Commit | Line | Data |
---|---|---|
fd8e198c AC |
1 | GPIO Descriptor Driver Interface |
2 | ================================ | |
3 | ||
4 | This document serves as a guide for GPIO chip drivers writers. Note that it | |
5 | describes the new descriptor-based interface. For a description of the | |
6 | deprecated integer-based GPIO interface please refer to gpio-legacy.txt. | |
7 | ||
8 | Each GPIO controller driver needs to include the following header, which defines | |
9 | the structures used to define a GPIO driver: | |
10 | ||
11 | #include <linux/gpio/driver.h> | |
12 | ||
13 | ||
14 | Internal Representation of GPIOs | |
15 | ================================ | |
16 | ||
17 | Inside a GPIO driver, individual GPIOs are identified by their hardware number, | |
18 | which is a unique number between 0 and n, n being the number of GPIOs managed by | |
19 | the chip. This number is purely internal: the hardware number of a particular | |
20 | GPIO descriptor is never made visible outside of the driver. | |
21 | ||
22 | On top of this internal number, each GPIO also need to have a global number in | |
23 | the integer GPIO namespace so that it can be used with the legacy GPIO | |
24 | interface. Each chip must thus have a "base" number (which can be automatically | |
25 | assigned), and for each GPIO the global number will be (base + hardware number). | |
26 | Although the integer representation is considered deprecated, it still has many | |
27 | users and thus needs to be maintained. | |
28 | ||
29 | So for example one platform could use numbers 32-159 for GPIOs, with a | |
30 | controller defining 128 GPIOs at a "base" of 32 ; while another platform uses | |
31 | numbers 0..63 with one set of GPIO controllers, 64-79 with another type of GPIO | |
32 | controller, and on one particular board 80-95 with an FPGA. The numbers need not | |
33 | be contiguous; either of those platforms could also use numbers 2000-2063 to | |
34 | identify GPIOs in a bank of I2C GPIO expanders. | |
35 | ||
36 | ||
37 | Controller Drivers: gpio_chip | |
38 | ============================= | |
39 | ||
40 | In the gpiolib framework each GPIO controller is packaged as a "struct | |
41 | gpio_chip" (see linux/gpio/driver.h for its complete definition) with members | |
42 | common to each controller of that type: | |
43 | ||
44 | - methods to establish GPIO direction | |
45 | - methods used to access GPIO values | |
46 | - method to return the IRQ number associated to a given GPIO | |
47 | - flag saying whether calls to its methods may sleep | |
48 | - optional debugfs dump method (showing extra state like pullup config) | |
49 | - optional base number (will be automatically assigned if omitted) | |
50 | - label for diagnostics and GPIOs mapping using platform data | |
51 | ||
52 | The code implementing a gpio_chip should support multiple instances of the | |
53 | controller, possibly using the driver model. That code will configure each | |
54 | gpio_chip and issue gpiochip_add(). Removing a GPIO controller should be rare; | |
55 | use gpiochip_remove() when it is unavoidable. | |
56 | ||
57 | Most often a gpio_chip is part of an instance-specific structure with state not | |
58 | exposed by the GPIO interfaces, such as addressing, power management, and more. | |
59 | Chips such as codecs will have complex non-GPIO state. | |
60 | ||
61 | Any debugfs dump method should normally ignore signals which haven't been | |
62 | requested as GPIOs. They can use gpiochip_is_requested(), which returns either | |
63 | NULL or the label associated with that GPIO when it was requested. | |
64 | ||
99adc059 LW |
65 | |
66 | GPIO drivers providing IRQs | |
67 | --------------------------- | |
68 | It is custom that GPIO drivers (GPIO chips) are also providing interrupts, | |
69 | most often cascaded off a parent interrupt controller, and in some special | |
70 | cases the GPIO logic is melded with a SoC's primary interrupt controller. | |
71 | ||
72 | The IRQ portions of the GPIO block are implemented using an irqchip, using | |
73 | the header <linux/irq.h>. So basically such a driver is utilizing two sub- | |
74 | systems simultaneously: gpio and irq. | |
75 | ||
90887db8 LW |
76 | GPIO irqchips usually fall in one of two categories: |
77 | ||
78 | * CHAINED GPIO irqchips: these are usually the type that is embedded on | |
79 | an SoC. This means that there is a fast IRQ handler for the GPIOs that | |
