pinctrl: SPEAr: Add plgpio driver

Most of SPEAr SoCs, which support pinctrl, can configure & use
pads as gpio. This patch adds plgpio driver for configuring
these pads as gpio.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

drivers/pinctrl/spear/Kconfig

@@ -41,4 +41,11 @@ config PINCTRL_SPEAR1340
 	depends on MACH_SPEAR1340
 	select PINCTRL_SPEAR
 
+config PINCTRL_SPEAR_PLGPIO
+	bool "SPEAr SoC PLGPIO Controller"
+	depends on GPIOLIB && PINCTRL_SPEAR
+	help
+	  Say yes here to support PLGPIO controller on ST Microelectronics SPEAr
+	  SoCs.
+
 endif

drivers/pinctrl/spear/Makefile

 # SPEAr pinmux support
 
+obj-$(CONFIG_PINCTRL_SPEAR_PLGPIO)	+= pinctrl-plgpio.o
 obj-$(CONFIG_PINCTRL_SPEAR)	+= pinctrl-spear.o
 obj-$(CONFIG_PINCTRL_SPEAR3XX)	+= pinctrl-spear3xx.o
 obj-$(CONFIG_PINCTRL_SPEAR300)	+= pinctrl-spear300.o

drivers/pinctrl/spear/pinctrl-plgpio.c

+/*
+ * SPEAr platform PLGPIO driver
+ *
+ * Copyright (C) 2012 ST Microelectronics
+ * Viresh Kumar <viresh.kumar@linaro.org>
+ *
+ * This file is licensed under the terms of the GNU General Public
+ * License version 2. This program is licensed "as is" without any
+ * warranty of any kind, whether express or implied.
+ */
+
+#include <linux/clk.h>
+#include <linux/err.h>
+#include <linux/gpio.h>
+#include <linux/io.h>
+#include <linux/irq.h>
+#include <linux/irqdomain.h>
+#include <linux/module.h>
+#include <linux/pinctrl/consumer.h>
+#include <linux/platform_device.h>
+#include <linux/pm.h>
+#include <linux/spinlock.h>
+#include <asm/mach/irq.h>
+
+#define MAX_GPIO_PER_REG		32
+#define PIN_OFFSET(pin)			(pin % MAX_GPIO_PER_REG)
+#define REG_OFFSET(base, reg, pin)	(base + reg + (pin / MAX_GPIO_PER_REG) \
+							* sizeof(int *))
+
+/*
+ * plgpio pins in all machines are not one to one mapped, bitwise with registers
+ * bits. These set of macros define register masks for which below functions
+ * (pin_to_offset and offset_to_pin) are required to be called.
+ */
+#define PTO_ENB_REG		0x001
+#define PTO_WDATA_REG		0x002
+#define PTO_DIR_REG		0x004
+#define PTO_IE_REG		0x008
+#define PTO_RDATA_REG		0x010
+#define PTO_MIS_REG		0x020
+
+struct plgpio_regs {
+	u32 enb;		/* enable register */
+	u32 wdata;		/* write data register */
+	u32 dir;		/* direction set register */
+	u32 rdata;		/* read data register */
+	u32 ie;			/* interrupt enable register */
+	u32 mis;		/* mask interrupt status register */
+	u32 eit;		/* edge interrupt type */
+};
+
+/*
+ * struct plgpio: plgpio driver specific structure
+ *
+ * lock: lock for guarding gpio registers
+ * base: base address of plgpio block
+ * irq_base: irq number of plgpio0
+ * chip: gpio framework specific chip information structure
+ * p2o: function ptr for pin to offset conversion. This is required only for
+ *	machines where mapping b/w pin and offset is not 1-to-1.
+ * o2p: function ptr for offset to pin conversion. This is required only for
+ *	machines where mapping b/w pin and offset is not 1-to-1.
