On Fri, Jul 08, 2016 at 12:45:59AM +0300, Andy Shevchenko wrote: > Intel Merrifield platform has a special GPIO controller to drive pads when they > are muxed in corresponding mode. > > Intel Merrifield GPIO IP is slightly different here and there in comparison to > the older Intel MID platforms. These differences include in particular the > shaked register offsets, specific support of level triggered interrupts and > wake capable sources, as well as a pinctrl which is a separate IP. > > Instead of uglifying existing driver I decide to provide a new one slightly > based on gpio-intel-mid.c. So, anyone can easily compare what changes are > happened to be here. > > Signed-off-by: Andy Shevchenko <andriy.shevchenko@xxxxxxxxxxxxxxx> Looks pretty good to me. I have a couple of comments see below. > --- > drivers/gpio/Kconfig | 7 + > drivers/gpio/Makefile | 1 + > drivers/gpio/gpio-merrifield.c | 450 +++++++++++++++++++++++++++++++++++++++++ > 3 files changed, 458 insertions(+) > create mode 100644 drivers/gpio/gpio-merrifield.c > > diff --git a/drivers/gpio/Kconfig b/drivers/gpio/Kconfig > index 3054f5f..98dd47a 100644 > --- a/drivers/gpio/Kconfig > +++ b/drivers/gpio/Kconfig > @@ -1044,6 +1044,13 @@ config GPIO_INTEL_MID > help > Say Y here to support Intel MID GPIO. > > +config GPIO_MERRIFIELD > + tristate "Intel Merrifield GPIO support" > + depends on X86_INTEL_MID > + select GPIOLIB_IRQCHIP > + help > + Say Y here to support Intel Merrifield GPIO. > + > config GPIO_ML_IOH > tristate "OKI SEMICONDUCTOR ML7213 IOH GPIO support" > select GENERIC_IRQ_CHIP > diff --git a/drivers/gpio/Makefile b/drivers/gpio/Makefile > index 6e111fc..2a035ed 100644 > --- a/drivers/gpio/Makefile > +++ b/drivers/gpio/Makefile > @@ -64,6 +64,7 @@ obj-$(CONFIG_GPIO_MAX732X) += gpio-max732x.o > obj-$(CONFIG_GPIO_MAX77620) += gpio-max77620.o > obj-$(CONFIG_GPIO_MB86S7X) += gpio-mb86s7x.o > obj-$(CONFIG_GPIO_MENZ127) += gpio-menz127.o > +obj-$(CONFIG_GPIO_MERRIFIELD) += gpio-merrifield.o > obj-$(CONFIG_GPIO_MC33880) += gpio-mc33880.o > obj-$(CONFIG_GPIO_MC9S08DZ60) += gpio-mc9s08dz60.o > obj-$(CONFIG_GPIO_MCP23S08) += gpio-mcp23s08.o > diff --git a/drivers/gpio/gpio-merrifield.c b/drivers/gpio/gpio-merrifield.c > new file mode 100644 > index 0000000..f399707 > --- /dev/null > +++ b/drivers/gpio/gpio-merrifield.c > @@ -0,0 +1,450 @@ > +/* > + * Intel Merrifield SoC GPIO driver > + * > + * Copyright (c) 2016 Intel Corporation. > + * Author: Andy Shevchenko <andriy.shevchenko@xxxxxxxxxxxxxxx> > + * > + * This program is free software; you can redistribute it and/or modify > + * it under the terms of the GNU General Public License version 2 as > + * published by the Free Software Foundation. > + */ > + > +#include <linux/bitops.h> > +#include <linux/gpio/driver.h> > +#include <linux/init.h> > +#include <linux/interrupt.h> > +#include <linux/io.h> > +#include <linux/module.h> > +#include <linux/pci.h> > +#include <linux/pinctrl/consumer.h> > +#include <linux/slab.h> slab.h really? > + > +#define GCCR 0x000 /* controller configuration */ > +#define GPLR 0x004 /* pin level r/o */ > +#define GPDR 0x01c /* pin direction */ > +#define GPSR 0x034 /* pin