Skip to contentSkip to conversation

watchdog: sama5d4: make use of timeout-secs provided in devicetree

PATCH In review Marcus Folkesson 7 participants 22 messages Feb 11, 2018 – Mar 17, 2018

Subsystems (7)
lore
Share & tools
lore
#0 lore raw

PATCH v3 1/7 watchdog: sama5d4: make use of timeout-secs provided in devicetree

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

Following best practice described in Documentation/watchdog/watchdog-kernel-api.txt to make use of the parameter logic.

---

Notes:
v3:
- Use wdd->timeout instead of wdt_timeout when print out timout in probe function.

 drivers/watchdog/sama5d4_wdt.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/watchdog/sama5d4_wdt.c +3 -3
index 0ae947c3d7bc..255169916dbb 100644
@@ -33,7 +33,7 @@ struct sama5d4_wdt {
 	unsigned long		last_ping;
 };
 
-static int wdt_timeout = WDT_DEFAULT_TIMEOUT;
+static int wdt_timeout;
 static bool nowayout = WATCHDOG_NOWAYOUT;
 
 module_param(wdt_timeout, int, 0);
@@ -212,7 +212,7 @@ static int sama5d4_wdt_probe(struct platform_device *pdev)
 		return -ENOMEM;
 
 	wdd = &wdt->wdd;
-	wdd->timeout = wdt_timeout;
+	wdd->timeout = WDT_DEFAULT_TIMEOUT;
 	wdd->info = &sama5d4_wdt_info;
 	wdd->ops = &sama5d4_wdt_ops;
 	wdd->min_timeout = MIN_WDT_TIMEOUT;
@@ -273,7 +273,7 @@ static int sama5d4_wdt_probe(struct platform_device *pdev)
 	platform_set_drvdata(pdev, wdt);
 
 	dev_info(&pdev->dev, "initialized (timeout = %d sec, nowayout = %d)\n",
-		 wdt_timeout, nowayout);
+		 wdd->timeout, nowayout);
 
 	return 0;
 }
Signature · 1 line
2.15.1
#1 lore raw

PATCH v3 2/7 watchdog: sunxi: allow setting timeout in devicetree

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

By following best practice described in Documentation/watchdog/watchdog-kernel-api.txt, it also let us to set timout-sec property in devicetree.

---
 Documentation/devicetree/bindings/watchdog/sunxi-wdt.txt | 4 ++++
 drivers/watchdog/sunxi_wdt.c                             | 2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)
Documentation/devicetree/bindings/watchdog/sunxi-wdt.txt +4 -0
index 62dd5baad70e..49900e72f6b1 100644
@@ -6,9 +6,13 @@ Required properties:
                "allwinner,sun6i-a31-wdt"
 - reg : Specifies base physical address and size of the registers.
 
+Optional properties:
+- timeout-sec : Contains the watchdog timeout in seconds
+
 Example:
 
 wdt: watchdog at 1c20c90 {
 	compatible = "allwinner,sun4i-a10-wdt";
 	reg = <0x01c20c90 0x10>;
+	timeout-sec = <10>;
 };
drivers/watchdog/sunxi_wdt.c +1 -1
index 9728fa32c357..55f166bec0ca 100644
@@ -39,7 +39,7 @@
 #define DRV_VERSION		"1.0"
 
 static bool nowayout = WATCHDOG_NOWAYOUT;
-static unsigned int timeout = WDT_MAX_TIMEOUT;
+static unsigned int timeout;
 
 /*
  * This structure stores the register offsets for different variants
Signature · 1 line
2.15.1
maintainer Reviewed-by on v3 2/7
#2 lore raw

Re: [PATCH v3 2/7] watchdog: sunxi: allow setting timeout in devicetree

Quoted 7 lines · Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

By following best practice described in Documentation/watchdog/watchdog-kernel-api.txt, it also let us to set timout-sec property in devicetree.

Quoted 6 lines
---
 Documentation/devicetree/bindings/watchdog/sunxi-wdt.txt | 4 ++++
 drivers/watchdog/sunxi_wdt.c                             | 2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)
Reviewed-by
#3 lore raw

Re: [PATCH v3 2/7] watchdog: sunxi: allow setting timeout in devicetree

Quoted 8 lines · Rob Herring
Quoted 7 lines · Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

By following best practice described in Documentation/watchdog/watchdog-kernel-api.txt, it also let us to set timout-sec property in devicetree.

Quoted 8 lines
Quoted 6 lines
---
 Documentation/devicetree/bindings/watchdog/sunxi-wdt.txt | 4 ++++
 drivers/watchdog/sunxi_wdt.c                             | 2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)
#4 lore raw

PATCH v3 3/7 watchdog: sirfsoc: allow setting timeout in devicetree

In reply to Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

By following best practice described in Documentation/watchdog/watchdog-kernel-api.txt, it also let us to set timout-sec property in devicetree.

