + ngpio - 1]) order.
*
* Return -EBUSY if the new chip overlaps with some other chip's integer
* space.
*/
static int gpiodev_add_to_list(struct gpio_device *gdev)
{
struct gpio_device *prev, *next;
if (list_empty(&gpio_devices)) {
/* initial entry in list */
list_add_tail(&gdev->list, &gpio_devices);
return 0;
}
next = list_entry(gpio_devices.next, struct gpio_device, list);
if (gdev->base + gdev->ngpio <= next->base) {
/* add before first entry */
list_add(&gdev->list, &gpio_devices);
return 0;
}
prev = list_entry(gpio_devices.prev, struct gpio_device, list);
if (prev->base + prev->ngpio <= gdev->base) {
/* add behind last entry */
list_add_tail(&gdev->list, &gpio_devices);
return 0;
}
list_for_each_entry_safe(prev, next, &gpio_devices, list) {
/* at the end of the list */
if (&next->list == &gpio_devices)
break;
/* add between prev and next */
if (prev->base + prev->ngpio <= gdev->base
&& gdev->base + gdev->ngpio <= next->base) {
list_add(&gdev->list, &prev->list);
return 0;
}
}
dev_err(&gdev->dev, "GPIO integer space overlap, cannot add chip\n");
return -EBUSY;
}
/*
* Convert a GPIO name to its descriptor
*/
static struct gpio_desc *gpio_name_to_desc(const char * const name)
{
struct gpio_device *gdev;
unsigned long flags;
spin_lock_irqsave(&gpio_lock, flags);
list_for_each_entry(gdev, &gpio_devices, list) {
int i;
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
for (i = 0; i != gdev->ngpio; ++i) {
struct gpio_desc *desc = &gdev->descs[i];
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
if (!desc->name || !name)
continue;
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
if (!strcmp(desc->name, name)) {
spin_unlock_irqrestore(&gpio_lock, flags);
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
return desc;
}
}
}
spin_unlock_irqrestore(&gpio_lock, flags);
return NULL;
}
/*
* Takes the names from gc->names and checks if they are all unique. If they
* are, they are assigned to their gpio descriptors.
*
* Warning if one of the names is already used for a different GPIO.
*/
static int gpiochip_set_desc_names(struct gpio_chip *gc)
{
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
struct gpio_device *gdev = gc->gpiodev;
int i;
if (!gc->names)
return 0;
/* First check all names if they are unique */
for (i = 0; i != gc->ngpio; ++i) {
struct gpio_desc *gpio;
gpio = gpio_name_to_desc(gc->names[i]);
if (gpio)
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
dev_warn(&gdev->dev,
"Detected name collision for GPIO name '%s'\n",
gc->names[i]);
}
/* Then add all names to the GPIO descriptors */
for (i = 0; i != gc->ngpio; ++i)
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
gdev->descs[i].name = gc->names[i];
return 0;
}
gpio: userspace ABI for reading/writing GPIO lines This adds a userspace ABI for reading and writing GPIO lines. The mechanism returns an anonymous file handle to a request to read/write n offsets from a gpiochip. This file handle in turn accepts two ioctl()s: one that reads and one that writes values to the selected lines. - Handles can be requested as input/output, active low, open drain, open source, however when you issue a request for n lines with GPIO_GET_LINEHANDLE_IOCTL, they must all have the same flags, i.e. all inputs or all outputs, all open drain etc. If a granular control of the flags for each line is desired, they need to be requested individually, not in a batch. - The GPIOHANDLE_GET_LINE_VALUES_IOCTL read ioctl() can be issued also to output lines to verify that the hardware is in the expected state. - It reads and writes up to GPIOHANDLES_MAX lines at once, utilizing the .set_multiple() call in the driver if possible, making the call efficient if several lines can be written with a single register update. The limitation of GPIOHANDLES_MAX to 64 lines is done under the assumption that we may expect hardware that can issue a transaction updating 64 bits at an instant but unlikely anything larger than that. ChangeLog v2->v3: - Use gpiod_get_value_cansleep() so we support also slowpath GPIO drivers. - Fix up the UAPI docs kerneldoc. - Allocate the anonymous fd last, so that the release function don't get called until that point of something fails. After this point, skip the errorpath. ChangeLog v1->v2: - Handle ioctl_compat() properly based on a similar patch to the other ioctl() handling code. - Use _IOWR() as we pass pointers both in and out of the ioctl() - Use kmalloc() and kfree() for the linehandled, do not try to be fancy with devm_* it doesn't work the way I thought. - Fix const-correctness on the linehandle name field. Acked-by: Michael Welling <mwelling@ieee.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-04-26 16:35:29 +08:00
/*
* GPIO line handle management
*/
/**
* struct linehandle_state - contains the state of a userspace handle
* @gdev: the GPIO device the handle pertains to
* @label: consumer label used to tag descriptors
* @descs: the GPIO descriptors held by this handle
* @numdescs: the number of descriptors held in the descs array
*/
struct linehandle_state {
struct gpio_device *gdev;
const char *label;
struct gpio_desc *descs[GPIOHANDLES_MAX];
u32 numdescs;
};
#define GPIOHANDLE_REQUEST_VALID_FLAGS \
(GPIOHANDLE_REQUEST_INPUT | \
GPIOHANDLE_REQUEST_OUTPUT | \
GPIOHANDLE_REQUEST_ACTIVE_LOW | \
GPIOHANDLE_REQUEST_OPEN_DRAIN | \
GPIOHANDLE_REQUEST_OPEN_SOURCE)
gpio: userspace ABI for reading/writing GPIO lines This adds a userspace ABI for reading and writing GPIO lines. The mechanism returns an anonymous file handle to a request to read/write n offsets from a gpiochip. This file handle in turn accepts two ioctl()s: one that reads and one that writes values to the selected lines. - Handles can be requested as input/output, active low, open drain, open source, however when you issue a request for n lines with GPIO_GET_LINEHANDLE_IOCTL, they must all have the same flags, i.e. all inputs or all outputs, all open drain etc. If a granular control of the flags for each line is desired, they need to be requested individually, not in a batch. - The GPIOHANDLE_GET_LINE_VALUES_IOCTL read ioctl() can be issued also to output lines to verify that the hardware is in the expected state. - It reads and writes up to GPIOHANDLES_MAX lines at once, utilizing the .set_multiple() call in the driver if possible, making the call efficient if several lines can be written with a single register update. The limitation of GPIOHANDLES_MAX to 64 lines is done under the assumption that we may expect hardware that can issue a transaction updating 64 bits at an instant but unlikely anything larger than that. ChangeLog v2->v3: - Use gpiod_get_value_cansleep() so we support also slowpath GPIO drivers. - Fix up the UAPI docs kerneldoc. - Allocate the anonymous fd last, so that the release function don't get called until that point of something fails. After this point, skip the errorpath. ChangeLog v1->v2: - Handle ioctl_compat() properly based on a similar patch to the other ioctl() handling code. - Use _IOWR() as we pass pointers both in and out of the ioctl() - Use kmalloc() and kfree() for the linehandled, do not try to be fancy with devm_* it doesn't work the way I thought. - Fix const-correctness on the linehandle name field. Acked-by: Michael Welling <mwelling@ieee.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-04-26 16:35:29 +08:00
static long linehandle_ioctl(struct file *filep, unsigned int cmd,
unsigned long arg)
{
struct linehandle_state *lh = filep->private_data;
void __user *ip = (void __user *)arg;
struct gpiohandle_data ghd;
int vals[GPIOHANDLES_MAX];
gpio: userspace ABI for reading/writing GPIO lines This adds a userspace ABI for reading and writing GPIO lines. The mechanism returns an anonymous file handle to a request to read/write n offsets from a gpiochip. This file handle in turn accepts two ioctl()s: one that reads and one that writes values to the selected lines. - Handles can be requested as input/output, active low, open drain, open source, however when you issue a request for n lines with GPIO_GET_LINEHANDLE_IOCTL, they must all have the same flags, i.e. all inputs or all outputs, all open drain etc. If a granular control of the flags for each line is desired, they need to be requested individually, not in a batch. - The GPIOHANDLE_GET_LINE_VALUES_IOCTL read ioctl() can be issued also to output lines to verify that the hardware is in the expected state. - It reads and writes up to GPIOHANDLES_MAX lines at once, utilizing the .set_multiple() call in the driver if possible, making the call efficient if several lines can be written with a single register update. The limitation of GPIOHANDLES_MAX to 64 lines is done under the assumption that we may expect hardware that can issue a transaction updating 64 bits at an instant but unlikely anything larger than that. ChangeLog v2->v3: - Use gpiod_get_value_cansleep() so we support also slowpath GPIO drivers. - Fix up the UAPI docs kerneldoc. - Allocate the anonymous fd last, so that the release function don't get called until that point of something fails. After this point, skip the errorpath. ChangeLog v1->v2: - Handle ioctl_compat() properly based on a similar patch to the other ioctl() handling code. - Use _IOWR() as we pass pointers both in and out of the ioctl() - Use kmalloc() and kfree() for the linehandled, do not try to be fancy with devm_* it doesn't work the way I thought. - Fix const-correctness on the linehandle name field. Acked-by: Michael Welling <mwelling@ieee.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-04-26 16:35:29 +08:00
int i;
if (cmd == GPIOHANDLE_GET_LINE_VALUES_IOCTL) {
/* TODO: check if descriptors are really input */
int ret = gpiod_get_array_value_complex(false,
true,
lh->numdescs,
lh->descs,
vals);
if (ret)
return ret;
gpio: userspace ABI for reading/writing GPIO lines This adds a userspace ABI for reading and writing GPIO lines. The mechanism returns an anonymous file handle to a request to read/write n offsets from a gpiochip. This file handle in turn accepts two ioctl()s: one that reads and one that writes values to the selected lines. - Handles can be requested as input/output, active low, open drain, open source, however when you issue a request for n lines with GPIO_GET_LINEHANDLE_IOCTL, they must all have the same flags, i.e. all inputs or all outputs, all open drain etc. If a granular control of the flags for each line is desired, they need to be requested individually, not in a batch. - The GPIOHANDLE_GET_LINE_VALUES_IOCTL read ioctl() can be issued also to output lines to verify that the hardware is in the expected state. - It reads and writes up to GPIOHANDLES_MAX lines at once, utilizing the .set_multiple() call in the driver if possible, making the call efficient if several lines can be written with a single register update. The limitation of GPIOHANDLES_MAX to 64 lines is done under the assumption that we may expect hardware that can issue a transaction updating 64 bits at an instant but unlikely anything larger than that. ChangeLog v2->v3: - Use gpiod_get_value_cansleep() so we support also slowpath GPIO drivers. - Fix up the UAPI docs kerneldoc. - Allocate the anonymous fd last, so that the release function don't get called until that point of something fails. After this point, skip the errorpath. ChangeLog v1->v2: - Handle ioctl_compat() properly based on a similar patch to the other ioctl() handling code. - Use _IOWR() as we pass pointers both in and out of the ioctl() - Use kmalloc() and kfree() for the linehandled, do not try to be fancy with devm_* it doesn't work the way I thought. - Fix const-correctness on the linehandle name field. Acked-by: Michael Welling <mwelling@ieee.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-04-26 16:35:29 +08:00
memset(&ghd, 0, sizeof(ghd));
for (i = 0; i < lh->numdescs; i++)
ghd.values[i] = vals[i];
gpio: userspace ABI for reading/writing GPIO lines This adds a userspace ABI for reading and writing GPIO lines. The mechanism returns an anonymous file handle to a request to read/write n offsets from a gpiochip. This file handle in turn accepts two ioctl()s: one that reads and one that writes values to the selected lines. - Handles can be requested as input/output, active low, open drain, open source, however when you issue a request for n lines with GPIO_GET_LINEHANDLE_IOCTL, they must all have the same flags, i.e. all inputs or all outputs, all open drain etc. If a granular control of the flags for each line is desired, they need to be requested individually, not in a batch. - The GPIOHANDLE_GET_LINE_VALUES_IOCTL read ioctl() can be issued also to output lines to verify that the hardware is in the expected state. - It reads and writes up to GPIOHANDLES_MAX lines at once, utilizing the .set_multiple() call in the driver if possible, making the call efficient if several lines can be written with a single register update. The limitation of GPIOHANDLES_MAX to 64 lines is done under the assumption that we may expect hardware that can issue a transaction updating 64 bits at an instant but unlikely anything larger than that. ChangeLog v2->v3: - Use gpiod_get_value_cansleep() so we support also slowpath GPIO drivers. - Fix up the UAPI docs kerneldoc. - Allocate the anonymous fd last, so that the release function don't get called until that point of something fails. After this point, skip the errorpath. ChangeLog v1->v2: - Handle ioctl_compat() properly based on a similar patch to the other ioctl() handling code. - Use _IOWR() as we pass pointers both in and out of the ioctl() - Use kmalloc() and kfree() for the linehandled, do not try to be fancy with devm_* it doesn't work the way I thought. - Fix const-correctness on the linehandle name field. Acked-by: Michael Welling <mwelling@ieee.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-04-26 16:35:29 +08:00
if (copy_to_user(ip, &ghd, sizeof(ghd)))
return -EFAULT;
return 0;
} else if (cmd == GPIOHANDLE_SET_LINE_VALUES_IOCTL) {
/* TODO: check if descriptors are really output */
if (copy_from_user(&ghd, ip, sizeof(ghd)))
return -EFAULT;
/* Clamp all values to [0,1] */
for (i = 0; i < lh->numdescs; i++)
vals[i] = !!ghd.values[i];
/* Reuse the array setting function */
gpiod_set_array_value_complex(false,
true,
lh->numdescs,
lh->descs,
vals);
return 0;
}
return -EINVAL;
}
#ifdef CONFIG_COMPAT
static long linehandle_ioctl_compat(struct file *filep, unsigned int cmd,
unsigned long arg)
{
return linehandle_ioctl(filep, cmd, (unsigned long)compat_ptr(arg));
}
#endif
static int linehandle_release(struct inode *inode, struct file *filep)
{
struct linehandle_state *lh = filep->private_data;
struct gpio_device *gdev = lh->gdev;
int i;
for (i = 0; i < lh->numdescs; i++)
gpiod_free(lh->descs[i]);
kfree(lh->label);
kfree(lh);
put_device(&gdev->dev);
return 0;
}
static const struct file_operations linehandle_fileops = {
.release = linehandle_release,
.owner = THIS_MODULE,
.llseek = noop_llseek,
.unlocked_ioctl = linehandle_ioctl,
#ifdef CONFIG_COMPAT
.compat_ioctl = linehandle_ioctl_compat,
#endif
};
static int linehandle_create(struct gpio_device *gdev, void __user *ip)
{
struct gpiohandle_request handlereq;
struct linehandle_state *lh;
gpio: GPIO_GET_LINE{HANDLE,EVENT}_IOCTL: Fix file descriptor leak When allocating a new line handle or event a file is allocated that it is associated to. The file is attached to a file descriptor of the current process and the file descriptor is returned to userspace using copy_to_user(). If this copy operation fails the line handle or event allocation is aborted, all acquired resources are freed and an error is returned. But the file struct is not freed and left attached to the userspace application and even though the file descriptor number was not copied it is trivial to guess. If a userspace application performs a IOCTL on such a left over file descriptor it will trigger a use-after-free and if the file descriptor is closed (latest when the application exits) a double-free is triggered. anon_inode_getfd() performs 3 tasks, allocate a file struct, allocate a file descriptor for the current process and install the file struct in the file descriptor. As soon as the file struct is installed in the file descriptor it is accessible by userspace (even if the IOCTL itself hasn't completed yet), this means uninstalling the fd on the error path is not an option, since userspace might already got a reference to the file. Instead anon_inode_getfd() needs to be broken into its individual steps. The allocation of the file struct and file descriptor is done first, then the copy_to_user() is executed and only if it succeeds the file is installed. Since the file struct is reference counted it can not be just freed, but its reference needs to be dropped, which will also call the release() callback, which will free the state attached to the file. So in this case the normal error cleanup path should not be taken. Cc: stable@vger.kernel.org Fixes: d932cd49182f ("gpio: free handles in fringe cases") Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-10-24 19:59:15 +08:00
struct file *file;
gpio: userspace ABI for reading/writing GPIO lines This adds a userspace ABI for reading and writing GPIO lines. The mechanism returns an anonymous file handle to a request to read/write n offsets from a gpiochip. This file handle in turn accepts two ioctl()s: one that reads and one that writes values to the selected lines. - Handles can be requested as input/output, active low, open drain, open source, however when you issue a request for n lines with GPIO_GET_LINEHANDLE_IOCTL, they must all have the same flags, i.e. all inputs or all outputs, all open drain etc. If a granular control of the flags for each line is desired, they need to be requested individually, not in a batch. - The GPIOHANDLE_GET_LINE_VALUES_IOCTL read ioctl() can be issued also to output lines to verify that the hardware is in the expected state. - It reads and writes up to GPIOHANDLES_MAX lines at once, utilizing the .set_multiple() call in the driver if possible, making the call efficient if several lines can be written with a single register update. The limitation of GPIOHANDLES_MAX to 64 lines is done under the assumption that we may expect hardware that can issue a transaction updating 64 bits at an instant but unlikely anything larger than that. ChangeLog v2->v3: - Use gpiod_get_value_cansleep() so we support also slowpath GPIO drivers. - Fix up the UAPI docs kerneldoc. - Allocate the anonymous fd last, so that the release function don't get called until that point of something fails. After this point, skip the errorpath. ChangeLog v1->v2: - Handle ioctl_compat() properly based on a similar patch to the other ioctl() handling code. - Use _IOWR() as we pass pointers both in and out of the ioctl() - Use kmalloc() and kfree() for the linehandled, do not try to be fancy with devm_* it doesn't work the way I thought. - Fix const-correctness on the linehandle name field. Acked-by: Michael Welling <mwelling@ieee.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-04-26 16:35:29 +08:00
int fd, i, ret;
u32 lflags;
gpio: userspace ABI for reading/writing GPIO lines This adds a userspace ABI for reading and writing GPIO lines. The mechanism returns an anonymous file handle to a request to read/write n offsets from a gpiochip. This file handle in turn accepts two ioctl()s: one that reads and one that writes values to the selected lines. - Handles can be requested as input/output, active low, open drain, open source, however when you issue a request for n lines with GPIO_GET_LINEHANDLE_IOCTL, they must all have the same flags, i.e. all inputs or all outputs, all open drain etc. If a granular control of the flags for each line is desired, they need to be requested individually, not in a batch. - The GPIOHANDLE_GET_LINE_VALUES_IOCTL read ioctl() can be issued also to output lines to verify that the hardware is in the expected state. - It reads and writes up to GPIOHANDLES_MAX lines at once, utilizing the .set_multiple() call in the driver if possible, making the call efficient if several lines can be written with a single register update. The limitation of GPIOHANDLES_MAX to 64 lines is done under the assumption that we may expect hardware that can issue a transaction updating 64 bits at an instant but unlikely anything larger than that. ChangeLog v2->v3: - Use gpiod_get_value_cansleep() so we support also slowpath GPIO drivers. - Fix up the UAPI docs kerneldoc. - Allocate the anonymous fd last, so that the release function don't get called until that point of something fails. After this point, skip the errorpath. ChangeLog v1->v2: - Handle ioctl_compat() properly based on a similar patch to the other ioctl() handling code. - Use _IOWR() as we pass pointers both in and out of the ioctl() - Use kmalloc() and kfree() for the linehandled, do not try to be fancy with devm_* it doesn't work the way I thought. - Fix const-correctness on the linehandle name field. Acked-by: Michael Welling <mwelling@ieee.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-04-26 16:35:29 +08:00
if (copy_from_user(&handlereq, ip, sizeof(handlereq)))
return -EFAULT;
if ((handlereq.lines == 0) || (handlereq.lines > GPIOHANDLES_MAX))
return -EINVAL;
lflags = handlereq.flags;
/* Return an error if an unknown flag is set */
if (lflags & ~GPIOHANDLE_REQUEST_VALID_FLAGS)
return -EINVAL;
/*
* Do not allow OPEN_SOURCE & OPEN_DRAIN flags in a single request. If
* the hardware actually supports enabling both at the same time the
* electrical result would be disastrous.
*/
if ((lflags & GPIOHANDLE_REQUEST_OPEN_DRAIN) &&
(lflags & GPIOHANDLE_REQUEST_OPEN_SOURCE))
return -EINVAL;
/* OPEN_DRAIN and OPEN_SOURCE flags only make sense for output mode. */
if (!(lflags & GPIOHANDLE_REQUEST_OUTPUT) &&
((lflags & GPIOHANDLE_REQUEST_OPEN_DRAIN) ||
(lflags & GPIOHANDLE_REQUEST_OPEN_SOURCE)))
return -EINVAL;
gpio: userspace ABI for reading/writing GPIO lines This adds a userspace ABI for reading and writing GPIO lines. The mechanism returns an anonymous file handle to a request to read/write n offsets from a gpiochip. This file handle in turn accepts two ioctl()s: one that reads and one that writes values to the selected lines. - Handles can be requested as input/output, active low, open drain, open source, however when you issue a request for n lines with GPIO_GET_LINEHANDLE_IOCTL, they must all have the same flags, i.e. all inputs or all outputs, all open drain etc. If a granular control of the flags for each line is desired, they need to be requested individually, not in a batch. - The GPIOHANDLE_GET_LINE_VALUES_IOCTL read ioctl() can be issued also to output lines to verify that the hardware is in the expected state. - It reads and writes up to GPIOHANDLES_MAX lines at once, utilizing the .set_multiple() call in the driver if possible, making the call efficient if several lines can be written with a single register update. The limitation of GPIOHANDLES_MAX to 64 lines is done under the assumption that we may expect hardware that can issue a transaction updating 64 bits at an instant but unlikely anything larger than that. ChangeLog v2->v3: - Use gpiod_get_value_cansleep() so we support also slowpath GPIO drivers. - Fix up the UAPI docs kerneldoc. - Allocate the anonymous fd last, so that the release function don't get called until that point of something fails. After this point, skip the errorpath. ChangeLog v1->v2: - Handle ioctl_compat() properly based on a similar patch to the other ioctl() handling code. - Use _IOWR() as we pass pointers both in and out of the ioctl() - Use kmalloc() and kfree() for the linehandled, do not try to be fancy with devm_* it doesn't work the way I thought. - Fix const-correctness on the linehandle name field. Acked-by: Michael Welling <mwelling@ieee.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-04-26 16:35:29 +08:00
lh = kzalloc(sizeof(*lh), GFP_KERNEL);
if (!lh)
return -ENOMEM;
lh->gdev = gdev;
get_device(&gdev->dev);
/* Make sure this is terminated */
handlereq.consumer_label[sizeof(handlereq.consumer_label)-1] = '\0';
if (strlen(handlereq.consumer_label)) {
lh->label = kstrdup(handlereq.consumer_label,
GFP_KERNEL);
if (!lh->label) {
ret = -ENOMEM;
goto out_free_lh;
}
}
/* Request each GPIO */
for (i = 0; i < handlereq.lines; i++) {
u32 offset = handlereq.lineoffsets[i];
struct gpio_desc *desc;
if (offset >= gdev->ngpio) {
ret = -EINVAL;
goto out_free_descs;
}
gpio: userspace ABI for reading/writing GPIO lines This adds a userspace ABI for reading and writing GPIO lines. The mechanism returns an anonymous file handle to a request to read/write n offsets from a gpiochip. This file handle in turn accepts two ioctl()s: one that reads and one that writes values to the selected lines. - Handles can be requested as input/output, active low, open drain, open source, however when you issue a request for n lines with GPIO_GET_LINEHANDLE_IOCTL, they must all have the same flags, i.e. all inputs or all outputs, all open drain etc. If a granular control of the flags for each line is desired, they need to be requested individually, not in a batch. - The GPIOHANDLE_GET_LINE_VALUES_IOCTL read ioctl() can be issued also to output lines to verify that the hardware is in the expected state. - It reads and writes up to GPIOHANDLES_MAX lines at once, utilizing the .set_multiple() call in the driver if possible, making the call efficient if several lines can be written with a single register update. The limitation of GPIOHANDLES_MAX to 64 lines is done under the assumption that we may expect hardware that can issue a transaction updating 64 bits at an instant but unlikely anything larger than that. ChangeLog v2->v3: - Use gpiod_get_value_cansleep() so we support also slowpath GPIO drivers. - Fix up the UAPI docs kerneldoc. - Allocate the anonymous fd last, so that the release function don't get called until that point of something fails. After this point, skip the errorpath. ChangeLog v1->v2: - Handle ioctl_compat() properly based on a similar patch to the other ioctl() handling code. - Use _IOWR() as we pass pointers both in and out of the ioctl() - Use kmalloc() and kfree() for the linehandled, do not try to be fancy with devm_* it doesn't work the way I thought. - Fix const-correctness on the linehandle name field. Acked-by: Michael Welling <mwelling@ieee.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-04-26 16:35:29 +08:00
desc = &gdev->descs[offset];
ret = gpiod_request(desc, lh->label);
if (ret)
goto out_free_descs;
lh->descs[i] = desc;
if (lflags & GPIOHANDLE_REQUEST_ACTIVE_LOW)
set_bit(FLAG_ACTIVE_LOW, &desc->flags);
if (lflags & GPIOHANDLE_REQUEST_OPEN_DRAIN)
set_bit(FLAG_OPEN_DRAIN, &desc->flags);
if (lflags & GPIOHANDLE_REQUEST_OPEN_SOURCE)
set_bit(FLAG_OPEN_SOURCE, &desc->flags);
ret = gpiod_set_transitory(desc, false);
if (ret < 0)
goto out_free_descs;
gpio: userspace ABI for reading/writing GPIO lines This adds a userspace ABI for reading and writing GPIO lines. The mechanism returns an anonymous file handle to a request to read/write n offsets from a gpiochip. This file handle in turn accepts two ioctl()s: one that reads and one that writes values to the selected lines. - Handles can be requested as input/output, active low, open drain, open source, however when you issue a request for n lines with GPIO_GET_LINEHANDLE_IOCTL, they must all have the same flags, i.e. all inputs or all outputs, all open drain etc. If a granular control of the flags for each line is desired, they need to be requested individually, not in a batch. - The GPIOHANDLE_GET_LINE_VALUES_IOCTL read ioctl() can be issued also to output lines to verify that the hardware is in the expected state. - It reads and writes up to GPIOHANDLES_MAX lines at once, utilizing the .set_multiple() call in the driver if possible, making the call efficient if several lines can be written with a single register update. The limitation of GPIOHANDLES_MAX to 64 lines is done under the assumption that we may expect hardware that can issue a transaction updating 64 bits at an instant but unlikely anything larger than that. ChangeLog v2->v3: - Use gpiod_get_value_cansleep() so we support also slowpath GPIO drivers. - Fix up the UAPI docs kerneldoc. - Allocate the anonymous fd last, so that the release function don't get called until that point of something fails. After this point, skip the errorpath. ChangeLog v1->v2: - Handle ioctl_compat() properly based on a similar patch to the other ioctl() handling code. - Use _IOWR() as we pass pointers both in and out of the ioctl() - Use kmalloc() and kfree() for the linehandled, do not try to be fancy with devm_* it doesn't work the way I thought. - Fix const-correctness on the linehandle name field. Acked-by: Michael Welling <mwelling@ieee.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-04-26 16:35:29 +08:00
/*
* Lines have to be requested explicitly for input
* or output, else the line will be treated "as is".