set w/o */ > +#define GPCR 0x04c /* pin clear w/o */ > +#define GRER 0x064 /* rising edge detect */ > +#define GFER 0x07c /* falling edge detect */ > +#define GFBR 0x094 /* glitch filter */ > +#define GIMR 0x0ac /* interrupt mask */ > +#define GISR 0x0c4 /* interrupt type */ > +#define GITR 0x300 /* input type */ > +#define GLPR 0x318 /* level input polarity */ > +#define GWMR 0x400 /* wake mask */ > +#define GWSR 0x418 /* wake source */ > +#define GSIR 0xc00 /* secure input */ > + > +/* Intel Merrifield has 192 GPIO pins */ > +#define MRFLD_NGPIO 192 > + > +struct mrfld_gpio_pinrange { > + unsigned gpio_base; > + unsigned pin_base; > + unsigned npins; unsigned int > +}; > + > +#define GPIO_PINRANGE(gstart, gend, pstart) \ > + { \ > + .gpio_base = (gstart), \ > + .pin_base = (pstart), \ > + .npins = (gend) - (gstart) + 1, \ > + } > + > +struct mrfld_gpio { > + struct gpio_chip chip; > + void __iomem *reg_base; > + raw_spinlock_t lock; > + struct device *dev; > +}; > + > +static const struct mrfld_gpio_pinrange mrfld_gpio_ranges[] = { > + GPIO_PINRANGE(0, 11, 146), > + GPIO_PINRANGE(12, 13, 144), > + GPIO_PINRANGE(14, 15, 35), > + GPIO_PINRANGE(16, 16, 164), > + GPIO_PINRANGE(17, 18, 105), > + GPIO_PINRANGE(19, 22, 101), > + GPIO_PINRANGE(23, 30, 107), > + GPIO_PINRANGE(32, 43, 67), > + GPIO_PINRANGE(44, 63, 195), > + GPIO_PINRANGE(64, 67, 140), > + GPIO_PINRANGE(68, 69, 165), > + GPIO_PINRANGE(70, 71, 65), > + GPIO_PINRANGE(72, 76, 228), > + GPIO_PINRANGE(77, 86, 37), > + GPIO_PINRANGE(87, 87, 48), > + GPIO_PINRANGE(88, 88, 47), > + GPIO_PINRANGE(89, 96, 49), > + GPIO_PINRANGE(97, 97, 34), > + GPIO_PINRANGE(102, 119, 83), > + GPIO_PINRANGE(120, 123, 79), > + GPIO_PINRANGE(124, 135, 115), > + GPIO_PINRANGE(137, 142, 158), > + GPIO_PINRANGE(154, 163, 24), > + GPIO_PINRANGE(164, 176, 215), > + GPIO_PINRANGE(177, 189, 127), > + GPIO_PINRANGE(190, 191, 178), > +}; > + > +#define MRFLD_PINCTRL_DEV_NAME "pinctrl-merrifield" No, use the name directly instead of inventing silly macros like this. > + > +static void __iomem *gpio_reg(struct gpio_chip *chip, unsigned int offset, > + unsigned int reg_type_offset) > +{ > + struct mrfld_gpio *priv = gpiochip_get_data(chip); > + u8 reg = offset / 32; > + > + return priv->reg_base + reg_type_offset + reg * 4; > +} > + > +static int mrfld_gpio_get(struct gpio_chip *chip, unsigned int offset) > +{ > + void __iomem *gplr = gpio_reg(chip, offset, GPLR); > + > + return !!(readl(gplr) & BIT(offset % 32)); > +} > + > +static void mrfld_gpio_set(struct gpio_chip *chip, unsigned int offset, > + int value) > +{ > + void __iomem *gpsr, *gpcr; > + > + if (value) { > + gpsr = gpio_reg(chip, offset, GPSR); > + writel(BIT(offset % 32), gpsr); > + } else { > + gpcr = gpio_reg(chip, offset, GPCR); > + writel(BIT(offset % 32), gpcr); > + } > +} > + > +static int mrfld_gpio_direction_input(struct gpio_chip *chip, > + unsigned int offset) > +{ > + struct mrfld_gpio *priv = gpiochip_get_data(chip); > + void __iomem *gpdr = gpio_reg(chip, offset, GPDR); > + unsigned long flags; > + u32 value; > + > + raw_spin_lock_irqsave(&priv->lock, flags); > + value = readl(gpdr); > + value &= ~BIT(offset % 