Skip to contentSkip to conversation

dt-bindings: watchdog: sama5d4_wdt: add microchip,sam9x60-wdt compatible

PATCH In review [email protected] 3 participants 8 messages Oct 21, 2019 – Nov 7, 2019

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH v2 1/2 dt-bindings: watchdog: sama5d4_wdt: add microchip,sam9x60-wdt compatible

From: Eugen Hristev <[email protected]>

The Atmel sama5d4_wdt needs to be compatible with microchip,sam9x60-wdt The sama5d4_wdt driver is updated to work with both hardware blocks (sama5d4/sama5d2 and sam9x60 based blocks)

---
 Documentation/devicetree/bindings/watchdog/atmel-sama5d4-wdt.txt | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
Documentation/devicetree/bindings/watchdog/atmel-sama5d4-wdt.txt +1 -1
index 4fec1e3..44727fc 100644
@@ -1,7 +1,7 @@
 * Atmel SAMA5D4 Watchdog Timer (WDT) Controller
 
 Required properties:
-- compatible: "atmel,sama5d4-wdt"
+- compatible: "atmel,sama5d4-wdt" or "microchip,sam9x60-wdt"
 - reg: base physical address and length of memory mapped region.
 
 Optional properties:
Signature · 1 line
2.7.4
#1 lore raw

PATCH v2 2/2 watchdog: sama5d4_wdt: addition of sam9x60 compatible watchdog

From: Eugen Hristev <[email protected]>

Add support for SAM9X60 WDT into sama5d4_wdt.
This means that this driver will have a platform data that will hold differences.
Added definitions of different bits.
The ops functions will differentiate between the two hardware blocks.

---

Hello,

This is a rework of the separate sam9x60 watchdog driver into a single driver that supports both hardware blocks (sam9x60 and sama5d4)
This was done as suggested on the original patches on the mailing list.

Thanks,
Eugen

 drivers/watchdog/at91sam9_wdt.h |  14 +++++
 drivers/watchdog/sama5d4_wdt.c  | 127 +++++++++++++++++++++++++++++++---------
 2 files changed, 112 insertions(+), 29 deletions(-)
drivers/watchdog/at91sam9_wdt.h +14 -0
index 390941c..7a053fd 100644
@@ -20,7 +20,10 @@
 #define AT91_WDT_MR		0x04			/* Watchdog Mode Register */
 #define		AT91_WDT_WDV		(0xfff << 0)		/* Counter Value */
 #define			AT91_WDT_SET_WDV(x)	((x) & AT91_WDT_WDV)
+#define		AT91_SAM9X60_PERIODRST	BIT(4)		/* Period Reset */
+#define		AT91_SAM9X60_RPTHRST	BIT(5)		/* Minimum Restart Period */
 #define		AT91_WDT_WDFIEN		(1     << 12)		/* Fault Interrupt Enable */
+#define		AT91_SAM9X60_WDDIS		BIT(12)		/* Disable */
 #define		AT91_WDT_WDRSTEN	(1     << 13)		/* Reset Processor */
 #define		AT91_WDT_WDRPROC	(1     << 14)		/* Timer Restart */
 #define		AT91_WDT_WDDIS		(1     << 15)		/* Watchdog Disable */
@@ -33,4 +36,15 @@
 #define		AT91_WDT_WDUNF		(1 << 0)		/* Watchdog Underflow */
 #define		AT91_WDT_WDERR		(1 << 1)		/* Watchdog Error */
 
+#define AT91_SAM9X60_VR		0x08			/* Watchdog Timer Value Register */
+
+#define AT91_SAM9X60_WLR		0x0c
+#define		AT91_SAM9X60_COUNTER	(0xfff << 0)		/* Watchdog Period Value */
+#define		AT91_SAM9X60_SET_COUNTER(x)	((x) & AT91_SAM9X60_COUNTER)
+
+#define AT91_SAM9X60_IER		0x14			/* Interrupt Enable Register */
+#define		AT91_SAM9X60_PERINT		BIT(0)		/* Period Interrupt Enable */
+#define AT91_SAM9X60_IDR		0x18			/* Interrupt Disable Register */
+#define AT91_SAM9X60_ISR		0x1c			/* Interrupt Status Register */
+
 #endif
drivers/watchdog/sama5d4_wdt.c +98 -29

231 lines not on this page · raw message

Signature · 1 line
2.7.4
maintainer
#2 lore raw

Re: [PATCH v2 2/2] watchdog: sama5d4_wdt: addition of sam9x60 compatible watchdog

Quoted 44 lines · On Mon, Oct 21, 2019 at 09:14:09AM +0000, [email protected] wrote:

From: Eugen Hristev <[email protected]>

Add support for SAM9X60 WDT into sama5d4_wdt.
This means that this driver will have a platform data that will hold differences.
Added definitions of different bits.
The ops functions will differentiate between the two hardware blocks.

---

Hello,

This is a rework of the separate sam9x60 watchdog driver into a single driver that supports both hardware blocks (sam9x60 and sama5d4)
This was done as suggested on the original patches on the mailing list.

