Commit | Line | Data |
---|---|---|
4c20386c DB |
1 | GPIO Interfaces |
2 | ||
3 | This provides an overview of GPIO access conventions on Linux. | |
4 | ||
7560fa60 DB |
5 | These calls use the gpio_* naming prefix. No other calls should use that |
6 | prefix, or the related __gpio_* prefix. | |
7 | ||
4c20386c DB |
8 | |
9 | What is a GPIO? | |
10 | =============== | |
11 | A "General Purpose Input/Output" (GPIO) is a flexible software-controlled | |
12 | digital signal. They are provided from many kinds of chip, and are familiar | |
13 | to Linux developers working with embedded and custom hardware. Each GPIO | |
14 | represents a bit connected to a particular pin, or "ball" on Ball Grid Array | |
15 | (BGA) packages. Board schematics show which external hardware connects to | |
16 | which GPIOs. Drivers can be written generically, so that board setup code | |
17 | passes such pin configuration data to drivers. | |
18 | ||
19 | System-on-Chip (SOC) processors heavily rely on GPIOs. In some cases, every | |
20 | non-dedicated pin can be configured as a GPIO; and most chips have at least | |
21 | several dozen of them. Programmable logic devices (like FPGAs) can easily | |
22 | provide GPIOs; multifunction chips like power managers, and audio codecs | |
23 | often have a few such pins to help with pin scarcity on SOCs; and there are | |
24 | also "GPIO Expander" chips that connect using the I2C or SPI serial busses. | |
25 | Most PC southbridges have a few dozen GPIO-capable pins (with only the BIOS | |
26 | firmware knowing how they're used). | |
27 | ||
28 | The exact capabilities of GPIOs vary between systems. Common options: | |
29 | ||
30 | - Output values are writable (high=1, low=0). Some chips also have | |
31 | options about how that value is driven, so that for example only one | |
32 | value might be driven ... supporting "wire-OR" and similar schemes | |
1668be71 | 33 | for the other value (notably, "open drain" signaling). |
4c20386c DB |
34 | |
35 | - Input values are likewise readable (1, 0). Some chips support readback | |
36 | of pins configured as "output", which is very useful in such "wire-OR" | |
37 | cases (to support bidirectional signaling). GPIO controllers may have | |
7c2db759 | 38 | input de-glitch/debounce logic, sometimes with software controls. |
4c20386c DB |
39 | |
40 | - Inputs can often be used as IRQ signals, often edge triggered but | |
41 | sometimes level triggered. Such IRQs may be configurable as system | |
42 | wakeup events, to wake the system from a low power state. | |
43 | ||
44 | - Usually a GPIO will be configurable as either input or output, as needed | |
45 | by different product boards; single direction ones exist too. | |
46 | ||
47 | - Most GPIOs can be accessed while holding spinlocks, but those accessed | |
48 | through a serial bus normally can't. Some systems support both types. | |
49 | ||
50 | On a given board each GPIO is used for one specific purpose like monitoring | |
51 | MMC/SD card insertion/removal, detecting card writeprotect status, driving | |
52 | a LED, configuring a transceiver, bitbanging a serial bus, poking a hardware | |
53 | watchdog, sensing a switch, and so on. | |
54 | ||
55 | ||
56 | GPIO conventions | |
57 | ================ | |
58 | Note that this is called a "convention" because you don't need to do it this | |
59 | way, and it's no crime if you don't. There **are** cases where portability | |
60 | is not the main issue; GPIOs are often used for the kind of board-specific | |
61 | glue logic that may even change between board revisions, and can't ever be | |
62 | used on a board that's wired differently. Only least-common-denominator | |
63 | functionality can be very portable. Other features are platform-specific, | |
64 | and that can be critical for glue logic. | |
65 | ||
7c2db759 | 66 | Plus, this doesn't require any implementation framework, just an interface. |
4c20386c DB |
67 | One platform might implement it as simple inline functions accessing chip |
68 | registers; another might implement it by delegating through abstractions | |
7c2db759 DB |
69 | used for several very different kinds of GPIO controller. (There is some |
70 | optional code supporting such an implementation strategy, described later | |
71 | in this document, but drivers acting as clients to the GPIO interface must | |
72 | not care how it's implemented.) | |
4c20386c DB |
73 | |
74 | That said, if the convention is supported on their platform, drivers should | |
7560fa60 DB |