32); > + writel(value, gpdr); > + raw_spin_unlock_irqrestore(&priv->lock, flags); > + > + return 0; > +} > + > +static int mrfld_gpio_direction_output(struct gpio_chip *chip, > + unsigned int offset, int value) > +{ > + struct mrfld_gpio *priv = gpiochip_get_data(chip); > + void __iomem *gpdr = gpio_reg(chip, offset, GPDR); > + unsigned long flags; > + > + mrfld_gpio_set(chip, offset, value); > + > + raw_spin_lock_irqsave(&priv->lock, flags); > + value = readl(gpdr); > + value |= BIT(offset % 32); > + writel(value, gpdr); > + raw_spin_unlock_irqrestore(&priv->lock, flags); > + > + return 0; > +} > + > +static void mrfld_irq_ack(struct irq_data *d) > +{ > + struct mrfld_gpio *priv = irq_data_get_irq_chip_data(d); > + u32 gpio = irqd_to_hwirq(d); > + void __iomem *gisr = gpio_reg(&priv->chip, gpio, GISR); > + unsigned long flags; > + > + if (gpio >= priv->chip.ngpio) Can this even happen? Ditto for other functions having similar check. > + return; > + > + raw_spin_lock_irqsave(&priv->lock, flags); > + > + writel(BIT(gpio % 32), gisr); Do you really need to take the lock for single register write? > + > + raw_spin_unlock_irqrestore(&priv->lock, flags); > +} > + > +static void mrfld_irq_unmask_mask(struct irq_data *d, bool unmask) > +{ > + struct mrfld_gpio *priv = irq_data_get_irq_chip_data(d); > + u32 gpio = irqd_to_hwirq(d); > + void __iomem *gimr = gpio_reg(&priv->chip, gpio, GIMR); > + unsigned long flags; > + u32 value; > + > + if (gpio >= priv->chip.ngpio) > + return; > + > + raw_spin_lock_irqsave(&priv->lock, flags); > + > + if (unmask) > + value = readl(gimr) | BIT(gpio % 32); > + else > + value = readl(gimr) & ~BIT(gpio % 32); > + writel(value, gimr); > + > + raw_spin_unlock_irqrestore(&priv->lock, flags); > +} > + > +static void mrfld_irq_mask(struct irq_data *d) > +{ > + mrfld_irq_unmask_mask(d, false); > +} > + > +static void mrfld_irq_unmask(struct irq_data *d) > +{ > + mrfld_irq_unmask_mask(d, true); > +} > + > +static int mrfld_irq_set_type(struct irq_data *d, unsigned int type) > +{ > + struct gpio_chip *gc = irq_data_get_irq_chip_data(d); > + struct mrfld_gpio *priv = gpiochip_get_data(gc); > + u32 gpio = irqd_to_hwirq(d); > + void __iomem *grer = gpio_reg(&priv->chip, gpio, GRER); > + void __iomem *gfer = gpio_reg(&priv->chip, gpio, GFER); > + void __iomem *gitr = gpio_reg(&priv->chip, gpio, GITR); > + void __iomem *glpr = gpio_reg(&priv->chip, gpio, GLPR); > + unsigned long flags; > + u32 value; > + > + if (gpio >= priv->chip.ngpio) > + return -EINVAL; > + > + raw_spin_lock_irqsave(&priv->lock, flags); > + > + if (type & IRQ_TYPE_EDGE_RISING) > + value = readl(grer) | BIT(gpio % 32); > + else > + value = readl(grer) & ~BIT(gpio % 32); > + writel(value, grer); > + > + if (type & IRQ_TYPE_EDGE_FALLING) > + value = readl(gfer) | BIT(gpio % 32); > + else > + value = readl(gfer) & ~BIT(gpio % 32); > + writel(value, gfer); > + > + /* > + * To prevent glitches from triggering an unintended level interrupt, > + * configure GLPR register first and then configure GITR. > + */ > + if (type & IRQ_TYPE_LEVEL_LOW) > + value = readl(glpr) | BIT(gpio % 32); > + else > + value = readl(glpr) & ~BIT(gpio % 