80 | gets called in a chain from the parent IRQ handler, most typically the | |
81 | system interrupt controller. This means the GPIO irqchip is registered | |
82 | using irq_set_chained_handler() or the corresponding | |
83 | gpiochip_set_chained_irqchip() helper function, and the GPIO irqchip | |
84 | handler will be called immediately from the parent irqchip, while | |
85 | holding the IRQs disabled. The GPIO irqchip will then end up calling | |
86 | something like this sequence in its interrupt handler: | |
87 | ||
88 | static irqreturn_t tc3589x_gpio_irq(int irq, void *data) | |
89 | chained_irq_enter(...); | |
90 | generic_handle_irq(...); | |
91 | chained_irq_exit(...); | |
92 | ||
93 | Chained GPIO irqchips typically can NOT set the .can_sleep flag on | |
94 | struct gpio_chip, as everything happens directly in the callbacks. | |
95 | ||
677b2ff4 LW |
96 | NOTE: chained IRQ handlers are usually not good for real time. If you |
97 | are submitting a new driver or refactoring a driver for real time compliance, | |
98 | consider using creating a nested/threaded irqchip instead, see below. | |
99 | ||
100 | * NESTED THREADED GPIO irqchips: these are traditionally off-chip GPIO | |
101 | expanders and any other GPIO irqchip residing on the other side of a | |
102 | sleeping bus. Of course such drivers that need slow bus traffic to read | |
103 | out IRQ status and similar, traffic which may in turn incur other IRQs to | |
104 | happen, cannot be handled in a quick IRQ handler with IRQs disabled. | |
105 | ||
106 | With the introduction of real time support in the Linux kernel, also other | |
107 | GPIO irqchips are encouraged to use a nested and threaded IRQ handler. | |
108 | Doing so makes the interrupts naturally preemptible on a real time | |
109 | setup, which means the system can easily be configured for real time with | |
110 | a (usually negligable) performance loss. | |
111 | ||
112 | These drivers spawn a thread and then mask the parent IRQ line until the | |
113 | interrupt is handled by the driver. The hallmark of this driver is to call | |
114 | something like this in its interrupt handler: | |
90887db8 LW |
115 | |
116 | static irqreturn_t tc3589x_gpio_irq(int irq, void *data) | |
117 | ... | |
118 | handle_nested_irq(irq); | |
677b2ff4 LW |
119 | OR |
120 | generic_handle_irq(irq); | |
90887db8 | 121 | |
677b2ff4 LW |
122 | Threaded GPIO irqchips should set the .can_sleep flag on struct gpio_chip |
123 | to true if they are e.g. accessing the chip over I2C or SPI, indicating that | |
124 | this chip may sleep when accessing the GPIOs. irqchips that are just made | |
125 | threaded to be preemptible and thus real time compliant need not do this: | |
126 | preemption is not sleeping. | |
90887db8 LW |
127 | |
128 | To help out in handling the set-up and management of GPIO irqchips and the | |
129 | associated irqdomain and resource allocation callbacks, the gpiolib has | |
130 | some helpers that can be enabled by selecting the GPIOLIB_IRQCHIP Kconfig | |
131 | symbol: | |
132 | ||
133 | * gpiochip_irqchip_add(): adds an irqchip to a gpiochip. It will pass | |
134 | the struct gpio_chip* for the chip to all IRQ callbacks, so the callbacks | |
135 | need to embed the gpio_chip in its state container and obtain a pointer | |
136 | to the container using container_of(). | |
137 | (See Documentation/driver-model/design-patterns.txt) | |
138 | ||
139 | * gpiochip_set_chained_irqchip(): sets up a chained irq handler for a | |
140 | gpio_chip from a parent IRQ and passes the struct gpio_chip* as handler | |
141 | data. (Notice handler data, since the irqchip data is likely used by the | |
3f97d5fc | 142 | parent irqchip!) This is for the chained type of chip. This is also used |
677b2ff4 | 143 | to set up a threaded/nested irqchip if NULL is passed as handler. |
90887db8 LW |
144 | |
145 | To use the helpers please keep the following in mind: | |
146 | ||
147 | - Make sure to assign all relevant members of the struct gpio_chip so that | |
148 | the irqchip can initialize. E.g. .dev and .can_sleep shall be set up | |
149 | properly. | |
150 | ||
99adc059 LW |
151 | It is legal for any IRQ consumer to request an IRQ from any irqchip no matter |
152 | if that is a combined GPIO+IRQ driver. The basic premise is that gpio_chip and | |
153 | irq_chip are orthogonal, and offering their services independent of each | |
154 | other. | |
155 | ||
