[PATCH v5 2/2] gpio: add a reusable generic gpio_chip using regmap
There are quite a lot simple GPIO controller which are using regmap to access the hardware. This driver tries to be a base to unify existing code into one place. This won't cover everything but it should be a good starting point. It does not implement its own irq_chip because there is already a generic one for regmap based devices. Instead, the irq_chip will be instantiated in the parent driver and its irq domain will be associate to this driver. For now it consists of the usual registers, like set (and an optional clear) data register, an input register and direction registers. Out-of-the-box, it supports consecutive register mappings and mappings where the registers have gaps between them with a linear mapping between GPIO offset and bit position. For weirder mappings the user can register its own .xlate(). Signed-off-by: Michael Walle --- drivers/gpio/Kconfig| 4 + drivers/gpio/Makefile | 1 + drivers/gpio/gpio-regmap.c | 352 include/linux/gpio-regmap.h | 70 +++ 4 files changed, 427 insertions(+) create mode 100644 drivers/gpio/gpio-regmap.c create mode 100644 include/linux/gpio-regmap.h diff --git a/drivers/gpio/Kconfig b/drivers/gpio/Kconfig index 7d077be10a0f..bcacd9c74aa8 100644 --- a/drivers/gpio/Kconfig +++ b/drivers/gpio/Kconfig @@ -73,6 +73,10 @@ config GPIO_GENERIC depends on HAS_IOMEM # Only for IOMEM drivers tristate +config GPIO_REGMAP + depends on REGMAP + tristate + # put drivers in the right section, in alphabetical order # This symbol is selected by both I2C and SPI expanders diff --git a/drivers/gpio/Makefile b/drivers/gpio/Makefile index 65bf3940e33c..1e4894e0bf0f 100644 --- a/drivers/gpio/Makefile +++ b/drivers/gpio/Makefile @@ -12,6 +12,7 @@ obj-$(CONFIG_GPIO_SYSFS) += gpiolib-sysfs.o obj-$(CONFIG_GPIO_ACPI)+= gpiolib-acpi.o # Device drivers. Generally keep list sorted alphabetically +obj-$(CONFIG_GPIO_REGMAP) += gpio-regmap.o obj-$(CONFIG_GPIO_GENERIC) += gpio-generic.o # directly supported by gpio-generic diff --git a/drivers/gpio/gpio-regmap.c b/drivers/gpio/gpio-regmap.c new file mode 100644 index ..5060ca865276 --- /dev/null +++ b/drivers/gpio/gpio-regmap.c @@ -0,0 +1,352 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * regmap based generic GPIO driver + * + * Copyright 2019 Michael Walle + */ + +#include +#include +#include +#include +#include + +struct gpio_regmap { + struct device *parent; + struct regmap *regmap; + struct gpio_chip gpio_chip; + + int reg_stride; + int ngpio_per_reg; + unsigned int reg_dat_base; + unsigned int reg_set_base; + unsigned int reg_clr_base; + unsigned int reg_dir_in_base; + unsigned int reg_dir_out_base; + + int (*reg_mask_xlate)(struct gpio_regmap *gpio, unsigned int base, + unsigned int offset, unsigned int *reg, + unsigned int *mask); + + void *driver_data; +}; + +static unsigned int gpio_regmap_addr(unsigned int addr) +{ + return (addr == GPIO_REGMAP_ADDR_ZERO) ? 0 : addr; +} + +/** + * gpio_regmap_simple_xlate() - translate base/offset to reg/mask + * + * Use a simple linear mapping to translate the offset to the bitmask. + */ +static int gpio_regmap_simple_xlate(struct gpio_regmap *gpio, + unsigned int base, unsigned int offset, + unsigned int *reg, unsigned int *mask) +{ + unsigned int line = offset % gpio->ngpio_per_reg; + unsigned int stride = offset / gpio->ngpio_per_reg; + + *reg = base + stride * gpio->reg_stride; + *mask = BIT(line); + + return 0; +} + +static int gpio_regmap_get(struct gpio_chip *chip, unsigned int offset) +{ + struct gpio_regmap *gpio = gpiochip_get_data(chip); + unsigned int base, val, reg, mask; + int ret; + + /* we might not have an output register if we are input only */ + if (gpio->reg_dat_base) + base = gpio_regmap_addr(gpio->reg_dat_base); + else + base = gpio_regmap_addr(gpio->reg_set_base); + + ret = gpio->reg_mask_xlate(gpio, base, offset, ®, &mask); + if (ret) + return ret; + + ret = regmap_read(gpio->regmap, reg, &val); + if (ret) + return ret; + + return (val & mask) ? 