+ * p2o_regs: mask of registers for which p2o and o2p are applicable
+ * regs: register offsets
+ * csave_regs: context save registers for standby/sleep/hibernate cases
+ */
+struct plgpio {
+	spinlock_t		lock;
+	void __iomem		*base;
+	struct clk		*clk;
+	unsigned		irq_base;
+	struct irq_domain	*irq_domain;
+	struct gpio_chip	chip;
+	int			(*p2o)(int pin);	/* pin_to_offset */
+	int			(*o2p)(int offset);	/* offset_to_pin */
+	u32			p2o_regs;
+	struct plgpio_regs	regs;
+#ifdef CONFIG_PM
+	struct plgpio_regs	*csave_regs;
+#endif
+};
+
+/* register manipulation inline functions */
+static inline u32 is_plgpio_set(void __iomem *base, u32 pin, u32 reg)
+{
+	u32 offset = PIN_OFFSET(pin);
+	void __iomem *reg_off = REG_OFFSET(base, reg, pin);
+	u32 val = readl_relaxed(reg_off);
+
+	return !!(val & (1 << offset));
+}
+
+static inline void plgpio_reg_set(void __iomem *base, u32 pin, u32 reg)
+{
+	u32 offset = PIN_OFFSET(pin);
+	void __iomem *reg_off = REG_OFFSET(base, reg, pin);
+	u32 val = readl_relaxed(reg_off);
+
+	writel_relaxed(val | (1 << offset), reg_off);
+}
+
+static inline void plgpio_reg_reset(void __iomem *base, u32 pin, u32 reg)
+{
+	u32 offset = PIN_OFFSET(pin);
+	void __iomem *reg_off = REG_OFFSET(base, reg, pin);
+	u32 val = readl_relaxed(reg_off);
+
+	writel_relaxed(val & ~(1 << offset), reg_off);
+}
+
+/* gpio framework specific routines */
+static int plgpio_direction_input(struct gpio_chip *chip, unsigned offset)
+{
+	struct plgpio *plgpio = container_of(chip, struct plgpio, chip);
+	unsigned long flags;
+
+	/* get correct offset for "offset" pin */
+	if (plgpio->p2o && (plgpio->p2o_regs & PTO_DIR_REG)) {
+		offset = plgpio->p2o(offset);
+		if (offset == -1)
+			return -EINVAL;
+	}
+
+	spin_lock_irqsave(&plgpio->lock, flags);
+	plgpio_reg_set(plgpio->base, offset, plgpio->regs.dir);
+	spin_unlock_irqrestore(&plgpio->lock, flags);
+
+	return 0;
+}
+
+static int plgpio_direction_output(struct gpio_chip *chip, unsigned offset,
+		int value)
+{
+	struct plgpio *plgpio = container_of(chip, struct plgpio, chip);
+	unsigned long flags;
+	unsigned dir_offset = offset, wdata_offset = offset, tmp;
+
+	/* get correct offset for "offset" pin */
+	if (plgpio->p2o && (plgpio->p2o_regs & (PTO_DIR_REG | PTO_WDATA_REG))) {
+		tmp = plgpio->p2o(offset);
+		if (tmp == -1)
+			return -EINVAL;
+
+		if (plgpio->p2o_regs & PTO_DIR_REG)
+			dir_offset = tmp;
+		if (plgpio->p2o_regs & PTO_WDATA_REG)
+			wdata_offset = tmp;
+	}
+
+	spin_lock_irqsave(&plgpio->lock, flags);
+	if (value)
+		plgpio_reg_set(plgpio->base, wdata_offset,
+				plgpio->regs.wdata);
+	else
+		plgpio_reg_reset(plgpio->base, wdata_offset,
+				plgpio->regs.wdata);
+
+	plgpio_reg_reset(plgpio->base, dir_offset, plgpio->regs.dir);
+	spin_unlock_irqrestore(&plgpio->lock, flags);
+
+	return 0;
+}
+
+static int plgpio_get_value(struct gpio_chip *chip, unsigned offset)
+{
+	struct plgpio *plgpio = container_of(chip, struct plgpio, chip);
+
+	if (offset >= chip->ngpio)