75 | use it when possible. Platforms must declare GENERIC_GPIO support in their |
76 | Kconfig (boolean true), and provide an <asm/gpio.h> file. Drivers that can't | |
77 | work without standard GPIO calls should have Kconfig entries which depend | |
78 | on GENERIC_GPIO. The GPIO calls are available, either as "real code" or as | |
79 | optimized-away stubs, when drivers use the include file: | |
4c20386c | 80 | |
7560fa60 | 81 | #include <linux/gpio.h> |
4c20386c DB |
82 | |
83 | If you stick to this convention then it'll be easier for other developers to | |
84 | see what your code is doing, and help maintain it. | |
85 | ||
a0a99835 DB |
86 | Note that these operations include I/O barriers on platforms which need to |
87 | use them; drivers don't need to add them explicitly. | |
88 | ||
4c20386c DB |
89 | |
90 | Identifying GPIOs | |
91 | ----------------- | |
92 | GPIOs are identified by unsigned integers in the range 0..MAX_INT. That | |
93 | reserves "negative" numbers for other purposes like marking signals as | |
f5de6111 DB |
94 | "not available on this board", or indicating faults. Code that doesn't |
95 | touch the underlying hardware treats these integers as opaque cookies. | |
4c20386c DB |
96 | |
97 | Platforms define how they use those integers, and usually #define symbols | |
98 | for the GPIO lines so that board-specific setup code directly corresponds | |
99 | to the relevant schematics. In contrast, drivers should only use GPIO | |
100 | numbers passed to them from that setup code, using platform_data to hold | |
101 | board-specific pin configuration data (along with other board specific | |
102 | data they need). That avoids portability problems. | |
103 | ||
104 | So for example one platform uses numbers 32-159 for GPIOs; while another | |
105 | uses numbers 0..63 with one set of GPIO controllers, 64-79 with another | |
106 | type of GPIO controller, and on one particular board 80-95 with an FPGA. | |
107 | The numbers need not be contiguous; either of those platforms could also | |
108 | use numbers 2000-2063 to identify GPIOs in a bank of I2C GPIO expanders. | |
109 | ||
e6de1808 GL |
110 | If you want to initialize a structure with an invalid GPIO number, use |
111 | some negative number (perhaps "-EINVAL"); that will never be valid. To | |
112 | test if a number could reference a GPIO, you may use this predicate: | |
113 | ||
114 | int gpio_is_valid(int number); | |
115 | ||
116 | A number that's not valid will be rejected by calls which may request | |
117 | or free GPIOs (see below). Other numbers may also be rejected; for | |
118 | example, a number might be valid but unused on a given board. | |
119 | ||
4c20386c DB |
120 | Whether a platform supports multiple GPIO controllers is currently a |
121 | platform-specific implementation issue. | |
122 | ||
123 | ||
124 | Using GPIOs | |
125 | ----------- | |
8a0cecff DB |
126 | The first thing a system should do with a GPIO is allocate it, using |
127 | the gpio_request() call; see later. | |
128 | ||
129 | One of the next things to do with a GPIO, often in board setup code when | |
4c20386c DB |
130 | setting up a platform_device using the GPIO, is mark its direction: |
131 | ||
132 | /* set as input or output, returning 0 or negative errno */ | |
133 | int gpio_direction_input(unsigned gpio); | |
28735a72 | 134 | int gpio_direction_output(unsigned gpio, int value); |
4c20386c DB |
135 | |
136 | The return value is zero for success, else a negative errno. It should | |
137 | be checked, since the get/set calls don't have error returns and since | |
83c6590c DB |
138 | misconfiguration is possible. You should normally issue these calls from |
139 | a task context. However, for spinlock-safe GPIOs it's OK to use them | |
140 | before tasking is enabled, as part of early board setup. | |
4c20386c | 141 | |
28735a72 DB |
142 | For output GPIOs, the value provided becomes the initial output value. |
143 | This helps avoid signal glitching during system startup. | |
144 | ||
7c2db759 DB |
145 | For compatibility with legacy interfaces to GPIOs, setting the direction |
146 | of a GPIO implicitly requests that GPIO (see below) if it has not been | |
8a0cecff DB |
147 | requested already. That compatibility is being removed from the optional |
148 | gpiolib framework. | |
7c2db759 | 149 | |
4c20386c DB |
150 | Setting the direction can fail if the GPIO number is invalid, or when |
151 | that particular GPIO can't be used in that mode. It's generally a bad | |
152 | idea to rely on boot firmware to have set the direction correctly, since | |