1 : 0; +} + +static void gpio_regmap_set(struct gpio_chip *chip, unsigned int offset, + int val) +{ + struct gpio_regmap *gpio = gpiochip_get_data(chip); + unsigned int base = gpio_regmap_addr(gpio->reg_set_base); + unsigned int reg, mask; + + gpio->reg_mask_xlate(gpio, base, offset, ®, &mask); + if (val) + regmap_update_bits(gpio->regmap, reg, mask, mask); + else + regmap_update_bits(gpio->regmap, reg, mask, 0); +} + +static void gpio_regma
Re: [PATCH v5 2/2] gpio: add a reusable generic gpio_chip using regmap
Am 2020-05-28 05:58, schrieb Michael Walle: There are quite a lot simple GPIO controller which are using regmap to access the hardware. This driver tries to be a base to unify existing code into one place. This won't cover everything but it should be a good starting point. It does not implement its own irq_chip because there is already a generic one for regmap based devices. Instead, the irq_chip will be instantiated in the parent driver and its irq domain will be associate to this driver. For now it consists of the usual registers, like set (and an optional clear) data register, an input register and direction registers. Out-of-the-box, it supports consecutive register mappings and mappings where the registers have gaps between them with a linear mapping between GPIO offset and bit position. For weirder mappings the user can register its own .xlate(). Signed-off-by: Michael Walle --- drivers/gpio/Kconfig| 4 + drivers/gpio/Makefile | 1 + drivers/gpio/gpio-regmap.c | 352 include/linux/gpio-regmap.h | 70 +++ 4 files changed, 427 insertions(+) create mode 100644 drivers/gpio/gpio-regmap.c create mode 100644 include/linux/gpio-regmap.h [..] --- /dev/null +++ b/include/linux/gpio-regmap.h @@ -0,0 +1,70 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ + +#ifndef _LINUX_GPIO_REGMAP_H +#define _LINUX_GPIO_REGMAP_H + +struct gpio_regmap; + +#define GPIO_REGMAP_ADDR_ZERO ((unsigned long)(-1)) +#define GPIO_REGMAP_ADDR(addr) ((addr) ? : GPIO_REGMAP_ADDR_ZERO) + +/** + * struct gpio_regmap_config - Description of a generic regmap gpio_chip. + * + * @parent:The parent device + * @regmap:The regmap used to access the registers + * given, the name of the device is used + * @label: (Optional) Descriptive name for GPIO controller. + * If not given, the name of the device is used. + * @ngpio: Number of GPIOs + * @names: (Optional) Array of names for gpios + * @reg_dat_base: (Optional) (in) register base address + * @reg_set_base: (Optional) set register base address + * @reg_clr_base: (Optional) clear register base address + * @reg_dir_in_base: (Optional) in setting register base address + * @reg_dir_out_base: (Optional) out setting register base address + * @reg_stride:(Optional) May be set if the registers (of the + * same type, dat, set, etc) are not consecutive. + * @ngpio_per_reg: Number of GPIOs per register + * @irq_domain:(Optional) IRQ domain if the controller is + * interrupt-capable + * @reg_mask_xlate: (Optional) Translates base address and GPIO + * offset to a register/bitmask pair. If not + * given the default gpio_regmap_simple_xlate() + * is used. + * + * The reg_mask_xlate translates a given base address and GPIO offset to + * register and mask pair. The base address is one of the given reg_*_base. + * + * All base addresses may have the special value GPIO_REGMAP_ADDR_ZERO + * which forces the address to the value 0. + */ +struct gpio_regmap_config { + struct device *parent; + struct regmap *regmap; + + const char *label; + int ngpio; damn.. I shouldn't send patches early in the morning. I've forgot to actually enable my GPIO driver before compiling. And of course.. it