+		return -EINVAL;
+
+	/* get correct offset for "offset" pin */
+	if (plgpio->p2o && (plgpio->p2o_regs & PTO_RDATA_REG)) {
+		offset = plgpio->p2o(offset);
+		if (offset == -1)
+			return -EINVAL;
+	}
+
+	return is_plgpio_set(plgpio->base, offset, plgpio->regs.rdata);
+}
+
+static void plgpio_set_value(struct gpio_chip *chip, unsigned offset, int value)
+{
+	struct plgpio *plgpio = container_of(chip, struct plgpio, chip);
+
+	if (offset >= chip->ngpio)
+		return;
+
+	/* get correct offset for "offset" pin */
+	if (plgpio->p2o && (plgpio->p2o_regs & PTO_WDATA_REG)) {
+		offset = plgpio->p2o(offset);
+		if (offset == -1)
+			return;
+	}
+
+	if (value)
+		plgpio_reg_set(plgpio->base, offset, plgpio->regs.wdata);
+	else
+		plgpio_reg_reset(plgpio->base, offset, plgpio->regs.wdata);
+}
+
+static int plgpio_request(struct gpio_chip *chip, unsigned offset)
+{
+	struct plgpio *plgpio = container_of(chip, struct plgpio, chip);
+	int gpio = chip->base + offset;
+	unsigned long flags;
+	int ret = 0;
+
+	if (offset >= chip->ngpio)
+		return -EINVAL;
+
+	ret = pinctrl_request_gpio(gpio);
+	if (ret)
+		return ret;
+
+	if (!IS_ERR(plgpio->clk)) {
+		ret = clk_prepare_enable(plgpio->clk);
+		if (ret)
+			goto err0;
+	}
+
+	if (plgpio->regs.enb == -1)
+		return 0;
+
+	/*
+	 * put gpio in IN mode before enabling it. This make enabling gpio safe
+	 */
+	ret = plgpio_direction_input(chip, offset);
+	if (ret)
+		goto err1;
+
+	/* get correct offset for "offset" pin */
+	if (plgpio->p2o && (plgpio->p2o_regs & PTO_ENB_REG)) {
+		offset = plgpio->p2o(offset);
+		if (offset == -1) {
+			ret = -EINVAL;
+			goto err1;
+		}
+	}
+
+	spin_lock_irqsave(&plgpio->lock, flags);
+	plgpio_reg_set(plgpio->base, offset, plgpio->regs.enb);
+	spin_unlock_irqrestore(&plgpio->lock, flags);
+	return 0;
+
+err1:
+	clk_disable_unprepare(plgpio->clk);
+err0:
+	pinctrl_free_gpio(gpio);
+	return ret;
+}
+
+static void plgpio_free(struct gpio_chip *chip, unsigned offset)
+{
+	struct plgpio *plgpio = container_of(chip, struct plgpio, chip);
+	int gpio = chip->base + offset;
+	unsigned long flags;
+
+	if (offset >= chip->ngpio)
+		return;
+
+	if (plgpio->regs.enb == -1)
+		goto disable_clk;
+
+	/* get correct offset for "offset" pin */
+	if (plgpio->p2o && (plgpio->p2o_regs & PTO_ENB_REG)) {
+		offset = plgpio->p2o(offset);
+		if (offset == -1)
+			return;
+	}
+
+	spin_lock_irqsave(&plgpio->lock, flags);
+	plgpio_reg_reset(plgpio->base, offset, plgpio->regs.enb);
+	spin_unlock_irqrestore(&plgpio->lock, flags);
+
+disable_clk:
+	if (!IS_ERR(plgpio->clk))
+		clk_disable_unprepare(plgpio->clk);
+
+	pinctrl_free_gpio(gpio);
+}
+
+static int plgpio_to_irq(struct gpio_chip *chip, unsigned offset)
+{
+	struct plgpio *plgpio = container_of(chip, struct plgpio, chip);
+
+	if (plgpio->irq_base < 0)
+		return -EINVAL;
+
+	return irq_find_mapping(plgpio->irq_domain, offset);
+}
+
+/* PLGPIO IRQ */
+static void plgpio_irq_disable(struct irq_data *d)
+{