*/
if (lflags & GPIOHANDLE_REQUEST_OUTPUT) {
int val = !!handlereq.default_values[i];
ret = gpiod_direction_output(desc, val);
if (ret)
goto out_free_descs;
} else if (lflags & GPIOHANDLE_REQUEST_INPUT) {
ret = gpiod_direction_input(desc);
if (ret)
goto out_free_descs;
}
dev_dbg(&gdev->dev, "registered chardev handle for line %d\n",
offset);
}
/* Let i point at the last handle */
i--;
gpio: userspace ABI for reading/writing GPIO lines This adds a userspace ABI for reading and writing GPIO lines. The mechanism returns an anonymous file handle to a request to read/write n offsets from a gpiochip. This file handle in turn accepts two ioctl()s: one that reads and one that writes values to the selected lines. - Handles can be requested as input/output, active low, open drain, open source, however when you issue a request for n lines with GPIO_GET_LINEHANDLE_IOCTL, they must all have the same flags, i.e. all inputs or all outputs, all open drain etc. If a granular control of the flags for each line is desired, they need to be requested individually, not in a batch. - The GPIOHANDLE_GET_LINE_VALUES_IOCTL read ioctl() can be issued also to output lines to verify that the hardware is in the expected state. - It reads and writes up to GPIOHANDLES_MAX lines at once, utilizing the .set_multiple() call in the driver if possible, making the call efficient if several lines can be written with a single register update. The limitation of GPIOHANDLES_MAX to 64 lines is done under the assumption that we may expect hardware that can issue a transaction updating 64 bits at an instant but unlikely anything larger than that. ChangeLog v2->v3: - Use gpiod_get_value_cansleep() so we support also slowpath GPIO drivers. - Fix up the UAPI docs kerneldoc. - Allocate the anonymous fd last, so that the release function don't get called until that point of something fails. After this point, skip the errorpath. ChangeLog v1->v2: - Handle ioctl_compat() properly based on a similar patch to the other ioctl() handling code. - Use _IOWR() as we pass pointers both in and out of the ioctl() - Use kmalloc() and kfree() for the linehandled, do not try to be fancy with devm_* it doesn't work the way I thought. - Fix const-correctness on the linehandle name field. Acked-by: Michael Welling <mwelling@ieee.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-04-26 16:35:29 +08:00
lh->numdescs = handlereq.lines;
gpio: GPIO_GET_LINE{HANDLE,EVENT}_IOCTL: Fix file descriptor leak When allocating a new line handle or event a file is allocated that it is associated to. The file is attached to a file descriptor of the current process and the file descriptor is returned to userspace using copy_to_user(). If this copy operation fails the line handle or event allocation is aborted, all acquired resources are freed and an error is returned. But the file struct is not freed and left attached to the userspace application and even though the file descriptor number was not copied it is trivial to guess. If a userspace application performs a IOCTL on such a left over file descriptor it will trigger a use-after-free and if the file descriptor is closed (latest when the application exits) a double-free is triggered. anon_inode_getfd() performs 3 tasks, allocate a file struct, allocate a file descriptor for the current process and install the file struct in the file descriptor. As soon as the file struct is installed in the file descriptor it is accessible by userspace (even if the IOCTL itself hasn't completed yet), this means uninstalling the fd on the error path is not an option, since userspace might already got a reference to the file. Instead anon_inode_getfd() needs to be broken into its individual steps. The allocation of the file struct and file descriptor is done first, then the copy_to_user() is executed and only if it succeeds the file is installed. Since the file struct is reference counted it can not be just freed, but its reference needs to be dropped, which will also call the release() callback, which will free the state attached to the file. So in this case the normal error cleanup path should not be taken. Cc: stable@vger.kernel.org Fixes: d932cd49182f ("gpio: free handles in fringe cases") Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-10-24 19:59:15 +08:00
fd = get_unused_fd_flags(O_RDONLY | O_CLOEXEC);
gpio: userspace ABI for reading/writing GPIO lines This adds a userspace ABI for reading and writing GPIO lines. The mechanism returns an anonymous file handle to a request to read/write n offsets from a gpiochip. This file handle in turn accepts two ioctl()s: one that reads and one that writes values to the selected lines. - Handles can be requested as input/output, active low, open drain, open source, however when you issue a request for n lines with GPIO_GET_LINEHANDLE_IOCTL, they must all have the same flags, i.e. all inputs or all outputs, all open drain etc. If a granular control of the flags for each line is desired, they need to be requested individually, not in a batch. - The GPIOHANDLE_GET_LINE_VALUES_IOCTL read ioctl() can be issued also to output lines to verify that the hardware is in the expected state. - It reads and writes up to GPIOHANDLES_MAX lines at once, utilizing the .set_multiple() call in the driver if possible, making the call efficient if several lines can be written with a single register update. The limitation of GPIOHANDLES_MAX to 64 lines is done under the assumption that we may expect hardware that can issue a transaction updating 64 bits at an instant but unlikely anything larger than that. ChangeLog v2->v3: - Use gpiod_get_value_cansleep() so we support also slowpath GPIO drivers. - Fix up the UAPI docs kerneldoc. - Allocate the anonymous fd last, so that the release function don't get called until that point of something fails. After this point, skip the errorpath. ChangeLog v1->v2: - Handle ioctl_compat() properly based on a similar patch to the other ioctl() handling code. - Use _IOWR() as we pass pointers both in and out of the ioctl() - Use kmalloc() and kfree() for the linehandled, do not try to be fancy with devm_* it doesn't work the way I thought. - Fix const-correctness on the linehandle name field. Acked-by: Michael Welling <mwelling@ieee.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-04-26 16:35:29 +08:00
if (fd < 0) {
ret = fd;
goto out_free_descs;
}
gpio: GPIO_GET_LINE{HANDLE,EVENT}_IOCTL: Fix file descriptor leak When allocating a new line handle or event a file is allocated that it is associated to. The file is attached to a file descriptor of the current process and the file descriptor is returned to userspace using copy_to_user(). If this copy operation fails the line handle or event allocation is aborted, all acquired resources are freed and an error is returned. But the file struct is not freed and left attached to the userspace application and even though the file descriptor number was not copied it is trivial to guess. If a userspace application performs a IOCTL on such a left over file descriptor it will trigger a use-after-free and if the file descriptor is closed (latest when the application exits) a double-free is triggered. anon_inode_getfd() performs 3 tasks, allocate a file struct, allocate a file descriptor for the current process and install the file struct in the file descriptor. As soon as the file struct is installed in the file descriptor it is accessible by userspace (even if the IOCTL itself hasn't completed yet), this means uninstalling the fd on the error path is not an option, since userspace might already got a reference to the file. Instead anon_inode_getfd() needs to be broken into its individual steps. The allocation of the file struct and file descriptor is done first, then the copy_to_user() is executed and only if it succeeds the file is installed. Since the file struct is reference counted it can not be just freed, but its reference needs to be dropped, which will also call the release() callback, which will free the state attached to the file. So in this case the normal error cleanup path should not be taken. Cc: stable@vger.kernel.org Fixes: d932cd49182f ("gpio: free handles in fringe cases") Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-10-24 19:59:15 +08:00
file = anon_inode_getfile("gpio-linehandle",
&linehandle_fileops,
lh,
O_RDONLY | O_CLOEXEC);
if (IS_ERR(file)) {
ret = PTR_ERR(file);
goto out_put_unused_fd;
}
gpio: userspace ABI for reading/writing GPIO lines This adds a userspace ABI for reading and writing GPIO lines. The mechanism returns an anonymous file handle to a request to read/write n offsets from a gpiochip. This file handle in turn accepts two ioctl()s: one that reads and one that writes values to the selected lines. - Handles can be requested as input/output, active low, open drain, open source, however when you issue a request for n lines with GPIO_GET_LINEHANDLE_IOCTL, they must all have the same flags, i.e. all inputs or all outputs, all open drain etc. If a granular control of the flags for each line is desired, they need to be requested individually, not in a batch. - The GPIOHANDLE_GET_LINE_VALUES_IOCTL read ioctl() can be issued also to output lines to verify that the hardware is in the expected state. - It reads and writes up to GPIOHANDLES_MAX lines at once, utilizing the .set_multiple() call in the driver if possible, making the call efficient if several lines can be written with a single register update. The limitation of GPIOHANDLES_MAX to 64 lines is done under the assumption that we may expect hardware that can issue a transaction updating 64 bits at an instant but unlikely anything larger than that. ChangeLog v2->v3: - Use gpiod_get_value_cansleep() so we support also slowpath GPIO drivers. - Fix up the UAPI docs kerneldoc. - Allocate the anonymous fd last, so that the release function don't get called until that point of something fails. After this point, skip the errorpath. ChangeLog v1->v2: - Handle ioctl_compat() properly based on a similar patch to the other ioctl() handling code. - Use _IOWR() as we pass pointers both in and out of the ioctl() - Use kmalloc() and kfree() for the linehandled, do not try to be fancy with devm_* it doesn't work the way I thought. - Fix const-correctness on the linehandle name field. Acked-by: Michael Welling <mwelling@ieee.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-04-26 16:35:29 +08:00
handlereq.fd = fd;
if (copy_to_user(ip, &handlereq, sizeof(handlereq))) {
gpio: GPIO_GET_LINE{HANDLE,EVENT}_IOCTL: Fix file descriptor leak When allocating a new line handle or event a file is allocated that it is associated to. The file is attached to a file descriptor of the current process and the file descriptor is returned to userspace using copy_to_user(). If this copy operation fails the line handle or event allocation is aborted, all acquired resources are freed and an error is returned. But the file struct is not freed and left attached to the userspace application and even though the file descriptor number was not copied it is trivial to guess. If a userspace application performs a IOCTL on such a left over file descriptor it will trigger a use-after-free and if the file descriptor is closed (latest when the application exits) a double-free is triggered. anon_inode_getfd() performs 3 tasks, allocate a file struct, allocate a file descriptor for the current process and install the file struct in the file descriptor. As soon as the file struct is installed in the file descriptor it is accessible by userspace (even if the IOCTL itself hasn't completed yet), this means uninstalling the fd on the error path is not an option, since userspace might already got a reference to the file. Instead anon_inode_getfd() needs to be broken into its individual steps. The allocation of the file struct and file descriptor is done first, then the copy_to_user() is executed and only if it succeeds the file is installed. Since the file struct is reference counted it can not be just freed, but its reference needs to be dropped, which will also call the release() callback, which will free the state attached to the file. So in this case the normal error cleanup path should not be taken. Cc: stable@vger.kernel.org Fixes: d932cd49182f ("gpio: free handles in fringe cases") Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-10-24 19:59:15 +08:00
/*
* fput() will trigger the release() callback, so do not go onto
* the regular error cleanup path here.
*/
fput(file);
put_unused_fd(fd);
return -EFAULT;
}
gpio: userspace ABI for reading/writing GPIO lines This adds a userspace ABI for reading and writing GPIO lines. The mechanism returns an anonymous file handle to a request to read/write n offsets from a gpiochip. This file handle in turn accepts two ioctl()s: one that reads and one that writes values to the selected lines. - Handles can be requested as input/output, active low, open drain, open source, however when you issue a request for n lines with GPIO_GET_LINEHANDLE_IOCTL, they must all have the same flags, i.e. all inputs or all outputs, all open drain etc. If a granular control of the flags for each line is desired, they need to be requested individually, not in a batch. - The GPIOHANDLE_GET_LINE_VALUES_IOCTL read ioctl() can be issued also to output lines to verify that the hardware is in the expected state. - It reads and writes up to GPIOHANDLES_MAX lines at once, utilizing the .set_multiple() call in the driver if possible, making the call efficient if several lines can be written with a single register update. The limitation of GPIOHANDLES_MAX to 64 lines is done under the assumption that we may expect hardware that can issue a transaction updating 64 bits at an instant but unlikely anything larger than that. ChangeLog v2->v3: - Use gpiod_get_value_cansleep() so we support also slowpath GPIO drivers. - Fix up the UAPI docs kerneldoc. - Allocate the anonymous fd last, so that the release function don't get called until that point of something fails. After this point, skip the errorpath. ChangeLog v1->v2: - Handle ioctl_compat() properly based on a similar patch to the other ioctl() handling code. - Use _IOWR() as we pass pointers both in and out of the ioctl() - Use kmalloc() and kfree() for the linehandled, do not try to be fancy with devm_* it doesn't work the way I thought. - Fix const-correctness on the linehandle name field. Acked-by: Michael Welling <mwelling@ieee.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-04-26 16:35:29 +08:00
gpio: GPIO_GET_LINE{HANDLE,EVENT}_IOCTL: Fix file descriptor leak When allocating a new line handle or event a file is allocated that it is associated to. The file is attached to a file descriptor of the current process and the file descriptor is returned to userspace using copy_to_user(). If this copy operation fails the line handle or event allocation is aborted, all acquired resources are freed and an error is returned. But the file struct is not freed and left attached to the userspace application and even though the file descriptor number was not copied it is trivial to guess. If a userspace application performs a IOCTL on such a left over file descriptor it will trigger a use-after-free and if the file descriptor is closed (latest when the application exits) a double-free is triggered. anon_inode_getfd() performs 3 tasks, allocate a file struct, allocate a file descriptor for the current process and install the file struct in the file descriptor. As soon as the file struct is installed in the file descriptor it is accessible by userspace (even if the IOCTL itself hasn't completed yet), this means uninstalling the fd on the error path is not an option, since userspace might already got a reference to the file. Instead anon_inode_getfd() needs to be broken into its individual steps. The allocation of the file struct and file descriptor is done first, then the copy_to_user() is executed and only if it succeeds the file is installed. Since the file struct is reference counted it can not be just freed, but its reference needs to be dropped, which will also call the release() callback, which will free the state attached to the file. So in this case the normal error cleanup path should not be taken. Cc: stable@vger.kernel.org Fixes: d932cd49182f ("gpio: free handles in fringe cases") Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-10-24 19:59:15 +08:00
fd_install(fd, file);
gpio: userspace ABI for reading/writing GPIO lines This adds a userspace ABI for reading and writing GPIO lines. The mechanism returns an anonymous file handle to a request to read/write n offsets from a gpiochip. This file handle in turn accepts two ioctl()s: one that reads and one that writes values to the selected lines. - Handles can be requested as input/output, active low, open drain, open source, however when you issue a request for n lines with GPIO_GET_LINEHANDLE_IOCTL, they must all have the same flags, i.e. all inputs or all outputs, all open drain etc. If a granular control of the flags for each line is desired, they need to be requested individually, not in a batch. - The GPIOHANDLE_GET_LINE_VALUES_IOCTL read ioctl() can be issued also to output lines to verify that the hardware is in the expected state. - It reads and writes up to GPIOHANDLES_MAX lines at once, utilizing the .set_multiple() call in the driver if possible, making the call efficient if several lines can be written with a single register update. The limitation of GPIOHANDLES_MAX to 64 lines is done under the assumption that we may expect hardware that can issue a transaction updating 64 bits at an instant but unlikely anything larger than that. ChangeLog v2->v3: - Use gpiod_get_value_cansleep() so we support also slowpath GPIO drivers. - Fix up the UAPI docs kerneldoc. - Allocate the anonymous fd last, so that the release function don't get called until that point of something fails. After this point, skip the errorpath. ChangeLog v1->v2: - Handle ioctl_compat() properly based on a similar patch to the other ioctl() handling code. - Use _IOWR() as we pass pointers both in and out of the ioctl() - Use kmalloc() and kfree() for the linehandled, do not try to be fancy with devm_* it doesn't work the way I thought. - Fix const-correctness on the linehandle name field. Acked-by: Michael Welling <mwelling@ieee.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-04-26 16:35:29 +08:00
dev_dbg(&gdev->dev, "registered chardev handle for %d lines\n",
lh->numdescs);
return 0;
gpio: GPIO_GET_LINE{HANDLE,EVENT}_IOCTL: Fix file descriptor leak When allocating a new line handle or event a file is allocated that it is associated to. The file is attached to a file descriptor of the current process and the file descriptor is returned to userspace using copy_to_user(). If this copy operation fails the line handle or event allocation is aborted, all acquired resources are freed and an error is returned. But the file struct is not freed and left attached to the userspace application and even though the file descriptor number was not copied it is trivial to guess. If a userspace application performs a IOCTL on such a left over file descriptor it will trigger a use-after-free and if the file descriptor is closed (latest when the application exits) a double-free is triggered. anon_inode_getfd() performs 3 tasks, allocate a file struct, allocate a file descriptor for the current process and install the file struct in the file descriptor. As soon as the file struct is installed in the file descriptor it is accessible by userspace (even if the IOCTL itself hasn't completed yet), this means uninstalling the fd on the error path is not an option, since userspace might already got a reference to the file. Instead anon_inode_getfd() needs to be broken into its individual steps. The allocation of the file struct and file descriptor is done first, then the copy_to_user() is executed and only if it succeeds the file is installed. Since the file struct is reference counted it can not be just freed, but its reference needs to be dropped, which will also call the release() callback, which will free the state attached to the file. So in this case the normal error cleanup path should not be taken. Cc: stable@vger.kernel.org Fixes: d932cd49182f ("gpio: free handles in fringe cases") Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-10-24 19:59:15 +08:00
out_put_unused_fd:
put_unused_fd(fd);
gpio: userspace ABI for reading/writing GPIO lines This adds a userspace ABI for reading and writing GPIO lines. The mechanism returns an anonymous file handle to a request to read/write n offsets from a gpiochip. This file handle in turn accepts two ioctl()s: one that reads and one that writes values to the selected lines. - Handles can be requested as input/output, active low, open drain, open source, however when you issue a request for n lines with GPIO_GET_LINEHANDLE_IOCTL, they must all have the same flags, i.e. all inputs or all outputs, all open drain etc. If a granular control of the flags for each line is desired, they need to be requested individually, not in a batch. - The GPIOHANDLE_GET_LINE_VALUES_IOCTL read ioctl() can be issued also to output lines to verify that the hardware is in the expected state. - It reads and writes up to GPIOHANDLES_MAX lines at once, utilizing the .set_multiple() call in the driver if possible, making the call efficient if several lines can be written with a single register update. The limitation of GPIOHANDLES_MAX to 64 lines is done under the assumption that we may expect hardware that can issue a transaction updating 64 bits at an instant but unlikely anything larger than that. ChangeLog v2->v3: - Use gpiod_get_value_cansleep() so we support also slowpath GPIO drivers. - Fix up the UAPI docs kerneldoc. - Allocate the anonymous fd last, so that the release function don't get called until that point of something fails. After this point, skip the errorpath. ChangeLog v1->v2: - Handle ioctl_compat() properly based on a similar patch to the other ioctl() handling code. - Use _IOWR() as we pass pointers both in and out of the ioctl() - Use kmalloc() and kfree() for the linehandled, do not try to be fancy with devm_* it doesn't work the way I thought. - Fix const-correctness on the linehandle name field. Acked-by: Michael Welling <mwelling@ieee.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-04-26 16:35:29 +08:00
out_free_descs:
for (; i >= 0; i--)
gpiod_free(lh->descs[i]);
kfree(lh->label);
out_free_lh:
kfree(lh);
put_device(&gdev->dev);
return ret;
}
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
/*
* GPIO line event management
*/
/**
* struct lineevent_state - contains the state of a userspace event
* @gdev: the GPIO device the event pertains to
* @label: consumer label used to tag descriptors
* @desc: the GPIO descriptor held by this event
* @eflags: the event flags this line was requested with
* @irq: the interrupt that trigger in response to events on this GPIO
* @wait: wait queue that handles blocking reads of events
* @events: KFIFO for the GPIO events
* @read_lock: mutex lock to protect reads from colliding with adding
* new events to the FIFO
*/
struct lineevent_state {
struct gpio_device *gdev;
const char *label;
struct gpio_desc *desc;
u32 eflags;
int irq;
wait_queue_head_t wait;
DECLARE_KFIFO(events, struct gpioevent_data, 16);
struct mutex read_lock;
};
#define GPIOEVENT_REQUEST_VALID_FLAGS \
(GPIOEVENT_REQUEST_RISING_EDGE | \
GPIOEVENT_REQUEST_FALLING_EDGE)
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
static unsigned int lineevent_poll(struct file *filep,
struct poll_table_struct *wait)
{
struct lineevent_state *le = filep->private_data;
unsigned int events = 0;
poll_wait(filep, &le->wait, wait);
if (!kfifo_is_empty(&le->events))
events = POLLIN | POLLRDNORM;
return events;
}
static ssize_t lineevent_read(struct file *filep,
char __user *buf,
size_t count,
loff_t *f_ps)
{
struct lineevent_state *le = filep->private_data;
unsigned int copied;
int ret;
if (count < sizeof(struct gpioevent_data))
return -EINVAL;
do {
if (kfifo_is_empty(&le->events)) {
if (filep->f_flags & O_NONBLOCK)
return -EAGAIN;
ret = wait_event_interruptible(le->wait,
!kfifo_is_empty(&le->events));
if (ret)
return ret;
}
if (mutex_lock_interruptible(&le->read_lock))
return -ERESTARTSYS;
ret = kfifo_to_user(&le->events, buf, count, &copied);
mutex_unlock(&le->read_lock);
if (ret)
return ret;
/*
* If we couldn't read anything from the fifo (a different
* thread might have been faster) we either return -EAGAIN if
* the file descriptor is non-blocking, otherwise we go back to
* sleep and wait for more data to arrive.
*/
if (copied == 0 && (filep->f_flags & O_NONBLOCK))
return -EAGAIN;
} while (copied == 0);
return copied;
}
static int lineevent_release(struct inode *inode, struct file *filep)
{
struct lineevent_state *le = filep->private_data;
struct gpio_device *gdev = le->gdev;
free_irq(le->irq, le);
gpiod_free(le->desc);
kfree(le->label);
kfree(le);
put_device(&gdev->dev);
return 0;
}
static long lineevent_ioctl(struct file *filep, unsigned int cmd,
unsigned long arg)
{
struct lineevent_state *le = filep->private_data;
void __user *ip = (void __user *)arg;
struct gpiohandle_data ghd;
/*
* We can get the value for an event line but not set it,
* because it is input by definition.