---
 Documentation/devicetree/bindings/watchdog/sirfsoc_wdt.txt | 4 ++++
 drivers/watchdog/sirfsoc_wdt.c                             | 2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)
Documentation/devicetree/bindings/watchdog/sirfsoc_wdt.txt +4 -0
index 9cbc76c89b2b..0dce5e3100b4 100644
@@ -5,10 +5,14 @@ Required properties:
 - reg: Address range of tick timer/WDT register set
 - interrupts: interrupt number to the cpu
 
+Optional properties:
+- timeout-sec : Contains the watchdog timeout in seconds
+
 Example:
 
 timer at b0020000 {
 	compatible = "sirf,prima2-tick";
 	reg = <0xb0020000 0x1000>;
 	interrupts = <0>;
+	timeout-sec = <30>;
 };
drivers/watchdog/sirfsoc_wdt.c +1 -1
index 4eea351e09b0..ac0c9d2c4aee 100644
@@ -29,7 +29,7 @@
 #define SIRFSOC_WDT_MAX_TIMEOUT		(10 * 60)	/* 10 mins */
 #define SIRFSOC_WDT_DEFAULT_TIMEOUT	30		/* 30 secs */
 
-static unsigned int timeout = SIRFSOC_WDT_DEFAULT_TIMEOUT;
+static unsigned int timeout;
 static bool nowayout = WATCHDOG_NOWAYOUT;
 
 module_param(timeout, uint, 0);
Signature · 1 line
2.15.1
maintainer
#5 lore raw

Re: [PATCH v3 3/7] watchdog: sirfsoc: allow setting timeout in devicetree

Quoted 6 lines · Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

By following best practice described in Documentation/watchdog/watchdog-kernel-api.txt, it also let us to set timout-sec property in devicetree.

typo

Quoted 32 lines
---
 Documentation/devicetree/bindings/watchdog/sirfsoc_wdt.txt | 4 ++++
 drivers/watchdog/sirfsoc_wdt.c                             | 2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)
Documentation/devicetree/bindings/watchdog/sirfsoc_wdt.txt +4 -0
index 9cbc76c89b2b..0dce5e3100b4 100644
@@ -5,10 +5,14 @@ Required properties:
 - reg: Address range of tick timer/WDT register set
 - interrupts: interrupt number to the cpu
 
+Optional properties:
+- timeout-sec : Contains the watchdog timeout in seconds
+
 Example:
 
 timer at b0020000 {
 	compatible = "sirf,prima2-tick";
 	reg = <0xb0020000 0x1000>;
 	interrupts = <0>;
+	timeout-sec = <30>;
 };
drivers/watchdog/sirfsoc_wdt.c +0 -0
index 4eea351e09b0..ac0c9d2c4aee 100644

drivers/watchdog/sirfsoc_wdt.c line 32

Quoted 5 lines
drivers/watchdog/sirfsoc_wdt.c +1 -1
@@ -29,7 +29,7 @@
 #define SIRFSOC_WDT_MAX_TIMEOUT		(10 * 60)	/* 10 mins */
 #define SIRFSOC_WDT_DEFAULT_TIMEOUT	30		/* 30 secs */
 
-static unsigned int timeout = SIRFSOC_WDT_DEFAULT_TIMEOUT;
+static unsigned int timeout;

If you have an old dtb, then you still need the default.

Quoted 6 lines
drivers/watchdog/sirfsoc_wdt.c +0 -0
 static bool nowayout = WATCHDOG_NOWAYOUT;
 
 module_param(timeout, uint, 0);
Signature · 1 line
2.15.1
maintainer
#6 lore raw

Re: [PATCH v3 3/7] watchdog: sirfsoc: allow setting timeout in devicetree

Quoted 45 lines · Rob Herring
Quoted 6 lines · Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

By following best practice described in Documentation/watchdog/watchdog-kernel-api.txt, it also let us to set timout-sec property in devicetree.

typo

Quoted 35 lines
---
  Documentation/devicetree/bindings/watchdog/sirfsoc_wdt.txt | 4 ++++
  drivers/watchdog/sirfsoc_wdt.c                             | 2 +-
  2 files changed, 5 insertions(+), 1 deletion(-)
Documentation/devicetree/bindings/watchdog/sirfsoc_wdt.txt +4 -0
index 9cbc76c89b2b..0dce5e3100b4 100644
@@ -5,10 +5,14 @@ Required properties:
  - reg: Address range of tick timer/WDT register set
  - interrupts: interrupt number to the cpu
 
+Optional properties:
+- timeout-sec : Contains the watchdog timeout in seconds
+
  Example:
 
  timer at b0020000 {
  	compatible = "sirf,prima2-tick";
  	reg = <0xb0020000 0x1000>;
  	interrupts = <0>;
+	timeout-sec = <30>;
  };
drivers/watchdog/sirfsoc_wdt.c +0 -0
index 4eea351e09b0..ac0c9d2c4aee 100644
@@ -29,7 +29,7 @@
  #define SIRFSOC_WDT_MAX_TIMEOUT		(10 * 60)	/* 10 mins */
  #define SIRFSOC_WDT_DEFAULT_TIMEOUT	30		/* 30 secs */
 
Quoted 5 lines

drivers/watchdog/sirfsoc_wdt.c line 32

Quoted 2 lines
drivers/watchdog/sirfsoc_wdt.c +1 -1
-static unsigned int timeout = SIRFSOC_WDT_DEFAULT_TIMEOUT;
+static unsigned int timeout;

If you have an old dtb, then you still need the default.