+	struct plgpio *plgpio = irq_data_get_irq_chip_data(d);
+	int offset = d->irq - plgpio->irq_base;
+	unsigned long flags;
+
+	/* get correct offset for "offset" pin */
+	if (plgpio->p2o && (plgpio->p2o_regs & PTO_IE_REG)) {
+		offset = plgpio->p2o(offset);
+		if (offset == -1)
+			return;
+	}
+
+	spin_lock_irqsave(&plgpio->lock, flags);
+	plgpio_reg_set(plgpio->base, offset, plgpio->regs.ie);
+	spin_unlock_irqrestore(&plgpio->lock, flags);
+}
+
+static void plgpio_irq_enable(struct irq_data *d)
+{
+	struct plgpio *plgpio = irq_data_get_irq_chip_data(d);
+	int offset = d->irq - plgpio->irq_base;
+	unsigned long flags;
+
+	/* get correct offset for "offset" pin */
+	if (plgpio->p2o && (plgpio->p2o_regs & PTO_IE_REG)) {
+		offset = plgpio->p2o(offset);
+		if (offset == -1)
+			return;
+	}
+
+	spin_lock_irqsave(&plgpio->lock, flags);
+	plgpio_reg_reset(plgpio->base, offset, plgpio->regs.ie);
+	spin_unlock_irqrestore(&plgpio->lock, flags);
+}
+
+static int plgpio_irq_set_type(struct irq_data *d, unsigned trigger)
+{
+	struct plgpio *plgpio = irq_data_get_irq_chip_data(d);
+	int offset = d->irq - plgpio->irq_base;
+	void __iomem *reg_off;
+	unsigned int supported_type = 0, val;
+
+	if (offset >= plgpio->chip.ngpio)
+		return -EINVAL;
+
+	if (plgpio->regs.eit == -1)
+		supported_type = IRQ_TYPE_LEVEL_HIGH;
+	else
+		supported_type = IRQ_TYPE_EDGE_RISING | IRQ_TYPE_EDGE_FALLING;
+
+	if (!(trigger & supported_type))
+		return -EINVAL;
+
+	if (plgpio->regs.eit == -1)
+		return 0;
+
+	reg_off = REG_OFFSET(plgpio->base, plgpio->regs.eit, offset);
+	val = readl_relaxed(reg_off);
+
+	offset = PIN_OFFSET(offset);
+	if (trigger & IRQ_TYPE_EDGE_RISING)
+		writel_relaxed(val | (1 << offset), reg_off);
+	else
+		writel_relaxed(val & ~(1 << offset), reg_off);
+
+	return 0;
+}
+
+static struct irq_chip plgpio_irqchip = {
+	.name		= "PLGPIO",
+	.irq_enable	= plgpio_irq_enable,
+	.irq_disable	= plgpio_irq_disable,
+	.irq_set_type	= plgpio_irq_set_type,
+};
+
+static void plgpio_irq_handler(unsigned irq, struct irq_desc *desc)
+{
+	struct plgpio *plgpio = irq_get_handler_data(irq);
+	struct irq_chip *irqchip = irq_desc_get_chip(desc);
+	int regs_count, count, pin, offset, i = 0;
+	unsigned long pending;
+
+	count = plgpio->chip.ngpio;
+	regs_count = DIV_ROUND_UP(count, MAX_GPIO_PER_REG);
+
+	chained_irq_enter(irqchip, desc);
+	/* check all plgpio MIS registers for a possible interrupt */
+	for (; i < regs_count; i++) {
+		pending = readl_relaxed(plgpio->base + plgpio->regs.mis +
+				i * sizeof(int *));
+		if (!pending)
+			continue;
+
+		/* clear interrupts */
+		writel_relaxed(~pending, plgpio->base + plgpio->regs.mis +
+				i * sizeof(int *));
+		/*
+		 * clear extra bits in last register having gpios < MAX/REG
+		 * ex: Suppose there are max 102 plgpios. then last register
+		 * must have only (102 - MAX_GPIO_PER_REG * 3) = 6 relevant bits
+		 * so, we must not take other 28 bits into consideration for
+		 * checking interrupt. so clear those bits.