*/
if (cmd == GPIOHANDLE_GET_LINE_VALUES_IOCTL) {
int val;
memset(&ghd, 0, sizeof(ghd));
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
val = gpiod_get_value_cansleep(le->desc);
if (val < 0)
return val;
ghd.values[0] = val;
if (copy_to_user(ip, &ghd, sizeof(ghd)))
return -EFAULT;
return 0;
}
return -EINVAL;
}
#ifdef CONFIG_COMPAT
static long lineevent_ioctl_compat(struct file *filep, unsigned int cmd,
unsigned long arg)
{
return lineevent_ioctl(filep, cmd, (unsigned long)compat_ptr(arg));
}
#endif
static const struct file_operations lineevent_fileops = {
.release = lineevent_release,
.read = lineevent_read,
.poll = lineevent_poll,
.owner = THIS_MODULE,
.llseek = noop_llseek,
.unlocked_ioctl = lineevent_ioctl,
#ifdef CONFIG_COMPAT
.compat_ioctl = lineevent_ioctl_compat,
#endif
};
static irqreturn_t lineevent_irq_thread(int irq, void *p)
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
{
struct lineevent_state *le = p;
struct gpioevent_data ge;
int ret, level;
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
ge.timestamp = ktime_get_real_ns();
level = gpiod_get_value_cansleep(le->desc);
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
if (le->eflags & GPIOEVENT_REQUEST_RISING_EDGE
&& le->eflags & GPIOEVENT_REQUEST_FALLING_EDGE) {
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
if (level)
/* Emit low-to-high event */
ge.id = GPIOEVENT_EVENT_RISING_EDGE;
else
/* Emit high-to-low event */
ge.id = GPIOEVENT_EVENT_FALLING_EDGE;
} else if (le->eflags & GPIOEVENT_REQUEST_RISING_EDGE && level) {
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
/* Emit low-to-high event */
ge.id = GPIOEVENT_EVENT_RISING_EDGE;
} else if (le->eflags & GPIOEVENT_REQUEST_FALLING_EDGE && !level) {
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
/* Emit high-to-low event */
ge.id = GPIOEVENT_EVENT_FALLING_EDGE;
} else {
return IRQ_NONE;
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
}
ret = kfifo_put(&le->events, ge);
if (ret != 0)
wake_up_poll(&le->wait, POLLIN);
return IRQ_HANDLED;
}
static int lineevent_create(struct gpio_device *gdev, void __user *ip)
{
struct gpioevent_request eventreq;
struct lineevent_state *le;
struct gpio_desc *desc;
gpio: GPIO_GET_LINE{HANDLE,EVENT}_IOCTL: Fix file descriptor leak When allocating a new line handle or event a file is allocated that it is associated to. The file is attached to a file descriptor of the current process and the file descriptor is returned to userspace using copy_to_user(). If this copy operation fails the line handle or event allocation is aborted, all acquired resources are freed and an error is returned. But the file struct is not freed and left attached to the userspace application and even though the file descriptor number was not copied it is trivial to guess. If a userspace application performs a IOCTL on such a left over file descriptor it will trigger a use-after-free and if the file descriptor is closed (latest when the application exits) a double-free is triggered. anon_inode_getfd() performs 3 tasks, allocate a file struct, allocate a file descriptor for the current process and install the file struct in the file descriptor. As soon as the file struct is installed in the file descriptor it is accessible by userspace (even if the IOCTL itself hasn't completed yet), this means uninstalling the fd on the error path is not an option, since userspace might already got a reference to the file. Instead anon_inode_getfd() needs to be broken into its individual steps. The allocation of the file struct and file descriptor is done first, then the copy_to_user() is executed and only if it succeeds the file is installed. Since the file struct is reference counted it can not be just freed, but its reference needs to be dropped, which will also call the release() callback, which will free the state attached to the file. So in this case the normal error cleanup path should not be taken. Cc: stable@vger.kernel.org Fixes: d932cd49182f ("gpio: free handles in fringe cases") Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-10-24 19:59:15 +08:00
struct file *file;
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
u32 offset;
u32 lflags;
u32 eflags;
int fd;
int ret;
int irqflags = 0;
if (copy_from_user(&eventreq, ip, sizeof(eventreq)))
return -EFAULT;
le = kzalloc(sizeof(*le), GFP_KERNEL);
if (!le)
return -ENOMEM;
le->gdev = gdev;
get_device(&gdev->dev);
/* Make sure this is terminated */
eventreq.consumer_label[sizeof(eventreq.consumer_label)-1] = '\0';
if (strlen(eventreq.consumer_label)) {
le->label = kstrdup(eventreq.consumer_label,
GFP_KERNEL);
if (!le->label) {
ret = -ENOMEM;
goto out_free_le;
}
}
offset = eventreq.lineoffset;
lflags = eventreq.handleflags;
eflags = eventreq.eventflags;
if (offset >= gdev->ngpio) {
ret = -EINVAL;
goto out_free_label;
}
/* Return an error if a unknown flag is set */
if ((lflags & ~GPIOHANDLE_REQUEST_VALID_FLAGS) ||
(eflags & ~GPIOEVENT_REQUEST_VALID_FLAGS)) {
ret = -EINVAL;
goto out_free_label;
}
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
/* This is just wrong: we don't look for events on output lines */
if (lflags & GPIOHANDLE_REQUEST_OUTPUT) {
ret = -EINVAL;
goto out_free_label;
}
desc = &gdev->descs[offset];
ret = gpiod_request(desc, le->label);
if (ret)
goto out_free_desc;
le->desc = desc;
le->eflags = eflags;
if (lflags & GPIOHANDLE_REQUEST_ACTIVE_LOW)
set_bit(FLAG_ACTIVE_LOW, &desc->flags);
if (lflags & GPIOHANDLE_REQUEST_OPEN_DRAIN)
set_bit(FLAG_OPEN_DRAIN, &desc->flags);
if (lflags & GPIOHANDLE_REQUEST_OPEN_SOURCE)
set_bit(FLAG_OPEN_SOURCE, &desc->flags);
ret = gpiod_direction_input(desc);
if (ret)
goto out_free_desc;
le->irq = gpiod_to_irq(desc);
if (le->irq <= 0) {
ret = -ENODEV;
goto out_free_desc;
}
if (eflags & GPIOEVENT_REQUEST_RISING_EDGE)
irqflags |= IRQF_TRIGGER_RISING;
if (eflags & GPIOEVENT_REQUEST_FALLING_EDGE)
irqflags |= IRQF_TRIGGER_FALLING;
irqflags |= IRQF_ONESHOT;
irqflags |= IRQF_SHARED;
INIT_KFIFO(le->events);
init_waitqueue_head(&le->wait);
mutex_init(&le->read_lock);
/* Request a thread to read the events */
ret = request_threaded_irq(le->irq,
NULL,
lineevent_irq_thread,
irqflags,
le->label,
le);
if (ret)
goto out_free_desc;
gpio: GPIO_GET_LINE{HANDLE,EVENT}_IOCTL: Fix file descriptor leak When allocating a new line handle or event a file is allocated that it is associated to. The file is attached to a file descriptor of the current process and the file descriptor is returned to userspace using copy_to_user(). If this copy operation fails the line handle or event allocation is aborted, all acquired resources are freed and an error is returned. But the file struct is not freed and left attached to the userspace application and even though the file descriptor number was not copied it is trivial to guess. If a userspace application performs a IOCTL on such a left over file descriptor it will trigger a use-after-free and if the file descriptor is closed (latest when the application exits) a double-free is triggered. anon_inode_getfd() performs 3 tasks, allocate a file struct, allocate a file descriptor for the current process and install the file struct in the file descriptor. As soon as the file struct is installed in the file descriptor it is accessible by userspace (even if the IOCTL itself hasn't completed yet), this means uninstalling the fd on the error path is not an option, since userspace might already got a reference to the file. Instead anon_inode_getfd() needs to be broken into its individual steps. The allocation of the file struct and file descriptor is done first, then the copy_to_user() is executed and only if it succeeds the file is installed. Since the file struct is reference counted it can not be just freed, but its reference needs to be dropped, which will also call the release() callback, which will free the state attached to the file. So in this case the normal error cleanup path should not be taken. Cc: stable@vger.kernel.org Fixes: d932cd49182f ("gpio: free handles in fringe cases") Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-10-24 19:59:15 +08:00
fd = get_unused_fd_flags(O_RDONLY | O_CLOEXEC);
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
if (fd < 0) {
ret = fd;
goto out_free_irq;
}
gpio: GPIO_GET_LINE{HANDLE,EVENT}_IOCTL: Fix file descriptor leak When allocating a new line handle or event a file is allocated that it is associated to. The file is attached to a file descriptor of the current process and the file descriptor is returned to userspace using copy_to_user(). If this copy operation fails the line handle or event allocation is aborted, all acquired resources are freed and an error is returned. But the file struct is not freed and left attached to the userspace application and even though the file descriptor number was not copied it is trivial to guess. If a userspace application performs a IOCTL on such a left over file descriptor it will trigger a use-after-free and if the file descriptor is closed (latest when the application exits) a double-free is triggered. anon_inode_getfd() performs 3 tasks, allocate a file struct, allocate a file descriptor for the current process and install the file struct in the file descriptor. As soon as the file struct is installed in the file descriptor it is accessible by userspace (even if the IOCTL itself hasn't completed yet), this means uninstalling the fd on the error path is not an option, since userspace might already got a reference to the file. Instead anon_inode_getfd() needs to be broken into its individual steps. The allocation of the file struct and file descriptor is done first, then the copy_to_user() is executed and only if it succeeds the file is installed. Since the file struct is reference counted it can not be just freed, but its reference needs to be dropped, which will also call the release() callback, which will free the state attached to the file. So in this case the normal error cleanup path should not be taken. Cc: stable@vger.kernel.org Fixes: d932cd49182f ("gpio: free handles in fringe cases") Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-10-24 19:59:15 +08:00
file = anon_inode_getfile("gpio-event",
&lineevent_fileops,
le,
O_RDONLY | O_CLOEXEC);
if (IS_ERR(file)) {
ret = PTR_ERR(file);
goto out_put_unused_fd;
}
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
eventreq.fd = fd;
if (copy_to_user(ip, &eventreq, sizeof(eventreq))) {
gpio: GPIO_GET_LINE{HANDLE,EVENT}_IOCTL: Fix file descriptor leak When allocating a new line handle or event a file is allocated that it is associated to. The file is attached to a file descriptor of the current process and the file descriptor is returned to userspace using copy_to_user(). If this copy operation fails the line handle or event allocation is aborted, all acquired resources are freed and an error is returned. But the file struct is not freed and left attached to the userspace application and even though the file descriptor number was not copied it is trivial to guess. If a userspace application performs a IOCTL on such a left over file descriptor it will trigger a use-after-free and if the file descriptor is closed (latest when the application exits) a double-free is triggered. anon_inode_getfd() performs 3 tasks, allocate a file struct, allocate a file descriptor for the current process and install the file struct in the file descriptor. As soon as the file struct is installed in the file descriptor it is accessible by userspace (even if the IOCTL itself hasn't completed yet), this means uninstalling the fd on the error path is not an option, since userspace might already got a reference to the file. Instead anon_inode_getfd() needs to be broken into its individual steps. The allocation of the file struct and file descriptor is done first, then the copy_to_user() is executed and only if it succeeds the file is installed. Since the file struct is reference counted it can not be just freed, but its reference needs to be dropped, which will also call the release() callback, which will free the state attached to the file. So in this case the normal error cleanup path should not be taken. Cc: stable@vger.kernel.org Fixes: d932cd49182f ("gpio: free handles in fringe cases") Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-10-24 19:59:15 +08:00
/*
* fput() will trigger the release() callback, so do not go onto
* the regular error cleanup path here.
*/
fput(file);
put_unused_fd(fd);
return -EFAULT;
}
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
gpio: GPIO_GET_LINE{HANDLE,EVENT}_IOCTL: Fix file descriptor leak When allocating a new line handle or event a file is allocated that it is associated to. The file is attached to a file descriptor of the current process and the file descriptor is returned to userspace using copy_to_user(). If this copy operation fails the line handle or event allocation is aborted, all acquired resources are freed and an error is returned. But the file struct is not freed and left attached to the userspace application and even though the file descriptor number was not copied it is trivial to guess. If a userspace application performs a IOCTL on such a left over file descriptor it will trigger a use-after-free and if the file descriptor is closed (latest when the application exits) a double-free is triggered. anon_inode_getfd() performs 3 tasks, allocate a file struct, allocate a file descriptor for the current process and install the file struct in the file descriptor. As soon as the file struct is installed in the file descriptor it is accessible by userspace (even if the IOCTL itself hasn't completed yet), this means uninstalling the fd on the error path is not an option, since userspace might already got a reference to the file. Instead anon_inode_getfd() needs to be broken into its individual steps. The allocation of the file struct and file descriptor is done first, then the copy_to_user() is executed and only if it succeeds the file is installed. Since the file struct is reference counted it can not be just freed, but its reference needs to be dropped, which will also call the release() callback, which will free the state attached to the file. So in this case the normal error cleanup path should not be taken. Cc: stable@vger.kernel.org Fixes: d932cd49182f ("gpio: free handles in fringe cases") Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-10-24 19:59:15 +08:00
fd_install(fd, file);
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
return 0;
gpio: GPIO_GET_LINE{HANDLE,EVENT}_IOCTL: Fix file descriptor leak When allocating a new line handle or event a file is allocated that it is associated to. The file is attached to a file descriptor of the current process and the file descriptor is returned to userspace using copy_to_user(). If this copy operation fails the line handle or event allocation is aborted, all acquired resources are freed and an error is returned. But the file struct is not freed and left attached to the userspace application and even though the file descriptor number was not copied it is trivial to guess. If a userspace application performs a IOCTL on such a left over file descriptor it will trigger a use-after-free and if the file descriptor is closed (latest when the application exits) a double-free is triggered. anon_inode_getfd() performs 3 tasks, allocate a file struct, allocate a file descriptor for the current process and install the file struct in the file descriptor. As soon as the file struct is installed in the file descriptor it is accessible by userspace (even if the IOCTL itself hasn't completed yet), this means uninstalling the fd on the error path is not an option, since userspace might already got a reference to the file. Instead anon_inode_getfd() needs to be broken into its individual steps. The allocation of the file struct and file descriptor is done first, then the copy_to_user() is executed and only if it succeeds the file is installed. Since the file struct is reference counted it can not be just freed, but its reference needs to be dropped, which will also call the release() callback, which will free the state attached to the file. So in this case the normal error cleanup path should not be taken. Cc: stable@vger.kernel.org Fixes: d932cd49182f ("gpio: free handles in fringe cases") Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-10-24 19:59:15 +08:00
out_put_unused_fd:
put_unused_fd(fd);
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
out_free_irq:
free_irq(le->irq, le);
out_free_desc:
gpiod_free(le->desc);
out_free_label:
kfree(le->label);
out_free_le:
kfree(le);
put_device(&gdev->dev);
return ret;
}
/*
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
* gpio_ioctl() - ioctl handler for the GPIO chardev
*/
static long gpio_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
{
struct gpio_device *gdev = filp->private_data;
struct gpio_chip *chip = gdev->chip;
void __user *ip = (void __user *)arg;
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
/* We fail any subsequent ioctl():s when the chip is gone */
if (!chip)
return -ENODEV;
/* Fill in the struct and pass to userspace */
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
if (cmd == GPIO_GET_CHIPINFO_IOCTL) {
struct gpiochip_info chipinfo;
memset(&chipinfo, 0, sizeof(chipinfo));
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
strncpy(chipinfo.name, dev_name(&gdev->dev),
sizeof(chipinfo.name));
chipinfo.name[sizeof(chipinfo.name)-1] = '\0';
strncpy(chipinfo.label, gdev->label,
sizeof(chipinfo.label));
chipinfo.label[sizeof(chipinfo.label)-1] = '\0';
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
chipinfo.lines = gdev->ngpio;
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
if (copy_to_user(ip, &chipinfo, sizeof(chipinfo)))
return -EFAULT;
return 0;
} else if (cmd == GPIO_GET_LINEINFO_IOCTL) {
struct gpioline_info lineinfo;
struct gpio_desc *desc;
if (copy_from_user(&lineinfo, ip, sizeof(lineinfo)))
return -EFAULT;
if (lineinfo.line_offset >= gdev->ngpio)
return -EINVAL;
desc = &gdev->descs[lineinfo.line_offset];
if (desc->name) {
strncpy(lineinfo.name, desc->name,
sizeof(lineinfo.name));
lineinfo.name[sizeof(lineinfo.name)-1] = '\0';
} else {
lineinfo.name[0] = '\0';
}
if (desc->label) {
strncpy(lineinfo.consumer, desc->label,
sizeof(lineinfo.consumer));
lineinfo.consumer[sizeof(lineinfo.consumer)-1] = '\0';
} else {
lineinfo.consumer[0] = '\0';
}
/*
* Userspace only need to know that the kernel is using
* this GPIO so it can't use it.
*/
lineinfo.flags = 0;
if (test_bit(FLAG_REQUESTED, &desc->flags) ||
test_bit(FLAG_IS_HOGGED, &desc->flags) ||
test_bit(FLAG_USED_AS_IRQ, &desc->flags) ||
test_bit(FLAG_EXPORT, &desc->flags) ||
test_bit(FLAG_SYSFS, &desc->flags))
lineinfo.flags |= GPIOLINE_FLAG_KERNEL;
if (test_bit(FLAG_IS_OUT, &desc->flags))
lineinfo.flags |= GPIOLINE_FLAG_IS_OUT;
if (test_bit(FLAG_ACTIVE_LOW, &desc->flags))
lineinfo.flags |= GPIOLINE_FLAG_ACTIVE_LOW;
if (test_bit(FLAG_OPEN_DRAIN, &desc->flags))
lineinfo.flags |= GPIOLINE_FLAG_OPEN_DRAIN;
if (test_bit(FLAG_OPEN_SOURCE, &desc->flags))
lineinfo.flags |= GPIOLINE_FLAG_OPEN_SOURCE;
if (copy_to_user(ip, &lineinfo, sizeof(lineinfo)))
return -EFAULT;
return 0;
gpio: userspace ABI for reading/writing GPIO lines This adds a userspace ABI for reading and writing GPIO lines. The mechanism returns an anonymous file handle to a request to read/write n offsets from a gpiochip. This file handle in turn accepts two ioctl()s: one that reads and one that writes values to the selected lines. - Handles can be requested as input/output, active low, open drain, open source, however when you issue a request for n lines with GPIO_GET_LINEHANDLE_IOCTL, they must all have the same flags, i.e. all inputs or all outputs, all open drain etc. If a granular control of the flags for each line is desired, they need to be requested individually, not in a batch. - The GPIOHANDLE_GET_LINE_VALUES_IOCTL read ioctl() can be issued also to output lines to verify that the hardware is in the expected state. - It reads and writes up to GPIOHANDLES_MAX lines at once, utilizing the .set_multiple() call in the driver if possible, making the call efficient if several lines can be written with a single register update. The limitation of GPIOHANDLES_MAX to 64 lines is done under the assumption that we may expect hardware that can issue a transaction updating 64 bits at an instant but unlikely anything larger than that. ChangeLog v2->v3: - Use gpiod_get_value_cansleep() so we support also slowpath GPIO drivers. - Fix up the UAPI docs kerneldoc. - Allocate the anonymous fd last, so that the release function don't get called until that point of something fails. After this point, skip the errorpath. ChangeLog v1->v2: - Handle ioctl_compat() properly based on a similar patch to the other ioctl() handling code. - Use _IOWR() as we pass pointers both in and out of the ioctl() - Use kmalloc() and kfree() for the linehandled, do not try to be fancy with devm_* it doesn't work the way I thought. - Fix const-correctness on the linehandle name field. Acked-by: Michael Welling <mwelling@ieee.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-04-26 16:35:29 +08:00
} else if (cmd == GPIO_GET_LINEHANDLE_IOCTL) {
return linehandle_create(gdev, ip);
gpio: userspace ABI for reading GPIO line events This adds an ABI for listening to events on GPIO lines. The mechanism returns an anonymous file handle to a request to listen to a specific offset on a specific gpiochip. To fetch the stream of events from the file handle, userspace simply reads an event. - Events can be requested with the same flags as ordinary handles, i.e. open drain or open source. An ioctl() call GPIO_GET_LINEEVENT_IOCTL is issued indicating the desired line. - Events can be requested for falling edge events, rising edge events, or both. - All events are timestamped using the kernel real time nanosecond timestamp (the same as is used by IIO). - The supplied consumer label will appear in "lsgpio" listings of the lines, and in /proc/interrupts as the mechanism will request an interrupt from the gpio chip. - Events are not supported on gpiochips that do not serve interrupts (no legal .to_irq() call). The event interrupt is threaded to avoid any realtime problems. - It is possible to also directly read the current value of the registered GPIO line by issuing the same GPIOHANDLE_GET_LINE_VALUES_IOCTL as used by the line handles. Setting the value is not supported: we do not listen to events on output lines. This ABI is strongly influenced by Industrial I/O and surpasses the old sysfs ABI by providing proper precision timestamps, making it possible to set flags like open drain, and put consumer names on the GPIO lines. Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-06-02 17:30:15 +08:00
} else if (cmd == GPIO_GET_LINEEVENT_IOCTL) {
return lineevent_create(gdev, ip);
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
}
return -EINVAL;
}
#ifdef CONFIG_COMPAT
static long gpio_ioctl_compat(struct file *filp, unsigned int cmd,
unsigned long arg)
{
return gpio_ioctl(filp, cmd, (unsigned long)compat_ptr(arg));
}
#endif
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
/**
* gpio_chrdev_open() - open the chardev for ioctl operations
* @inode: inode for this chardev
* @filp: file struct for storing private data
* Returns 0 on success
*/
static int gpio_chrdev_open(struct inode *inode, struct file *filp)
{
struct gpio_device *gdev = container_of(inode->i_cdev,
struct gpio_device, chrdev);
/* Fail on open if the backing gpiochip is gone */
if (!gdev->chip)
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
return -ENODEV;
get_device(&gdev->dev);
filp->private_data = gdev;
gpio: chardev: Return error for seek operations The GPIO chardev is used for management tasks (allocating line and event handles) and does neither support read() nor write() operations. Hence it does not make much sense to allow seek operations. Currently the chardev uses noop_llseek() for its seek implementation. This function does not move the pointer and simply returns the current position (always 0 for the GPIO chardev). noop_llseek() is primarily meant for devices that can not support seek, but where there might be a user that depends on the seek() operation succeeding. For newly added devices that can not support seek operations it is recommended to use no_llseek(), which will return an error. For more information see commit 6038f373a3dc ("llseek: automatically add .llseek fop"). Unfortunately this was overlooked when the GPIO chardev ABI was introduced. But it is highly unlikely that since then userspace applications have appeared that rely on being able to perform non-failing seek operations on a GPIO chardev file descriptor. So it should be safe to change from noop_llseel() to no_seek(). Also use nonseekable_open() in the chardev open() callback to clear the FMODE_SEEK, FMODE_PREAD and FMODE_PWRITE flags from the file. Neither of these should be set on a file that does not support seek operations. Cc: stable@vger.kernel.org Fixes: 3c702e9987e2 ("gpio: add a userspace chardev ABI for GPIOs") Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-11-30 20:05:21 +08:00
return nonseekable_open(inode, filp);
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
}
/**
* gpio_chrdev_release() - close chardev after ioctl operations
* @inode: inode for this chardev
* @filp: file struct for storing private data
* Returns 0 on success
*/
static int gpio_chrdev_release(struct inode *inode, struct file *filp)
{
struct gpio_device *gdev = container_of(inode->i_cdev,
struct gpio_device, chrdev);
put_device(&gdev->dev);
return 0;
}
static const struct file_operations gpio_fileops = {
.release = gpio_chrdev_release,
.open = gpio_chrdev_open,
.owner = THIS_MODULE,
gpio: chardev: Return error for seek operations The GPIO chardev is used for management tasks (allocating line and event handles) and does neither support read() nor write() operations. Hence it does not make much sense to allow seek operations. Currently the chardev uses noop_llseek() for its seek implementation. This function does not move the pointer and simply returns the current position (always 0 for the GPIO chardev). noop_llseek() is primarily meant for devices that can not support seek, but where there might be a user that depends on the seek() operation succeeding. For newly added devices that can not support seek operations it is recommended to use no_llseek(), which will return an error. For more information see commit 6038f373a3dc ("llseek: automatically add .llseek fop"). Unfortunately this was overlooked when the GPIO chardev ABI was introduced. But it is highly unlikely that since then userspace applications have appeared that rely on being able to perform non-failing seek operations on a GPIO chardev file descriptor. So it should be safe to change from noop_llseel() to no_seek(). Also use nonseekable_open() in the chardev open() callback to clear the FMODE_SEEK, FMODE_PREAD and FMODE_PWRITE flags from the file. Neither of these should be set on a file that does not support seek operations. Cc: stable@vger.kernel.org Fixes: 3c702e9987e2 ("gpio: add a userspace chardev ABI for GPIOs") Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-11-30 20:05:21 +08:00
.llseek = no_llseek,
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
.unlocked_ioctl = gpio_ioctl,
#ifdef CONFIG_COMPAT
.compat_ioctl = gpio_ioctl_compat,
#endif
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
};
static void gpiodevice_release(struct device *dev)
{
struct gpio_device *gdev = dev_get_drvdata(dev);
list_del(&gdev->list);
ida_simple_remove(&gpio_ida, gdev->id);
kfree(gdev->label);
kfree(gdev->descs);
kfree(gdev);
}
static int gpiochip_setup_dev(struct gpio_device *gdev)
{
int status;
cdev_init(&gdev->chrdev, &gpio_fileops);
gdev->chrdev.owner = THIS_MODULE;
gdev->dev.devt = MKDEV(MAJOR(gpio_devt), gdev->id);
status = cdev_device_add(&gdev->chrdev, &gdev->dev);
if (status)
return status;
chip_dbg(gdev->chip, "added GPIO chardev (%d:%d)\n",
MAJOR(gpio_devt), gdev->id);
status = gpiochip_sysfs_register(gdev);
if (status)
goto err_remove_device;
/* From this point, the .release() function cleans up gpio_device */
gdev->dev.release = gpiodevice_release;
pr_debug("%s: registered GPIOs %d to %d on device: %s (%s)\n",
__func__, gdev->base, gdev->base + gdev->ngpio - 1,
dev_name(&gdev->dev), gdev->chip->label ? : "generic");
return 0;
err_remove_device:
cdev_device_del(&gdev->chrdev, &gdev->dev);
return status;
}
static void gpiochip_setup_devs(void)
{
struct gpio_device *gdev;
int err;
list_for_each_entry(gdev, &gpio_devices, list) {
err = gpiochip_setup_dev(gdev);
if (err)
pr_err("%s: Failed to initialize gpio device (%d)\n",
dev_name(&gdev->dev), err);
}
}
int gpiochip_add_data_with_key(struct gpio_chip *chip, void *data,
struct lock_class_key *key)
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
{
unsigned long flags;
int status = 0;
unsigned i;
int base = chip->base;
struct gpio_device *gdev;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/*
* First: allocate and populate the internal stat container, and
* set up the struct device.
*/
gdev = kzalloc(sizeof(*gdev), GFP_KERNEL);
if (!gdev)
return -ENOMEM;
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
gdev->dev.bus = &gpio_bus_type;
gdev->chip = chip;
chip->gpiodev = gdev;
if (chip->parent) {
gdev->dev.parent = chip->parent;
gdev->dev.of_node = chip->parent->of_node;
}
#ifdef CONFIG_OF_GPIO
/* If the gpiochip has an assigned OF node this takes precedence */
if (chip->of_node)
gdev->dev.of_node = chip->of_node;
#endif
gdev->id = ida_simple_get(&gpio_ida, 0, 0, GFP_KERNEL);
if (gdev->id < 0) {
status = gdev->id;
goto err_free_gdev;
}
dev_set_name(&gdev->dev, "gpiochip%d", gdev->id);
device_initialize(&gdev->dev);
dev_set_drvdata(&gdev->dev, gdev);
if (chip->parent && chip->parent->driver)
gdev->owner = chip->parent->driver->owner;
else if (chip->owner)
/* TODO: remove chip->owner */
gdev->owner = chip->owner;
else
gdev->owner = THIS_MODULE;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
gdev->descs = kcalloc(chip->ngpio, sizeof(gdev->descs[0]), GFP_KERNEL);
if (!gdev->descs) {
status = -ENOMEM;
goto err_free_gdev;
}
if (chip->ngpio == 0) {
chip_err(chip, "tried to insert a GPIO chip with zero lines\n");
status = -EINVAL;
goto err_free_descs;
}
if (chip->label)
gdev->label = kstrdup(chip->label, GFP_KERNEL);
else
gdev->label = kstrdup("unknown", GFP_KERNEL);
if (!gdev->label) {
status = -ENOMEM;
goto err_free_descs;
}
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
gdev->ngpio = chip->ngpio;
gdev->data = data;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
spin_lock_irqsave(&gpio_lock, flags);
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
/*
* TODO: this allocates a Linux GPIO number base in the global
* GPIO numberspace for this chip. In the long run we want to
* get *rid* of this numberspace and use only descriptors, but
* it may be a pipe dream. It will not happen before we get rid
* of the sysfs interface anyways.