Thanks,
Eugen

 drivers/watchdog/at91sam9_wdt.h |  14 +++++
 drivers/watchdog/sama5d4_wdt.c  | 127 +++++++++++++++++++++++++++++++---------
 2 files changed, 112 insertions(+), 29 deletions(-)
drivers/watchdog/at91sam9_wdt.h +4 -0
index 390941c..7a053fd 100644
@@ -20,7 +20,10 @@
 #define AT91_WDT_MR		0x04			/* Watchdog Mode Register */
 #define		AT91_WDT_WDV		(0xfff << 0)		/* Counter Value */
 #define			AT91_WDT_SET_WDV(x)	((x) & AT91_WDT_WDV)
+#define		AT91_SAM9X60_PERIODRST	BIT(4)		/* Period Reset */
+#define		AT91_SAM9X60_RPTHRST	BIT(5)		/* Minimum Restart Period */
 #define		AT91_WDT_WDFIEN		(1     << 12)		/* Fault Interrupt Enable */
+#define		AT91_SAM9X60_WDDIS		BIT(12)		/* Disable */
 #define		AT91_WDT_WDRSTEN	(1     << 13)		/* Reset Processor */
 #define		AT91_WDT_WDRPROC	(1     << 14)		/* Timer Restart */
 #define		AT91_WDT_WDDIS		(1     << 15)		/* Watchdog Disable */
@@ -33,4 +36,15 @@
 #define		AT91_WDT_WDUNF		(1 << 0)		/* Watchdog Underflow */
 #define		AT91_WDT_WDERR		(1 << 1)		/* Watchdog Error */
 
+#define AT91_SAM9X60_VR		0x08			/* Watchdog Timer Value Register */

drivers/watchdog/at91sam9_wdt.h line 49

Quoted 10 lines
drivers/watchdog/at91sam9_wdt.h +10 -0
+
+#define AT91_SAM9X60_WLR		0x0c
+#define		AT91_SAM9X60_COUNTER	(0xfff << 0)		/* Watchdog Period Value */
+#define		AT91_SAM9X60_SET_COUNTER(x)	((x) & AT91_SAM9X60_COUNTER)
+
+#define AT91_SAM9X60_IER		0x14			/* Interrupt Enable Register */
+#define		AT91_SAM9X60_PERINT		BIT(0)		/* Period Interrupt Enable */
+#define AT91_SAM9X60_IDR		0x18			/* Interrupt Disable Register */
+#define AT91_SAM9X60_ISR		0x1c			/* Interrupt Status Register */
+

Those tabs are getting messy, and the mix of BIT() and shift is messy too. Mind cleaning it up a bit ? Especially, two tabs after #define doesn't really add value (use two spaces), and use BIT() throughout or not at all.

Quoted 32 lines
drivers/watchdog/at91sam9_wdt.h +0 -0
 #endif
drivers/watchdog/sama5d4_wdt.c +6 -2
index d193a60..b92afd7 100644
@@ -2,7 +2,7 @@
 /*
  * Driver for Atmel SAMA5D4 Watchdog Timer
  *
- * Copyright (C) 2015 Atmel Corporation
+ * Copyright (C) 2015-2019 Microchip Technology Inc. and its subsidiaries
  */
 
 #include <linux/delay.h>
@@ -11,6 +11,7 @@
 #include <linux/kernel.h>
 #include <linux/module.h>
 #include <linux/of.h>
+#include <linux/of_device.h>
 #include <linux/of_irq.h>
 #include <linux/platform_device.h>
 #include <linux/reboot.h>
@@ -25,11 +26,18 @@
 
 #define WDT_SEC2TICKS(s)	((s) ? (((s) << 8) - 1) : 0)
 
+struct sama5d4_wdt_data {
+	const unsigned int		sam9x60_support;
+};
+
 struct sama5d4_wdt {
-	struct watchdog_device	wdd;

drivers/watchdog/sama5d4_wdt.c line 40

Quoted 10 lines
drivers/watchdog/sama5d4_wdt.c +7 -3
-	void __iomem		*reg_base;
-	u32			mr;
-	unsigned long		last_ping;
+	struct watchdog_device		wdd;
+	const struct sama5d4_wdt_data	*data;
+	void __iomem			*reg_base;
+	u32				mr;
+	u32				ir;
+	unsigned long			last_ping;
+	unsigned int			need_irq:1;

This can be a bool. Making it a bit just adds complexity to the code.

Quoted 156 lines
drivers/watchdog/sama5d4_wdt.c +64 -16
 };
 
 static int wdt_timeout;
@@ -78,7 +86,12 @@ static int sama5d4_wdt_start(struct watchdog_device *wdd)
 {
 	struct sama5d4_wdt *wdt = watchdog_get_drvdata(wdd);
 
-	wdt->mr &= ~AT91_WDT_WDDIS;
+	if (wdt->data->sam9x60_support) {
+		writel_relaxed(wdt->ir, wdt->reg_base + AT91_SAM9X60_IER);
+		wdt->mr &= ~AT91_SAM9X60_WDDIS;
+	} else {
+		wdt->mr &= ~AT91_WDT_WDDIS;
+	}
 	wdt_write(wdt, AT91_WDT_MR, wdt->mr);
 
 	return 0;
@@ -88,7 +101,12 @@ static int sama5d4_wdt_stop(struct watchdog_device *wdd)
 {
 	struct sama5d4_wdt *wdt = watchdog_get_drvdata(wdd);
 