153 | it probably wasn't validated to do more than boot Linux. (Similarly, | |
154 | that board setup code probably needs to multiplex that pin as a GPIO, | |
155 | and configure pullups/pulldowns appropriately.) | |
156 | ||
157 | ||
158 | Spinlock-Safe GPIO access | |
159 | ------------------------- | |
160 | Most GPIO controllers can be accessed with memory read/write instructions. | |
161 | That doesn't need to sleep, and can safely be done from inside IRQ handlers. | |
7c2db759 | 162 | (That includes hardirq contexts on RT kernels.) |
4c20386c DB |
163 | |
164 | Use these calls to access such GPIOs: | |
165 | ||
166 | /* GPIO INPUT: return zero or nonzero */ | |
167 | int gpio_get_value(unsigned gpio); | |
168 | ||
169 | /* GPIO OUTPUT */ | |
170 | void gpio_set_value(unsigned gpio, int value); | |
171 | ||
172 | The values are boolean, zero for low, nonzero for high. When reading the | |
173 | value of an output pin, the value returned should be what's seen on the | |
174 | pin ... that won't always match the specified output value, because of | |
7c2db759 | 175 | issues including open-drain signaling and output latencies. |
4c20386c DB |
176 | |
177 | The get/set calls have no error returns because "invalid GPIO" should have | |
be1ff386 | 178 | been reported earlier from gpio_direction_*(). However, note that not all |
4c20386c | 179 | platforms can read the value of output pins; those that can't should always |
f5de6111 DB |
180 | return zero. Also, using these calls for GPIOs that can't safely be accessed |
181 | without sleeping (see below) is an error. | |
4c20386c | 182 | |
f5de6111 | 183 | Platform-specific implementations are encouraged to optimize the two |
4c20386c DB |
184 | calls to access the GPIO value in cases where the GPIO number (and for |
185 | output, value) are constant. It's normal for them to need only a couple | |
186 | of instructions in such cases (reading or writing a hardware register), | |
187 | and not to need spinlocks. Such optimized calls can make bitbanging | |
188 | applications a lot more efficient (in both space and time) than spending | |
189 | dozens of instructions on subroutine calls. | |
190 | ||
191 | ||
192 | GPIO access that may sleep | |
193 | -------------------------- | |
194 | Some GPIO controllers must be accessed using message based busses like I2C | |
195 | or SPI. Commands to read or write those GPIO values require waiting to | |
196 | get to the head of a queue to transmit a command and get its response. | |
197 | This requires sleeping, which can't be done from inside IRQ handlers. | |
198 | ||
199 | Platforms that support this type of GPIO distinguish them from other GPIOs | |
7c2db759 | 200 | by returning nonzero from this call (which requires a valid GPIO number, |
8a0cecff | 201 | which should have been previously allocated with gpio_request): |
4c20386c DB |
202 | |
203 | int gpio_cansleep(unsigned gpio); | |
204 | ||
205 | To access such GPIOs, a different set of accessors is defined: | |
206 | ||
207 | /* GPIO INPUT: return zero or nonzero, might sleep */ | |
208 | int gpio_get_value_cansleep(unsigned gpio); | |
209 | ||
210 | /* GPIO OUTPUT, might sleep */ | |
211 | void gpio_set_value_cansleep(unsigned gpio, int value); | |
212 | ||
213 | Other than the fact that these calls might sleep, and will not be ignored | |
214 | for GPIOs that can't be accessed from IRQ handlers, these calls act the | |
215 | same as the spinlock-safe calls. | |
216 | ||
217 | ||
8a0cecff DB |
218 | Claiming and Releasing GPIOs |
219 | ---------------------------- | |
4c20386c | 220 | To help catch system configuration errors, two calls are defined. |
4c20386c DB |
221 | |
222 | /* request GPIO, returning 0 or negative errno. | |
223 | * non-null labels may be useful for diagnostics. | |
224 | */ | |
225 | int gpio_request(unsigned gpio, const char *label); | |
226 | ||
227 | /* release previously-claimed GPIO */ | |
228 | void gpio_free(unsigned gpio); | |
229 | ||
230 | Passing invalid GPIO numbers to gpio_request() will fail, as will requesting | |
231 | GPIOs that have already been claimed with that call. The return value of | |
83c6590c DB |
232 | gpio_request() must be checked. You should normally issue these calls from |
233 | a task context. However, for spinlock-safe GPIOs it's OK to request GPIOs | |
234 | before tasking is enabled, as part of early board setup. | |
4c20386c DB |
235 | |
236 | These calls serve two basic purposes. One is marking the signals which | |