*/
if (base < 0) {
base = gpiochip_find_base(chip->ngpio);
if (base < 0) {
status = base;
spin_unlock_irqrestore(&gpio_lock, flags);
goto err_free_label;
}
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
/*
* TODO: it should not be necessary to reflect the assigned
* base outside of the GPIO subsystem. Go over drivers and
* see if anyone makes use of this, else drop this and assign
* a poison instead.
*/
chip->base = base;
}
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
gdev->base = base;
status = gpiodev_add_to_list(gdev);
if (status) {
spin_unlock_irqrestore(&gpio_lock, flags);
goto err_free_label;
}
spin_unlock_irqrestore(&gpio_lock, flags);
for (i = 0; i < chip->ngpio; i++) {
struct gpio_desc *desc = &gdev->descs[i];
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
desc->gdev = gdev;
/*
* REVISIT: most hardware initializes GPIOs as inputs
* (often with pullups enabled) so power usage is
* minimized. Linux code should set the gpio direction
* first thing; but until it does, and in case
* chip->get_direction is not set, we may expose the
* wrong direction in sysfs.
*/
if (chip->get_direction) {
/*
* If we have .get_direction, set up the initial
* direction flag from the hardware.
*/
int dir = chip->get_direction(chip, i);
if (!dir)
set_bit(FLAG_IS_OUT, &desc->flags);
} else if (!chip->direction_input) {
/*
* If the chip lacks the .direction_input callback
* we logically assume all lines are outputs.
*/
set_bit(FLAG_IS_OUT, &desc->flags);
}
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
}
#ifdef CONFIG_PINCTRL
INIT_LIST_HEAD(&gdev->pin_ranges);
#endif
status = gpiochip_set_desc_names(chip);
if (status)
goto err_remove_from_list;
status = gpiochip_irqchip_init_valid_mask(chip);
if (status)
goto err_remove_from_list;
status = gpiochip_add_irqchip(chip, key);
if (status)
goto err_remove_chip;
status = of_gpiochip_add(chip);
if (status)
goto err_remove_chip;
acpi_gpiochip_add(chip);
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
/*
* By first adding the chardev, and then adding the device,
* we get a device node entry in sysfs under
* /sys/bus/gpio/devices/gpiochipN/dev that can be used for
* coldplug of device nodes and other udev business.
* We can do this only if gpiolib has been initialized.
* Otherwise, defer until later.
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
*/
if (gpiolib_initialized) {
status = gpiochip_setup_dev(gdev);
if (status)
goto err_remove_chip;
}
return 0;
err_remove_chip:
acpi_gpiochip_remove(chip);
gpiochip_free_hogs(chip);
of_gpiochip_remove(chip);
gpiochip_irqchip_free_valid_mask(chip);
err_remove_from_list:
spin_lock_irqsave(&gpio_lock, flags);
list_del(&gdev->list);
spin_unlock_irqrestore(&gpio_lock, flags);
err_free_label:
kfree(gdev->label);
err_free_descs:
kfree(gdev->descs);
err_free_gdev:
ida_simple_remove(&gpio_ida, gdev->id);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/* failures here can mean systems won't boot... */
pr_err("%s: GPIOs %d..%d (%s) failed to register\n", __func__,
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
gdev->base, gdev->base + gdev->ngpio - 1,
chip->label ? : "generic");
kfree(gdev);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
return status;
}
EXPORT_SYMBOL_GPL(gpiochip_add_data_with_key);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/**
* gpiochip_get_data() - get per-subdriver data for the chip
* @chip: GPIO chip
*
* Returns:
* The per-subdriver data for the chip.
*/
void *gpiochip_get_data(struct gpio_chip *chip)
{
return chip->gpiodev->data;
}
EXPORT_SYMBOL_GPL(gpiochip_get_data);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/**
* gpiochip_remove() - unregister a gpio_chip
* @chip: the chip to unregister
*
* A gpio_chip with any GPIOs still requested may not be removed.
*/
void gpiochip_remove(struct gpio_chip *chip)
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
{
struct gpio_device *gdev = chip->gpiodev;
struct gpio_desc *desc;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
unsigned long flags;
unsigned i;
bool requested = false;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/* FIXME: should the legacy sysfs handling be moved to gpio_device? */
gpiochip_sysfs_unregister(gdev);
gpiochip_free_hogs(chip);
/* Numb the device, cancelling all outstanding operations */
gdev->chip = NULL;
gpiochip_irqchip_remove(chip);
acpi_gpiochip_remove(chip);
gpiochip_remove_pin_ranges(chip);
of_gpiochip_remove(chip);
/*
* We accept no more calls into the driver from this point, so
* NULL the driver data pointer
*/
gdev->data = NULL;
spin_lock_irqsave(&gpio_lock, flags);
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
for (i = 0; i < gdev->ngpio; i++) {
desc = &gdev->descs[i];
if (test_bit(FLAG_REQUESTED, &desc->flags))
requested = true;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
}
spin_unlock_irqrestore(&gpio_lock, flags);
if (requested)
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
dev_crit(&gdev->dev,
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>
2015-11-04 16:56:26 +08:00
"REMOVING GPIOCHIP WITH GPIOS STILL REQUESTED\n");
/*
* The gpiochip side puts its use of the device to rest here:
* if there are no userspace clients, the chardev and device will
* be removed, else it will be dangling until the last user is
* gone.
*/
cdev_device_del(&gdev->chrdev, &gdev->dev);
put_device(&gdev->dev);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
}
EXPORT_SYMBOL_GPL(gpiochip_remove);
static void devm_gpio_chip_release(struct device *dev, void *res)
{
struct gpio_chip *chip = *(struct gpio_chip **)res;
gpiochip_remove(chip);
}
static int devm_gpio_chip_match(struct device *dev, void *res, void *data)
{
struct gpio_chip **r = res;
if (!r || !*r) {
WARN_ON(!r || !*r);
return 0;
}
return *r == data;
}
/**
* devm_gpiochip_add_data() - Resource manager piochip_add_data()
* @dev: the device pointer on which irq_chip belongs to.
* @chip: the chip to register, with chip->base initialized
* @data: driver-private data associated with this chip
*
* Context: potentially before irqs will work
*
* The gpio chip automatically be released when the device is unbound.
*
* Returns:
* A negative errno if the chip can't be registered, such as because the
* chip->base is invalid or already associated with a different chip.
* Otherwise it returns zero as a success code.
*/
int devm_gpiochip_add_data(struct device *dev, struct gpio_chip *chip,
void *data)
{
struct gpio_chip **ptr;
int ret;
ptr = devres_alloc(devm_gpio_chip_release, sizeof(*ptr),
GFP_KERNEL);
if (!ptr)
return -ENOMEM;
ret = gpiochip_add_data(chip, data);
if (ret < 0) {
devres_free(ptr);
return ret;
}
*ptr = chip;
devres_add(dev, ptr);
return 0;
}
EXPORT_SYMBOL_GPL(devm_gpiochip_add_data);
/**
* devm_gpiochip_remove() - Resource manager of gpiochip_remove()
* @dev: device for which which resource was allocated
* @chip: the chip to remove
*
* A gpio_chip with any GPIOs still requested may not be removed.
*/
void devm_gpiochip_remove(struct device *dev, struct gpio_chip *chip)
{
int ret;
ret = devres_release(dev, devm_gpio_chip_release,
devm_gpio_chip_match, chip);
WARN_ON(ret);
}
EXPORT_SYMBOL_GPL(devm_gpiochip_remove);
/**
* gpiochip_find() - iterator for locating a specific gpio_chip
* @data: data to pass to match function
* @match: Callback function to check gpio_chip
*
* Similar to bus_find_device. It returns a reference to a gpio_chip as
* determined by a user supplied @match callback. The callback should return
* 0 if the device doesn't match and non-zero if it does. If the callback is
* non-zero, this function will return to the caller and not iterate over any
* more gpio_chips.
*/
struct gpio_chip *gpiochip_find(void *data,
int (*match)(struct gpio_chip *chip,
void *data))
{
struct gpio_device *gdev;
struct gpio_chip *chip = NULL;
unsigned long flags;
spin_lock_irqsave(&gpio_lock, flags);
list_for_each_entry(gdev, &gpio_devices, list)
if (gdev->chip && match(gdev->chip, data)) {
chip = gdev->chip;
break;
}
spin_unlock_irqrestore(&gpio_lock, flags);
return chip;
}
EXPORT_SYMBOL_GPL(gpiochip_find);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
static int gpiochip_match_name(struct gpio_chip *chip, void *data)
{
const char *name = data;
return !strcmp(chip->label, name);
}
static struct gpio_chip *find_chip_by_name(const char *name)
{
return gpiochip_find((void *)name, gpiochip_match_name);
}
#ifdef CONFIG_GPIOLIB_IRQCHIP
/*
* The following is irqchip helper code for gpiochips.
*/
static int gpiochip_irqchip_init_valid_mask(struct gpio_chip *gpiochip)
{
if (!gpiochip->irq.need_valid_mask)
return 0;
gpiochip->irq.valid_mask = kcalloc(BITS_TO_LONGS(gpiochip->ngpio),
sizeof(long), GFP_KERNEL);
if (!gpiochip->irq.valid_mask)
return -ENOMEM;
/* Assume by default all GPIOs are valid */
bitmap_fill(gpiochip->irq.valid_mask, gpiochip->ngpio);
return 0;
}
static void gpiochip_irqchip_free_valid_mask(struct gpio_chip *gpiochip)
{
kfree(gpiochip->irq.valid_mask);
gpiochip->irq.valid_mask = NULL;
}
static bool gpiochip_irqchip_irq_valid(const struct gpio_chip *gpiochip,
unsigned int offset)
{
/* No mask means all valid */
if (likely(!gpiochip->irq.valid_mask))
return true;
return test_bit(offset, gpiochip->irq.valid_mask);
}
/**
gpio: simplify adding threaded interrupts This tries to simplify the use of CONFIG_GPIOLIB_IRQCHIP when using threaded interrupts: add a new call gpiochip_irqchip_add_nested() to indicate that we're dealing with a nested rather than a chained irqchip, then create a separate gpiochip_set_nested_irqchip() to mirror the gpiochip_set_chained_irqchip() call to connect the parent and child interrupts. In the nested case gpiochip_set_nested_irqchip() does nothing more than call irq_set_parent() on each valid child interrupt, which has little semantic effect in the kernel, but this is probably still formally correct. Update all drivers using nested interrupts to use gpiochip_irqchip_add_nested() so we can now see clearly which these users are. The DLN2 driver can drop its specific hack with .irq_not_threaded as we now recognize whether a chip is threaded or not from its use of gpiochip_irqchip_add_nested() signature rather than from inspecting .can_sleep. We rename the .irq_parent to .irq_chained_parent since this parent IRQ is only really kept around for the chained interrupt handlers. Cc: Lars Poeschel <poeschel@lemonage.de> Cc: Octavian Purdila <octavian.purdila@intel.com> Cc: Daniel Baluta <daniel.baluta@intel.com> Cc: Bin Gao <bin.gao@linux.intel.com> Cc: Mika Westerberg <mika.westerberg@linux.intel.com> Cc: Ajay Thomas <ajay.thomas.david.rajamanickam@intel.com> Cc: Semen Protsenko <semen.protsenko@globallogic.com> Cc: Alexander Stein <alexander.stein@systec-electronic.com> Cc: Phil Reid <preid@electromag.com.au> Cc: Bartosz Golaszewski <bgolaszewski@baylibre.com> Cc: Patrice Chotard <patrice.chotard@st.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-11-24 17:57:25 +08:00
* gpiochip_set_cascaded_irqchip() - connects a cascaded irqchip to a gpiochip
* @gpiochip: the gpiochip to set the irqchip chain to
* @irqchip: the irqchip to chain to the gpiochip
* @parent_irq: the irq number corresponding to the parent IRQ for this
* chained irqchip
* @parent_handler: the parent interrupt handler for the accumulated IRQ
* coming out of the gpiochip. If the interrupt is nested rather than
* cascaded, pass NULL in this handler argument
*/
gpio: simplify adding threaded interrupts This tries to simplify the use of CONFIG_GPIOLIB_IRQCHIP when using threaded interrupts: add a new call gpiochip_irqchip_add_nested() to indicate that we're dealing with a nested rather than a chained irqchip, then create a separate gpiochip_set_nested_irqchip() to mirror the gpiochip_set_chained_irqchip() call to connect the parent and child interrupts. In the nested case gpiochip_set_nested_irqchip() does nothing more than call irq_set_parent() on each valid child interrupt, which has little semantic effect in the kernel, but this is probably still formally correct. Update all drivers using nested interrupts to use gpiochip_irqchip_add_nested() so we can now see clearly which these users are. The DLN2 driver can drop its specific hack with .irq_not_threaded as we now recognize whether a chip is threaded or not from its use of gpiochip_irqchip_add_nested() signature rather than from inspecting .can_sleep. We rename the .irq_parent to .irq_chained_parent since this parent IRQ is only really kept around for the chained interrupt handlers. Cc: Lars Poeschel <poeschel@lemonage.de> Cc: Octavian Purdila <octavian.purdila@intel.com> Cc: Daniel Baluta <daniel.baluta@intel.com> Cc: Bin Gao <bin.gao@linux.intel.com> Cc: Mika Westerberg <mika.westerberg@linux.intel.com> Cc: Ajay Thomas <ajay.thomas.david.rajamanickam@intel.com> Cc: Semen Protsenko <semen.protsenko@globallogic.com> Cc: Alexander Stein <alexander.stein@systec-electronic.com> Cc: Phil Reid <preid@electromag.com.au> Cc: Bartosz Golaszewski <bgolaszewski@baylibre.com> Cc: Patrice Chotard <patrice.chotard@st.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-11-24 17:57:25 +08:00
static void gpiochip_set_cascaded_irqchip(struct gpio_chip *gpiochip,
struct irq_chip *irqchip,
unsigned int parent_irq,
gpio: simplify adding threaded interrupts This tries to simplify the use of CONFIG_GPIOLIB_IRQCHIP when using threaded interrupts: add a new call gpiochip_irqchip_add_nested() to indicate that we're dealing with a nested rather than a chained irqchip, then create a separate gpiochip_set_nested_irqchip() to mirror the gpiochip_set_chained_irqchip() call to connect the parent and child interrupts. In the nested case gpiochip_set_nested_irqchip() does nothing more than call irq_set_parent() on each valid child interrupt, which has little semantic effect in the kernel, but this is probably still formally correct. Update all drivers using nested interrupts to use gpiochip_irqchip_add_nested() so we can now see clearly which these users are. The DLN2 driver can drop its specific hack with .irq_not_threaded as we now recognize whether a chip is threaded or not from its use of gpiochip_irqchip_add_nested() signature rather than from inspecting .can_sleep. We rename the .irq_parent to .irq_chained_parent since this parent IRQ is only really kept around for the chained interrupt handlers. Cc: Lars Poeschel <poeschel@lemonage.de> Cc: Octavian Purdila <octavian.purdila@intel.com> Cc: Daniel Baluta <daniel.baluta@intel.com> Cc: Bin Gao <bin.gao@linux.intel.com> Cc: Mika Westerberg <mika.westerberg@linux.intel.com> Cc: Ajay Thomas <ajay.thomas.david.rajamanickam@intel.com> Cc: Semen Protsenko <semen.protsenko@globallogic.com> Cc: Alexander Stein <alexander.stein@systec-electronic.com> Cc: Phil Reid <preid@electromag.com.au> Cc: Bartosz Golaszewski <bgolaszewski@baylibre.com> Cc: Patrice Chotard <patrice.chotard@st.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-11-24 17:57:25 +08:00
irq_flow_handler_t parent_handler)
{
unsigned int offset;
if (!gpiochip->irq.domain) {
chip_err(gpiochip, "called %s before setting up irqchip\n",
__func__);
return;
}
if (parent_handler) {
if (gpiochip->can_sleep) {
chip_err(gpiochip,
"you cannot have chained interrupts on a "
"chip that may sleep\n");
return;
}
/*
* The parent irqchip is already using the chip_data for this
* irqchip, so our callbacks simply use the handler_data.
*/
irq_set_chained_handler_and_data(parent_irq, parent_handler,
gpiochip);
gpiochip->irq.parents = &parent_irq;
gpiochip->irq.num_parents = 1;
}
/* Set the parent IRQ for all affected IRQs */
for (offset = 0; offset < gpiochip->ngpio; offset++) {
if (!gpiochip_irqchip_irq_valid(gpiochip, offset))
continue;
irq_set_parent(irq_find_mapping(gpiochip->irq.domain, offset),
parent_irq);
}
}
gpio: simplify adding threaded interrupts This tries to simplify the use of CONFIG_GPIOLIB_IRQCHIP when using threaded interrupts: add a new call gpiochip_irqchip_add_nested() to indicate that we're dealing with a nested rather than a chained irqchip, then create a separate gpiochip_set_nested_irqchip() to mirror the gpiochip_set_chained_irqchip() call to connect the parent and child interrupts. In the nested case gpiochip_set_nested_irqchip() does nothing more than call irq_set_parent() on each valid child interrupt, which has little semantic effect in the kernel, but this is probably still formally correct. Update all drivers using nested interrupts to use gpiochip_irqchip_add_nested() so we can now see clearly which these users are. The DLN2 driver can drop its specific hack with .irq_not_threaded as we now recognize whether a chip is threaded or not from its use of gpiochip_irqchip_add_nested() signature rather than from inspecting .can_sleep. We rename the .irq_parent to .irq_chained_parent since this parent IRQ is only really kept around for the chained interrupt handlers. Cc: Lars Poeschel <poeschel@lemonage.de> Cc: Octavian Purdila <octavian.purdila@intel.com> Cc: Daniel Baluta <daniel.baluta@intel.com> Cc: Bin Gao <bin.gao@linux.intel.com> Cc: Mika Westerberg <mika.westerberg@linux.intel.com> Cc: Ajay Thomas <ajay.thomas.david.rajamanickam@intel.com> Cc: Semen Protsenko <semen.protsenko@globallogic.com> Cc: Alexander Stein <alexander.stein@systec-electronic.com> Cc: Phil Reid <preid@electromag.com.au> Cc: Bartosz Golaszewski <bgolaszewski@baylibre.com> Cc: Patrice Chotard <patrice.chotard@st.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-11-24 17:57:25 +08:00
/**
* gpiochip_set_chained_irqchip() - connects a chained irqchip to a gpiochip
* @gpiochip: the gpiochip to set the irqchip chain to
* @irqchip: the irqchip to chain to the gpiochip
* @parent_irq: the irq number corresponding to the parent IRQ for this
* chained irqchip
* @parent_handler: the parent interrupt handler for the accumulated IRQ
* coming out of the gpiochip. If the interrupt is nested rather than
* cascaded, pass NULL in this handler argument
*/
void gpiochip_set_chained_irqchip(struct gpio_chip *gpiochip,
struct irq_chip *irqchip,
unsigned int parent_irq,
gpio: simplify adding threaded interrupts This tries to simplify the use of CONFIG_GPIOLIB_IRQCHIP when using threaded interrupts: add a new call gpiochip_irqchip_add_nested() to indicate that we're dealing with a nested rather than a chained irqchip, then create a separate gpiochip_set_nested_irqchip() to mirror the gpiochip_set_chained_irqchip() call to connect the parent and child interrupts. In the nested case gpiochip_set_nested_irqchip() does nothing more than call irq_set_parent() on each valid child interrupt, which has little semantic effect in the kernel, but this is probably still formally correct. Update all drivers using nested interrupts to use gpiochip_irqchip_add_nested() so we can now see clearly which these users are. The DLN2 driver can drop its specific hack with .irq_not_threaded as we now recognize whether a chip is threaded or not from its use of gpiochip_irqchip_add_nested() signature rather than from inspecting .can_sleep. We rename the .irq_parent to .irq_chained_parent since this parent IRQ is only really kept around for the chained interrupt handlers. Cc: Lars Poeschel <poeschel@lemonage.de> Cc: Octavian Purdila <octavian.purdila@intel.com> Cc: Daniel Baluta <daniel.baluta@intel.com> Cc: Bin Gao <bin.gao@linux.intel.com> Cc: Mika Westerberg <mika.westerberg@linux.intel.com> Cc: Ajay Thomas <ajay.thomas.david.rajamanickam@intel.com> Cc: Semen Protsenko <semen.protsenko@globallogic.com> Cc: Alexander Stein <alexander.stein@systec-electronic.com> Cc: Phil Reid <preid@electromag.com.au> Cc: Bartosz Golaszewski <bgolaszewski@baylibre.com> Cc: Patrice Chotard <patrice.chotard@st.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-11-24 17:57:25 +08:00
irq_flow_handler_t parent_handler)
{
if (gpiochip->irq.threaded) {
chip_err(gpiochip, "tried to chain a threaded gpiochip\n");
return;
}
gpio: simplify adding threaded interrupts This tries to simplify the use of CONFIG_GPIOLIB_IRQCHIP when using threaded interrupts: add a new call gpiochip_irqchip_add_nested() to indicate that we're dealing with a nested rather than a chained irqchip, then create a separate gpiochip_set_nested_irqchip() to mirror the gpiochip_set_chained_irqchip() call to connect the parent and child interrupts. In the nested case gpiochip_set_nested_irqchip() does nothing more than call irq_set_parent() on each valid child interrupt, which has little semantic effect in the kernel, but this is probably still formally correct. Update all drivers using nested interrupts to use gpiochip_irqchip_add_nested() so we can now see clearly which these users are. The DLN2 driver can drop its specific hack with .irq_not_threaded as we now recognize whether a chip is threaded or not from its use of gpiochip_irqchip_add_nested() signature rather than from inspecting .can_sleep. We rename the .irq_parent to .irq_chained_parent since this parent IRQ is only really kept around for the chained interrupt handlers. Cc: Lars Poeschel <poeschel@lemonage.de> Cc: Octavian Purdila <octavian.purdila@intel.com> Cc: Daniel Baluta <daniel.baluta@intel.com> Cc: Bin Gao <bin.gao@linux.intel.com> Cc: Mika Westerberg <mika.westerberg@linux.intel.com> Cc: Ajay Thomas <ajay.thomas.david.rajamanickam@intel.com> Cc: Semen Protsenko <semen.protsenko@globallogic.com> Cc: Alexander Stein <alexander.stein@systec-electronic.com> Cc: Phil Reid <preid@electromag.com.au> Cc: Bartosz Golaszewski <bgolaszewski@baylibre.com> Cc: Patrice Chotard <patrice.chotard@st.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-11-24 17:57:25 +08:00
gpiochip_set_cascaded_irqchip(gpiochip, irqchip, parent_irq,
parent_handler);
}
EXPORT_SYMBOL_GPL(gpiochip_set_chained_irqchip);
gpio: simplify adding threaded interrupts This tries to simplify the use of CONFIG_GPIOLIB_IRQCHIP when using threaded interrupts: add a new call gpiochip_irqchip_add_nested() to indicate that we're dealing with a nested rather than a chained irqchip, then create a separate gpiochip_set_nested_irqchip() to mirror the gpiochip_set_chained_irqchip() call to connect the parent and child interrupts. In the nested case gpiochip_set_nested_irqchip() does nothing more than call irq_set_parent() on each valid child interrupt, which has little semantic effect in the kernel, but this is probably still formally correct. Update all drivers using nested interrupts to use gpiochip_irqchip_add_nested() so we can now see clearly which these users are. The DLN2 driver can drop its specific hack with .irq_not_threaded as we now recognize whether a chip is threaded or not from its use of gpiochip_irqchip_add_nested() signature rather than from inspecting .can_sleep. We rename the .irq_parent to .irq_chained_parent since this parent IRQ is only really kept around for the chained interrupt handlers. Cc: Lars Poeschel <poeschel@lemonage.de> Cc: Octavian Purdila <octavian.purdila@intel.com> Cc: Daniel Baluta <daniel.baluta@intel.com> Cc: Bin Gao <bin.gao@linux.intel.com> Cc: Mika Westerberg <mika.westerberg@linux.intel.com> Cc: Ajay Thomas <ajay.thomas.david.rajamanickam@intel.com> Cc: Semen Protsenko <semen.protsenko@globallogic.com> Cc: Alexander Stein <alexander.stein@systec-electronic.com> Cc: Phil Reid <preid@electromag.com.au> Cc: Bartosz Golaszewski <bgolaszewski@baylibre.com> Cc: Patrice Chotard <patrice.chotard@st.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-11-24 17:57:25 +08:00
/**
* gpiochip_set_nested_irqchip() - connects a nested irqchip to a gpiochip
* @gpiochip: the gpiochip to set the irqchip nested handler to
* @irqchip: the irqchip to nest to the gpiochip
* @parent_irq: the irq number corresponding to the parent IRQ for this
* nested irqchip
*/
void gpiochip_set_nested_irqchip(struct gpio_chip *gpiochip,
struct irq_chip *irqchip,
unsigned int parent_irq)
gpio: simplify adding threaded interrupts This tries to simplify the use of CONFIG_GPIOLIB_IRQCHIP when using threaded interrupts: add a new call gpiochip_irqchip_add_nested() to indicate that we're dealing with a nested rather than a chained irqchip, then create a separate gpiochip_set_nested_irqchip() to mirror the gpiochip_set_chained_irqchip() call to connect the parent and child interrupts. In the nested case gpiochip_set_nested_irqchip() does nothing more than call irq_set_parent() on each valid child interrupt, which has little semantic effect in the kernel, but this is probably still formally correct. Update all drivers using nested interrupts to use gpiochip_irqchip_add_nested() so we can now see clearly which these users are. The DLN2 driver can drop its specific hack with .irq_not_threaded as we now recognize whether a chip is threaded or not from its use of gpiochip_irqchip_add_nested() signature rather than from inspecting .can_sleep. We rename the .irq_parent to .irq_chained_parent since this parent IRQ is only really kept around for the chained interrupt handlers. Cc: Lars Poeschel <poeschel@lemonage.de> Cc: Octavian Purdila <octavian.purdila@intel.com> Cc: Daniel Baluta <daniel.baluta@intel.com> Cc: Bin Gao <bin.gao@linux.intel.com> Cc: Mika Westerberg <mika.westerberg@linux.intel.com> Cc: Ajay Thomas <ajay.thomas.david.rajamanickam@intel.com> Cc: Semen Protsenko <semen.protsenko@globallogic.com> Cc: Alexander Stein <alexander.stein@systec-electronic.com> Cc: Phil Reid <preid@electromag.com.au> Cc: Bartosz Golaszewski <bgolaszewski@baylibre.com> Cc: Patrice Chotard <patrice.chotard@st.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-11-24 17:57:25 +08:00
{
gpiochip_set_cascaded_irqchip(gpiochip, irqchip, parent_irq,
NULL);
}
EXPORT_SYMBOL_GPL(gpiochip_set_nested_irqchip);
/**
* gpiochip_irq_map() - maps an IRQ into a GPIO irqchip
* @d: the irqdomain used by this irqchip
* @irq: the global irq number used by this GPIO irqchip irq
* @hwirq: the local IRQ/GPIO line offset on this gpiochip
*
* This function will set up the mapping for a certain IRQ line on a
* gpiochip by assigning the gpiochip as chip data, and using the irqchip
* stored inside the gpiochip.