-	wdt->mr |= AT91_WDT_WDDIS;
+	if (wdt->data->sam9x60_support) {
+		writel_relaxed(wdt->ir, wdt->reg_base + AT91_SAM9X60_IDR);
+		wdt->mr |= AT91_SAM9X60_WDDIS;
+	} else {
+		wdt->mr |= AT91_WDT_WDDIS;
+	}
 	wdt_write(wdt, AT91_WDT_MR, wdt->mr);
 
 	return 0;
@@ -109,6 +127,14 @@ static int sama5d4_wdt_set_timeout(struct watchdog_device *wdd,
 	struct sama5d4_wdt *wdt = watchdog_get_drvdata(wdd);
 	u32 value = WDT_SEC2TICKS(timeout);
 
+	if (wdt->data->sam9x60_support) {
+		wdt_write(wdt, AT91_SAM9X60_WLR,
+			  AT91_SAM9X60_SET_COUNTER(value));
+
+		wdd->timeout = timeout;
+		return 0;
+	}
+
 	wdt->mr &= ~AT91_WDT_WDV;
 	wdt->mr |= AT91_WDT_SET_WDV(value);
 
@@ -143,8 +169,14 @@ static const struct watchdog_ops sama5d4_wdt_ops = {
 static irqreturn_t sama5d4_wdt_irq_handler(int irq, void *dev_id)
 {
 	struct sama5d4_wdt *wdt = platform_get_drvdata(dev_id);
+	u32 reg;
 
-	if (wdt_read(wdt, AT91_WDT_SR)) {
+	if (wdt->data->sam9x60_support)
+		reg = wdt_read(wdt, AT91_SAM9X60_ISR);
+	else
+		reg = wdt_read(wdt, AT91_WDT_SR);
+
+	if (reg) {
 		pr_crit("Atmel Watchdog Software Reset\n");
 		emergency_restart();
 		pr_crit("Reboot didn't succeed\n");
@@ -157,13 +189,14 @@ static int of_sama5d4_wdt_init(struct device_node *np, struct sama5d4_wdt *wdt)
 {
 	const char *tmp;
 
-	wdt->mr = AT91_WDT_WDDIS;
+	if (wdt->data->sam9x60_support)
+		wdt->mr = AT91_SAM9X60_WDDIS;
+	else
+		wdt->mr = AT91_WDT_WDDIS;
 
 	if (!of_property_read_string(np, "atmel,watchdog-type", &tmp) &&
 	    !strcmp(tmp, "software"))
-		wdt->mr |= AT91_WDT_WDFIEN;
-	else
-		wdt->mr |= AT91_WDT_WDRSTEN;
+		wdt->need_irq = 1;
 
 	if (of_property_read_bool(np, "atmel,idle-halt"))
 		wdt->mr |= AT91_WDT_WDIDLEHLT;
@@ -176,21 +209,46 @@ static int of_sama5d4_wdt_init(struct device_node *np, struct sama5d4_wdt *wdt)
 
 static int sama5d4_wdt_init(struct sama5d4_wdt *wdt)
 {
-	u32 reg;
+	u32 reg, val;
+
+	val = WDT_SEC2TICKS(WDT_DEFAULT_TIMEOUT);
 	/*
 	 * When booting and resuming, the bootloader may have changed the
 	 * watchdog configuration.
 	 * If the watchdog is already running, we can safely update it.
 	 * Else, we have to disable it properly.
 	 */
-	if (wdt_enabled) {
-		wdt_write_nosleep(wdt, AT91_WDT_MR, wdt->mr);
-	} else {
+	if (!wdt_enabled) {
 		reg = wdt_read(wdt, AT91_WDT_MR);
-		if (!(reg & AT91_WDT_WDDIS))
+		if (wdt->data->sam9x60_support && (!(reg & AT91_SAM9X60_WDDIS)))
+			wdt_write_nosleep(wdt, AT91_WDT_MR,
+					  reg | AT91_SAM9X60_WDDIS);
+		else if (!wdt->data->sam9x60_support &&
+			 (!(reg & AT91_WDT_WDDIS)))
 			wdt_write_nosleep(wdt, AT91_WDT_MR,
 					  reg | AT91_WDT_WDDIS);
 	}
+
+	if (wdt->data->sam9x60_support) {
+		if (wdt->need_irq)
+			wdt->ir = AT91_SAM9X60_PERINT;
+		else
+			wdt->mr |= AT91_SAM9X60_PERIODRST;
+
+		wdt_write(wdt, AT91_SAM9X60_IER, wdt->ir);
+		wdt_write(wdt, AT91_SAM9X60_WLR, AT91_SAM9X60_SET_COUNTER(val));
+	} else {
+		wdt->mr |= AT91_WDT_SET_WDD(WDT_SEC2TICKS(MAX_WDT_TIMEOUT));
+		wdt->mr |= AT91_WDT_SET_WDV(val);
+
+		if (wdt->need_irq)
+			wdt->mr |= AT91_WDT_WDFIEN;
+		else
+			wdt->mr |= AT91_WDT_WDRSTEN;
+	}
+
+	wdt_write_nosleep(wdt, AT91_WDT_MR, wdt->mr);
+
 	return 0;
 }
 