237 | are actually in use as GPIOs, for better diagnostics; systems may have | |
238 | several hundred potential GPIOs, but often only a dozen are used on any | |
7c2db759 DB |
239 | given board. Another is to catch conflicts, identifying errors when |
240 | (a) two or more drivers wrongly think they have exclusive use of that | |
241 | signal, or (b) something wrongly believes it's safe to remove drivers | |
242 | needed to manage a signal that's in active use. That is, requesting a | |
243 | GPIO can serve as a kind of lock. | |
4c20386c | 244 | |
35e8bb51 DB |
245 | Some platforms may also use knowledge about what GPIOs are active for |
246 | power management, such as by powering down unused chip sectors and, more | |
247 | easily, gating off unused clocks. | |
248 | ||
4c20386c DB |
249 | Note that requesting a GPIO does NOT cause it to be configured in any |
250 | way; it just marks that GPIO as in use. Separate code must handle any | |
251 | pin setup (e.g. controlling which pin the GPIO uses, pullup/pulldown). | |
252 | ||
7c2db759 DB |
253 | Also note that it's your responsibility to have stopped using a GPIO |
254 | before you free it. | |
255 | ||
3e45f1d1 EM |
256 | Considering in most cases GPIOs are actually configured right after they |
257 | are claimed, three additional calls are defined: | |
258 | ||
259 | /* request a single GPIO, with initial configuration specified by | |
260 | * 'flags', identical to gpio_request() wrt other arguments and | |
261 | * return value | |
262 | */ | |
263 | int gpio_request_one(unsigned gpio, unsigned long flags, const char *label); | |
264 | ||
265 | /* request multiple GPIOs in a single call | |
266 | */ | |
267 | int gpio_request_array(struct gpio *array, size_t num); | |
268 | ||
269 | /* release multiple GPIOs in a single call | |
270 | */ | |
271 | void gpio_free_array(struct gpio *array, size_t num); | |
272 | ||
273 | where 'flags' is currently defined to specify the following properties: | |
274 | ||
275 | * GPIOF_DIR_IN - to configure direction as input | |
276 | * GPIOF_DIR_OUT - to configure direction as output | |
277 | ||
278 | * GPIOF_INIT_LOW - as output, set initial level to LOW | |
279 | * GPIOF_INIT_HIGH - as output, set initial level to HIGH | |
280 | ||
281 | since GPIOF_INIT_* are only valid when configured as output, so group valid | |
282 | combinations as: | |
283 | ||
284 | * GPIOF_IN - configure as input | |
285 | * GPIOF_OUT_INIT_LOW - configured as output, initial level LOW | |
286 | * GPIOF_OUT_INIT_HIGH - configured as output, initial level HIGH | |
287 | ||
288 | In the future, these flags can be extended to support more properties such | |
289 | as open-drain status. | |
290 | ||
291 | Further more, to ease the claim/release of multiple GPIOs, 'struct gpio' is | |
292 | introduced to encapsulate all three fields as: | |
293 | ||
294 | struct gpio { | |
295 | unsigned gpio; | |
296 | unsigned long flags; | |
297 | const char *label; | |
298 | }; | |
299 | ||
300 | A typical example of usage: | |
301 | ||
302 | static struct gpio leds_gpios[] = { | |
303 | { 32, GPIOF_OUT_INIT_HIGH, "Power LED" }, /* default to ON */ | |
304 | { 33, GPIOF_OUT_INIT_LOW, "Green LED" }, /* default to OFF */ | |
305 | { 34, GPIOF_OUT_INIT_LOW, "Red LED" }, /* default to OFF */ | |
306 | { 35, GPIOF_OUT_INIT_LOW, "Blue LED" }, /* default to OFF */ | |
307 | { ... }, | |
308 | }; | |
309 | ||
310 | err = gpio_request_one(31, GPIOF_IN, "Reset Button"); | |
311 | if (err) | |
312 | ... | |
313 | ||
314 | err = gpio_request_array(leds_gpios, ARRAY_SIZE(leds_gpios)); | |
315 | if (err) | |
316 | ... | |
317 | ||
318 | gpio_free_array(leds_gpios, ARRAY_SIZE(leds_gpios)); | |
319 | ||
4c20386c DB |
320 | |
321 | GPIOs mapped to IRQs | |
322 | -------------------- | |
323 | GPIO numbers are unsigned integers; so are IRQ numbers. These make up | |
324 | two logically distinct namespaces (GPIO 0 need not use IRQ 0). You can | |
325 | map between them using calls like: | |
326 | ||
327 | /* map GPIO numbers to IRQ numbers */ | |
328 | int gpio_to_irq(unsigned gpio); | |
329 | ||
0f6d504e | 330 | /* map IRQ numbers to GPIO numbers (avoid using this) */ |
4c20386c DB |
331 | int irq_to_gpio(unsigned irq); |
332 | ||
333 | Those return either the corresponding number in the other namespace, or | |
334 | else a negative errno code if the mapping can't be done. (For example, | |
7c2db759 | 335 | some GPIOs can't be used as IRQs.) It is an unchecked error to use a GPIO |
be1ff386 | 336 | number that wasn't set up as an input using gpio_direction_input(), or |