156 | gpiod_to_irq() is just a convenience function to figure out the IRQ for a | |
157 | certain GPIO line and should not be relied upon to have been called before | |
158 | the IRQ is used. | |
159 | ||
160 | So always prepare the hardware and make it ready for action in respective | |
161 | callbacks from the GPIO and irqchip APIs. Do not rely on gpiod_to_irq() having | |
162 | been called first. | |
163 | ||
164 | This orthogonality leads to ambiguities that we need to solve: if there is | |
165 | competition inside the subsystem which side is using the resource (a certain | |
166 | GPIO line and register for example) it needs to deny certain operations and | |
167 | keep track of usage inside of the gpiolib subsystem. This is why the API | |
168 | below exists. | |
169 | ||
170 | ||
fd8e198c AC |
171 | Locking IRQ usage |
172 | ----------------- | |
173 | Input GPIOs can be used as IRQ signals. When this happens, a driver is requested | |
174 | to mark the GPIO as being used as an IRQ: | |
175 | ||
e3a2e878 | 176 | int gpiochip_lock_as_irq(struct gpio_chip *chip, unsigned int offset) |
fd8e198c AC |
177 | |
178 | This will prevent the use of non-irq related GPIO APIs until the GPIO IRQ lock | |
179 | is released: | |
180 | ||
e3a2e878 | 181 | void gpiochip_unlock_as_irq(struct gpio_chip *chip, unsigned int offset) |
99adc059 LW |
182 | |
183 | When implementing an irqchip inside a GPIO driver, these two functions should | |
184 | typically be called in the .startup() and .shutdown() callbacks from the | |
185 | irqchip. | |
f7d4ad98 GR |
186 | |
187 | ||
677b2ff4 LW |
188 | Real-Time compliance for GPIO IRQ chips |
189 | --------------------------------------- | |
190 | ||
191 | Any provider of irqchips needs to be carefully tailored to support Real Time | |
192 | preemption. It is desireable that all irqchips in the GPIO subsystem keep this | |
193 | in mind and does the proper testing to assure they are real time-enabled. The | |
194 | following is a checklist to follow when preparing a driver for real | |
195 | time-compliance: | |
196 | ||
197 | - Nominally use raw_spinlock_t in the IRQ context path of the IRQ handler as | |
198 | we do not want these sections to be preempted. | |
199 | ||
200 | - Do NOT use chained_irq_enter() or chained_irq_exit() in the IRQ handler, | |
201 | as we want the hotpath to be preemptible. | |
202 | ||
203 | - Instead use nested IRQs and generic handlers such as handle_bad_irq(), | |
204 | handle_level_irq() and handle_edge_irq(). Consequentally the handler | |
205 | argument of gpiochip_set_chained_irqchip() should be NULL when using the | |
206 | gpiolib irqchip helpers. | |
207 | ||
208 | - Nominally set all handlers to handle_bad_irq() in the setup call, then | |
209 | set the handler to handle_level_irq() and/or handle_edge_irq() in the irqchip | |
210 | .set_type() callback depending on what your controller supports. | |
211 | ||
212 | - If you need to use the pm_runtime_get*()/pm_runtime_put*() callbacks in some | |
213 | of the irqchip callbacks, these should be moved to the .irq_bus_lock() | |
214 | and .irq_bus_unlock() callbacks respectively, as these are the only | |
215 | slowpath callbacks on an irqchip. Create the callbacks if need be. | |
216 | ||
217 | - Test your driver with the apropriate in-kernel real time test cases for both | |
218 | level and edge IRQs. | |
219 | ||
220 | ||
f7d4ad98 GR |
221 | Requesting self-owned GPIO pins |
222 | ------------------------------- | |
223 | ||
224 | Sometimes it is useful to allow a GPIO chip driver to request its own GPIO | |
225 | descriptors through the gpiolib API. Using gpio_request() for this purpose | |
226 | does not help since it pins the module to the kernel forever (it calls | |
227 | try_module_get()). A GPIO driver can use the following functions instead | |
228 | to request and free descriptors without being pinned to the kernel forever. | |
229 | ||
abdc08a3 AC |
230 | struct gpio_desc *gpiochip_request_own_desc(struct gpio_desc *desc, |
231 | const char *label) | |
f7d4ad98 GR |
232 | |
233 | void gpiochip_free_own_desc(struct gpio_desc *desc) | |
234 | ||
235 | Descriptors requested with gpiochip_request_own_desc() must be released with | |
236 | gpiochip_free_own_desc(). | |
237 | ||
238 | These functions must be used with care since they do not affect module use | |
239 | count. Do not use the functions to request gpio descriptors not owned by the | |
240 | calling driver. |