@@ -201,13 +259,14 @@ static int sama5d4_wdt_probe(struct platform_device *pdev)
 	struct sama5d4_wdt *wdt;
 	void __iomem *regs;
 	u32 irq = 0;
-	u32 timeout;
 	int ret;
 
 	wdt = devm_kzalloc(dev, sizeof(*wdt), GFP_KERNEL);
 	if (!wdt)
 		return -ENOMEM;
 
+	wdt->data = of_device_get_match_data(&pdev->dev);
+
 	wdd = &wdt->wdd;
 	wdd->timeout = WDT_DEFAULT_TIMEOUT;
 	wdd->info = &sama5d4_wdt_info;
@@ -224,15 +283,17 @@ static int sama5d4_wdt_probe(struct platform_device *pdev)
 
 	wdt->reg_base = regs;
 
-	irq = irq_of_parse_and_map(dev->of_node, 0);
-	if (!irq)
-		dev_warn(dev, "failed to get IRQ from DT\n");

drivers/watchdog/sama5d4_wdt.c static int sama5d4_wdt_probe(struct platform_device *pdev) · line 293

Quoted 10 lines
drivers/watchdog/sama5d4_wdt.c +4 -2
-
 	ret = of_sama5d4_wdt_init(dev->of_node, wdt);
 	if (ret)
 		return ret;
 
-	if ((wdt->mr & AT91_WDT_WDFIEN) && irq) {
+	irq = irq_of_parse_and_map(dev->of_node, 0);
+	if (!irq) {
+		dev_warn(dev, "failed to get IRQ from DT\n");
+		wdt->need_irq = 0;

Does it make sense to ignore that ?

Quoted 19 lines
drivers/watchdog/sama5d4_wdt.c +3 -5
+	}
+
+	if (wdt->need_irq) {
 		ret = devm_request_irq(dev, irq, sama5d4_wdt_irq_handler,
 				       IRQF_SHARED | IRQF_IRQPOLL |
 				       IRQF_NO_SUSPEND, pdev->name, pdev);
@@ -244,11 +305,6 @@ static int sama5d4_wdt_probe(struct platform_device *pdev)
 
 	watchdog_init_timeout(wdd, wdt_timeout, dev);
 
-	timeout = WDT_SEC2TICKS(wdd->timeout);
-
-	wdt->mr |= AT91_WDT_SET_WDD(WDT_SEC2TICKS(MAX_WDT_TIMEOUT));
-	wdt->mr |= AT91_WDT_SET_WDV(timeout);
-
 	ret = sama5d4_wdt_init(wdt);
 	if (ret)
 		return ret;

drivers/watchdog/sama5d4_wdt.c static int sama5d4_wdt_probe(struct platform_device *pdev) · line 331

Quoted 8 lines
drivers/watchdog/sama5d4_wdt.c +5 -0
@@ -268,8 +324,21 @@ static int sama5d4_wdt_probe(struct platform_device *pdev)
 	return 0;
 }
 
+static struct sama5d4_wdt_data sama5d4_config;
+
+static struct sama5d4_wdt_data sam9x60_config = {
+	.sam9x60_support = 1,
+};

Unless there is reason to believe that there will be other configuration data, please just assign the flag value directly to .data and add a variable to struct sama5d4_wdt to access it. Please make that variable a bool.

Quoted 14 lines
drivers/watchdog/sama5d4_wdt.c +9 -1
+
 static const struct of_device_id sama5d4_wdt_of_match[] = {
-	{ .compatible = "atmel,sama5d4-wdt", },
+	{
+		.compatible = "atmel,sama5d4-wdt",
+		.data = &sama5d4_config,
+	},
+	{
+		.compatible = "microchip,sam9x60-wdt",
+		.data = &sam9x60_config,
+	},
 	{ }
 };
 MODULE_DEVICE_TABLE(of, sama5d4_wdt_of_match);
#3 lore raw

Re: [PATCH v2 2/2] watchdog: sama5d4_wdt: addition of sam9x60 compatible watchdog

Quoted 267 lines · Guenter Roeck
Quoted 44 lines · On Mon, Oct 21, 2019 at 09:14:09AM +0000, [email protected] wrote:

From: Eugen Hristev <[email protected]>

Add support for SAM9X60 WDT into sama5d4_wdt.
This means that this driver will have a platform data that will hold differences.
Added definitions of different bits.
The ops functions will differentiate between the two hardware blocks.

---

Hello,

This is a rework of the separate sam9x60 watchdog driver into a single driver that supports both hardware blocks (sam9x60 and sama5d4)
This was done as suggested on the original patches on the mailing list.