4c20386c DB |
337 | to use an IRQ number that didn't originally come from gpio_to_irq(). |
338 | ||
339 | These two mapping calls are expected to cost on the order of a single | |
340 | addition or subtraction. They're not allowed to sleep. | |
341 | ||
342 | Non-error values returned from gpio_to_irq() can be passed to request_irq() | |
343 | or free_irq(). They will often be stored into IRQ resources for platform | |
344 | devices, by the board-specific initialization code. Note that IRQ trigger | |
345 | options are part of the IRQ interface, e.g. IRQF_TRIGGER_FALLING, as are | |
346 | system wakeup capabilities. | |
347 | ||
348 | Non-error values returned from irq_to_gpio() would most commonly be used | |
f5de6111 | 349 | with gpio_get_value(), for example to initialize or update driver state |
0f6d504e DB |
350 | when the IRQ is edge-triggered. Note that some platforms don't support |
351 | this reverse mapping, so you should avoid using it. | |
4c20386c DB |
352 | |
353 | ||
1668be71 DB |
354 | Emulating Open Drain Signals |
355 | ---------------------------- | |
356 | Sometimes shared signals need to use "open drain" signaling, where only the | |
357 | low signal level is actually driven. (That term applies to CMOS transistors; | |
358 | "open collector" is used for TTL.) A pullup resistor causes the high signal | |
359 | level. This is sometimes called a "wire-AND"; or more practically, from the | |
360 | negative logic (low=true) perspective this is a "wire-OR". | |
361 | ||
362 | One common example of an open drain signal is a shared active-low IRQ line. | |
363 | Also, bidirectional data bus signals sometimes use open drain signals. | |
364 | ||
365 | Some GPIO controllers directly support open drain outputs; many don't. When | |
366 | you need open drain signaling but your hardware doesn't directly support it, | |
367 | there's a common idiom you can use to emulate it with any GPIO pin that can | |
368 | be used as either an input or an output: | |
369 | ||
370 | LOW: gpio_direction_output(gpio, 0) ... this drives the signal | |
371 | and overrides the pullup. | |
372 | ||
373 | HIGH: gpio_direction_input(gpio) ... this turns off the output, | |
374 | so the pullup (or some other device) controls the signal. | |
375 | ||
376 | If you are "driving" the signal high but gpio_get_value(gpio) reports a low | |
377 | value (after the appropriate rise time passes), you know some other component | |
378 | is driving the shared signal low. That's not necessarily an error. As one | |
379 | common example, that's how I2C clocks are stretched: a slave that needs a | |
380 | slower clock delays the rising edge of SCK, and the I2C master adjusts its | |
381 | signaling rate accordingly. | |
382 | ||
4c20386c DB |
383 | |
384 | What do these conventions omit? | |
385 | =============================== | |
386 | One of the biggest things these conventions omit is pin multiplexing, since | |
387 | this is highly chip-specific and nonportable. One platform might not need | |
388 | explicit multiplexing; another might have just two options for use of any | |
389 | given pin; another might have eight options per pin; another might be able | |
390 | to route a given GPIO to any one of several pins. (Yes, those examples all | |
391 | come from systems that run Linux today.) | |
392 | ||
393 | Related to multiplexing is configuration and enabling of the pullups or | |
394 | pulldowns integrated on some platforms. Not all platforms support them, | |
395 | or support them in the same way; and any given board might use external | |
396 | pullups (or pulldowns) so that the on-chip ones should not be used. | |
7c2db759 | 397 | (When a circuit needs 5 kOhm, on-chip 100 kOhm resistors won't do.) |
7560fa60 DB |
398 | Likewise drive strength (2 mA vs 20 mA) and voltage (1.8V vs 3.3V) is a |
399 | platform-specific issue, as are models like (not) having a one-to-one | |
400 | correspondence between configurable pins and GPIOs. | |
4c20386c DB |
401 | |
402 | There are other system-specific mechanisms that are not specified here, | |
403 | like the aforementioned options for input de-glitching and wire-OR output. | |
404 | Hardware may support reading or writing GPIOs in gangs, but that's usually | |
f5de6111 | 405 | configuration dependent: for GPIOs sharing the same bank. (GPIOs are |
4c20386c | 406 | commonly grouped in banks of 16 or 32, with a given SOC having several such |
7c2db759 DB |
407 | banks.) Some systems can trigger IRQs from output GPIOs, or read values |