+		 */
+		count = count - i * MAX_GPIO_PER_REG;
+		if (count < MAX_GPIO_PER_REG)
+			pending &= (1 << count) - 1;
+
+		for_each_set_bit(offset, &pending, MAX_GPIO_PER_REG) {
+			/* get correct pin for "offset" */
+			if (plgpio->o2p && (plgpio->p2o_regs & PTO_MIS_REG)) {
+				pin = plgpio->o2p(offset);
+				if (pin == -1)
+					continue;
+			} else
+				pin = offset;
+
+			/* get correct irq line number */
+			pin = i * MAX_GPIO_PER_REG + pin;
+			generic_handle_irq(plgpio_to_irq(&plgpio->chip, pin));
+		}
+	}
+	chained_irq_exit(irqchip, desc);
+}
+
+/*
+ * pin to offset and offset to pin converter functions
+ *
+ * In spear310 there is inconsistency among bit positions in plgpio regiseters,
+ * for different plgpio pins. For example: for pin 27, bit offset is 23, pin
+ * 28-33 are not supported, pin 95 has offset bit 95, bit 100 has offset bit 1
+ */
+static int spear310_p2o(int pin)
+{
+	int offset = pin;
+
+	if (pin <= 27)
+		offset += 4;
+	else if (pin <= 33)
+		offset = -1;
+	else if (pin <= 97)
+		offset -= 2;
+	else if (pin <= 101)
+		offset = 101 - pin;
+	else
+		offset = -1;
+
+	return offset;
+}
+
+int spear310_o2p(int offset)
+{
+	if (offset <= 3)
+		return 101 - offset;
+	else if (offset <= 31)
+		return offset - 4;
+	else
+		return offset + 2;
+}
+
+static int __devinit plgpio_probe_dt(struct platform_device *pdev,
+		struct plgpio *plgpio)
+{
+	struct device_node *np = pdev->dev.of_node;
+	int ret = -EINVAL;
+	u32 val;
+
+	if (of_machine_is_compatible("st,spear310")) {
+		plgpio->p2o = spear310_p2o;
+		plgpio->o2p = spear310_o2p;
+		plgpio->p2o_regs = PTO_WDATA_REG | PTO_DIR_REG | PTO_IE_REG |
+			PTO_RDATA_REG | PTO_MIS_REG;
+	}
+
+	if (!of_property_read_u32(np, "st-plgpio,ngpio", &val)) {
+		plgpio->chip.ngpio = val;
+	} else {
+		dev_err(&pdev->dev, "DT: Invalid ngpio field\n");
+		goto end;
+	}
+
+	if (!of_property_read_u32(np, "st-plgpio,enb-reg", &val))
+		plgpio->regs.enb = val;
+	else
+		plgpio->regs.enb = -1;
+
+	if (!of_property_read_u32(np, "st-plgpio,wdata-reg", &val)) {
+		plgpio->regs.wdata = val;
+	} else {
+		dev_err(&pdev->dev, "DT: Invalid wdata reg\n");
+		goto end;
+	}
+
+	if (!of_property_read_u32(np, "st-plgpio,dir-reg", &val)) {
+		plgpio->regs.dir = val;
+	} else {
+		dev_err(&pdev->dev, "DT: Invalid dir reg\n");
+		goto end;
+	}
+
+	if (!of_property_read_u32(np, "st-plgpio,ie-reg", &val)) {
+		plgpio->regs.ie = val;
+	} else {
+		dev_err(&pdev->dev, "DT: Invalid ie reg\n");
+		goto end;
+	}
+
+	if (!of_property_read_u32(np, "st-plgpio,rdata-reg", &val)) {
+		plgpio->regs.rdata = val;
+	} else {
+		dev_err(&pdev->dev, "DT: Invalid rdata reg\n");
+		goto end;
+	}
+
+	if (!of_property_read_u32(np, "st-plgpio,mis-reg", &val)) {
+		plgpio->regs.mis = val;
+	} else {
+		dev_err(&pdev->dev, "DT: Invalid mis reg\n");
+		goto end;
+	}
+
+	if (!of_property_read_u32(np, "st-plgpio,eit-reg", &val))
+		plgpio->regs.eit = val;
+	else
+		plgpio->regs.eit = -1;
+
+	return 0;
+
+end:
+	return ret;
+}
+static int __devinit plgpio_probe(struct platform_device *pdev)
+{
+	struct device_node *np = pdev->dev.of_node;
+	struct plgpio *plgpio;
+	struct resource *res;
+	int ret, irq, i;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res) {
+		dev_err(&pdev->dev, "invalid IORESOURCE_MEM\n");
+		return -EBUSY;
+	}
+
+	plgpio = devm_kzalloc(&pdev->dev, sizeof(*plgpio), GFP_KERNEL);
+	if (!plgpio) {
+		dev_err(&pdev->dev, "memory allocation fail\n");
+		return -ENOMEM;
+	}
+
+	plgpio->base = devm_request_and_ioremap(&pdev->dev, res);
+	if (!plgpio->base) {
+		dev_err(&pdev->dev, "request and ioremap fail\n");