Thanks,
Eugen

  drivers/watchdog/at91sam9_wdt.h |  14 +++++
  drivers/watchdog/sama5d4_wdt.c  | 127 +++++++++++++++++++++++++++++++---------
  2 files changed, 112 insertions(+), 29 deletions(-)
drivers/watchdog/at91sam9_wdt.h +4 -0
index 390941c..7a053fd 100644
@@ -20,7 +20,10 @@
  #define AT91_WDT_MR		0x04			/* Watchdog Mode Register */
  #define		AT91_WDT_WDV		(0xfff << 0)		/* Counter Value */
  #define			AT91_WDT_SET_WDV(x)	((x) & AT91_WDT_WDV)
+#define		AT91_SAM9X60_PERIODRST	BIT(4)		/* Period Reset */
+#define		AT91_SAM9X60_RPTHRST	BIT(5)		/* Minimum Restart Period */
  #define		AT91_WDT_WDFIEN		(1     << 12)		/* Fault Interrupt Enable */
+#define		AT91_SAM9X60_WDDIS		BIT(12)		/* Disable */
  #define		AT91_WDT_WDRSTEN	(1     << 13)		/* Reset Processor */
  #define		AT91_WDT_WDRPROC	(1     << 14)		/* Timer Restart */
  #define		AT91_WDT_WDDIS		(1     << 15)		/* Watchdog Disable */
@@ -33,4 +36,15 @@
  #define		AT91_WDT_WDUNF		(1 << 0)		/* Watchdog Underflow */
  #define		AT91_WDT_WDERR		(1 << 1)		/* Watchdog Error */
 
+#define AT91_SAM9X60_VR		0x08			/* Watchdog Timer Value Register */

drivers/watchdog/at91sam9_wdt.h line 49

Quoted 10 lines
drivers/watchdog/at91sam9_wdt.h +10 -0
+
+#define AT91_SAM9X60_WLR		0x0c
+#define		AT91_SAM9X60_COUNTER	(0xfff << 0)		/* Watchdog Period Value */
+#define		AT91_SAM9X60_SET_COUNTER(x)	((x) & AT91_SAM9X60_COUNTER)
+
+#define AT91_SAM9X60_IER		0x14			/* Interrupt Enable Register */
+#define		AT91_SAM9X60_PERINT		BIT(0)		/* Period Interrupt Enable */
+#define AT91_SAM9X60_IDR		0x18			/* Interrupt Disable Register */
+#define AT91_SAM9X60_ISR		0x1c			/* Interrupt Status Register */
+

Those tabs are getting messy, and the mix of BIT() and shift is messy too. Mind cleaning it up a bit ? Especially, two tabs after #define doesn't really add value (use two spaces), and use BIT() throughout or not at all.

Quoted 32 lines
drivers/watchdog/at91sam9_wdt.h +0 -0
  #endif
drivers/watchdog/sama5d4_wdt.c +6 -2
index d193a60..b92afd7 100644
@@ -2,7 +2,7 @@
  /*
   * Driver for Atmel SAMA5D4 Watchdog Timer
   *
- * Copyright (C) 2015 Atmel Corporation
+ * Copyright (C) 2015-2019 Microchip Technology Inc. and its subsidiaries
   */
 
  #include <linux/delay.h>
@@ -11,6 +11,7 @@
  #include <linux/kernel.h>
  #include <linux/module.h>
  #include <linux/of.h>
+#include <linux/of_device.h>
  #include <linux/of_irq.h>
  #include <linux/platform_device.h>
  #include <linux/reboot.h>
@@ -25,11 +26,18 @@
 
  #define WDT_SEC2TICKS(s)	((s) ? (((s) << 8) - 1) : 0)
 
+struct sama5d4_wdt_data {
+	const unsigned int		sam9x60_support;
+};
+
  struct sama5d4_wdt {
-	struct watchdog_device	wdd;

drivers/watchdog/sama5d4_wdt.c line 40

Quoted 10 lines
drivers/watchdog/sama5d4_wdt.c +7 -3
-	void __iomem		*reg_base;
-	u32			mr;
-	unsigned long		last_ping;
+	struct watchdog_device		wdd;
+	const struct sama5d4_wdt_data	*data;
+	void __iomem			*reg_base;
+	u32				mr;
+	u32				ir;
+	unsigned long			last_ping;
+	unsigned int			need_irq:1;

This can be a bool. Making it a bit just adds complexity to the code.

Quoted 161 lines
drivers/watchdog/sama5d4_wdt.c +64 -17
  };
 
  static int wdt_timeout;
@@ -78,7 +86,12 @@ static int sama5d4_wdt_start(struct watchdog_device *wdd)
  {
  	struct sama5d4_wdt *wdt = watchdog_get_drvdata(wdd);
 
-	wdt->mr &= ~AT91_WDT_WDDIS;
+	if (wdt->data->sam9x60_support) {
+		writel_relaxed(wdt->ir, wdt->reg_base + AT91_SAM9X60_IER);
+		wdt->mr &= ~AT91_SAM9X60_WDDIS;
+	} else {
+		wdt->mr &= ~AT91_WDT_WDDIS;
+	}
  	wdt_write(wdt, AT91_WDT_MR, wdt->mr);
 
  	return 0;
@@ -88,7 +101,12 @@ static int sama5d4_wdt_stop(struct watchdog_device *wdd)
  {
  	struct sama5d4_wdt *wdt = watchdog_get_drvdata(wdd);
 