32); > + writel(value, glpr); > + > + if (type & IRQ_TYPE_LEVEL_MASK) { > + value = readl(gitr) | BIT(gpio % 32); > + writel(value, gitr); > + > + irq_set_handler_locked(d, handle_level_irq); > + } else if (type & IRQ_TYPE_EDGE_BOTH) { > + value = readl(gitr) & ~BIT(gpio % 32); > + writel(value, gitr); > + > + irq_set_handler_locked(d, handle_edge_irq); > + } > + > + raw_spin_unlock_irqrestore(&priv->lock, flags); > + > + return 0; > +} > + > +static int mrfld_irq_set_wake(struct irq_data *d, unsigned int on) > +{ > + struct gpio_chip *gc = irq_data_get_irq_chip_data(d); > + struct mrfld_gpio *priv = gpiochip_get_data(gc); > + u32 gpio = irqd_to_hwirq(d); > + void __iomem *gwmr = gpio_reg(&priv->chip, gpio, GWMR); > + void __iomem *gwsr = gpio_reg(&priv->chip, gpio, GWSR); > + u32 value; > + > + if (gpio >= priv->chip.ngpio) > + return -EINVAL; > + > + /* Clear the existing wake status */ > + writel(BIT(gpio % 32), gwsr); > + > + if (on) > + value = readl(gwmr) | BIT(gpio % 32); > + else > + value = readl(gwmr) & ~BIT(gpio % 32); > + writel(value, gwmr); > + > + dev_dbg(priv->dev, "%sable wake for pin %u\n", on ? "en" : "dis", gpio); pin -> gpio > + return 0; > +} > + > +static struct irq_chip mrfld_irqchip = { > + .name = "gpio-merrifield", > + .irq_ack = mrfld_irq_ack, > + .irq_mask = mrfld_irq_mask, > + .irq_unmask = mrfld_irq_unmask, > + .irq_set_type = mrfld_irq_set_type, > + .irq_set_wake = mrfld_irq_set_wake, > +}; > + > +static const struct pci_device_id mrfld_gpio_ids[] = { > + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x1199) }, PCI_VDEVICE(INTEL, 0x1199), > + { } > +}; > +MODULE_DEVICE_TABLE(pci, mrfld_gpio_ids); > + > +static void mrfld_irq_handler(struct irq_desc *desc) > +{ > + struct gpio_chip *gc = irq_desc_get_handler_data(desc); > + struct mrfld_gpio *priv = gpiochip_get_data(gc); > + struct irq_data *data = irq_desc_get_irq_data(desc); > + struct irq_chip *chip = irq_data_get_irq_chip(data); > + unsigned long base, gpio; > + > + /* Check GPIO controller to check which pin triggered the interrupt */ > + for (base = 0; base < priv->chip.ngpio; base += 32) { > + void __iomem *gisr = gpio_reg(&priv->chip, base, GISR); > + void __iomem *gimr = gpio_reg(&priv->chip, base, GIMR); > + unsigned long pending, enabled; > + > + pending = readl(gisr); > + enabled = readl(gimr); > + > + /* Only interrupts that are enabled */ > + pending &= enabled; > + > + for_each_set_bit(gpio, &pending, 32) { > + unsigned int irq; > + > + irq = irq_find_mapping(gc->irqdomain, base + gpio); > + generic_handle_irq(irq); > + } > + } > + > + chip->irq_eoi(data); For chained irqs you can use chained_irq_enter()/exit() instead. > +} > + > +static void mrfld_irq_init_hw(struct mrfld_gpio *priv) > +{ > + void __iomem *reg; > + unsigned base; unsigned int Ditto for other places. > + > + for (base = 0; base < priv->chip.ngpio; base += 32) { > + /* Clear the rising-edge detect register */ > + reg = gpio_reg(&priv->chip, base, GRER); > + writel(0, reg); > + /* Clear the falling-edge detect register */ > + reg = gpio_reg(&priv->chip, base, GFER); > + writel(0, reg); > + } > +} > + > +static int mrfld_gpio_probe(struct pci_dev *pdev, const struct