+		return -ENOMEM;
+	}
+
+	ret = plgpio_probe_dt(pdev, plgpio);
+	if (ret) {
+		dev_err(&pdev->dev, "DT probe failed\n");
+		return ret;
+	}
+
+	plgpio->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(plgpio->clk))
+		dev_warn(&pdev->dev, "clk_get() failed, work without it\n");
+
+#ifdef CONFIG_PM
+	plgpio->csave_regs = devm_kzalloc(&pdev->dev,
+			sizeof(*plgpio->csave_regs) *
+			DIV_ROUND_UP(plgpio->chip.ngpio, MAX_GPIO_PER_REG),
+			GFP_KERNEL);
+	if (!plgpio->csave_regs) {
+		dev_err(&pdev->dev, "csave registers memory allocation fail\n");
+		return -ENOMEM;
+	}
+#endif
+
+	platform_set_drvdata(pdev, plgpio);
+	spin_lock_init(&plgpio->lock);
+
+	plgpio->irq_base = -1;
+	plgpio->chip.base = -1;
+	plgpio->chip.request = plgpio_request;
+	plgpio->chip.free = plgpio_free;
+	plgpio->chip.direction_input = plgpio_direction_input;
+	plgpio->chip.direction_output = plgpio_direction_output;
+	plgpio->chip.get = plgpio_get_value;
+	plgpio->chip.set = plgpio_set_value;
+	plgpio->chip.to_irq = plgpio_to_irq;
+	plgpio->chip.label = dev_name(&pdev->dev);
+	plgpio->chip.dev = &pdev->dev;
+	plgpio->chip.owner = THIS_MODULE;
+
+	ret = gpiochip_add(&plgpio->chip);
+	if (ret) {
+		dev_err(&pdev->dev, "unable to add gpio chip\n");
+		return ret;
+	}
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_info(&pdev->dev, "irqs not supported\n");
+		return 0;
+	}
+
+	plgpio->irq_base = irq_alloc_descs(-1, 0, plgpio->chip.ngpio, 0);
+	if (IS_ERR_VALUE(plgpio->irq_base)) {
+		/* we would not support irq for gpio */
+		dev_warn(&pdev->dev, "couldn't allocate irq base\n");
+		return 0;
+	}
+
+	plgpio->irq_domain = irq_domain_add_legacy(np, plgpio->chip.ngpio,
+			plgpio->irq_base, 0, &irq_domain_simple_ops, NULL);
+	if (WARN_ON(!plgpio->irq_domain)) {
+		dev_err(&pdev->dev, "irq domain init failed\n");
+		irq_free_descs(plgpio->irq_base, plgpio->chip.ngpio);
+		ret = -ENXIO;
+		goto remove_gpiochip;
+	}
+
+	irq_set_chained_handler(irq, plgpio_irq_handler);
+	for (i = 0; i < plgpio->chip.ngpio; i++) {
+		irq_set_chip_and_handler(i + plgpio->irq_base, &plgpio_irqchip,
+				handle_simple_irq);
+		set_irq_flags(i + plgpio->irq_base, IRQF_VALID);
+		irq_set_chip_data(i + plgpio->irq_base, plgpio);
+	}
+
+	irq_set_handler_data(irq, plgpio);
+	dev_info(&pdev->dev, "PLGPIO registered with IRQs\n");
+
+	return 0;
+
+remove_gpiochip:
+	dev_info(&pdev->dev, "Remove gpiochip\n");
+	if (gpiochip_remove(&plgpio->chip))
+		dev_err(&pdev->dev, "unable to remove gpiochip\n");
+
+	return ret;
+}
+
+#ifdef CONFIG_PM
+static int plgpio_suspend(struct device *dev)
+{
+	struct plgpio *plgpio = dev_get_drvdata(dev);
+	int i, reg_count = DIV_ROUND_UP(plgpio->chip.ngpio, MAX_GPIO_PER_REG);
+	void __iomem *off;
+
+	for (i = 0; i < reg_count; i++) {
+		off = plgpio->base + i * sizeof(int *);
+
+		if (plgpio->regs.enb != -1)
+			plgpio->csave_regs[i].enb =
+				readl_relaxed(plgpio->regs.enb + off);
+		if (plgpio->regs.eit != -1)
+			plgpio->csave_regs[i].eit =
+				readl_relaxed(plgpio->regs.eit + off);
+		plgpio->csave_regs[i].wdata = readl_relaxed(plgpio->regs.wdata +
+				off);
+		plgpio->csave_regs[i].dir = readl_relaxed(plgpio->regs.dir +
+				off);
+		plgpio->csave_regs[i].ie = readl_relaxed(plgpio->regs.ie + off);
+	}
+
+	return 0;
+}
+
+/*
+ * This is used to correct the values in end registers. End registers contain
+ * extra bits that might be used for other purpose in platform. So, we shouldn't
+ * overwrite these bits. This macro, reads given register again, preserves other
+ * bit values (non-plgpio bits), and retain captured value (plgpio bits).