408 | from pins not managed as GPIOs. Code relying on such mechanisms will | |
409 | necessarily be nonportable. | |
4c20386c | 410 | |
7c2db759 | 411 | Dynamic definition of GPIOs is not currently standard; for example, as |
4c20386c DB |
412 | a side effect of configuring an add-on board with some GPIO expanders. |
413 | ||
7c2db759 DB |
414 | |
415 | GPIO implementor's framework (OPTIONAL) | |
416 | ======================================= | |
417 | As noted earlier, there is an optional implementation framework making it | |
418 | easier for platforms to support different kinds of GPIO controller using | |
d8f388d8 | 419 | the same programming interface. This framework is called "gpiolib". |
7c2db759 DB |
420 | |
421 | As a debugging aid, if debugfs is available a /sys/kernel/debug/gpio file | |
422 | will be found there. That will list all the controllers registered through | |
423 | this framework, and the state of the GPIOs currently in use. | |
424 | ||
425 | ||
426 | Controller Drivers: gpio_chip | |
427 | ----------------------------- | |
428 | In this framework each GPIO controller is packaged as a "struct gpio_chip" | |
429 | with information common to each controller of that type: | |
430 | ||
431 | - methods to establish GPIO direction | |
432 | - methods used to access GPIO values | |
433 | - flag saying whether calls to its methods may sleep | |
434 | - optional debugfs dump method (showing extra state like pullup config) | |
435 | - label for diagnostics | |
436 | ||
437 | There is also per-instance data, which may come from device.platform_data: | |
438 | the number of its first GPIO, and how many GPIOs it exposes. | |
439 | ||
440 | The code implementing a gpio_chip should support multiple instances of the | |
441 | controller, possibly using the driver model. That code will configure each | |
442 | gpio_chip and issue gpiochip_add(). Removing a GPIO controller should be | |
443 | rare; use gpiochip_remove() when it is unavoidable. | |
444 | ||
445 | Most often a gpio_chip is part of an instance-specific structure with state | |
446 | not exposed by the GPIO interfaces, such as addressing, power management, | |
bfc9dcab | 447 | and more. Chips such as codecs will have complex non-GPIO state. |
7c2db759 DB |
448 | |
449 | Any debugfs dump method should normally ignore signals which haven't been | |
450 | requested as GPIOs. They can use gpiochip_is_requested(), which returns | |
451 | either NULL or the label associated with that GPIO when it was requested. | |
452 | ||
453 | ||
454 | Platform Support | |
455 | ---------------- | |
7444a72e MB |
456 | To support this framework, a platform's Kconfig will "select" either |
457 | ARCH_REQUIRE_GPIOLIB or ARCH_WANT_OPTIONAL_GPIOLIB | |
7c2db759 DB |
458 | and arrange that its <asm/gpio.h> includes <asm-generic/gpio.h> and defines |
459 | three functions: gpio_get_value(), gpio_set_value(), and gpio_cansleep(). | |
460 | They may also want to provide a custom value for ARCH_NR_GPIOS. | |
461 | ||
7444a72e MB |
462 | ARCH_REQUIRE_GPIOLIB means that the gpio-lib code will always get compiled |
463 | into the kernel on that architecture. | |
464 | ||
465 | ARCH_WANT_OPTIONAL_GPIOLIB means the gpio-lib code defaults to off and the user | |
466 | can enable it and build it into the kernel optionally. | |
467 | ||
468 | If neither of these options are selected, the platform does not support | |
469 | GPIOs through GPIO-lib and the code cannot be enabled by the user. | |
470 | ||
7c2db759 DB |
471 | Trivial implementations of those functions can directly use framework |
472 | code, which always dispatches through the gpio_chip: | |
473 | ||
474 | #define gpio_get_value __gpio_get_value | |
475 | #define gpio_set_value __gpio_set_value | |
476 | #define gpio_cansleep __gpio_cansleep | |
477 | ||
478 | Fancier implementations could instead define those as inline functions with | |
479 | logic optimizing access to specific SOC-based GPIOs. For example, if the | |
480 | referenced GPIO is the constant "12", getting or setting its value could | |
481 | cost as little as two or three instructions, never sleeping. When such an | |
482 | optimization is not possible those calls must delegate to the framework | |
483 | code, costing at least a few dozen instructions. For bitbanged I/O, such | |
484 | instruction savings can be significant. | |
485 | ||
486 | For SOCs, platform-specific code defines and registers gpio_chip instances | |
487 | for each bank of on-chip GPIOs. Those GPIOs should be numbered/labeled to | |