doesn't compile. const char *const *names; is missing here. So I have to send a v6 anyway. Let's wait what Linus and Bert says about the location of the gpio-regmap.h header. Sorry for the noise. -michael + + unsigned int reg_dat_base; + unsigned int reg_set_base; + unsigned int reg_clr_base; + unsigned int reg_dir_in_base; + unsigned int reg_dir_out_base; + int reg_stride; + int ngpio_per_reg; + struct irq_domain *irq_domain; + + int (*reg_mask_xlate)(struct gpio_regmap *gpio, unsigned int base, + unsigned int offset, unsigned int *reg, + unsigned int *mask); +}; + +struct gpio_regmap *gpio_regmap_register(const struct gpio_regmap_config *config); +void gpio_regmap_unregister(struct gpio_regmap *gpio); +struct gpio_regmap *devm_gpio_regmap_register(struct device *dev, + const struct gpio_regmap_config *config); +void gpio_regmap_set_drvdata(struct gpio_regmap *gpio, void *data); +void *gpio_regmap_get_drvdata(struct gpio_regmap *gpio); + +#endif /* _LINUX_GPIO_REGMAP_H */
Re: [PATCH v5 2/2] gpio: add a reusable generic gpio_chip using regmap
On Thu, May 28, 2020 at 7:04 AM Michael Walle wrote: > > There are quite a lot simple GPIO controller which are using regmap to > access the hardware. This driver tries to be a base to unify existing > code into one place. This won't cover everything but it should be a good > starting point. > > It does not implement its own irq_chip because there is already a > generic one for regmap based devices. Instead, the irq_chip will be > instantiated in the parent driver and its irq domain will be associate > to this driver. > > For now it consists of the usual registers, like set (and an optional > clear) data register, an input register and direction registers. > Out-of-the-box, it supports consecutive register mappings and mappings > where the registers have gaps between them with a linear mapping between > GPIO offset and bit position. For weirder mappings the user can register > its own .xlate(). More comments from me below. ... > # Device drivers. Generally keep list sorted alphabetically Hmm... > +obj-$(CONFIG_GPIO_REGMAP) += gpio-regmap.o > obj-$(CONFIG_GPIO_GENERIC) += gpio-generic.o ...is it? ... > + * Copyright 2019 Michael Walle 2020? ... > +#include > +#include Yes, I would like to see this as gpio/regmap.h (in gpio folder). ... > +static unsigned int gpio_regmap_addr(unsigned int addr) > +{ > + return (addr == GPIO_REGMAP_ADDR_ZERO) ? 0 : addr; I would prefer rather to have if (...) return 0; return addr; here, but any of them fine. > +} ... > +/** > + * gpio_regmap_simple_xlate() - translate base/offset to reg/mask Don't you get plenty of complains from kernel doc validation script? You forgot to describe all function parameters here. > + * > + * Use a simple linear mapping to translate the offset to the bitmask. > + */ ... > + return (val & mask) ? 1 : 0; Hmm... many (new!) GPIO drivers are using !! instead of ternary. Can we do the same here? ... > +static int gpio_regmap_get_direction(struct gpio_chip *chip, > +unsigned int offset) > +{ > + if (gpio->reg_dir_out_base) { > + base = gpio_regmap_addr(gpio->reg_dir_out_base); > + invert = 0; > + } else if (gpio->reg_dir_in_base) { > + base = gpio_regmap_addr(gpio->reg_dir_in_base); > + invert = 1; > + } else { > + return GPIO_LINE_DIRECTION_IN; Hmm... Doesn't it an erroneous case and we basically shouldn't be here? > + } > + if (!!(val & mask) ^ invert) > + return GPIO_LINE_DIRECTION_OUT; > + else Redundant 'else'. > + return GPIO_LINE_DIRECTION_IN; > +} > +static int gpio_regmap_set_direction(struct gpio_chip *chip, > +unsigned int offset, bool output) > +{ > + if (gpio->reg_dir_out_base) { > + base = gpio_regmap_addr(gpio->reg_dir_out_base); > + invert = 0; > + } else if (gpio->reg_dir_in_base) { > + base = gpio_regmap_addr(gpio->reg_dir_in_base); > + invert = 1; > + } else { > + return 0; Question as above. > + } > + if (!invert) > + val = (output) ? mask : 0; > + else > + val = (output) ? 