+ */
+#define plgpio_prepare_reg(__reg, _off, _mask, _tmp)		\
+{								\
+	_tmp = readl_relaxed(plgpio->regs.__reg + _off);		\
+	_tmp &= ~_mask;						\
+	plgpio->csave_regs[i].__reg =				\
+		_tmp | (plgpio->csave_regs[i].__reg & _mask);	\
+}
+
+static int plgpio_resume(struct device *dev)
+{
+	struct plgpio *plgpio = dev_get_drvdata(dev);
+	int i, reg_count = DIV_ROUND_UP(plgpio->chip.ngpio, MAX_GPIO_PER_REG);
+	void __iomem *off;
+	u32 mask, tmp;
+
+	for (i = 0; i < reg_count; i++) {
+		off = plgpio->base + i * sizeof(int *);
+
+		if (i == reg_count - 1) {
+			mask = (1 << (plgpio->chip.ngpio - i *
+						MAX_GPIO_PER_REG)) - 1;
+
+			if (plgpio->regs.enb != -1)
+				plgpio_prepare_reg(enb, off, mask, tmp);
+
+			if (plgpio->regs.eit != -1)
+				plgpio_prepare_reg(eit, off, mask, tmp);
+
+			plgpio_prepare_reg(wdata, off, mask, tmp);
+			plgpio_prepare_reg(dir, off, mask, tmp);
+			plgpio_prepare_reg(ie, off, mask, tmp);
+		}
+
+		writel_relaxed(plgpio->csave_regs[i].wdata, plgpio->regs.wdata +
+				off);
+		writel_relaxed(plgpio->csave_regs[i].dir, plgpio->regs.dir +
+				off);
+
+		if (plgpio->regs.eit != -1)
+			writel_relaxed(plgpio->csave_regs[i].eit,
+					plgpio->regs.eit + off);
+
+		writel_relaxed(plgpio->csave_regs[i].ie, plgpio->regs.ie + off);
+
+		if (plgpio->regs.enb != -1)
+			writel_relaxed(plgpio->csave_regs[i].enb,
+					plgpio->regs.enb + off);
+	}
+
+	return 0;
+}
+#endif
+
+static SIMPLE_DEV_PM_OPS(plgpio_dev_pm_ops, plgpio_suspend, plgpio_resume);
+
+static const struct of_device_id plgpio_of_match[] = {
+	{ .compatible = "st,spear-plgpio" },
+	{}
+};
+MODULE_DEVICE_TABLE(of, plgpio_of_match);
+
+static struct platform_driver plgpio_driver = {
+	.probe = plgpio_probe,
+	.driver = {
+		.owner = THIS_MODULE,
+		.name = "spear-plgpio",
+		.pm = &plgpio_dev_pm_ops,
+		.of_match_table = of_match_ptr(plgpio_of_match),
+	},
+};
+
+static int __init plgpio_init(void)
+{
+	return platform_driver_register(&plgpio_driver);
+}
+subsys_initcall(plgpio_init);
+
+MODULE_AUTHOR("Viresh Kumar <viresh.kumar@linaro.org>");
+MODULE_DESCRIPTION("ST Microlectronics SPEAr PLGPIO driver");
+MODULE_LICENSE("GPL");