-	wdt->mr |= AT91_WDT_WDDIS;
+	if (wdt->data->sam9x60_support) {
+		writel_relaxed(wdt->ir, wdt->reg_base + AT91_SAM9X60_IDR);
+		wdt->mr |= AT91_SAM9X60_WDDIS;
+	} else {
+		wdt->mr |= AT91_WDT_WDDIS;
+	}
  	wdt_write(wdt, AT91_WDT_MR, wdt->mr);
 
  	return 0;
@@ -109,6 +127,14 @@ static int sama5d4_wdt_set_timeout(struct watchdog_device *wdd,
  	struct sama5d4_wdt *wdt = watchdog_get_drvdata(wdd);
  	u32 value = WDT_SEC2TICKS(timeout);
 
+	if (wdt->data->sam9x60_support) {
+		wdt_write(wdt, AT91_SAM9X60_WLR,
+			  AT91_SAM9X60_SET_COUNTER(value));
+
+		wdd->timeout = timeout;
+		return 0;
+	}
+
  	wdt->mr &= ~AT91_WDT_WDV;
  	wdt->mr |= AT91_WDT_SET_WDV(value);
 
@@ -143,8 +169,14 @@ static const struct watchdog_ops sama5d4_wdt_ops = {
  static irqreturn_t sama5d4_wdt_irq_handler(int irq, void *dev_id)
  {
  	struct sama5d4_wdt *wdt = platform_get_drvdata(dev_id);
+	u32 reg;
 
-	if (wdt_read(wdt, AT91_WDT_SR)) {
+	if (wdt->data->sam9x60_support)
+		reg = wdt_read(wdt, AT91_SAM9X60_ISR);
+	else
+		reg = wdt_read(wdt, AT91_WDT_SR);
+
+	if (reg) {
  		pr_crit("Atmel Watchdog Software Reset\n");
  		emergency_restart();
  		pr_crit("Reboot didn't succeed\n");
@@ -157,13 +189,14 @@ static int of_sama5d4_wdt_init(struct device_node *np, struct sama5d4_wdt *wdt)
  {
  	const char *tmp;
 
-	wdt->mr = AT91_WDT_WDDIS;
+	if (wdt->data->sam9x60_support)
+		wdt->mr = AT91_SAM9X60_WDDIS;
+	else
+		wdt->mr = AT91_WDT_WDDIS;
 
  	if (!of_property_read_string(np, "atmel,watchdog-type", &tmp) &&
  	    !strcmp(tmp, "software"))
-		wdt->mr |= AT91_WDT_WDFIEN;
-	else
-		wdt->mr |= AT91_WDT_WDRSTEN;
+		wdt->need_irq = 1;
 
  	if (of_property_read_bool(np, "atmel,idle-halt"))
  		wdt->mr |= AT91_WDT_WDIDLEHLT;
@@ -176,21 +209,46 @@ static int of_sama5d4_wdt_init(struct device_node *np, struct sama5d4_wdt *wdt)
 
  static int sama5d4_wdt_init(struct sama5d4_wdt *wdt)
  {
-	u32 reg;
+	u32 reg, val;
+
+	val = WDT_SEC2TICKS(WDT_DEFAULT_TIMEOUT);
  	/*
  	 * When booting and resuming, the bootloader may have changed the
  	 * watchdog configuration.
  	 * If the watchdog is already running, we can safely update it.
  	 * Else, we have to disable it properly.
  	 */
-	if (wdt_enabled) {
-		wdt_write_nosleep(wdt, AT91_WDT_MR, wdt->mr);
-	} else {
+	if (!wdt_enabled) {
  		reg = wdt_read(wdt, AT91_WDT_MR);
-		if (!(reg & AT91_WDT_WDDIS))
+		if (wdt->data->sam9x60_support && (!(reg & AT91_SAM9X60_WDDIS)))
+			wdt_write_nosleep(wdt, AT91_WDT_MR,
+					  reg | AT91_SAM9X60_WDDIS);
+		else if (!wdt->data->sam9x60_support &&
+			 (!(reg & AT91_WDT_WDDIS)))
  			wdt_write_nosleep(wdt, AT91_WDT_MR,
  					  reg | AT91_WDT_WDDIS);
  	}
+
+	if (wdt->data->sam9x60_support) {
+		if (wdt->need_irq)
+			wdt->ir = AT91_SAM9X60_PERINT;
+		else
+			wdt->mr |= AT91_SAM9X60_PERIODRST;
+
+		wdt_write(wdt, AT91_SAM9X60_IER, wdt->ir);
+		wdt_write(wdt, AT91_SAM9X60_WLR, AT91_SAM9X60_SET_COUNTER(val));
+	} else {
+		wdt->mr |= AT91_WDT_SET_WDD(WDT_SEC2TICKS(MAX_WDT_TIMEOUT));
+		wdt->mr |= AT91_WDT_SET_WDV(val);
+
+		if (wdt->need_irq)
+			wdt->mr |= AT91_WDT_WDFIEN;
+		else
+			wdt->mr |= AT91_WDT_WDRSTEN;
+	}
+
+	wdt_write_nosleep(wdt, AT91_WDT_MR, wdt->mr);
+
  	return 0;
  }
 