488 | match chip vendor documentation, and directly match board schematics. They | |
489 | may well start at zero and go up to a platform-specific limit. Such GPIOs | |
490 | are normally integrated into platform initialization to make them always be | |
491 | available, from arch_initcall() or earlier; they can often serve as IRQs. | |
492 | ||
493 | ||
494 | Board Support | |
495 | ------------- | |
496 | For external GPIO controllers -- such as I2C or SPI expanders, ASICs, multi | |
497 | function devices, FPGAs or CPLDs -- most often board-specific code handles | |
498 | registering controller devices and ensures that their drivers know what GPIO | |
499 | numbers to use with gpiochip_add(). Their numbers often start right after | |
500 | platform-specific GPIOs. | |
501 | ||
502 | For example, board setup code could create structures identifying the range | |
503 | of GPIOs that chip will expose, and passes them to each GPIO expander chip | |
504 | using platform_data. Then the chip driver's probe() routine could pass that | |
505 | data to gpiochip_add(). | |
506 | ||
507 | Initialization order can be important. For example, when a device relies on | |
508 | an I2C-based GPIO, its probe() routine should only be called after that GPIO | |
509 | becomes available. That may mean the device should not be registered until | |
510 | calls for that GPIO can work. One way to address such dependencies is for | |
511 | such gpio_chip controllers to provide setup() and teardown() callbacks to | |
512 | board specific code; those board specific callbacks would register devices | |
d8f388d8 DB |
513 | once all the necessary resources are available, and remove them later when |
514 | the GPIO controller device becomes unavailable. | |
515 | ||
516 | ||
517 | Sysfs Interface for Userspace (OPTIONAL) | |
518 | ======================================== | |
519 | Platforms which use the "gpiolib" implementors framework may choose to | |
520 | configure a sysfs user interface to GPIOs. This is different from the | |
521 | debugfs interface, since it provides control over GPIO direction and | |
522 | value instead of just showing a gpio state summary. Plus, it could be | |
523 | present on production systems without debugging support. | |
524 | ||
19f59460 | 525 | Given appropriate hardware documentation for the system, userspace could |
d8f388d8 DB |
526 | know for example that GPIO #23 controls the write protect line used to |
527 | protect boot loader segments in flash memory. System upgrade procedures | |
528 | may need to temporarily remove that protection, first importing a GPIO, | |
529 | then changing its output state, then updating the code before re-enabling | |
530 | the write protection. In normal use, GPIO #23 would never be touched, | |
531 | and the kernel would have no need to know about it. | |
532 | ||
533 | Again depending on appropriate hardware documentation, on some systems | |
534 | userspace GPIO can be used to determine system configuration data that | |
535 | standard kernels won't know about. And for some tasks, simple userspace | |
536 | GPIO drivers could be all that the system really needs. | |
537 | ||
538 | Note that standard kernel drivers exist for common "LEDs and Buttons" | |
539 | GPIO tasks: "leds-gpio" and "gpio_keys", respectively. Use those | |
540 | instead of talking directly to the GPIOs; they integrate with kernel | |
541 | frameworks better than your userspace code could. | |
542 | ||
543 | ||
544 | Paths in Sysfs | |
545 | -------------- | |
546 | There are three kinds of entry in /sys/class/gpio: | |
547 | ||
548 | - Control interfaces used to get userspace control over GPIOs; | |
549 | ||
550 | - GPIOs themselves; and | |
551 | ||
552 | - GPIO controllers ("gpio_chip" instances). | |
553 | ||
554 | That's in addition to standard files including the "device" symlink. | |
555 | ||
556 | The control interfaces are write-only: | |
557 | ||
558 | /sys/class/gpio/ | |
559 | ||
560 | "export" ... Userspace may ask the kernel to export control of | |
561 | a GPIO to userspace by writing its number to this file. | |
562 | ||
563 | Example: "echo 19 > export" will create a "gpio19" node | |
564 | for GPIO #19, if that's not requested by kernel code. | |
565 | ||
566 | "unexport" ... Reverses the effect of exporting to userspace. | |
567 | ||
568 | Example: "echo 19 > unexport" will remove a "gpio19" | |
569 | node exported using the "export" file. | |
570 | ||
571 | GPIO signals have paths like /sys/class/gpio/gpio42/ (for GPIO #42) | |
572 | and have the following read/write attributes: | |