*/
int gpiochip_irq_map(struct irq_domain *d, unsigned int irq,
irq_hw_number_t hwirq)
{
struct gpio_chip *chip = d->host_data;
int err = 0;
gpiolib: allow gpio irqchip to map irqs dynamically Now IRQ mappings are always created for all (allowed) GPIOs in gpiochip in gpiochip_irqchip_add_key() which goes against the idea of SPARSE_IRQ and, as result, leads to: - increasing of memory consumption for IRQ descriptors most of which will never ever be used (espessially on platform with a high number of GPIOs). (sizeof(struct irq_desc) == 256 on my tested platforms) - imposibility to use GPIO irqchip APIs by gpio drivers when HW implements GPIO IRQ functionality as IRQ crossbar/router which has only limited number of IRQ outputs (example from [1], all GPIOs can be mapped on only 8 IRQs). Hence, remove static IRQ mapping code from gpiochip_irqchip_add_key() and instead replace irq_find_mapping() with irq_create_mapping() in gpiochip_to_irq(). Also add additional gpiochip_irqchip_irq_valid() calls in gpiochip_to_irq() and gpiochip_irq_map(). After this change gpio2irq mapping will happen the following way when GPIO irqchip APIs are used by gpio driver: - IRQ mappings will be created statically if driver passes first_irq>0 vlaue in gpiochip_irqchip_add_key(). - IRQ mappings will be created dynamically from gpio_to_irq() or of_irq_get(). Tested on am335x-evm and dra72-evm-revc. - dra72-evm-revc: number of created irq mappings decreased from 402 -> 135 Mem savings 267*256 = 68352 (66kB) - am335x-evm: number of created irq mappings decreased from 188 -> 63 Mem savings 125*256 = 32000 (31kB) [1] https://lkml.org/lkml/2017/6/15/428 Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2017-07-22 00:49:00 +08:00
if (!gpiochip_irqchip_irq_valid(chip, hwirq))
return -ENXIO;
irq_set_chip_data(irq, chip);
gpiolib: irqchip: use different lockdep class for each gpio irqchip Since IRQ chip helpers were introduced drivers lose ability to register separate lockdep classes for each registered GPIO IRQ chip and the gpiolib now is using shared lockdep class for all GPIO IRQ chips (gpiochip_irq_lock_class). As result, lockdep will produce warning when there are min two stacked GPIO chips and all of them are interrupt controllers. HW configuration which generates lockdep warning (TI dra7-evm): [SOC GPIO bankA.gpioX] <- irq - [pcf875x.gpioY] <- irq - DevZ.enable_irq_wake(pcf_gpioY_irq); The issue was reported in [1] and discussed [2]. ============================================= [ INFO: possible recursive locking detected ] 4.2.0-rc6-00013-g5d050ed-dirty #55 Not tainted --------------------------------------------- sh/63 is trying to acquire lock: (class){......}, at: [<c009b91c>] __irq_get_desc_lock+0x50/0x94 but task is already holding lock: (class){......}, at: [<c009b91c>] __irq_get_desc_lock+0x50/0x94 other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(class); lock(class); *** DEADLOCK *** May be due to missing lock nesting notation 7 locks held by sh/63: #0: (sb_writers#4){.+.+.+}, at: [<c016bbb8>] vfs_write+0x13c/0x164 #1: (&of->mutex){+.+.+.}, at: [<c01debf4>] kernfs_fop_write+0x4c/0x1a0 #2: (s_active#36){.+.+.+}, at: [<c01debfc>] kernfs_fop_write+0x54/0x1a0 #3: (pm_mutex){+.+.+.}, at: [<c009758c>] pm_suspend+0xec/0x4c4 #4: (&dev->mutex){......}, at: [<c03f77f8>] __device_suspend+0xd4/0x398 #5: (&gpio->lock){+.+.+.}, at: [<c009b940>] __irq_get_desc_lock+0x74/0x94 #6: (class){......}, at: [<c009b91c>] __irq_get_desc_lock+0x50/0x94 stack backtrace: CPU: 0 PID: 63 Comm: sh Not tainted 4.2.0-rc6-00013-g5d050ed-dirty #55 Hardware name: Generic DRA74X (Flattened Device Tree) [<c0016e24>] (unwind_backtrace) from [<c0013338>] (show_stack+0x10/0x14) [<c0013338>] (show_stack) from [<c05f6b24>] (dump_stack+0x84/0x9c) [<c05f6b24>] (dump_stack) from [<c00903f4>] (__lock_acquire+0x19c0/0x1e20) [<c00903f4>] (__lock_acquire) from [<c0091098>] (lock_acquire+0xa8/0x128) [<c0091098>] (lock_acquire) from [<c05fd61c>] (_raw_spin_lock_irqsave+0x38/0x4c) [<c05fd61c>] (_raw_spin_lock_irqsave) from [<c009b91c>] (__irq_get_desc_lock+0x50/0x94) [<c009b91c>] (__irq_get_desc_lock) from [<c009c4f4>] (irq_set_irq_wake+0x20/0xfc) [<c009c4f4>] (irq_set_irq_wake) from [<c0393ac4>] (pcf857x_irq_set_wake+0x24/0x54) [<c0393ac4>] (pcf857x_irq_set_wake) from [<c009c560>] (irq_set_irq_wake+0x8c/0xfc) [<c009c560>] (irq_set_irq_wake) from [<c04a02ac>] (gpio_keys_suspend+0x70/0xd4) [<c04a02ac>] (gpio_keys_suspend) from [<c03f6a00>] (dpm_run_callback+0x50/0x124) [<c03f6a00>] (dpm_run_callback) from [<c03f7830>] (__device_suspend+0x10c/0x398) [<c03f7830>] (__device_suspend) from [<c03f90f0>] (dpm_suspend+0x134/0x2f4) [<c03f90f0>] (dpm_suspend) from [<c0096e20>] (suspend_devices_and_enter+0xa8/0x728) [<c0096e20>] (suspend_devices_and_enter) from [<c00977cc>] (pm_suspend+0x32c/0x4c4) [<c00977cc>] (pm_suspend) from [<c0096060>] (state_store+0x64/0xb8) [<c0096060>] (state_store) from [<c01dec64>] (kernfs_fop_write+0xbc/0x1a0) [<c01dec64>] (kernfs_fop_write) from [<c016b280>] (__vfs_write+0x20/0xd8) [<c016b280>] (__vfs_write) from [<c016bb0c>] (vfs_write+0x90/0x164) [<c016bb0c>] (vfs_write) from [<c016c330>] (SyS_write+0x44/0x9c) [<c016c330>] (SyS_write) from [<c000f500>] (ret_fast_syscall+0x0/0x54) Lets fix it by using separate lockdep class for each registered GPIO IRQ Chip. This is done by wrapping gpiochip_irqchip_add call into macros. The implementation of this patch inspired by solution done by Nicolas Boichat for regmap [3] [1] http://www.spinics.net/lists/linux-gpio/msg05844.html [2] http://www.spinics.net/lists/linux-gpio/msg06021.html [3] http://www.spinics.net/lists/arm-kernel/msg429834.html Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Roger Quadros <rogerq@ti.com> Reported-by: Roger Quadros <rogerq@ti.com> Tested-by: Roger Quadros <rogerq@ti.com> Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-08-17 20:35:23 +08:00
/*
* This lock class tells lockdep that GPIO irqs are in a different
* category than their parents, so it won't report false recursion.
*/
irq_set_lockdep_class(irq, chip->irq.lock_key);
irq_set_chip_and_handler(irq, chip->irq.chip, chip->irq.handler);
gpio: simplify adding threaded interrupts This tries to simplify the use of CONFIG_GPIOLIB_IRQCHIP when using threaded interrupts: add a new call gpiochip_irqchip_add_nested() to indicate that we're dealing with a nested rather than a chained irqchip, then create a separate gpiochip_set_nested_irqchip() to mirror the gpiochip_set_chained_irqchip() call to connect the parent and child interrupts. In the nested case gpiochip_set_nested_irqchip() does nothing more than call irq_set_parent() on each valid child interrupt, which has little semantic effect in the kernel, but this is probably still formally correct. Update all drivers using nested interrupts to use gpiochip_irqchip_add_nested() so we can now see clearly which these users are. The DLN2 driver can drop its specific hack with .irq_not_threaded as we now recognize whether a chip is threaded or not from its use of gpiochip_irqchip_add_nested() signature rather than from inspecting .can_sleep. We rename the .irq_parent to .irq_chained_parent since this parent IRQ is only really kept around for the chained interrupt handlers. Cc: Lars Poeschel <poeschel@lemonage.de> Cc: Octavian Purdila <octavian.purdila@intel.com> Cc: Daniel Baluta <daniel.baluta@intel.com> Cc: Bin Gao <bin.gao@linux.intel.com> Cc: Mika Westerberg <mika.westerberg@linux.intel.com> Cc: Ajay Thomas <ajay.thomas.david.rajamanickam@intel.com> Cc: Semen Protsenko <semen.protsenko@globallogic.com> Cc: Alexander Stein <alexander.stein@systec-electronic.com> Cc: Phil Reid <preid@electromag.com.au> Cc: Bartosz Golaszewski <bgolaszewski@baylibre.com> Cc: Patrice Chotard <patrice.chotard@st.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-11-24 17:57:25 +08:00
/* Chips that use nested thread handlers have them marked */
if (chip->irq.threaded)
irq_set_nested_thread(irq, 1);
irq_set_noprobe(irq);
if (chip->irq.num_parents == 1)
err = irq_set_parent(irq, chip->irq.parents[0]);
else if (chip->irq.map)
err = irq_set_parent(irq, chip->irq.map[hwirq]);
if (err < 0)
return err;
/*
* No set-up of the hardware will happen if IRQ_TYPE_NONE
* is passed as default type.
*/
if (chip->irq.default_type != IRQ_TYPE_NONE)
irq_set_irq_type(irq, chip->irq.default_type);
return 0;
}
EXPORT_SYMBOL_GPL(gpiochip_irq_map);
void gpiochip_irq_unmap(struct irq_domain *d, unsigned int irq)
{
struct gpio_chip *chip = d->host_data;
if (chip->irq.threaded)
irq_set_nested_thread(irq, 0);
irq_set_chip_and_handler(irq, NULL, NULL);
irq_set_chip_data(irq, NULL);
}
EXPORT_SYMBOL_GPL(gpiochip_irq_unmap);
static const struct irq_domain_ops gpiochip_domain_ops = {
.map = gpiochip_irq_map,
.unmap = gpiochip_irq_unmap,
/* Virtually all GPIO irqchips are twocell:ed */
.xlate = irq_domain_xlate_twocell,
};
static int gpiochip_irq_reqres(struct irq_data *d)
{
struct gpio_chip *chip = irq_data_get_irq_chip_data(d);
if (!try_module_get(chip->gpiodev->owner))
return -ENODEV;
if (gpiochip_lock_as_irq(chip, d->hwirq)) {
chip_err(chip,
"unable to lock HW IRQ %lu for IRQ\n",
d->hwirq);
module_put(chip->gpiodev->owner);
return -EINVAL;
}
return 0;
}
static void gpiochip_irq_relres(struct irq_data *d)
{
struct gpio_chip *chip = irq_data_get_irq_chip_data(d);
gpiochip_unlock_as_irq(chip, d->hwirq);
module_put(chip->gpiodev->owner);
}
static int gpiochip_to_irq(struct gpio_chip *chip, unsigned offset)
{
gpiolib: allow gpio irqchip to map irqs dynamically Now IRQ mappings are always created for all (allowed) GPIOs in gpiochip in gpiochip_irqchip_add_key() which goes against the idea of SPARSE_IRQ and, as result, leads to: - increasing of memory consumption for IRQ descriptors most of which will never ever be used (espessially on platform with a high number of GPIOs). (sizeof(struct irq_desc) == 256 on my tested platforms) - imposibility to use GPIO irqchip APIs by gpio drivers when HW implements GPIO IRQ functionality as IRQ crossbar/router which has only limited number of IRQ outputs (example from [1], all GPIOs can be mapped on only 8 IRQs). Hence, remove static IRQ mapping code from gpiochip_irqchip_add_key() and instead replace irq_find_mapping() with irq_create_mapping() in gpiochip_to_irq(). Also add additional gpiochip_irqchip_irq_valid() calls in gpiochip_to_irq() and gpiochip_irq_map(). After this change gpio2irq mapping will happen the following way when GPIO irqchip APIs are used by gpio driver: - IRQ mappings will be created statically if driver passes first_irq>0 vlaue in gpiochip_irqchip_add_key(). - IRQ mappings will be created dynamically from gpio_to_irq() or of_irq_get(). Tested on am335x-evm and dra72-evm-revc. - dra72-evm-revc: number of created irq mappings decreased from 402 -> 135 Mem savings 267*256 = 68352 (66kB) - am335x-evm: number of created irq mappings decreased from 188 -> 63 Mem savings 125*256 = 32000 (31kB) [1] https://lkml.org/lkml/2017/6/15/428 Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2017-07-22 00:49:00 +08:00
if (!gpiochip_irqchip_irq_valid(chip, offset))
return -ENXIO;
return irq_create_mapping(chip->irq.domain, offset);
}
/**
* gpiochip_add_irqchip() - adds an IRQ chip to a GPIO chip
* @gpiochip: the GPIO chip to add the IRQ chip to
* @lock_key: lockdep class
*/
static int gpiochip_add_irqchip(struct gpio_chip *gpiochip,
struct lock_class_key *lock_key)
{
struct irq_chip *irqchip = gpiochip->irq.chip;
const struct irq_domain_ops *ops;
struct device_node *np;
unsigned int type;
unsigned int i;
if (!irqchip)
return 0;
if (gpiochip->irq.parent_handler && gpiochip->can_sleep) {
chip_err(gpiochip, "you cannot have chained interrupts on a "
"chip that may sleep\n");
return -EINVAL;
}
np = gpiochip->gpiodev->dev.of_node;
type = gpiochip->irq.default_type;
/*
* Specifying a default trigger is a terrible idea if DT or ACPI is
* used to configure the interrupts, as you may end up with
* conflicting triggers. Tell the user, and reset to NONE.
*/
if (WARN(np && type != IRQ_TYPE_NONE,
"%s: Ignoring %u default trigger\n", np->full_name, type))
type = IRQ_TYPE_NONE;
if (has_acpi_companion(gpiochip->parent) && type != IRQ_TYPE_NONE) {
acpi_handle_warn(ACPI_HANDLE(gpiochip->parent),
"Ignoring %u default trigger\n", type);
type = IRQ_TYPE_NONE;
}
gpiochip->to_irq = gpiochip_to_irq;
gpiochip->irq.default_type = type;
gpiochip->irq.lock_key = lock_key;
if (gpiochip->irq.domain_ops)
ops = gpiochip->irq.domain_ops;
else
ops = &gpiochip_domain_ops;
gpiochip->irq.domain = irq_domain_add_simple(np, gpiochip->ngpio,
gpiochip->irq.first,
ops, gpiochip);
if (!gpiochip->irq.domain)
return -EINVAL;
/*
* It is possible for a driver to override this, but only if the
* alternative functions are both implemented.
*/
if (!irqchip->irq_request_resources &&
!irqchip->irq_release_resources) {
irqchip->irq_request_resources = gpiochip_irq_reqres;
irqchip->irq_release_resources = gpiochip_irq_relres;
}
if (gpiochip->irq.parent_handler) {
void *data = gpiochip->irq.parent_handler_data ?: gpiochip;
for (i = 0; i < gpiochip->irq.num_parents; i++) {
/*
* The parent IRQ chip is already using the chip_data
* for this IRQ chip, so our callbacks simply use the
* handler_data.
*/
irq_set_chained_handler_and_data(gpiochip->irq.parents[i],
gpiochip->irq.parent_handler,
data);
}
}
acpi_gpiochip_request_interrupts(gpiochip);
return 0;
}
/**
* gpiochip_irqchip_remove() - removes an irqchip added to a gpiochip
* @gpiochip: the gpiochip to remove the irqchip from
*
* This is called only from gpiochip_remove()
*/
static void gpiochip_irqchip_remove(struct gpio_chip *gpiochip)
{
unsigned int offset;
acpi_gpiochip_free_interrupts(gpiochip);
if (gpiochip->irq.chip && gpiochip->irq.parent_handler) {
struct gpio_irq_chip *irq = &gpiochip->irq;
unsigned int i;
for (i = 0; i < irq->num_parents; i++)
irq_set_chained_handler_and_data(irq->parents[i],
NULL, NULL);
}
/* Remove all IRQ mappings and delete the domain */
if (gpiochip->irq.domain) {
unsigned int irq;
for (offset = 0; offset < gpiochip->ngpio; offset++) {
if (!gpiochip_irqchip_irq_valid(gpiochip, offset))
continue;
irq = irq_find_mapping(gpiochip->irq.domain, offset);
irq_dispose_mapping(irq);
}
irq_domain_remove(gpiochip->irq.domain);
}
if (gpiochip->irq.chip) {
gpiochip->irq.chip->irq_request_resources = NULL;
gpiochip->irq.chip->irq_release_resources = NULL;
gpiochip->irq.chip = NULL;
}
gpiochip_irqchip_free_valid_mask(gpiochip);
}
/**
* gpiochip_irqchip_add_key() - adds an irqchip to a gpiochip
* @gpiochip: the gpiochip to add the irqchip to
* @irqchip: the irqchip to add to the gpiochip
* @first_irq: if not dynamically assigned, the base (first) IRQ to
* allocate gpiochip irqs from
* @handler: the irq handler to use (often a predefined irq core function)
* @type: the default type for IRQs on this irqchip, pass IRQ_TYPE_NONE
* to have the core avoid setting up any default type in the hardware.
* @threaded: whether this irqchip uses a nested thread handler
gpiolib: irqchip: use different lockdep class for each gpio irqchip Since IRQ chip helpers were introduced drivers lose ability to register separate lockdep classes for each registered GPIO IRQ chip and the gpiolib now is using shared lockdep class for all GPIO IRQ chips (gpiochip_irq_lock_class). As result, lockdep will produce warning when there are min two stacked GPIO chips and all of them are interrupt controllers. HW configuration which generates lockdep warning (TI dra7-evm): [SOC GPIO bankA.gpioX] <- irq - [pcf875x.gpioY] <- irq - DevZ.enable_irq_wake(pcf_gpioY_irq); The issue was reported in [1] and discussed [2]. ============================================= [ INFO: possible recursive locking detected ] 4.2.0-rc6-00013-g5d050ed-dirty #55 Not tainted --------------------------------------------- sh/63 is trying to acquire lock: (class){......}, at: [<c009b91c>] __irq_get_desc_lock+0x50/0x94 but task is already holding lock: (class){......}, at: [<c009b91c>] __irq_get_desc_lock+0x50/0x94 other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(class); lock(class); *** DEADLOCK *** May be due to missing lock nesting notation 7 locks held by sh/63: #0: (sb_writers#4){.+.+.+}, at: [<c016bbb8>] vfs_write+0x13c/0x164 #1: (&of->mutex){+.+.+.}, at: [<c01debf4>] kernfs_fop_write+0x4c/0x1a0 #2: (s_active#36){.+.+.+}, at: [<c01debfc>] kernfs_fop_write+0x54/0x1a0 #3: (pm_mutex){+.+.+.}, at: [<c009758c>] pm_suspend+0xec/0x4c4 #4: (&dev->mutex){......}, at: [<c03f77f8>] __device_suspend+0xd4/0x398 #5: (&gpio->lock){+.+.+.}, at: [<c009b940>] __irq_get_desc_lock+0x74/0x94 #6: (class){......}, at: [<c009b91c>] __irq_get_desc_lock+0x50/0x94 stack backtrace: CPU: 0 PID: 63 Comm: sh Not tainted 4.2.0-rc6-00013-g5d050ed-dirty #55 Hardware name: Generic DRA74X (Flattened Device Tree) [<c0016e24>] (unwind_backtrace) from [<c0013338>] (show_stack+0x10/0x14) [<c0013338>] (show_stack) from [<c05f6b24>] (dump_stack+0x84/0x9c) [<c05f6b24>] (dump_stack) from [<c00903f4>] (__lock_acquire+0x19c0/0x1e20) [<c00903f4>] (__lock_acquire) from [<c0091098>] (lock_acquire+0xa8/0x128) [<c0091098>] (lock_acquire) from [<c05fd61c>] (_raw_spin_lock_irqsave+0x38/0x4c) [<c05fd61c>] (_raw_spin_lock_irqsave) from [<c009b91c>] (__irq_get_desc_lock+0x50/0x94) [<c009b91c>] (__irq_get_desc_lock) from [<c009c4f4>] (irq_set_irq_wake+0x20/0xfc) [<c009c4f4>] (irq_set_irq_wake) from [<c0393ac4>] (pcf857x_irq_set_wake+0x24/0x54) [<c0393ac4>] (pcf857x_irq_set_wake) from [<c009c560>] (irq_set_irq_wake+0x8c/0xfc) [<c009c560>] (irq_set_irq_wake) from [<c04a02ac>] (gpio_keys_suspend+0x70/0xd4) [<c04a02ac>] (gpio_keys_suspend) from [<c03f6a00>] (dpm_run_callback+0x50/0x124) [<c03f6a00>] (dpm_run_callback) from [<c03f7830>] (__device_suspend+0x10c/0x398) [<c03f7830>] (__device_suspend) from [<c03f90f0>] (dpm_suspend+0x134/0x2f4) [<c03f90f0>] (dpm_suspend) from [<c0096e20>] (suspend_devices_and_enter+0xa8/0x728) [<c0096e20>] (suspend_devices_and_enter) from [<c00977cc>] (pm_suspend+0x32c/0x4c4) [<c00977cc>] (pm_suspend) from [<c0096060>] (state_store+0x64/0xb8) [<c0096060>] (state_store) from [<c01dec64>] (kernfs_fop_write+0xbc/0x1a0) [<c01dec64>] (kernfs_fop_write) from [<c016b280>] (__vfs_write+0x20/0xd8) [<c016b280>] (__vfs_write) from [<c016bb0c>] (vfs_write+0x90/0x164) [<c016bb0c>] (vfs_write) from [<c016c330>] (SyS_write+0x44/0x9c) [<c016c330>] (SyS_write) from [<c000f500>] (ret_fast_syscall+0x0/0x54) Lets fix it by using separate lockdep class for each registered GPIO IRQ Chip. This is done by wrapping gpiochip_irqchip_add call into macros. The implementation of this patch inspired by solution done by Nicolas Boichat for regmap [3] [1] http://www.spinics.net/lists/linux-gpio/msg05844.html [2] http://www.spinics.net/lists/linux-gpio/msg06021.html [3] http://www.spinics.net/lists/arm-kernel/msg429834.html Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Roger Quadros <rogerq@ti.com> Reported-by: Roger Quadros <rogerq@ti.com> Tested-by: Roger Quadros <rogerq@ti.com> Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-08-17 20:35:23 +08:00
* @lock_key: lockdep class
*
* This function closely associates a certain irqchip with a certain
* gpiochip, providing an irq domain to translate the local IRQs to
* global irqs in the gpiolib core, and making sure that the gpiochip
* is passed as chip data to all related functions. Driver callbacks
* need to use gpiochip_get_data() to get their local state containers back
* from the gpiochip passed as chip data. An irqdomain will be stored
* in the gpiochip that shall be used by the driver to handle IRQ number
* translation. The gpiochip will need to be initialized and registered
* before calling this function.
*
* This function will handle two cell:ed simple IRQs and assumes all
* the pins on the gpiochip can generate a unique IRQ. Everything else
* need to be open coded.
*/
int gpiochip_irqchip_add_key(struct gpio_chip *gpiochip,
struct irq_chip *irqchip,
unsigned int first_irq,
irq_flow_handler_t handler,
unsigned int type,
bool threaded,
struct lock_class_key *lock_key)
{
struct device_node *of_node;
if (!gpiochip || !irqchip)
return -EINVAL;
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>
2015-11-04 16:56:26 +08:00
if (!gpiochip->parent) {
pr_err("missing gpiochip .dev parent pointer\n");
return -EINVAL;
}
gpiochip->irq.threaded = threaded;
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>
2015-11-04 16:56:26 +08:00
of_node = gpiochip->parent->of_node;
#ifdef CONFIG_OF_GPIO
/*
* If the gpiochip has an assigned OF node this takes precedence
* FIXME: get rid of this and use gpiochip->parent->of_node
* everywhere
*/
if (gpiochip->of_node)
of_node = gpiochip->of_node;
#endif
/*
* Specifying a default trigger is a terrible idea if DT or ACPI is
* used to configure the interrupts, as you may end-up with
* conflicting triggers. Tell the user, and reset to NONE.