@@ -201,13 +259,14 @@ static int sama5d4_wdt_probe(struct platform_device *pdev)
  	struct sama5d4_wdt *wdt;
  	void __iomem *regs;
  	u32 irq = 0;
-	u32 timeout;
  	int ret;
 
  	wdt = devm_kzalloc(dev, sizeof(*wdt), GFP_KERNEL);
  	if (!wdt)
  		return -ENOMEM;
 
+	wdt->data = of_device_get_match_data(&pdev->dev);
+
  	wdd = &wdt->wdd;
  	wdd->timeout = WDT_DEFAULT_TIMEOUT;
  	wdd->info = &sama5d4_wdt_info;
@@ -224,15 +283,17 @@ static int sama5d4_wdt_probe(struct platform_device *pdev)
 
  	wdt->reg_base = regs;
 
-	irq = irq_of_parse_and_map(dev->of_node, 0);
-	if (!irq)
-		dev_warn(dev, "failed to get IRQ from DT\n");
-
  	ret = of_sama5d4_wdt_init(dev->of_node, wdt);
  	if (ret)
  		return ret;
 
Quoted 7 lines

drivers/watchdog/sama5d4_wdt.c static int sama5d4_wdt_probe(struct platform_device *pdev) · line 293

Quoted 5 lines
drivers/watchdog/sama5d4_wdt.c +4 -1
-	if ((wdt->mr & AT91_WDT_WDFIEN) && irq) {
+	irq = irq_of_parse_and_map(dev->of_node, 0);
+	if (!irq) {
+		dev_warn(dev, "failed to get IRQ from DT\n");
+		wdt->need_irq = 0;

Does it make sense to ignore that ?

Hi Guenter,

Can you detail what exactly is ignored ?

Quoted 23 lines
Quoted 23 lines
drivers/watchdog/sama5d4_wdt.c +3 -5
+	}
+
+	if (wdt->need_irq) {
  		ret = devm_request_irq(dev, irq, sama5d4_wdt_irq_handler,
  				       IRQF_SHARED | IRQF_IRQPOLL |
  				       IRQF_NO_SUSPEND, pdev->name, pdev);
@@ -244,11 +305,6 @@ static int sama5d4_wdt_probe(struct platform_device *pdev)
 
  	watchdog_init_timeout(wdd, wdt_timeout, dev);
 
-	timeout = WDT_SEC2TICKS(wdd->timeout);
-
-	wdt->mr |= AT91_WDT_SET_WDD(WDT_SEC2TICKS(MAX_WDT_TIMEOUT));
-	wdt->mr |= AT91_WDT_SET_WDV(timeout);
-
  	ret = sama5d4_wdt_init(wdt);
  	if (ret)
  		return ret;
@@ -268,8 +324,21 @@ static int sama5d4_wdt_probe(struct platform_device *pdev)
  	return 0;
  }
 
Quoted 10 lines

drivers/watchdog/sama5d4_wdt.c static int sama5d4_wdt_probe(struct platform_device *pdev) · line 331

Quoted 5 lines
drivers/watchdog/sama5d4_wdt.c +5 -0
+static struct sama5d4_wdt_data sama5d4_config;
+
+static struct sama5d4_wdt_data sam9x60_config = {
+	.sam9x60_support = 1,
+};

Unless there is reason to believe that there will be other configuration data, please just assign the flag value directly to .data and add a variable to struct sama5d4_wdt to access it. Please make that variable a bool.

There will be more configuration data for future products, but not at this moment. Do the change or keep it this way ?

I will do the other changes as requested.
Thanks for reviewing,

Eugen

Quoted 17 lines
Quoted 15 lines
drivers/watchdog/sama5d4_wdt.c +9 -1
+
  static const struct of_device_id sama5d4_wdt_of_match[] = {
-	{ .compatible = "atmel,sama5d4-wdt", },
+	{
+		.compatible = "atmel,sama5d4-wdt",
+		.data = &sama5d4_config,
+	},
+	{
+		.compatible = "microchip,sam9x60-wdt",
+		.data = &sam9x60_config,
+	},
  	{ }
  };
  MODULE_DEVICE_TABLE(of, sama5d4_wdt_of_match);
maintainer
#4 lore raw

Re: [PATCH v2 2/2] watchdog: sama5d4_wdt: addition of sam9x60 compatible watchdog

Quoted 10 lines · On Thu, Nov 07, 2019 at 12:51:15PM +0000, [email protected] wrote:
Quoted 9 lines

drivers/watchdog/sama5d4_wdt.c static int sama5d4_wdt_probe(struct platform_device *pdev) · line 293

Quoted 7 lines
drivers/watchdog/sama5d4_wdt.c +4 -1
-	if ((wdt->mr & AT91_WDT_WDFIEN) && irq) {
+	irq = irq_of_parse_and_map(dev->of_node, 0);
+	if (!irq) {
+		dev_warn(dev, "failed to get IRQ from DT\n");
+		wdt->need_irq = 0;

Does it make sense to ignore that ?