0 : mask; Why not positive conditional? Also, too many parentheses. > + return regmap_update_bits(gpio->regmap, reg, mask, val); > +} ... > +/** > + * gpio_regmap_register() - Register a generic regmap GPIO controller > + * Extra blank line. > + * @gpio: gpio_regmap device to register > + * > + * Returns 0 on success or an errno on failure. > + */ ... > + if (!config->label) > + chip->label = dev_name(config->parent); > + else > + chip->label = config->label; Why not positive or here even ternary may look good chip->label = config->label ?: dev_name(config->parent); ... > + ret = gpiochip_irqchip_add_domain(chip, config->irq_domain); > + if (ret < 0) Does ' < 0' make sense? > + goto err_remove_gpiochip; ... > +/** > + * gpio_regmap_unregister() - Unregister a generic regmap GPIO controller > + * Extra blank line > + * @gpio: gpio_regmap device to unregister > + */ ... > +/** > + * devm_gpio_regmap_register() - resource managed gpio_regmap_register() > + * Ditto. > + * @dev: device that is registering this GPIO device > + * @gpio: gpio_regmap device to register > + * > + * Managed gpio_regmap_register(). For generic regmap GPIO device registered > by > + * this function, gpio_regmap_unregister() is automatically called on driver > + * detach. See gpio_regmap_register() for more information. > + */ ... > + gpio = gpio_regmap_register(config); > + Extra blank line. > + if (!IS_ERR(gpio)) { > + *ptr = gpio; > + devres_add(dev, ptr); > + } else { > + devres_free(ptr); >
Re: [PATCH v5 2/2] gpio: add a reusable generic gpio_chip using regmap
Am 2020-05-28 13:45, schrieb Andy Shevchenko: On Thu, May 28, 2020 at 7:04 AM Michael Walle wrote: There are quite a lot simple GPIO controller which are using regmap to access the hardware. This driver tries to be a base to unify existing code into one place. This won't cover everything but it should be a good starting point. It does not implement its own irq_chip because there is already a generic one for regmap based devices. Instead, the irq_chip will be instantiated in the parent driver and its irq domain will be associate to this driver. For now it consists of the usual registers, like set (and an optional clear) data register, an input register and direction registers. Out-of-the-box, it supports consecutive register mappings and mappings where the registers have gaps between them with a linear mapping between GPIO offset and bit position. For weirder mappings the user can register its own .xlate(). More comments from me below. Thanks for the review. ... # Device drivers. Generally keep list sorted alphabetically Hmm... +obj-$(CONFIG_GPIO_REGMAP) += gpio-regmap.o obj-$(CONFIG_GPIO_GENERIC) += gpio-generic.o ...is it? That's because gpio-regmap.o seems not be a driver and more of a generic thing (like gpio-generic.o) and gpio-generic.o has another rule two lines below and I don't want to put gpio-regmap.o in between. ... + * Copyright 2019 Michael Walle 2020? ... +#include +#include Yes, I would like to see this as gpio/regmap.h (in gpio folder). ... +static unsigned int gpio_regmap_addr(unsigned int addr) +{ + return (addr == GPIO_REGMAP_ADDR_ZERO) ? 0 : addr; I would prefer rather to have if (...) return 0; return addr; here, but any of them fine. yes looks nicer. +} ... +/** + * gpio_regmap_simple_xlate() - translate base/offset to reg/mask Don't you get plenty of complains from kernel doc validation script? now that I know there is one, yes. there are many complains. You forgot to describe all function parameters here. + * + * Use a simple linear mapping to translate the offset to the bitmask. + */ This is a leftover, I'm actually gonna remove it since its not exported anymore. ... + return (val & mask) ? 1 : 0; Hmm... many (new!) GPIO drivers are using !! instead of ternary. Can we do the same here? ok ... +static int gpio_regmap_get_direction(struct gpio_chip *chip, +unsigned int offset) +{ + if (gpio->reg_dir_out_base) { + base = gpio_regmap_addr(gpio->reg_dir_out_base); + invert = 0; + } else if (gpio->reg_dir_in_base) { + base = gpio_regmap_addr(gpio->reg_dir_in_base); + invert = 1; + } else { + return GPIO_LINE_DIRECTION_IN; Hmm... Doesn't it an erroneous case and we basically shouldn't be here? yeah, I'll return -EOPNOTSUPP. Better than just ignoring, right? + } + if (!!(val & mask) ^ invert) + return GPIO_LINE_DIRECTION_OUT; + else Redundant 'else'. IMHO, That looks really strange. like it has nothing to do with the if statement. I'd like to keep that one. + return GPIO_LINE_DIRECTION_IN; +} +static int gpio_regmap_set_direction(struct gpio_chip *chip, +unsigned int offset, bool output) +{ + if (gpio->reg_dir_out_base) { + base = gpio_regmap_addr(gpio->reg_dir_out_base); + invert = 0; + } else if (gpio->reg_dir_in_base) { + base = gpio_regmap_addr(gpio->reg_dir_in_base); + invert = 1; + } else { + return 0; Question as above. same answer ;) + } + if (!invert) + val = (output) ? mask : 0; + else + val = (output) ? 0 : mask; Why not positive conditional? Also, too many parentheses. ok + return regmap_update_bits(gpio->regmap, reg, mask, val); +} ... +/** + * gpio_regmap_register() - Register a generic regmap GPIO controller + * Extra blank line. didn't know that. so ok for all kind of these comments. + * @gpio: gpio_regmap device to register + * + * Returns 0 on success or an errno on failure. + */ ... + if (!config->label) + chip->label = dev_name(config->parent); + else + chip->label = config->label; Why not positive or here even ternary may look good chip->label = config->label ?: dev_name(config->parent); ok ... + ret = gpiochip_irqchip_add_domain(chip, config->irq_domain); + if (ret < 0) Does ' < 0' make sense? more or less, I'll change it to "if (ret)" + goto err_remove_gpiochip; ... +/** + * gpio_regmap_unregister() - Unregister a generic regmap GPIO controller + * Extra blank line + * @gpio: gpio_regmap device to unregister + */
Re: [PATCH v5 2/2] gpio: add a reusable generic gpio_chip using regmap
On Thu, May 28, 2020 at 4:00 PM Michael Walle wrote: > Am 2020-05-28 13:45, schrieb Andy Shevchenko: > > On Thu, May 28, 2020 at 7:04 AM Michael Walle wrote: > > More comments from me below. > > Thanks for the review. You are welcome! Thanks for doing this actually. (So, the not commented points I think you agreed with) ... > >> # Device drivers. Generally keep list sorted alphabetically > > > > Hmm... > > > >> +obj-$(CONFIG_GPIO_REGMAP) += gpio-regmap.o > >> obj-$(CONFIG_GPIO_GENERIC) += gpio-generic.o > > > > ...is it? > > That's because gpio-regmap.o seems not be a driver and more of a generic > thing (like gpio-generic.o) and gpio-generic.o has another rule two > lines > below and I don't want to put gpio-regmap.o in between. OK! ... > >> + if (gpio->reg_dir_out_base) { > >> + base = gpio_regmap_addr(gpio->reg_dir_out_base); > >> + invert = 0; > >> + } else if (gpio->reg_dir_in_base) { > >> + base = gpio_regmap_addr(gpio->reg_dir_in_base); > >> + invert = 1; > >> + } else { > > > >> + return GPIO_LINE_DIRECTION_IN; > > > > Hmm... Doesn't it an erroneous case and we basically shouldn't be here? > > yeah, I'll return -EOPNOTSUPP. Better than just ignoring, right? Yes, that's what I meant. ... > >> + if (!!(val & mask) ^ invert) > >> + return GPIO_LINE_DIRECTION_OUT; > > > >> + else > > > > Redundant 'else'. > > IMHO, That looks really strange. like it has nothing to do with the > if statement. I'd like to keep that one. We have many drivers done like that, but it's minor, so, up to you and maintainers. -- With Best Regards, Andy Shevchenko