pci_device_id *id) > +{ > + const struct mrfld_gpio_pinrange *range; > + struct mrfld_gpio *priv; > + void __iomem *base; > + unsigned int i; > + u32 gpio_base; > + u32 irq_base; u32 gpio_base, irq_base; > + int retval; > + > + retval = pcim_enable_device(pdev); > + if (retval) > + return retval; > + > + retval = pcim_iomap_regions(pdev, BIT(1) | BIT(0), pci_name(pdev)); > + if (retval) { > + dev_err(&pdev->dev, "I/O memory mapping error\n"); > + return retval; > + } > + > + base = pcim_iomap_table(pdev)[1]; > + > + irq_base = readl(base); > + gpio_base = readl(sizeof(u32) + base); > + > + /* Release the IO mapping, since we already get the info from BAR1 */ > + pcim_iounmap_regions(pdev, BIT(1)); > + > + priv = devm_kzalloc(&pdev->dev, sizeof(*priv), GFP_KERNEL); > + if (!priv) { > + dev_err(&pdev->dev, "can't allocate chip data\n"); > + return -ENOMEM; > + } > + > + priv->dev = &pdev->dev; > + priv->reg_base = pcim_iomap_table(pdev)[0]; > + > + priv->chip.label = dev_name(&pdev->dev); > + priv->chip.parent = &pdev->dev; > + priv->chip.request = gpiochip_generic_request; > + priv->chip.free = gpiochip_generic_free; > + priv->chip.direction_input = mrfld_gpio_direction_input; > + priv->chip.direction_output = mrfld_gpio_direction_output; > + priv->chip.get = mrfld_gpio_get; > + priv->chip.set = mrfld_gpio_set; > + priv->chip.base = gpio_base; > + priv->chip.ngpio = MRFLD_NGPIO; > + priv->chip.can_sleep = false; > + > + raw_spin_lock_init(&priv->lock); > + > + pci_set_drvdata(pdev, priv); > + retval = devm_gpiochip_add_data(&pdev->dev, &priv->chip, priv); > + if (retval) { > + dev_err(&pdev->dev, "gpiochip_add error %d\n", retval); > + return retval; > + } > + > + for (i = 0; i < ARRAY_SIZE(mrfld_gpio_ranges); i++) { > + range = &mrfld_gpio_ranges[i]; > + retval = gpiochip_add_pin_range(&priv->chip, > + MRFLD_PINCTRL_DEV_NAME, > + range->gpio_base, > + range->pin_base, > + range->npins); > + if (retval) { > + dev_err(&pdev->dev, "failed to add GPIO pin range\n"); > + return retval; > + } > + } > + > + retval = gpiochip_irqchip_add(&priv->chip, > + &mrfld_irqchip, > + irq_base, > + handle_simple_irq, > + IRQ_TYPE_NONE); > + if (retval) { > + dev_err(&pdev->dev, > + "could not connect irqchip to gpiochip\n"); Should fit into 80 cols. > + return retval; > + } > + > + mrfld_irq_init_hw(priv); > + > + gpiochip_set_chained_irqchip(&priv->chip, > + &mrfld_irqchip, > + pdev->irq, > + mrfld_irq_handler); Why not spell it like: gpiochip_set_chained_irqchip(&priv->chip, &mrfld_irqchip, pdev->irq, mrfld_irq_handler); Ditto for the gpiochip_add() above. > + > + return 0; > +} Move mrfld_gpio_ids and MODULE_DEVICE_TABLE here. > + > +static struct pci_driver mrfld_gpio_driver = { const? > + .name = "gpio-merrifield", > + .id_table = mrfld_gpio_ids, > + .probe = mrfld_gpio_probe, > +}; > + > +module_pci_driver(mrfld_gpio_driver); > + > +MODULE_AUTHOR("Andy Shevchenko <andriy.shevchenko@xxxxxxxxxxxxxxx>"); > +MODULE_DESCRIPTION("Intel Merrifield SoC GPIO driver"); > +MODULE_LICENSE("GPL v2"); > -- > 2.8.1 -- To unsubscribe from this list: send the line "unsubscribe linux-gpio" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html