Quoted 4 lines

Hi Guenter,

Can you detail what exactly is ignored ?

The missing interrupt.

Quoted 11 lines
Quoted 10 lines

drivers/watchdog/sama5d4_wdt.c static int sama5d4_wdt_probe(struct platform_device *pdev) · line 331

Quoted 5 lines
drivers/watchdog/sama5d4_wdt.c +5 -0
+static struct sama5d4_wdt_data sama5d4_config;
+
+static struct sama5d4_wdt_data sam9x60_config = {
+	.sam9x60_support = 1,
+};

Unless there is reason to believe that there will be other configuration data, please just assign the flag value directly to .data and add a variable to struct sama5d4_wdt to access it. Please make that variable a bool.

Quoted 3 lines

There will be more configuration data for future products, but not at this moment. Do the change or keep it this way ?

If not as part of this series, it is better to just assign the flag directly. If there are changes coming at a later time which indeed need a structure (with more than one object in it), that structure can be added at that time.

Guenter

#5 lore raw

Re: [PATCH v2 2/2] watchdog: sama5d4_wdt: addition of sam9x60 compatible watchdog

Quoted 11 lines · Guenter Roeck
Quoted 10 lines · On Thu, Nov 07, 2019 at 12:51:15PM +0000, [email protected] wrote:
Quoted 9 lines

drivers/watchdog/sama5d4_wdt.c static int sama5d4_wdt_probe(struct platform_device *pdev) · line 293

Quoted 7 lines
drivers/watchdog/sama5d4_wdt.c +4 -1
-	if ((wdt->mr & AT91_WDT_WDFIEN) && irq) {
+	irq = irq_of_parse_and_map(dev->of_node, 0);
+	if (!irq) {
+		dev_warn(dev, "failed to get IRQ from DT\n");
+		wdt->need_irq = 0;

Does it make sense to ignore that ?

Quoted 5 lines
Quoted 4 lines

Hi Guenter,

Can you detail what exactly is ignored ?

The missing interrupt.

Hi,

The interrupt is not mandatory for the work of the watchdog. So, if it's not available, we must act accordingly: just configure watchdog to reset the CPU.
If we have an IRQ, we can then act in the 'software-style' watchdog:
trigger an IRQ first and try to allow the system to a safe reboot. So , it's not ignoring, it's acting accordingly (later on, do not enable anything irq-related )

Quoted 12 lines
Quoted 12 lines
Quoted 11 lines

drivers/watchdog/sama5d4_wdt.c static int sama5d4_wdt_probe(struct platform_device *pdev) · line 331

Quoted 6 lines
drivers/watchdog/sama5d4_wdt.c +5 -0
+static struct sama5d4_wdt_data sama5d4_config;
+
+static struct sama5d4_wdt_data sam9x60_config = {
+	.sam9x60_support = 1,
+};

Unless there is reason to believe that there will be other configuration data, please just assign the flag value directly to .data and add a variable to struct sama5d4_wdt to access it. Please make that variable a bool.

Quoted 7 lines
Quoted 3 lines

There will be more configuration data for future products, but not at this moment. Do the change or keep it this way ?

If not as part of this series, it is better to just assign the flag directly. If there are changes coming at a later time which indeed need a structure (with more than one object in it), that structure can be added at that time.

Ok, I will change it in next version.

Thanks,
Eugen

Quoted 4 lines

Guenter

maintainer Acked-by on v2 1/2
#6 lore raw
In reply to [email protected]
Quoted 12 lines · On Mon, 21 Oct 2019 09:14:05 +0000, <[email protected]> wrote:

From: Eugen Hristev <[email protected]>

The Atmel sama5d4_wdt needs to be compatible with microchip,sam9x60-wdt The sama5d4_wdt driver is updated to work with both hardware blocks (sama5d4/sama5d2 and sam9x60 based blocks)

---
 Documentation/devicetree/bindings/watchdog/atmel-sama5d4-wdt.txt | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
maintainer Reviewed-by on v2 1/2
#7 lore raw
In reply to [email protected]
Quoted 8 lines · On Mon, Oct 21, 2019 at 09:14:05AM +0000, [email protected] wrote:

From: Eugen Hristev <[email protected]>

The Atmel sama5d4_wdt needs to be compatible with microchip,sam9x60-wdt The sama5d4_wdt driver is updated to work with both hardware blocks (sama5d4/sama5d2 and sam9x60 based blocks)

Quoted 17 lines
---
 Documentation/devicetree/bindings/watchdog/atmel-sama5d4-wdt.txt | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
Documentation/devicetree/bindings/watchdog/atmel-sama5d4-wdt.txt +1 -1
index 4fec1e3..44727fc 100644
@@ -1,7 +1,7 @@
 * Atmel SAMA5D4 Watchdog Timer (WDT) Controller
 
 Required properties:
-- compatible: "atmel,sama5d4-wdt"
+- compatible: "atmel,sama5d4-wdt" or "microchip,sam9x60-wdt"
 - reg: base physical address and length of memory mapped region.
 
 Optional properties: