sysfs.rst 6.4 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167
  1. GPIO Sysfs Interface for Userspace
  2. ==================================
  3. .. warning::
  4. THIS ABI IS DEPRECATED, THE ABI DOCUMENTATION HAS BEEN MOVED TO
  5. Documentation/ABI/obsolete/sysfs-gpio AND NEW USERSPACE CONSUMERS
  6. ARE SUPPOSED TO USE THE CHARACTER DEVICE ABI. THIS OLD SYSFS ABI WILL
  7. NOT BE DEVELOPED (NO NEW FEATURES), IT WILL JUST BE MAINTAINED.
  8. Refer to the examples in tools/gpio/* for an introduction to the new
  9. character device ABI. Also see the userspace header in
  10. include/uapi/linux/gpio.h
  11. The deprecated sysfs ABI
  12. ------------------------
  13. Platforms which use the "gpiolib" implementors framework may choose to
  14. configure a sysfs user interface to GPIOs. This is different from the
  15. debugfs interface, since it provides control over GPIO direction and
  16. value instead of just showing a gpio state summary. Plus, it could be
  17. present on production systems without debugging support.
  18. Given appropriate hardware documentation for the system, userspace could
  19. know for example that GPIO #23 controls the write protect line used to
  20. protect boot loader segments in flash memory. System upgrade procedures
  21. may need to temporarily remove that protection, first importing a GPIO,
  22. then changing its output state, then updating the code before re-enabling
  23. the write protection. In normal use, GPIO #23 would never be touched,
  24. and the kernel would have no need to know about it.
  25. Again depending on appropriate hardware documentation, on some systems
  26. userspace GPIO can be used to determine system configuration data that
  27. standard kernels won't know about. And for some tasks, simple userspace
  28. GPIO drivers could be all that the system really needs.
  29. DO NOT ABUSE SYSFS TO CONTROL HARDWARE THAT HAS PROPER KERNEL DRIVERS.
  30. PLEASE READ THE DOCUMENT AT Documentation/driver-api/gpio/drivers-on-gpio.rst
  31. TO AVOID REINVENTING KERNEL WHEELS IN USERSPACE. I MEAN IT. REALLY.
  32. Paths in Sysfs
  33. --------------
  34. There are three kinds of entries in /sys/class/gpio:
  35. - Control interfaces used to get userspace control over GPIOs;
  36. - GPIOs themselves; and
  37. - GPIO controllers ("gpio_chip" instances).
  38. That's in addition to standard files including the "device" symlink.
  39. The control interfaces are write-only:
  40. /sys/class/gpio/
  41. "export" ...
  42. Userspace may ask the kernel to export control of
  43. a GPIO to userspace by writing its number to this file.
  44. Example: "echo 19 > export" will create a "gpio19" node
  45. for GPIO #19, if that's not requested by kernel code.
  46. "unexport" ...
  47. Reverses the effect of exporting to userspace.
  48. Example: "echo 19 > unexport" will remove a "gpio19"
  49. node exported using the "export" file.
  50. GPIO signals have paths like /sys/class/gpio/gpio42/ (for GPIO #42)
  51. and have the following read/write attributes:
  52. /sys/class/gpio/gpioN/
  53. "direction" ...
  54. reads as either "in" or "out". This value may
  55. normally be written. Writing as "out" defaults to
  56. initializing the value as low. To ensure glitch free
  57. operation, values "low" and "high" may be written to
  58. configure the GPIO as an output with that initial value.
  59. Note that this attribute *will not exist* if the kernel
  60. doesn't support changing the direction of a GPIO, or
  61. it was exported by kernel code that didn't explicitly
  62. allow userspace to reconfigure this GPIO's direction.
  63. "value" ...
  64. reads as either 0 (low) or 1 (high). If the GPIO
  65. is configured as an output, this value may be written;
  66. any nonzero value is treated as high.
  67. If the pin can be configured as interrupt-generating interrupt
  68. and if it has been configured to generate interrupts (see the
  69. description of "edge"), you can poll(2) on that file and
  70. poll(2) will return whenever the interrupt was triggered. If
  71. you use poll(2), set the events POLLPRI and POLLERR. If you
  72. use select(2), set the file descriptor in exceptfds. After
  73. poll(2) returns, either lseek(2) to the beginning of the sysfs
  74. file and read the new value or close the file and re-open it
  75. to read the value.
  76. "edge" ...
  77. reads as either "none", "rising", "falling", or
  78. "both". Write these strings to select the signal edge(s)
  79. that will make poll(2) on the "value" file return.
  80. This file exists only if the pin can be configured as an
  81. interrupt generating input pin.
  82. "active_low" ...
  83. reads as either 0 (false) or 1 (true). Write
  84. any nonzero value to invert the value attribute both
  85. for reading and writing. Existing and subsequent
  86. poll(2) support configuration via the edge attribute
  87. for "rising" and "falling" edges will follow this
  88. setting.
  89. GPIO controllers have paths like /sys/class/gpio/gpiochip42/ (for the
  90. controller implementing GPIOs starting at #42) and have the following
  91. read-only attributes:
  92. /sys/class/gpio/gpiochipN/
  93. "base" ...
  94. same as N, the first GPIO managed by this chip
  95. "label" ...
  96. provided for diagnostics (not always unique)
  97. "ngpio" ...
  98. how many GPIOs this manages (N to N + ngpio - 1)
  99. Board documentation should in most cases cover what GPIOs are used for
  100. what purposes. However, those numbers are not always stable; GPIOs on
  101. a daughtercard might be different depending on the base board being used,
  102. or other cards in the stack. In such cases, you may need to use the
  103. gpiochip nodes (possibly in conjunction with schematics) to determine
  104. the correct GPIO number to use for a given signal.
  105. Exporting from Kernel code
  106. --------------------------
  107. Kernel code can explicitly manage exports of GPIOs which have already been
  108. requested using gpio_request()::
  109. /* export the GPIO to userspace */
  110. int gpiod_export(struct gpio_desc *desc, bool direction_may_change);
  111. /* reverse gpio_export() */
  112. void gpiod_unexport(struct gpio_desc *desc);
  113. /* create a sysfs link to an exported GPIO node */
  114. int gpiod_export_link(struct device *dev, const char *name,
  115. struct gpio_desc *desc);
  116. After a kernel driver requests a GPIO, it may only be made available in
  117. the sysfs interface by gpiod_export(). The driver can control whether the
  118. signal direction may change. This helps drivers prevent userspace code
  119. from accidentally clobbering important system state.
  120. This explicit exporting can help with debugging (by making some kinds
  121. of experiments easier), or can provide an always-there interface that's
  122. suitable for documenting as part of a board support package.
  123. After the GPIO has been exported, gpiod_export_link() allows creating
  124. symlinks from elsewhere in sysfs to the GPIO sysfs node. Drivers can
  125. use this to provide the interface under their own device in sysfs with
  126. a descriptive name.