*/
if (WARN(of_node && type != IRQ_TYPE_NONE,
"%pOF: Ignoring %d default trigger\n", of_node, type))
type = IRQ_TYPE_NONE;
if (has_acpi_companion(gpiochip->parent) && type != IRQ_TYPE_NONE) {
acpi_handle_warn(ACPI_HANDLE(gpiochip->parent),
"Ignoring %d default trigger\n", type);
type = IRQ_TYPE_NONE;
}
gpiochip->irq.chip = irqchip;
gpiochip->irq.handler = handler;
gpiochip->irq.default_type = type;
gpiochip->to_irq = gpiochip_to_irq;
gpiochip->irq.lock_key = lock_key;
gpiochip->irq.domain = irq_domain_add_simple(of_node,
gpiochip->ngpio, first_irq,
&gpiochip_domain_ops, gpiochip);
if (!gpiochip->irq.domain) {
gpiochip->irq.chip = NULL;
return -EINVAL;
}
/*
* It is possible for a driver to override this, but only if the
* alternative functions are both implemented.
*/
if (!irqchip->irq_request_resources &&
!irqchip->irq_release_resources) {
irqchip->irq_request_resources = gpiochip_irq_reqres;
irqchip->irq_release_resources = gpiochip_irq_relres;
}
acpi_gpiochip_request_interrupts(gpiochip);
return 0;
}
EXPORT_SYMBOL_GPL(gpiochip_irqchip_add_key);
#else /* CONFIG_GPIOLIB_IRQCHIP */
static inline int gpiochip_add_irqchip(struct gpio_chip *gpiochip,
struct lock_class_key *key)
{
return 0;
}
static void gpiochip_irqchip_remove(struct gpio_chip *gpiochip) {}
static inline int gpiochip_irqchip_init_valid_mask(struct gpio_chip *gpiochip)
{
return 0;
}
static inline void gpiochip_irqchip_free_valid_mask(struct gpio_chip *gpiochip)
{ }
#endif /* CONFIG_GPIOLIB_IRQCHIP */
/**
* gpiochip_generic_request() - request the gpio function for a pin
* @chip: the gpiochip owning the GPIO
* @offset: the offset of the GPIO to request for GPIO function
*/
int gpiochip_generic_request(struct gpio_chip *chip, unsigned offset)
{
return pinctrl_gpio_request(chip->gpiodev->base + offset);
}
EXPORT_SYMBOL_GPL(gpiochip_generic_request);
/**
* gpiochip_generic_free() - free the gpio function from a pin
* @chip: the gpiochip to request the gpio function for
* @offset: the offset of the GPIO to free from GPIO function
*/
void gpiochip_generic_free(struct gpio_chip *chip, unsigned offset)
{
pinctrl_gpio_free(chip->gpiodev->base + offset);
}
EXPORT_SYMBOL_GPL(gpiochip_generic_free);
/**
* gpiochip_generic_config() - apply configuration for a pin
* @chip: the gpiochip owning the GPIO
* @offset: the offset of the GPIO to apply the configuration
* @config: the configuration to be applied
*/
int gpiochip_generic_config(struct gpio_chip *chip, unsigned offset,
unsigned long config)
{
return pinctrl_gpio_set_config(chip->gpiodev->base + offset, config);
}
EXPORT_SYMBOL_GPL(gpiochip_generic_config);
#ifdef CONFIG_PINCTRL
/**
* gpiochip_add_pingroup_range() - add a range for GPIO <-> pin mapping
* @chip: the gpiochip to add the range for
* @pctldev: the pin controller to map to
* @gpio_offset: the start offset in the current gpio_chip number space
* @pin_group: name of the pin group inside the pin controller
*/
int gpiochip_add_pingroup_range(struct gpio_chip *chip,
struct pinctrl_dev *pctldev,
unsigned int gpio_offset, const char *pin_group)
{
struct gpio_pin_range *pin_range;
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
struct gpio_device *gdev = chip->gpiodev;
int ret;
pin_range = kzalloc(sizeof(*pin_range), GFP_KERNEL);
if (!pin_range) {
chip_err(chip, "failed to allocate pin ranges\n");
return -ENOMEM;
}
/* Use local offset as range ID */
pin_range->range.id = gpio_offset;
pin_range->range.gc = chip;
pin_range->range.name = chip->label;
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
pin_range->range.base = gdev->base + gpio_offset;
pin_range->pctldev = pctldev;
ret = pinctrl_get_group_pins(pctldev, pin_group,
&pin_range->range.pins,
&pin_range->range.npins);
if (ret < 0) {
kfree(pin_range);
return ret;
}
pinctrl_add_gpio_range(pctldev, &pin_range->range);
chip_dbg(chip, "created GPIO range %d->%d ==> %s PINGRP %s\n",
gpio_offset, gpio_offset + pin_range->range.npins - 1,
pinctrl_dev_get_devname(pctldev), pin_group);
list_add_tail(&pin_range->node, &gdev->pin_ranges);
return 0;
}
EXPORT_SYMBOL_GPL(gpiochip_add_pingroup_range);
/**
* gpiochip_add_pin_range() - add a range for GPIO <-> pin mapping
* @chip: the gpiochip to add the range for
* @pinctl_name: the dev_name() of the pin controller to map to
* @gpio_offset: the start offset in the current gpio_chip number space
* @pin_offset: the start offset in the pin controller number space
* @npins: the number of pins from the offset of each pin space (GPIO and
* pin controller) to accumulate in this range
*
* Returns:
* 0 on success, or a negative error-code on failure.
*/
int gpiochip_add_pin_range(struct gpio_chip *chip, const char *pinctl_name,
unsigned int gpio_offset, unsigned int pin_offset,
unsigned int npins)
{
struct gpio_pin_range *pin_range;
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
struct gpio_device *gdev = chip->gpiodev;
int ret;
pin_range = kzalloc(sizeof(*pin_range), GFP_KERNEL);
if (!pin_range) {
chip_err(chip, "failed to allocate pin ranges\n");
return -ENOMEM;
}
/* Use local offset as range ID */
pin_range->range.id = gpio_offset;
pin_range->range.gc = chip;
pin_range->range.name = chip->label;
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
pin_range->range.base = gdev->base + gpio_offset;
pin_range->range.pin_base = pin_offset;
pin_range->range.npins = npins;
pin_range->pctldev = pinctrl_find_and_add_gpio_range(pinctl_name,
&pin_range->range);
if (IS_ERR(pin_range->pctldev)) {
ret = PTR_ERR(pin_range->pctldev);
chip_err(chip, "could not create pin range\n");
kfree(pin_range);
return ret;
}
chip_dbg(chip, "created GPIO range %d->%d ==> %s PIN %d->%d\n",
gpio_offset, gpio_offset + npins - 1,
pinctl_name,
pin_offset, pin_offset + npins - 1);
list_add_tail(&pin_range->node, &gdev->pin_ranges);
return 0;
}
EXPORT_SYMBOL_GPL(gpiochip_add_pin_range);
/**
* gpiochip_remove_pin_ranges() - remove all the GPIO <-> pin mappings
* @chip: the chip to remove all the mappings for
*/
void gpiochip_remove_pin_ranges(struct gpio_chip *chip)
{
struct gpio_pin_range *pin_range, *tmp;
struct gpio_device *gdev = chip->gpiodev;
list_for_each_entry_safe(pin_range, tmp, &gdev->pin_ranges, node) {
list_del(&pin_range->node);
pinctrl_remove_gpio_range(pin_range->pctldev,
&pin_range->range);
kfree(pin_range);
}
}
EXPORT_SYMBOL_GPL(gpiochip_remove_pin_ranges);
#endif /* CONFIG_PINCTRL */
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/* These "optional" allocation calls help prevent drivers from stomping
* on each other, and help provide better diagnostics in debugfs.
* They're called even less than the "set direction" calls.
*/
static int gpiod_request_commit(struct gpio_desc *desc, const char *label)
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
{
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
struct gpio_chip *chip = desc->gdev->chip;
int status;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
unsigned long flags;
spin_lock_irqsave(&gpio_lock, flags);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/* NOTE: gpio_request() can be called in early boot,
* before IRQs are enabled, for non-sleeping (SOC) GPIOs.
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
*/
if (test_and_set_bit(FLAG_REQUESTED, &desc->flags) == 0) {
desc_set_label(desc, label ? : "?");
status = 0;
} else {
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
status = -EBUSY;
goto done;
}
if (chip->request) {
/* chip->request may sleep */
spin_unlock_irqrestore(&gpio_lock, flags);
status = chip->request(chip, gpio_chip_hwgpio(desc));
spin_lock_irqsave(&gpio_lock, flags);
if (status < 0) {
desc_set_label(desc, NULL);
clear_bit(FLAG_REQUESTED, &desc->flags);
goto done;
}
}
if (chip->get_direction) {
/* chip->get_direction may sleep */
spin_unlock_irqrestore(&gpio_lock, flags);
gpiod_get_direction(desc);
spin_lock_irqsave(&gpio_lock, flags);
}
done:
spin_unlock_irqrestore(&gpio_lock, flags);
return status;
}
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
/*
* This descriptor validation needs to be inserted verbatim into each
* function taking a descriptor, so we need to use a preprocessor
* macro to avoid endless duplication. If the desc is NULL it is an
* optional GPIO and calls should just bail out.
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
*/
#define VALIDATE_DESC(desc) do { \
if (!desc) \
return 0; \
if (IS_ERR(desc)) { \
pr_warn("%s: invalid GPIO (errorpointer)\n", __func__); \
return PTR_ERR(desc); \
} \
if (!desc->gdev) { \
pr_warn("%s: invalid GPIO (no device)\n", __func__); \
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
return -EINVAL; \
} \
if ( !desc->gdev->chip ) { \
dev_warn(&desc->gdev->dev, \
"%s: backing chip is gone\n", __func__); \
return 0; \
} } while (0)
#define VALIDATE_DESC_VOID(desc) do { \
if (!desc) \
return; \
if (IS_ERR(desc)) { \
pr_warn("%s: invalid GPIO (errorpointer)\n", __func__); \
return; \
} \
if (!desc->gdev) { \
pr_warn("%s: invalid GPIO (no device)\n", __func__); \
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
return; \
} \
if (!desc->gdev->chip) { \
dev_warn(&desc->gdev->dev, \
"%s: backing chip is gone\n", __func__); \
return; \
} } while (0)
int gpiod_request(struct gpio_desc *desc, const char *label)
{
int status = -EPROBE_DEFER;
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
struct gpio_device *gdev;
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
VALIDATE_DESC(desc);
gdev = desc->gdev;
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
if (try_module_get(gdev->owner)) {
status = gpiod_request_commit(desc, label);
if (status < 0)
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
module_put(gdev->owner);
else
get_device(&gdev->dev);
}
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
if (status)
gpiod_dbg(desc, "%s: status %d\n", __func__, status);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
return status;
}
static bool gpiod_free_commit(struct gpio_desc *desc)
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
{
bool ret = false;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
unsigned long flags;
struct gpio_chip *chip;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
might_sleep();
gpiod_unexport(desc);
gpio: sysfs interface This adds a simple sysfs interface for GPIOs. /sys/class/gpio /export ... asks the kernel to export a GPIO to userspace /unexport ... to return a GPIO to the kernel /gpioN ... for each exported GPIO #N /value ... always readable, writes fail for input GPIOs /direction ... r/w as: in, out (default low); write high, low /gpiochipN ... for each gpiochip; #N is its first GPIO /base ... (r/o) same as N /label ... (r/o) descriptive, not necessarily unique /ngpio ... (r/o) number of GPIOs; numbered N .. N+(ngpio - 1) GPIOs claimed by kernel code may be exported by its owner using a new gpio_export() call, which should be most useful for driver debugging. Such exports may optionally be done without a "direction" attribute. Userspace may ask to take over a GPIO by writing to a sysfs control file, helping to cope with incomplete board support or other "one-off" requirements that don't merit full kernel support: echo 23 > /sys/class/gpio/export ... will gpio_request(23, "sysfs") and gpio_export(23); use /sys/class/gpio/gpio-23/direction to (re)configure it, when that GPIO can be used as both input and output. echo 23 > /sys/class/gpio/unexport ... will gpio_free(23), when it was exported as above The extra D-space footprint is a few hundred bytes, except for the sysfs resources associated with each exported GPIO. The additional I-space footprint is about two thirds of the current size of gpiolib (!). Since no /dev node creation is involved, no "udev" support is needed. Related changes: * This adds a device pointer to "struct gpio_chip". When GPIO providers initialize that, sysfs gpio class devices become children of that device instead of being "virtual" devices. * The (few) gpio_chip providers which have such a device node have been updated. * Some gpio_chip drivers also needed to update their module "owner" field ... for which missing kerneldoc was added. * Some gpio_chips don't support input GPIOs. Those GPIOs are now flagged appropriately when the chip is registered. Based on previous patches, and discussion both on and off LKML. A Documentation/ABI/testing/sysfs-gpio update is ready to submit once this merges to mainline. [akpm@linux-foundation.org: a few maintenance build fixes] Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Guennadi Liakhovetski <g.liakhovetski@pengutronix.de> Cc: Greg KH <greg@kroah.com> Cc: Kay Sievers <kay.sievers@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 16:46:07 +08:00
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
spin_lock_irqsave(&gpio_lock, flags);
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
chip = desc->gdev->chip;
if (chip && test_bit(FLAG_REQUESTED, &desc->flags)) {
if (chip->free) {
spin_unlock_irqrestore(&gpio_lock, flags);
might_sleep_if(chip->can_sleep);
chip->free(chip, gpio_chip_hwgpio(desc));
spin_lock_irqsave(&gpio_lock, flags);
}
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
desc_set_label(desc, NULL);
clear_bit(FLAG_ACTIVE_LOW, &desc->flags);
clear_bit(FLAG_REQUESTED, &desc->flags);
clear_bit(FLAG_OPEN_DRAIN, &desc->flags);
clear_bit(FLAG_OPEN_SOURCE, &desc->flags);
clear_bit(FLAG_IS_HOGGED, &desc->flags);
ret = true;
}
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
spin_unlock_irqrestore(&gpio_lock, flags);
return ret;
}
void gpiod_free(struct gpio_desc *desc)
{
if (desc && desc->gdev && gpiod_free_commit(desc)) {
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
module_put(desc->gdev->owner);
put_device(&desc->gdev->dev);
} else {
WARN_ON(extra_checks);
}
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
}
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/**
* gpiochip_is_requested - return string iff signal was requested
* @chip: controller managing the signal
* @offset: of signal within controller's 0..(ngpio - 1) range
*
* Returns NULL if the GPIO is not currently requested, else a string.
* The string returned is the label passed to gpio_request(); if none has been
* passed it is a meaningless, non-NULL constant.
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
*
* This function is for use by GPIO controller drivers. The label can
* help with diagnostics, and knowing that the signal is used as a GPIO
* can help avoid accidentally multiplexing it to another controller.
*/
const char *gpiochip_is_requested(struct gpio_chip *chip, unsigned offset)
{
struct gpio_desc *desc;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
if (offset >= chip->ngpio)
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
return NULL;
desc = &chip->gpiodev->descs[offset];
if (test_bit(FLAG_REQUESTED, &desc->flags) == 0)
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
return NULL;
return desc->label;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
}
EXPORT_SYMBOL_GPL(gpiochip_is_requested);
/**
* gpiochip_request_own_desc - Allow GPIO chip to request its own descriptor
* @chip: GPIO chip
* @hwnum: hardware number of the GPIO for which to request the descriptor
* @label: label for the GPIO
*
* Function allows GPIO chip drivers to request and use their own GPIO
* descriptors via gpiolib API. Difference to gpiod_request() is that this
* function will not increase reference count of the GPIO chip module. This
* allows the GPIO chip module to be unloaded as needed (we assume that the
* GPIO chip driver handles freeing the GPIOs it has requested).
*
* Returns:
* A pointer to the GPIO descriptor, or an ERR_PTR()-encoded negative error
* code on failure.
*/
struct gpio_desc *gpiochip_request_own_desc(struct gpio_chip *chip, u16 hwnum,
const char *label)
{
struct gpio_desc *desc = gpiochip_get_desc(chip, hwnum);
int err;
if (IS_ERR(desc)) {
chip_err(chip, "failed to get GPIO descriptor\n");
return desc;
}
err = gpiod_request_commit(desc, label);
if (err < 0)
return ERR_PTR(err);
return desc;
}
EXPORT_SYMBOL_GPL(gpiochip_request_own_desc);
/**
* gpiochip_free_own_desc - Free GPIO requested by the chip driver
* @desc: GPIO descriptor to free
*
* Function frees the given GPIO requested previously with
* gpiochip_request_own_desc().
*/
void gpiochip_free_own_desc(struct gpio_desc *desc)
{
if (desc)
gpiod_free_commit(desc);
}
EXPORT_SYMBOL_GPL(gpiochip_free_own_desc);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
/*
* Drivers MUST set GPIO direction before making get/set calls. In
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
* some cases this is done in early boot, before IRQs are enabled.
*
* As a rule these aren't called more than once (except for drivers
* using the open-drain emulation idiom) so these are natural places
* to accumulate extra debugging checks. Note that we can't (yet)
* rely on gpio_request() having been called beforehand.
*/
/**
* gpiod_direction_input - set the GPIO direction to input
* @desc: GPIO to set to input
*
* Set the direction of the passed GPIO to input, such as gpiod_get_value() can
* be called safely on it.
*
* Return 0 in case of success, else an error code.
*/
int gpiod_direction_input(struct gpio_desc *desc)
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
{
struct gpio_chip *chip;
int status = -EINVAL;
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
VALIDATE_DESC(desc);
chip = desc->gdev->chip;
if (!chip->get || !chip->direction_input) {
gpiod_warn(desc,
"%s: missing get() or direction_input() operations\n",
__func__);
return -EIO;
}
status = chip->direction_input(chip, gpio_chip_hwgpio(desc));
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
if (status == 0)
clear_bit(FLAG_IS_OUT, &desc->flags);
trace_gpio_direction(desc_to_gpio(desc), 1, status);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
return status;
}
EXPORT_SYMBOL_GPL(gpiod_direction_input);
static int gpio_set_drive_single_ended(struct gpio_chip *gc, unsigned offset,
enum pin_config_param mode)
{
unsigned long config = { PIN_CONF_PACKED(mode, 0) };
return gc->set_config ? gc->set_config(gc, offset, config) : -ENOTSUPP;
}
static int gpiod_direction_output_raw_commit(struct gpio_desc *desc, int value)
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
{
struct gpio_chip *gc = desc->gdev->chip;
int val = !!value;
int ret;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
if (!gc->set || !gc->direction_output) {
gpiod_warn(desc,
"%s: missing set() or direction_output() operations\n",
__func__);
return -EIO;
}
ret = gc->direction_output(gc, gpio_chip_hwgpio(desc), val);
if (!ret)
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
set_bit(FLAG_IS_OUT, &desc->flags);
trace_gpio_value(desc_to_gpio(desc), 0, val);
trace_gpio_direction(desc_to_gpio(desc), 0, ret);
return ret;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
}
/**
* gpiod_direction_output_raw - set the GPIO direction to output
* @desc: GPIO to set to output
* @value: initial output value of the GPIO
*
* Set the direction of the passed GPIO to output, such as gpiod_set_value() can
* be called safely on it. The initial value of the output must be specified
* as raw value on the physical line without regard for the ACTIVE_LOW status.
*
* Return 0 in case of success, else an error code.
*/
int gpiod_direction_output_raw(struct gpio_desc *desc, int value)
{
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
VALIDATE_DESC(desc);
return gpiod_direction_output_raw_commit(desc, value);
}
EXPORT_SYMBOL_GPL(gpiod_direction_output_raw);
/**
* gpiod_direction_output - set the GPIO direction to output
* @desc: GPIO to set to output
* @value: initial output value of the GPIO
*
* Set the direction of the passed GPIO to output, such as gpiod_set_value() can
* be called safely on it. The initial value of the output must be specified
* as the logical value of the GPIO, i.e. taking its ACTIVE_LOW status into
* account.
*
* Return 0 in case of success, else an error code.
*/
int gpiod_direction_output(struct gpio_desc *desc, int value)
{
struct gpio_chip *gc = desc->gdev->chip;
int ret;
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
VALIDATE_DESC(desc);
if (test_bit(FLAG_ACTIVE_LOW, &desc->flags))
value = !value;
else
value = !!value;
/* GPIOs used for IRQs shall not be set as output */
if (test_bit(FLAG_USED_AS_IRQ, &desc->flags)) {
gpiod_err(desc,
"%s: tried to set a GPIO tied to an IRQ as output\n",
__func__);
return -EIO;
}
if (test_bit(FLAG_OPEN_DRAIN, &desc->flags)) {
/* First see if we can enable open drain in hardware */
ret = gpio_set_drive_single_ended(gc, gpio_chip_hwgpio(desc),
PIN_CONFIG_DRIVE_OPEN_DRAIN);
if (!ret)
goto set_output_value;
/* Emulate open drain by not actively driving the line high */
if (value)
return gpiod_direction_input(desc);
}
else if (test_bit(FLAG_OPEN_SOURCE, &desc->flags)) {
ret = gpio_set_drive_single_ended(gc, gpio_chip_hwgpio(desc),
PIN_CONFIG_DRIVE_OPEN_SOURCE);
if (!ret)
goto set_output_value;
/* Emulate open source by not actively driving the line low */
if (!value)
return gpiod_direction_input(desc);
} else {
gpio_set_drive_single_ended(gc, gpio_chip_hwgpio(desc),
PIN_CONFIG_DRIVE_PUSH_PULL);
}
set_output_value:
return gpiod_direction_output_raw_commit(desc, value);
}
EXPORT_SYMBOL_GPL(gpiod_direction_output);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
2010-05-27 05:42:23 +08:00
/**
* gpiod_set_debounce - sets @debounce time for a GPIO
* @desc: descriptor of the GPIO for which to set debounce time
* @debounce: debounce time in microseconds
*
* Returns:
* 0 on success, %-ENOTSUPP if the controller doesn't support setting the
* debounce time.
2010-05-27 05:42:23 +08:00
*/
int gpiod_set_debounce(struct gpio_desc *desc, unsigned debounce)
2010-05-27 05:42:23 +08:00
{
struct gpio_chip *chip;
unsigned long config;
2010-05-27 05:42:23 +08:00
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
VALIDATE_DESC(desc);
chip = desc->gdev->chip;
if (!chip->set || !chip->set_config) {
gpiod_dbg(desc,
"%s: missing set() or set_config() operations\n",
__func__);
return -ENOTSUPP;
}
config = pinconf_to_config_packed(PIN_CONFIG_INPUT_DEBOUNCE, debounce);
return chip->set_config(chip, gpio_chip_hwgpio(desc), config);
2010-05-27 05:42:23 +08:00
}
EXPORT_SYMBOL_GPL(gpiod_set_debounce);
/**
* gpiod_set_transitory - Lose or retain GPIO state on suspend or reset
* @desc: descriptor of the GPIO for which to configure persistence
* @transitory: True to lose state on suspend or reset, false for persistence
*
* Returns:
* 0 on success, otherwise a negative error code.
*/
int gpiod_set_transitory(struct gpio_desc *desc, bool transitory)
{
struct gpio_chip *chip;
unsigned long packed;
int gpio;
int rc;
/*
* Handle FLAG_TRANSITORY first, enabling queries to gpiolib for
* persistence state.
*/
if (transitory)
set_bit(FLAG_TRANSITORY, &desc->flags);
else
clear_bit(FLAG_TRANSITORY, &desc->flags);
/* If the driver supports it, set the persistence state now */
chip = desc->gdev->chip;
if (!chip->set_config)
return 0;
packed = pinconf_to_config_packed(PIN_CONFIG_PERSIST_STATE,
!transitory);
gpio = gpio_chip_hwgpio(desc);
rc = chip->set_config(chip, gpio, packed);
if (rc == -ENOTSUPP) {
dev_dbg(&desc->gdev->dev, "Persistence not supported for GPIO %d\n",
gpio);
return 0;
}
return rc;
}
EXPORT_SYMBOL_GPL(gpiod_set_transitory);
/**
* gpiod_is_active_low - test whether a GPIO is active-low or not
* @desc: the gpio descriptor to test
*
* Returns 1 if the GPIO is active-low, 0 otherwise.
*/
int gpiod_is_active_low(const struct gpio_desc *desc)
{
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
VALIDATE_DESC(desc);
return test_bit(FLAG_ACTIVE_LOW, &desc->flags);
}
EXPORT_SYMBOL_GPL(gpiod_is_active_low);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/* I/O calls are only valid after configuration completed; the relevant
* "is this a valid GPIO" error checks should already have been done.
*
* "Get" operations are often inlinable as reading a pin value register,
* and masking the relevant bit in that register.
*
* When "set" operations are inlinable, they involve writing that mask to
* one register to set a low value, or a different register to set it high.
* Otherwise locking is needed, so there may be little value to inlining.
*
*------------------------------------------------------------------------
*
* IMPORTANT!!! The hot paths -- get/set value -- assume that callers
* have requested the GPIO. That can include implicit requesting by
* a direction setting call. Marking a gpio as requested locks its chip
* in memory, guaranteeing that these table lookups need no more locking
* and that gpiochip_remove() will fail.
*
* REVISIT when debugging, consider adding some instrumentation to ensure
* that the GPIO was actually requested.
*/
static int gpiod_get_raw_value_commit(const struct gpio_desc *desc)
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
{
struct gpio_chip *chip;
int offset;
int value;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
chip = desc->gdev->chip;
offset = gpio_chip_hwgpio(desc);
value = chip->get ? chip->get(chip, offset) : -EIO;
value = value < 0 ? value : !!value;
trace_gpio_value(desc_to_gpio(desc), 1, value);
return value;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
}
static int gpio_chip_get_multiple(struct gpio_chip *chip,
unsigned long *mask, unsigned long *bits)
{
if (chip->get_multiple) {
return chip->get_multiple(chip, mask, bits);
} else if (chip->get) {
int i, value;
for_each_set_bit(i, mask, chip->ngpio) {
value = chip->get(chip, i);
if (value < 0)
return value;
__assign_bit(i, bits, value);
}
return 0;
}
return -EIO;
}
int gpiod_get_array_value_complex(bool raw, bool can_sleep,
unsigned int array_size,
struct gpio_desc **desc_array,
int *value_array)
{
int i = 0;
while (i < array_size) {
struct gpio_chip *chip = desc_array[i]->gdev->chip;
unsigned long mask[BITS_TO_LONGS(chip->ngpio)];
unsigned long bits[BITS_TO_LONGS(chip->ngpio)];
int first, j, ret;
if (!can_sleep)
WARN_ON(chip->can_sleep);
/* collect all inputs belonging to the same chip */
first = i;
memset(mask, 0, sizeof(mask));
do {
const struct gpio_desc *desc = desc_array[i];
int hwgpio = gpio_chip_hwgpio(desc);
__set_bit(hwgpio, mask);
i++;
} while ((i < array_size) &&
(desc_array[i]->gdev->chip == chip));
ret = gpio_chip_get_multiple(chip, mask, bits);
if (ret)
return ret;
for (j = first; j < i; j++) {
const struct gpio_desc *desc = desc_array[j];
int hwgpio = gpio_chip_hwgpio(desc);
int value = test_bit(hwgpio, bits);
if (!raw && test_bit(FLAG_ACTIVE_LOW, &desc->flags))
value = !value;
value_array[j] = value;
trace_gpio_value(desc_to_gpio(desc), 1, value);
}
}
return 0;
}
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/**
* gpiod_get_raw_value() - return a gpio's raw value
* @desc: gpio whose value will be returned
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
*
* Return the GPIO's raw value, i.e. the value of the physical line disregarding
* its ACTIVE_LOW status, or negative errno on failure.
*
* This function should be called from contexts where we cannot sleep, and will
* complain if the GPIO chip functions potentially sleep.
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
*/
int gpiod_get_raw_value(const struct gpio_desc *desc)
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
{
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
VALIDATE_DESC(desc);
/* Should be using gpio_get_value_cansleep() */
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
WARN_ON(desc->gdev->chip->can_sleep);
return gpiod_get_raw_value_commit(desc);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
}
EXPORT_SYMBOL_GPL(gpiod_get_raw_value);
/**
* gpiod_get_value() - return a gpio's value
* @desc: gpio whose value will be returned
*
* Return the GPIO's logical value, i.e. taking the ACTIVE_LOW status into
* account, or negative errno on failure.
*
* This function should be called from contexts where we cannot sleep, and will
* complain if the GPIO chip functions potentially sleep.
*/
int gpiod_get_value(const struct gpio_desc *desc)
{
int value;
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
VALIDATE_DESC(desc);
/* Should be using gpio_get_value_cansleep() */
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
WARN_ON(desc->gdev->chip->can_sleep);
value = gpiod_get_raw_value_commit(desc);
if (value < 0)
return value;
if (test_bit(FLAG_ACTIVE_LOW, &desc->flags))
value = !value;
return value;
}
EXPORT_SYMBOL_GPL(gpiod_get_value);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/**
* gpiod_get_raw_array_value() - read raw values from an array of GPIOs
* @array_size: number of elements in the descriptor / value arrays
* @desc_array: array of GPIO descriptors whose values will be read
* @value_array: array to store the read values
*
* Read the raw values of the GPIOs, i.e. the values of the physical lines
* without regard for their ACTIVE_LOW status. Return 0 in case of success,
* else an error code.
*
* This function should be called from contexts where we cannot sleep,
* and it will complain if the GPIO chip functions potentially sleep.
*/
int gpiod_get_raw_array_value(unsigned int array_size,
struct gpio_desc **desc_array, int *value_array)
{
if (!desc_array)
return -EINVAL;
return gpiod_get_array_value_complex(true, false, array_size,
desc_array, value_array);
}
EXPORT_SYMBOL_GPL(gpiod_get_raw_array_value);
/**
* gpiod_get_array_value() - read values from an array of GPIOs
* @array_size: number of elements in the descriptor / value arrays
* @desc_array: array of GPIO descriptors whose values will be read
* @value_array: array to store the read values
*
* Read the logical values of the GPIOs, i.e. taking their ACTIVE_LOW status
* into account. Return 0 in case of success, else an error code.
*
* This function should be called from contexts where we cannot sleep,
* and it will complain if the GPIO chip functions potentially sleep.
*/
int gpiod_get_array_value(unsigned int array_size,
struct gpio_desc **desc_array, int *value_array)
{
if (!desc_array)
return -EINVAL;
return gpiod_get_array_value_complex(false, false, array_size,
desc_array, value_array);
}
EXPORT_SYMBOL_GPL(gpiod_get_array_value);
/*
* gpio_set_open_drain_value_commit() - Set the open drain gpio's value.
* @desc: gpio descriptor whose state need to be set.
* @value: Non-zero for setting it HIGH otherwise it will set to LOW.
*/
static void gpio_set_open_drain_value_commit(struct gpio_desc *desc, bool value)
{
int err = 0;
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
struct gpio_chip *chip = desc->gdev->chip;
int offset = gpio_chip_hwgpio(desc);
if (value) {
err = chip->direction_input(chip, offset);
if (!err)
clear_bit(FLAG_IS_OUT, &desc->flags);
} else {
err = chip->direction_output(chip, offset, 0);
if (!err)
set_bit(FLAG_IS_OUT, &desc->flags);
}
trace_gpio_direction(desc_to_gpio(desc), value, err);
if (err < 0)
gpiod_err(desc,
"%s: Error in set_value for open drain err %d\n",
__func__, err);
}
/*
* _gpio_set_open_source_value() - Set the open source gpio's value.
* @desc: gpio descriptor whose state need to be set.
* @value: Non-zero for setting it HIGH otherwise it will set to LOW.
*/
static void gpio_set_open_source_value_commit(struct gpio_desc *desc, bool value)
{
int err = 0;
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
struct gpio_chip *chip = desc->gdev->chip;
int offset = gpio_chip_hwgpio(desc);
if (value) {
err = chip->direction_output(chip, offset, 1);
if (!err)
set_bit(FLAG_IS_OUT, &desc->flags);
} else {
err = chip->direction_input(chip, offset);
if (!err)
clear_bit(FLAG_IS_OUT, &desc->flags);
}
trace_gpio_direction(desc_to_gpio(desc), !value, err);
if (err < 0)
gpiod_err(desc,
"%s: Error in set_value for open source err %d\n",
__func__, err);
}
static void gpiod_set_raw_value_commit(struct gpio_desc *desc, bool value)
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
{
struct gpio_chip *chip;
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
chip = desc->gdev->chip;
trace_gpio_value(desc_to_gpio(desc), 0, value);
chip->set(chip, gpio_chip_hwgpio(desc), value);
}
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
/*
* set multiple outputs on the same chip;
* use the chip's set_multiple function if available;
* otherwise set the outputs sequentially;
* @mask: bit mask array; one bit per output; BITS_PER_LONG bits per word
* defines which outputs are to be changed
* @bits: bit value array; one bit per output; BITS_PER_LONG bits per word
* defines the values the outputs specified by mask are to be set to
*/
static void gpio_chip_set_multiple(struct gpio_chip *chip,
unsigned long *mask, unsigned long *bits)
{
if (chip->set_multiple) {
chip->set_multiple(chip, mask, bits);
} else {
unsigned int i;
/* set outputs if the corresponding mask bit is set */
for_each_set_bit(i, mask, chip->ngpio)
chip->set(chip, i, test_bit(i, bits));
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
}
}
void gpiod_set_array_value_complex(bool raw, bool can_sleep,
unsigned int array_size,
struct gpio_desc **desc_array,
int *value_array)
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
{
int i = 0;
while (i < array_size) {
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
struct gpio_chip *chip = desc_array[i]->gdev->chip;
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
unsigned long mask[BITS_TO_LONGS(chip->ngpio)];
unsigned long bits[BITS_TO_LONGS(chip->ngpio)];
int count = 0;
if (!can_sleep)
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
WARN_ON(chip->can_sleep);
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
memset(mask, 0, sizeof(mask));
do {
struct gpio_desc *desc = desc_array[i];
int hwgpio = gpio_chip_hwgpio(desc);
int value = value_array[i];
if (!raw && test_bit(FLAG_ACTIVE_LOW, &desc->flags))
value = !value;
trace_gpio_value(desc_to_gpio(desc), 0, value);
/*
* collect all normal outputs belonging to the same chip
* open drain and open source outputs are set individually
*/
if (test_bit(FLAG_OPEN_DRAIN, &desc->flags) && !raw) {
gpio_set_open_drain_value_commit(desc, value);
} else if (test_bit(FLAG_OPEN_SOURCE, &desc->flags) && !raw) {
gpio_set_open_source_value_commit(desc, value);
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
} else {
__set_bit(hwgpio, mask);
if (value)
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
__set_bit(hwgpio, bits);
else
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
__clear_bit(hwgpio, bits);
count++;
}
i++;
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
} while ((i < array_size) &&
(desc_array[i]->gdev->chip == chip));
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
/* push collected bits to outputs */
if (count != 0)
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
gpio_chip_set_multiple(chip, mask, bits);
}
}
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/**
* gpiod_set_raw_value() - assign a gpio's raw value
* @desc: gpio whose value will be assigned
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
* @value: value to assign
*
* Set the raw value of the GPIO, i.e. the value of its physical line without
* regard for its ACTIVE_LOW status.
*
* This function should be called from contexts where we cannot sleep, and will
* complain if the GPIO chip functions potentially sleep.
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
*/
void gpiod_set_raw_value(struct gpio_desc *desc, int value)
{
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
VALIDATE_DESC_VOID(desc);
/* Should be using gpiod_set_value_cansleep() */
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
WARN_ON(desc->gdev->chip->can_sleep);
gpiod_set_raw_value_commit(desc, value);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
}
EXPORT_SYMBOL_GPL(gpiod_set_raw_value);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/**
* gpiod_set_value() - assign a gpio's value
* @desc: gpio whose value will be assigned
* @value: value to assign
*
* Set the logical value of the GPIO, i.e. taking its ACTIVE_LOW,
* OPEN_DRAIN and OPEN_SOURCE flags into account.
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
*
* This function should be called from contexts where we cannot sleep, and will
* complain if the GPIO chip functions potentially sleep.
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
*/
void gpiod_set_value(struct gpio_desc *desc, int value)
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
{
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
VALIDATE_DESC_VOID(desc);
/* Should be using gpiod_set_value_cansleep() */
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
WARN_ON(desc->gdev->chip->can_sleep);
if (test_bit(FLAG_ACTIVE_LOW, &desc->flags))
value = !value;
if (test_bit(FLAG_OPEN_DRAIN, &desc->flags))
gpio_set_open_drain_value_commit(desc, value);
else if (test_bit(FLAG_OPEN_SOURCE, &desc->flags))
gpio_set_open_source_value_commit(desc, value);
else
gpiod_set_raw_value_commit(desc, value);
}
EXPORT_SYMBOL_GPL(gpiod_set_value);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
/**
* gpiod_set_raw_array_value() - assign values to an array of GPIOs
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
* @array_size: number of elements in the descriptor / value arrays
* @desc_array: array of GPIO descriptors whose values will be assigned
* @value_array: array of values to assign
*
* Set the raw values of the GPIOs, i.e. the values of the physical lines
* without regard for their ACTIVE_LOW status.
*
* This function should be called from contexts where we cannot sleep, and will
* complain if the GPIO chip functions potentially sleep.
*/
void gpiod_set_raw_array_value(unsigned int array_size,
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
struct gpio_desc **desc_array, int *value_array)
{
if (!desc_array)
return;
gpiod_set_array_value_complex(true, false, array_size, desc_array,
value_array);
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
}
EXPORT_SYMBOL_GPL(gpiod_set_raw_array_value);
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
/**
* gpiod_set_array_value() - assign values to an array of GPIOs
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
* @array_size: number of elements in the descriptor / value arrays
* @desc_array: array of GPIO descriptors whose values will be assigned
* @value_array: array of values to assign
*
* Set the logical values of the GPIOs, i.e. taking their ACTIVE_LOW status
* into account.
*
* This function should be called from contexts where we cannot sleep, and will
* complain if the GPIO chip functions potentially sleep.
*/
void gpiod_set_array_value(unsigned int array_size,
struct gpio_desc **desc_array, int *value_array)
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
{
if (!desc_array)
return;
gpiod_set_array_value_complex(false, false, array_size, desc_array,
value_array);
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
}
EXPORT_SYMBOL_GPL(gpiod_set_array_value);
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/**
* gpiod_cansleep() - report whether gpio value access may sleep
* @desc: gpio to check
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
*
*/
int gpiod_cansleep(const struct gpio_desc *desc)
{
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
VALIDATE_DESC(desc);
return desc->gdev->chip->can_sleep;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
}
EXPORT_SYMBOL_GPL(gpiod_cansleep);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/**
* gpiod_to_irq() - return the IRQ corresponding to a GPIO
* @desc: gpio whose IRQ will be returned (already requested)
*
* Return the IRQ corresponding to the passed GPIO, or an error code in case of
* error.
*/
int gpiod_to_irq(const struct gpio_desc *desc)
{
struct gpio_chip *chip;
int offset;
/*
* Cannot VALIDATE_DESC() here as gpiod_to_irq() consumer semantics
* requires this function to not return zero on an invalid descriptor
* but rather a negative error number.
*/
if (!desc || IS_ERR(desc) || !desc->gdev || !desc->gdev->chip)
return -EINVAL;
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
chip = desc->gdev->chip;
offset = gpio_chip_hwgpio(desc);
if (chip->to_irq) {
int retirq = chip->to_irq(chip, offset);
/* Zero means NO_IRQ */
if (!retirq)
return -ENXIO;
return retirq;
}
return -ENXIO;
}
EXPORT_SYMBOL_GPL(gpiod_to_irq);
/**
* gpiochip_lock_as_irq() - lock a GPIO to be used as IRQ
* @chip: the chip the GPIO to lock belongs to
* @offset: the offset of the GPIO to lock as IRQ
*
* This is used directly by GPIO drivers that want to lock down
* a certain GPIO line to be used for IRQs.
*/
int gpiochip_lock_as_irq(struct gpio_chip *chip, unsigned int offset)
{
struct gpio_desc *desc;
desc = gpiochip_get_desc(chip, offset);
if (IS_ERR(desc))
return PTR_ERR(desc);
/*
* If it's fast: flush the direction setting if something changed
* behind our back
*/
if (!chip->can_sleep && chip->get_direction) {
int dir = chip->get_direction(chip, offset);
if (dir)
clear_bit(FLAG_IS_OUT, &desc->flags);
else
set_bit(FLAG_IS_OUT, &desc->flags);
}
if (test_bit(FLAG_IS_OUT, &desc->flags)) {
chip_err(chip,
"%s: tried to flag a GPIO set as output for IRQ\n",
__func__);
return -EIO;
}
set_bit(FLAG_USED_AS_IRQ, &desc->flags);
/*
* If the consumer has not set up a label (such as when the
* IRQ is referenced from .to_irq()) we set up a label here
* so it is clear this is used as an interrupt.
*/
if (!desc->label)
desc_set_label(desc, "interrupt");
return 0;
}
EXPORT_SYMBOL_GPL(gpiochip_lock_as_irq);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/**
* gpiochip_unlock_as_irq() - unlock a GPIO used as IRQ
* @chip: the chip the GPIO to lock belongs to
* @offset: the offset of the GPIO to lock as IRQ
*
* This is used directly by GPIO drivers that want to indicate
* that a certain GPIO is no longer used exclusively for IRQ.
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
*/
void gpiochip_unlock_as_irq(struct gpio_chip *chip, unsigned int offset)
{
struct gpio_desc *desc;
desc = gpiochip_get_desc(chip, offset);
if (IS_ERR(desc))
return;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
clear_bit(FLAG_USED_AS_IRQ, &desc->flags);
/* If we only had this marking, erase it */
if (desc->label && !strcmp(desc->label, "interrupt"))
desc_set_label(desc, NULL);
}
EXPORT_SYMBOL_GPL(gpiochip_unlock_as_irq);
bool gpiochip_line_is_irq(struct gpio_chip *chip, unsigned int offset)
{
if (offset >= chip->ngpio)
return false;
return test_bit(FLAG_USED_AS_IRQ, &chip->gpiodev->descs[offset].flags);
}
EXPORT_SYMBOL_GPL(gpiochip_line_is_irq);
bool gpiochip_line_is_open_drain(struct gpio_chip *chip, unsigned int offset)
{
if (offset >= chip->ngpio)
return false;
return test_bit(FLAG_OPEN_DRAIN, &chip->gpiodev->descs[offset].flags);
}
EXPORT_SYMBOL_GPL(gpiochip_line_is_open_drain);
bool gpiochip_line_is_open_source(struct gpio_chip *chip, unsigned int offset)
{
if (offset >= chip->ngpio)
return false;
return test_bit(FLAG_OPEN_SOURCE, &chip->gpiodev->descs[offset].flags);
}
EXPORT_SYMBOL_GPL(gpiochip_line_is_open_source);
bool gpiochip_line_is_persistent(struct gpio_chip *chip, unsigned int offset)
{
if (offset >= chip->ngpio)
return false;
return !test_bit(FLAG_TRANSITORY, &chip->gpiodev->descs[offset].flags);
}
EXPORT_SYMBOL_GPL(gpiochip_line_is_persistent);
/**
* gpiod_get_raw_value_cansleep() - return a gpio's raw value
* @desc: gpio whose value will be returned
*
* Return the GPIO's raw value, i.e. the value of the physical line disregarding
* its ACTIVE_LOW status, or negative errno on failure.
*
* This function is to be called from contexts that can sleep.
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
*/
int gpiod_get_raw_value_cansleep(const struct gpio_desc *desc)
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
{
might_sleep_if(extra_checks);
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
VALIDATE_DESC(desc);
return gpiod_get_raw_value_commit(desc);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
}
EXPORT_SYMBOL_GPL(gpiod_get_raw_value_cansleep);
/**
* gpiod_get_value_cansleep() - return a gpio's value
* @desc: gpio whose value will be returned
*
* Return the GPIO's logical value, i.e. taking the ACTIVE_LOW status into
* account, or negative errno on failure.
*
* This function is to be called from contexts that can sleep.
*/
int gpiod_get_value_cansleep(const struct gpio_desc *desc)
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
{
int value;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
might_sleep_if(extra_checks);
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
VALIDATE_DESC(desc);
value = gpiod_get_raw_value_commit(desc);
if (value < 0)
return value;
if (test_bit(FLAG_ACTIVE_LOW, &desc->flags))
value = !value;
return value;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
}
EXPORT_SYMBOL_GPL(gpiod_get_value_cansleep);
/**
* gpiod_get_raw_array_value_cansleep() - read raw values from an array of GPIOs
* @array_size: number of elements in the descriptor / value arrays
* @desc_array: array of GPIO descriptors whose values will be read
* @value_array: array to store the read values
*
* Read the raw values of the GPIOs, i.e. the values of the physical lines
* without regard for their ACTIVE_LOW status. Return 0 in case of success,
* else an error code.
*
* This function is to be called from contexts that can sleep.
*/
int gpiod_get_raw_array_value_cansleep(unsigned int array_size,
struct gpio_desc **desc_array,
int *value_array)
{
might_sleep_if(extra_checks);
if (!desc_array)
return -EINVAL;
return gpiod_get_array_value_complex(true, true, array_size,
desc_array, value_array);
}
EXPORT_SYMBOL_GPL(gpiod_get_raw_array_value_cansleep);
/**
* gpiod_get_array_value_cansleep() - read values from an array of GPIOs
* @array_size: number of elements in the descriptor / value arrays
* @desc_array: array of GPIO descriptors whose values will be read
* @value_array: array to store the read values
*
* Read the logical values of the GPIOs, i.e. taking their ACTIVE_LOW status
* into account. Return 0 in case of success, else an error code.
*
* This function is to be called from contexts that can sleep.
*/
int gpiod_get_array_value_cansleep(unsigned int array_size,
struct gpio_desc **desc_array,
int *value_array)
{
might_sleep_if(extra_checks);
if (!desc_array)
return -EINVAL;
return gpiod_get_array_value_complex(false, true, array_size,
desc_array, value_array);
}
EXPORT_SYMBOL_GPL(gpiod_get_array_value_cansleep);
/**
* gpiod_set_raw_value_cansleep() - assign a gpio's raw value
* @desc: gpio whose value will be assigned
* @value: value to assign
*
* Set the raw value of the GPIO, i.e. the value of its physical line without
* regard for its ACTIVE_LOW status.
*
* This function is to be called from contexts that can sleep.
*/
void gpiod_set_raw_value_cansleep(struct gpio_desc *desc, int value)
{
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
might_sleep_if(extra_checks);
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
VALIDATE_DESC_VOID(desc);
gpiod_set_raw_value_commit(desc, value);
}
EXPORT_SYMBOL_GPL(gpiod_set_raw_value_cansleep);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/**
* gpiod_set_value_cansleep() - assign a gpio's value
* @desc: gpio whose value will be assigned
* @value: value to assign
*
* Set the logical value of the GPIO, i.e. taking its ACTIVE_LOW status into
* account
*
* This function is to be called from contexts that can sleep.
*/
void gpiod_set_value_cansleep(struct gpio_desc *desc, int value)
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
{
might_sleep_if(extra_checks);
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
VALIDATE_DESC_VOID(desc);
if (test_bit(FLAG_ACTIVE_LOW, &desc->flags))
value = !value;
gpiod_set_raw_value_commit(desc, value);
}
EXPORT_SYMBOL_GPL(gpiod_set_value_cansleep);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
/**
* gpiod_set_raw_array_value_cansleep() - assign values to an array of GPIOs
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
* @array_size: number of elements in the descriptor / value arrays
* @desc_array: array of GPIO descriptors whose values will be assigned
* @value_array: array of values to assign
*
* Set the raw values of the GPIOs, i.e. the values of the physical lines
* without regard for their ACTIVE_LOW status.
*
* This function is to be called from contexts that can sleep.
*/
void gpiod_set_raw_array_value_cansleep(unsigned int array_size,
struct gpio_desc **desc_array,
int *value_array)
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
{
might_sleep_if(extra_checks);
if (!desc_array)
return;
gpiod_set_array_value_complex(true, true, array_size, desc_array,
value_array);
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
}
EXPORT_SYMBOL_GPL(gpiod_set_raw_array_value_cansleep);
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
/**
* gpiod_add_lookup_tables() - register GPIO device consumers
* @tables: list of tables of consumers to register
* @n: number of tables in the list
*/
void gpiod_add_lookup_tables(struct gpiod_lookup_table **tables, size_t n)
{
unsigned int i;
mutex_lock(&gpio_lookup_lock);
for (i = 0; i < n; i++)
list_add_tail(&tables[i]->list, &gpio_lookup_list);
mutex_unlock(&gpio_lookup_lock);
}
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
/**
* gpiod_set_array_value_cansleep() - assign values to an array of GPIOs
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
* @array_size: number of elements in the descriptor / value arrays
* @desc_array: array of GPIO descriptors whose values will be assigned
* @value_array: array of values to assign
*
* Set the logical values of the GPIOs, i.e. taking their ACTIVE_LOW status
* into account.
*
* This function is to be called from contexts that can sleep.
*/
void gpiod_set_array_value_cansleep(unsigned int array_size,
struct gpio_desc **desc_array,
int *value_array)
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
{
might_sleep_if(extra_checks);
if (!desc_array)
return;
gpiod_set_array_value_complex(false, true, array_size, desc_array,
value_array);
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
}
EXPORT_SYMBOL_GPL(gpiod_set_array_value_cansleep);
gpiolib: allow simultaneous setting of multiple GPIO outputs Introduce new functions gpiod_set_array & gpiod_set_raw_array to the consumer interface which allow setting multiple outputs with just one function call. Also add an optional set_multiple function to the driver interface. Without an implementation of that function in the chip driver outputs are set sequentially. Implementing the set_multiple function in a chip driver allows for: - Improved performance for certain use cases. The original motivation for this was the task of configuring an FPGA. In that specific case, where 9 GPIO lines have to be set many times, configuration time goes down from 48 s to 20 s when using the new function. - Simultaneous glitch-free setting of multiple pins on any kind of parallel bus attached to GPIOs provided they all reside on the same chip and bank. Limitations: Performance is only improved for normal high-low outputs. Open drain and open source outputs are always set separately from each other. Those kinds of outputs could probably be accelerated in a similar way if we could forgo the error checking when setting GPIO directions. Change log: v6: - rebase on current linux-gpio devel branch v5: - check can_sleep property per chip - remove superfluous checks - supplement documentation v4: - add gpiod_set_array function for setting logical values - change interface of the set_multiple driver function to use unsigned long as type for the bit fields - use generic bitops (which also use unsigned long for bit fields) - do not use ARCH_NR_GPIOS any more v3: - add documentation - change commit message v2: - use descriptor interface - allow arbitrary groups of GPIOs spanning multiple chips Signed-off-by: Rojhalat Ibrahim <imr@rtschenk.de> Reviewed-by: Alexandre Courbot <acourbot@nvidia.com> Reviewed-by: Mark Brown <broonie@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2014-11-05 00:12:06 +08:00
/**
* gpiod_add_lookup_table() - register GPIO device consumers
* @table: table of consumers to register
*/
void gpiod_add_lookup_table(struct gpiod_lookup_table *table)
{
mutex_lock(&gpio_lookup_lock);
list_add_tail(&table->list, &gpio_lookup_list);
mutex_unlock(&gpio_lookup_lock);
}
EXPORT_SYMBOL_GPL(gpiod_add_lookup_table);
/**
* gpiod_remove_lookup_table() - unregister GPIO device consumers
* @table: table of consumers to unregister
*/
void gpiod_remove_lookup_table(struct gpiod_lookup_table *table)
{
mutex_lock(&gpio_lookup_lock);
list_del(&table->list);
mutex_unlock(&gpio_lookup_lock);
}
EXPORT_SYMBOL_GPL(gpiod_remove_lookup_table);
static struct gpiod_lookup_table *gpiod_find_lookup_table(struct device *dev)
{
const char *dev_id = dev ? dev_name(dev) : NULL;
struct gpiod_lookup_table *table;
mutex_lock(&gpio_lookup_lock);
list_for_each_entry(table, &gpio_lookup_list, list) {
if (table->dev_id && dev_id) {
/*
* Valid strings on both ends, must be identical to have
* a match
*/
if (!strcmp(table->dev_id, dev_id))
goto found;
} else {
/*
* One of the pointers is NULL, so both must be to have
* a match
*/
if (dev_id == table->dev_id)
goto found;
}
}
table = NULL;
found:
mutex_unlock(&gpio_lookup_lock);
return table;
}
static struct gpio_desc *gpiod_find(struct device *dev, const char *con_id,
unsigned int idx,
enum gpio_lookup_flags *flags)
{
struct gpio_desc *desc = ERR_PTR(-ENOENT);
struct gpiod_lookup_table *table;
struct gpiod_lookup *p;
table = gpiod_find_lookup_table(dev);
if (!table)
return desc;
for (p = &table->table[0]; p->chip_label; p++) {
struct gpio_chip *chip;
/* idx must always match exactly */
if (p->idx != idx)
continue;
/* If the lookup entry has a con_id, require exact match */
if (p->con_id && (!con_id || strcmp(p->con_id, con_id)))
continue;
chip = find_chip_by_name(p->chip_label);
if (!chip) {
dev_err(dev, "cannot find GPIO chip %s\n",
p->chip_label);
return ERR_PTR(-ENODEV);
}
if (chip->ngpio <= p->chip_hwnum) {
dev_err(dev,
"requested GPIO %d is out of range [0..%d] for chip %s\n",
idx, chip->ngpio, chip->label);
return ERR_PTR(-EINVAL);
}
desc = gpiochip_get_desc(chip, p->chip_hwnum);
*flags = p->flags;
return desc;
}
return desc;
}
static int dt_gpio_count(struct device *dev, const char *con_id)
{
int ret;
char propname[32];
unsigned int i;
for (i = 0; i < ARRAY_SIZE(gpio_suffixes); i++) {
if (con_id)
snprintf(propname, sizeof(propname), "%s-%s",
con_id, gpio_suffixes[i]);
else
snprintf(propname, sizeof(propname), "%s",
gpio_suffixes[i]);
ret = of_gpio_named_count(dev->of_node, propname);
if (ret > 0)
break;
}
return ret ? ret : -ENOENT;
}
static int platform_gpio_count(struct device *dev, const char *con_id)
{
struct gpiod_lookup_table *table;
struct gpiod_lookup *p;
unsigned int count = 0;
table = gpiod_find_lookup_table(dev);
if (!table)
return -ENOENT;
for (p = &table->table[0]; p->chip_label; p++) {
if ((con_id && p->con_id && !strcmp(con_id, p->con_id)) ||
(!con_id && !p->con_id))
count++;
}
if (!count)
return -ENOENT;
return count;
}
/**
* gpiod_count - return the number of GPIOs associated with a device / function
* or -ENOENT if no GPIO has been assigned to the requested function
* @dev: GPIO consumer, can be NULL for system-global GPIOs
* @con_id: function within the GPIO consumer
*/
int gpiod_count(struct device *dev, const char *con_id)
{
int count = -ENOENT;
if (IS_ENABLED(CONFIG_OF) && dev && dev->of_node)
count = dt_gpio_count(dev, con_id);
else if (IS_ENABLED(CONFIG_ACPI) && dev && ACPI_HANDLE(dev))
count = acpi_gpio_count(dev, con_id);
if (count < 0)
count = platform_gpio_count(dev, con_id);
return count;
}
EXPORT_SYMBOL_GPL(gpiod_count);
/**
* gpiod_get - obtain a GPIO for a given GPIO function
* @dev: GPIO consumer, can be NULL for system-global GPIOs
* @con_id: function within the GPIO consumer
* @flags: optional GPIO initialization flags
*
* Return the GPIO descriptor corresponding to the function con_id of device
* dev, -ENOENT if no GPIO has been assigned to the requested function, or
* another IS_ERR() code if an error occurred while trying to acquire the GPIO.