573 | ||
574 | /sys/class/gpio/gpioN/ | |
575 | ||
576 | "direction" ... reads as either "in" or "out". This value may | |
577 | normally be written. Writing as "out" defaults to | |
578 | initializing the value as low. To ensure glitch free | |
579 | operation, values "low" and "high" may be written to | |
580 | configure the GPIO as an output with that initial value. | |
581 | ||
582 | Note that this attribute *will not exist* if the kernel | |
583 | doesn't support changing the direction of a GPIO, or | |
584 | it was exported by kernel code that didn't explicitly | |
585 | allow userspace to reconfigure this GPIO's direction. | |
586 | ||
587 | "value" ... reads as either 0 (low) or 1 (high). If the GPIO | |
588 | is configured as an output, this value may be written; | |
589 | any nonzero value is treated as high. | |
590 | ||
ff77c352 DG |
591 | "edge" ... reads as either "none", "rising", "falling", or |
592 | "both". Write these strings to select the signal edge(s) | |
593 | that will make poll(2) on the "value" file return. | |
594 | ||
595 | This file exists only if the pin can be configured as an | |
596 | interrupt generating input pin. | |
597 | ||
07697461 JN |
598 | "active_low" ... reads as either 0 (false) or 1 (true). Write |
599 | any nonzero value to invert the value attribute both | |
600 | for reading and writing. Existing and subsequent | |
601 | poll(2) support configuration via the edge attribute | |
602 | for "rising" and "falling" edges will follow this | |
603 | setting. | |
604 | ||
bfc9dcab | 605 | GPIO controllers have paths like /sys/class/gpio/gpiochip42/ (for the |
d8f388d8 DB |
606 | controller implementing GPIOs starting at #42) and have the following |
607 | read-only attributes: | |
608 | ||
609 | /sys/class/gpio/gpiochipN/ | |
610 | ||
611 | "base" ... same as N, the first GPIO managed by this chip | |
612 | ||
613 | "label" ... provided for diagnostics (not always unique) | |
614 | ||
615 | "ngpio" ... how many GPIOs this manges (N to N + ngpio - 1) | |
616 | ||
617 | Board documentation should in most cases cover what GPIOs are used for | |
618 | what purposes. However, those numbers are not always stable; GPIOs on | |
619 | a daughtercard might be different depending on the base board being used, | |
620 | or other cards in the stack. In such cases, you may need to use the | |
621 | gpiochip nodes (possibly in conjunction with schematics) to determine | |
622 | the correct GPIO number to use for a given signal. | |
623 | ||
624 | ||
625 | Exporting from Kernel code | |
626 | -------------------------- | |
627 | Kernel code can explicitly manage exports of GPIOs which have already been | |
628 | requested using gpio_request(): | |
629 | ||
630 | /* export the GPIO to userspace */ | |
631 | int gpio_export(unsigned gpio, bool direction_may_change); | |
632 | ||
633 | /* reverse gpio_export() */ | |
634 | void gpio_unexport(); | |
635 | ||
a4177ee7 JN |
636 | /* create a sysfs link to an exported GPIO node */ |
637 | int gpio_export_link(struct device *dev, const char *name, | |
638 | unsigned gpio) | |
639 | ||
07697461 JN |
640 | /* change the polarity of a GPIO node in sysfs */ |
641 | int gpio_sysfs_set_active_low(unsigned gpio, int value); | |
a4177ee7 | 642 | |
d8f388d8 DB |
643 | After a kernel driver requests a GPIO, it may only be made available in |
644 | the sysfs interface by gpio_export(). The driver can control whether the | |
645 | signal direction may change. This helps drivers prevent userspace code | |
646 | from accidentally clobbering important system state. | |
647 | ||
648 | This explicit exporting can help with debugging (by making some kinds | |
649 | of experiments easier), or can provide an always-there interface that's | |
650 | suitable for documenting as part of a board support package. | |
a4177ee7 JN |
651 | |
652 | After the GPIO has been exported, gpio_export_link() allows creating | |
653 | symlinks from elsewhere in sysfs to the GPIO sysfs node. Drivers can | |
654 | use this to provide the interface under their own device in sysfs with | |
655 | a descriptive name. | |
07697461 JN |
656 | |
657 | Drivers can use gpio_sysfs_set_active_low() to hide GPIO line polarity | |
658 | differences between boards from user space. This only affects the | |
659 | sysfs interface. Polarity change can be done both before and after | |
660 | gpio_export(), and previously enabled poll(2) support for either | |
661 | rising or falling edge will be reconfigured to follow this setting. |