*/
struct gpio_desc *__must_check gpiod_get(struct device *dev, const char *con_id,
enum gpiod_flags flags)
{
return gpiod_get_index(dev, con_id, 0, flags);
}
EXPORT_SYMBOL_GPL(gpiod_get);
/**
* gpiod_get_optional - obtain an optional GPIO for a given GPIO function
* @dev: GPIO consumer, can be NULL for system-global GPIOs
* @con_id: function within the GPIO consumer
* @flags: optional GPIO initialization flags
*
* This is equivalent to gpiod_get(), except that when no GPIO was assigned to
* the requested function it will return NULL. This is convenient for drivers
* that need to handle optional GPIOs.
*/
struct gpio_desc *__must_check gpiod_get_optional(struct device *dev,
const char *con_id,
enum gpiod_flags flags)
{
return gpiod_get_index_optional(dev, con_id, 0, flags);
}
EXPORT_SYMBOL_GPL(gpiod_get_optional);
/**
* gpiod_configure_flags - helper function to configure a given GPIO
* @desc: gpio whose value will be assigned
* @con_id: function within the GPIO consumer
* @lflags: gpio_lookup_flags - returned from of_find_gpio() or
* of_get_gpio_hog()
* @dflags: gpiod_flags - optional GPIO initialization flags
*
* Return 0 on success, -ENOENT if no GPIO has been assigned to the
* requested function and/or index, or another IS_ERR() code if an error
* occurred while trying to acquire the GPIO.
*/
int gpiod_configure_flags(struct gpio_desc *desc, const char *con_id,
unsigned long lflags, enum gpiod_flags dflags)
{
int status;
if (lflags & GPIO_ACTIVE_LOW)
set_bit(FLAG_ACTIVE_LOW, &desc->flags);
if (lflags & GPIO_OPEN_DRAIN)
set_bit(FLAG_OPEN_DRAIN, &desc->flags);
else if (dflags & GPIOD_FLAGS_BIT_OPEN_DRAIN) {
/*
* This enforces open drain mode from the consumer side.
* This is necessary for some busses like I2C, but the lookup
* should *REALLY* have specified them as open drain in the
* first place, so print a little warning here.
*/
set_bit(FLAG_OPEN_DRAIN, &desc->flags);
gpiod_warn(desc,
"enforced open drain please flag it properly in DT/ACPI DSDT/board file\n");
}
if (lflags & GPIO_OPEN_SOURCE)
set_bit(FLAG_OPEN_SOURCE, &desc->flags);
status = gpiod_set_transitory(desc, (lflags & GPIO_TRANSITORY));
if (status < 0)
return status;
/* No particular flag request, return here... */
if (!(dflags & GPIOD_FLAGS_BIT_DIR_SET)) {
pr_debug("no flags found for %s\n", con_id);
return 0;
}
/* Process flags */
if (dflags & GPIOD_FLAGS_BIT_DIR_OUT)
status = gpiod_direction_output(desc,
!!(dflags & GPIOD_FLAGS_BIT_DIR_VAL));
else
status = gpiod_direction_input(desc);
return status;
}
/**
* gpiod_get_index - obtain a GPIO from a multi-index GPIO function
* @dev: GPIO consumer, can be NULL for system-global GPIOs
* @con_id: function within the GPIO consumer
* @idx: index of the GPIO to obtain in the consumer
* @flags: optional GPIO initialization flags
*
* This variant of gpiod_get() allows to access GPIOs other than the first
* defined one for functions that define several GPIOs.
*
* Return a valid GPIO descriptor, -ENOENT if no GPIO has been assigned to the
* requested function and/or index, or another IS_ERR() code if an error
* occurred while trying to acquire the GPIO.
*/
struct gpio_desc *__must_check gpiod_get_index(struct device *dev,
const char *con_id,
unsigned int idx,
enum gpiod_flags flags)
{
struct gpio_desc *desc = NULL;
int status;
enum gpio_lookup_flags lookupflags = 0;
dev_dbg(dev, "GPIO lookup for consumer %s\n", con_id);
if (dev) {
/* Using device tree? */
if (IS_ENABLED(CONFIG_OF) && dev->of_node) {
dev_dbg(dev, "using device tree for GPIO lookup\n");
desc = of_find_gpio(dev, con_id, idx, &lookupflags);
} else if (ACPI_COMPANION(dev)) {
dev_dbg(dev, "using ACPI for GPIO lookup\n");
desc = acpi_find_gpio(dev, con_id, idx, &flags, &lookupflags);
}
}
/*
* Either we are not using DT or ACPI, or their lookup did not return
* a result. In that case, use platform lookup as a fallback.
*/
if (!desc || desc == ERR_PTR(-ENOENT)) {
dev_dbg(dev, "using lookup tables for GPIO lookup\n");
desc = gpiod_find(dev, con_id, idx, &lookupflags);
}
if (IS_ERR(desc)) {
dev_dbg(dev, "lookup for GPIO %s failed\n", con_id);
return desc;
}
status = gpiod_request(desc, con_id);
if (status < 0)
return ERR_PTR(status);
status = gpiod_configure_flags(desc, con_id, lookupflags, flags);
if (status < 0) {
dev_dbg(dev, "setup of GPIO %s failed\n", con_id);
gpiod_put(desc);
return ERR_PTR(status);
}
return desc;
}
EXPORT_SYMBOL_GPL(gpiod_get_index);
/**
* fwnode_get_named_gpiod - obtain a GPIO from firmware node
* @fwnode: handle of the firmware node
* @propname: name of the firmware property representing the GPIO
* @index: index of the GPIO to obtain in the consumer
* @dflags: GPIO initialization flags
* @label: label to attach to the requested GPIO
*
* This function can be used for drivers that get their configuration
* from firmware.
*
* Function properly finds the corresponding GPIO using whatever is the
* underlying firmware interface and then makes sure that the GPIO
* descriptor is requested before it is returned to the caller.
*
* Returns:
* On successful request the GPIO pin is configured in accordance with
* provided @dflags.
*
* In case of error an ERR_PTR() is returned.
*/
struct gpio_desc *fwnode_get_named_gpiod(struct fwnode_handle *fwnode,
const char *propname, int index,
enum gpiod_flags dflags,
const char *label)
{
struct gpio_desc *desc = ERR_PTR(-ENODEV);
unsigned long lflags = 0;
bool active_low = false;
bool single_ended = false;
bool open_drain = false;
bool transitory = false;
int ret;
if (!fwnode)
return ERR_PTR(-EINVAL);
if (is_of_node(fwnode)) {
enum of_gpio_flags flags;
desc = of_get_named_gpiod_flags(to_of_node(fwnode), propname,
index, &flags);
if (!IS_ERR(desc)) {
active_low = flags & OF_GPIO_ACTIVE_LOW;
single_ended = flags & OF_GPIO_SINGLE_ENDED;
open_drain = flags & OF_GPIO_OPEN_DRAIN;
transitory = flags & OF_GPIO_TRANSITORY;
}
} else if (is_acpi_node(fwnode)) {
struct acpi_gpio_info info;
desc = acpi_node_get_gpiod(fwnode, propname, index, &info);
if (!IS_ERR(desc)) {
active_low = info.polarity == GPIO_ACTIVE_LOW;
acpi_gpio_update_gpiod_flags(&dflags, &info);
}
}
if (IS_ERR(desc))
return desc;
ret = gpiod_request(desc, label);
if (ret)
return ERR_PTR(ret);
if (active_low)
lflags |= GPIO_ACTIVE_LOW;
if (single_ended) {
if (open_drain)
lflags |= GPIO_OPEN_DRAIN;
else
lflags |= GPIO_OPEN_SOURCE;
}
if (transitory)
lflags |= GPIO_TRANSITORY;
ret = gpiod_configure_flags(desc, propname, lflags, dflags);
if (ret < 0) {
gpiod_put(desc);
return ERR_PTR(ret);
}
return desc;
}
EXPORT_SYMBOL_GPL(fwnode_get_named_gpiod);
/**
* gpiod_get_index_optional - obtain an optional GPIO from a multi-index GPIO
* function
* @dev: GPIO consumer, can be NULL for system-global GPIOs
* @con_id: function within the GPIO consumer
* @index: index of the GPIO to obtain in the consumer
* @flags: optional GPIO initialization flags
*
* This is equivalent to gpiod_get_index(), except that when no GPIO with the
* specified index was assigned to the requested function it will return NULL.
* This is convenient for drivers that need to handle optional GPIOs.
*/
struct gpio_desc *__must_check gpiod_get_index_optional(struct device *dev,
const char *con_id,
unsigned int index,
enum gpiod_flags flags)
{
struct gpio_desc *desc;
desc = gpiod_get_index(dev, con_id, index, flags);
if (IS_ERR(desc)) {
if (PTR_ERR(desc) == -ENOENT)
return NULL;
}
return desc;
}
EXPORT_SYMBOL_GPL(gpiod_get_index_optional);
/**
* gpiod_hog - Hog the specified GPIO desc given the provided flags
* @desc: gpio whose value will be assigned
* @name: gpio line name
* @lflags: gpio_lookup_flags - returned from of_find_gpio() or
* of_get_gpio_hog()
* @dflags: gpiod_flags - optional GPIO initialization flags
*/
int gpiod_hog(struct gpio_desc *desc, const char *name,
unsigned long lflags, enum gpiod_flags dflags)
{
struct gpio_chip *chip;
struct gpio_desc *local_desc;
int hwnum;
int status;
chip = gpiod_to_chip(desc);
hwnum = gpio_chip_hwgpio(desc);
local_desc = gpiochip_request_own_desc(chip, hwnum, name);
if (IS_ERR(local_desc)) {
status = PTR_ERR(local_desc);
pr_err("requesting hog GPIO %s (chip %s, offset %d) failed, %d\n",
name, chip->label, hwnum, status);
return status;
}
status = gpiod_configure_flags(desc, name, lflags, dflags);
if (status < 0) {
pr_err("setup of hog GPIO %s (chip %s, offset %d) failed, %d\n",
name, chip->label, hwnum, status);
gpiochip_free_own_desc(desc);
return status;
}
/* Mark GPIO as hogged so it can be identified and removed later */
set_bit(FLAG_IS_HOGGED, &desc->flags);
pr_info("GPIO line %d (%s) hogged as %s%s\n",
desc_to_gpio(desc), name,
(dflags&GPIOD_FLAGS_BIT_DIR_OUT) ? "output" : "input",
(dflags&GPIOD_FLAGS_BIT_DIR_OUT) ?
(dflags&GPIOD_FLAGS_BIT_DIR_VAL) ? "/high" : "/low":"");
return 0;
}
/**
* gpiochip_free_hogs - Scan gpio-controller chip and release GPIO hog
* @chip: gpio chip to act on
*
* This is only used by of_gpiochip_remove to free hogged gpios
*/
static void gpiochip_free_hogs(struct gpio_chip *chip)
{
int id;
for (id = 0; id < chip->ngpio; id++) {
if (test_bit(FLAG_IS_HOGGED, &chip->gpiodev->descs[id].flags))
gpiochip_free_own_desc(&chip->gpiodev->descs[id]);
}
}
/**
* gpiod_get_array - obtain multiple GPIOs from a multi-index GPIO function
* @dev: GPIO consumer, can be NULL for system-global GPIOs
* @con_id: function within the GPIO consumer
* @flags: optional GPIO initialization flags
*
* This function acquires all the GPIOs defined under a given function.
*
* Return a struct gpio_descs containing an array of descriptors, -ENOENT if
* no GPIO has been assigned to the requested function, or another IS_ERR()
* code if an error occurred while trying to acquire the GPIOs.
*/
struct gpio_descs *__must_check gpiod_get_array(struct device *dev,
const char *con_id,
enum gpiod_flags flags)
{
struct gpio_desc *desc;
struct gpio_descs *descs;
int count;
count = gpiod_count(dev, con_id);
if (count < 0)
return ERR_PTR(count);
descs = kzalloc(sizeof(*descs) + sizeof(descs->desc[0]) * count,
GFP_KERNEL);
if (!descs)
return ERR_PTR(-ENOMEM);
for (descs->ndescs = 0; descs->ndescs < count; ) {
desc = gpiod_get_index(dev, con_id, descs->ndescs, flags);
if (IS_ERR(desc)) {
gpiod_put_array(descs);
return ERR_CAST(desc);
}
descs->desc[descs->ndescs] = desc;
descs->ndescs++;
}
return descs;
}
EXPORT_SYMBOL_GPL(gpiod_get_array);
/**
* gpiod_get_array_optional - obtain multiple GPIOs from a multi-index GPIO
* function
* @dev: GPIO consumer, can be NULL for system-global GPIOs
* @con_id: function within the GPIO consumer
* @flags: optional GPIO initialization flags
*
* This is equivalent to gpiod_get_array(), except that when no GPIO was
* assigned to the requested function it will return NULL.
*/
struct gpio_descs *__must_check gpiod_get_array_optional(struct device *dev,
const char *con_id,
enum gpiod_flags flags)
{
struct gpio_descs *descs;
descs = gpiod_get_array(dev, con_id, flags);
if (IS_ERR(descs) && (PTR_ERR(descs) == -ENOENT))
return NULL;
return descs;
}
EXPORT_SYMBOL_GPL(gpiod_get_array_optional);
/**
* gpiod_put - dispose of a GPIO descriptor
* @desc: GPIO descriptor to dispose of
*
* No descriptor can be used after gpiod_put() has been called on it.
*/
void gpiod_put(struct gpio_desc *desc)
{
gpiod_free(desc);
}
EXPORT_SYMBOL_GPL(gpiod_put);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
/**
* gpiod_put_array - dispose of multiple GPIO descriptors
* @descs: struct gpio_descs containing an array of descriptors
*/
void gpiod_put_array(struct gpio_descs *descs)
{
unsigned int i;
for (i = 0; i < descs->ndescs; i++)
gpiod_put(descs->desc[i]);
kfree(descs);
}
EXPORT_SYMBOL_GPL(gpiod_put_array);
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
static int __init gpiolib_dev_init(void)
{
int ret;
/* Register GPIO sysfs bus */
ret = bus_register(&gpio_bus_type);
if (ret < 0) {
pr_err("gpiolib: could not register GPIO bus type\n");
return ret;
}
ret = alloc_chrdev_region(&gpio_devt, 0, GPIO_DEV_MAX, "gpiochip");
if (ret < 0) {
pr_err("gpiolib: failed to allocate char dev region\n");
bus_unregister(&gpio_bus_type);
} else {
gpiolib_initialized = true;
gpiochip_setup_devs();
gpio: add a userspace chardev ABI for GPIOs A new chardev that is to be used for userspace GPIO access is added in this patch. It is intended to gradually replace the horribly broken sysfs ABI. Using a chardev has many upsides: - All operations are per-gpiochip, which is the actual device underlying the GPIOs, making us tie in to the kernel device model properly. - Hotpluggable GPIO controllers can come and go, as this kind of problem has been know to userspace for character devices since ages, and if a gpiochip handle is held in userspace we know we will break something, whereas the sysfs is stateless. - The one-value-per-file rule of sysfs is really hard to maintain when you want to twist more than one knob at a time, for example have in-kernel APIs to switch several GPIO lines at the same time, and this will be possible to do with a single ioctl() from userspace, saving a lot of context switching. We also need to add a new bus type for GPIO. This is necessary for example for userspace coldplug, where sysfs is traversed to find the boot-time device nodes and create the character devices in /dev. This new chardev ABI is *non* *optional* and can be counted on to be present in the future, emphasizing the preference of this ABI. The ABI only implements one single ioctl() to get the name and number of GPIO lines of a chip. Even this is debatable: see it as a minimal example for review. This ABI shall be ruthlessly reviewed and etched in stone. The old /sys/class/gpio is still optional to compile in, but will be deprecated. Unique device IDs are created using IDR, which is overkill and insanely scalable, but also well tested. Cc: Johan Hovold <johan@kernel.org> Cc: Michael Welling <mwelling@ieee.org> Cc: Markus Pargmann <mpa@pengutronix.de> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2015-10-21 21:29:53 +08:00
}
return ret;
}
core_initcall(gpiolib_dev_init);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
#ifdef CONFIG_DEBUG_FS
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
static void gpiolib_dbg_show(struct seq_file *s, struct gpio_device *gdev)
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
{
unsigned i;
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
struct gpio_chip *chip = gdev->chip;
unsigned gpio = gdev->base;
struct gpio_desc *gdesc = &gdev->descs[0];
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
int is_out;
int is_irq;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
for (i = 0; i < gdev->ngpio; i++, gpio++, gdesc++) {
if (!test_bit(FLAG_REQUESTED, &gdesc->flags)) {
if (gdesc->name) {
seq_printf(s, " gpio-%-3d (%-20.20s)\n",
gpio, gdesc->name);
}
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
continue;
}
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
gpiod_get_direction(gdesc);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
is_out = test_bit(FLAG_IS_OUT, &gdesc->flags);
is_irq = test_bit(FLAG_USED_AS_IRQ, &gdesc->flags);
seq_printf(s, " gpio-%-3d (%-20.20s|%-20.20s) %s %s %s",
gpio, gdesc->name ? gdesc->name : "", gdesc->label,
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
is_out ? "out" : "in ",
chip->get
? (chip->get(chip, i) ? "hi" : "lo")
: "? ",
is_irq ? "IRQ" : " ");
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
seq_printf(s, "\n");
}
}
static void *gpiolib_seq_start(struct seq_file *s, loff_t *pos)
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
{
unsigned long flags;
struct gpio_device *gdev = NULL;
loff_t index = *pos;
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
s->private = "";
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
spin_lock_irqsave(&gpio_lock, flags);
list_for_each_entry(gdev, &gpio_devices, list)
if (index-- == 0) {
spin_unlock_irqrestore(&gpio_lock, flags);
return gdev;
}
spin_unlock_irqrestore(&gpio_lock, flags);
return NULL;
}
static void *gpiolib_seq_next(struct seq_file *s, void *v, loff_t *pos)
{
unsigned long flags;
struct gpio_device *gdev = v;
void *ret = NULL;
spin_lock_irqsave(&gpio_lock, flags);
if (list_is_last(&gdev->list, &gpio_devices))
ret = NULL;
else
ret = list_entry(gdev->list.next, struct gpio_device, list);
spin_unlock_irqrestore(&gpio_lock, flags);
s->private = "\n";
++*pos;
return ret;
}
static void gpiolib_seq_stop(struct seq_file *s, void *v)
{
}
static int gpiolib_seq_show(struct seq_file *s, void *v)
{
struct gpio_device *gdev = v;
struct gpio_chip *chip = gdev->chip;
struct device *parent;
if (!chip) {
seq_printf(s, "%s%s: (dangling chip)", (char *)s->private,
dev_name(&gdev->dev));
return 0;
}
seq_printf(s, "%s%s: GPIOs %d-%d", (char *)s->private,
dev_name(&gdev->dev),
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
gdev->base, gdev->base + gdev->ngpio - 1);
parent = chip->parent;
if (parent)
seq_printf(s, ", parent: %s/%s",
parent->bus ? parent->bus->name : "no-bus",
dev_name(parent));
if (chip->label)
seq_printf(s, ", %s", chip->label);
if (chip->can_sleep)
seq_printf(s, ", can sleep");
seq_printf(s, ":\n");
if (chip->dbg_show)
chip->dbg_show(s, chip);
else
gpio: reflect base and ngpio into gpio_device Some information about the GPIO chip need to stay around also after the gpio_chip has been removed and only the gpio_device persist. The base and ngpio are such things, for example we don't want a new chip arriving to overlap the number space of a dangling gpio_device, and the chardev may still query the device for the number of lines etc. Note that the code that assigns base and insert gpio_device into the global list no longer check for a missing gpio_chip: we respect the number space allocated by any other gpio_device. As a consequence of the gdev being referenced directly from the gpio_desc, we need to verify it differently from all in-kernel API calls that fall through to direct queries to the gpio_chip vtable: we first check that desc is !NULL, then that desc->gdev is !NULL, then, if desc->gdev->chip is NULL, we *BAIL OUT* without any error, so as to manage the case where operations are requested on a device that is gone. These checks were non-uniform and partly missing in the past: so to simplify: create the macros VALIDATE_DESC() that will return -EINVAL if the desc or desc->gdev is missing and just 0 if the chip is gone, and conversely VALIDATE_DESC_VOID() for the case where the function does not return an error. By using these macros, we get warning messages about missing gdev with reference to the right function in the kernel log. Despite the macro business this simplifies the code and make it more readable than if we copy/paste the same descriptor checking code into all code ABI call sites (IMHO). Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2016-02-10 17:57:36 +08:00
gpiolib_dbg_show(s, gdev);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
return 0;
}
static const struct seq_operations gpiolib_seq_ops = {
.start = gpiolib_seq_start,
.next = gpiolib_seq_next,
.stop = gpiolib_seq_stop,
.show = gpiolib_seq_show,
};
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
static int gpiolib_open(struct inode *inode, struct file *file)
{
return seq_open(file, &gpiolib_seq_ops);
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
}
static const struct file_operations gpiolib_operations = {
.owner = THIS_MODULE,
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
.open = gpiolib_open,
.read = seq_read,
.llseek = seq_lseek,
.release = seq_release,
gpiolib: add gpio provider infrastructure Provide new implementation infrastructure that platforms may choose to use when implementing the GPIO programming interface. Platforms can update their GPIO support to use this. In many cases the incremental cost to access a non-inlined GPIO should be less than a dozen instructions, with the memory cost being about a page (total) of extra data and code. The upside is: * Providing two features which were "want to have (but OK to defer)" when GPIO interfaces were first discussed in November 2006: - A "struct gpio_chip" to plug in GPIOs that aren't directly supported by SOC platforms, but come from FPGAs or other multifunction devices using conventional device registers (like UCB-1x00 or SM501 GPIOs, and southbridges in PCs with more open specs than usual). - Full support for message-based GPIO expanders, where registers are accessed through sleeping I/O calls. Previous support for these "cansleep" calls was just stubs. (One example: the widely used pcf8574 I2C chips, with 8 GPIOs each.) * Including a non-stub implementation of the gpio_{request,free}() calls, making those calls much more useful. The diagnostic labels are also recorded given DEBUG_FS, so /sys/kernel/debug/gpio can show a snapshot of all GPIOs known to this infrastructure. The driver programming interfaces introduced in 2.6.21 do not change at all; this infrastructure is entirely below those covers. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Jean Delvare <khali@linux-fr.org> Cc: Eric Miao <eric.miao@marvell.com> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Philipp Zabel <philipp.zabel@gmail.com> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ben Gardner <bgardner@wabtec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:20 +08:00
};
static int __init gpiolib_debugfs_init(void)
{
/* /sys/kernel/debug/gpio */
(void) debugfs_create_file("gpio", S_IFREG | S_IRUGO,
NULL, NULL, &gpiolib_operations);
return 0;
}
subsys_initcall(gpiolib_debugfs_init);
#endif /* DEBUG_FS */