No. It is optional to start with, and the driver already has

static struct watchdog_device sirfsoc_wdd = {
.info = &sirfsoc_wdt_ident,
.ops = &sirfsoc_wdt_ops,
.timeout = SIRFSOC_WDT_DEFAULT_TIMEOUT, <------
.min_timeout = SIRFSOC_WDT_MIN_TIMEOUT,
.max_timeout = SIRFSOC_WDT_MAX_TIMEOUT,
};

Guenter

Quoted 7 lines
Quoted 6 lines
drivers/watchdog/sirfsoc_wdt.c +0 -0
  static bool nowayout = WATCHDOG_NOWAYOUT;
 
  module_param(timeout, uint, 0);
Signature · 1 line
2.15.1
maintainer Reviewed-by
#7 lore raw

Re: [PATCH v3 3/7] watchdog: sirfsoc: allow setting timeout in devicetree

Quoted 58 lines · Guenter Roeck
Quoted 45 lines · Rob Herring
Quoted 7 lines · Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

By following best practice described in Documentation/watchdog/watchdog-kernel-api.txt, it also let us to set timout-sec property in devicetree.

typo

Quoted 32 lines
---
  Documentation/devicetree/bindings/watchdog/sirfsoc_wdt.txt | 4 ++++
  drivers/watchdog/sirfsoc_wdt.c                             | 2 +-
  2 files changed, 5 insertions(+), 1 deletion(-)
Documentation/devicetree/bindings/watchdog/sirfsoc_wdt.txt +3 -0
index 9cbc76c89b2b..0dce5e3100b4 100644
@@ -5,10 +5,14 @@ Required properties:
  - reg: Address range of tick timer/WDT register set
  - interrupts: interrupt number to the cpu
  +Optional properties:
+- timeout-sec : Contains the watchdog timeout in seconds
+
  Example:
    timer at b0020000 {
        compatible = "sirf,prima2-tick";
        reg = <0xb0020000 0x1000>;
        interrupts = <0>;
+       timeout-sec = <30>;
  };
drivers/watchdog/sirfsoc_wdt.c +0 -0
index 4eea351e09b0..ac0c9d2c4aee 100644
@@ -29,7 +29,7 @@
Quoted 9 lines
Quoted 5 lines
drivers/watchdog/sirfsoc_wdt.c +0 -0
  #define SIRFSOC_WDT_MAX_TIMEOUT               (10 * 60)       /* 10

mins */
#define SIRFSOC_WDT_DEFAULT_TIMEOUT 30 /* 30 secs */
-static unsigned int timeout = SIRFSOC_WDT_DEFAULT_TIMEOUT;
+static unsigned int timeout;

If you have an old dtb, then you still need the default.

No. It is optional to start with, and the driver already has

Quoted 7 lines

static struct watchdog_device sirfsoc_wdd = {
.info = &sirfsoc_wdt_ident,
.ops = &sirfsoc_wdt_ops,
.timeout = SIRFSOC_WDT_DEFAULT_TIMEOUT, <------
.min_timeout = SIRFSOC_WDT_MIN_TIMEOUT,
.max_timeout = SIRFSOC_WDT_MAX_TIMEOUT,
};

Okay, then:

for the lot.

Rob

#8 lore raw

PATCH v3 4/7 watchdog: pnx4008: make use of timeout-secs provided in devicetree

In reply to Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

Following best practice described in Documentation/watchdog/watchdog-kernel-api.txt to make use of the parameter logic.

---
 drivers/watchdog/pnx4008_wdt.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/watchdog/pnx4008_wdt.c +1 -1
index 0529aed158a4..8e261799c84e 100644
@@ -78,7 +78,7 @@
 #define WDOG_COUNTER_RATE 13000000	/*the counter clock is 13 MHz fixed */
 
 static bool nowayout = WATCHDOG_NOWAYOUT;
-static unsigned int heartbeat = DEFAULT_HEARTBEAT;
+static unsigned int heartbeat;
 
 static DEFINE_SPINLOCK(io_lock);
 static void __iomem	*wdt_base;
Signature · 1 line
2.15.1
#9 lore raw

PATCH v3 5/7 watchdog: mtk: allow setting timeout in devicetree

In reply to Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

By following best practice described in Documentation/watchdog/watchdog-kernel-api.txt, it also let us to set timout-sec property in devicetree.

---
 Documentation/devicetree/bindings/watchdog/mtk-wdt.txt | 4 ++++
 drivers/watchdog/mtk_wdt.c                             | 2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)
Documentation/devicetree/bindings/watchdog/mtk-wdt.txt +4 -0
index 5b38a30e608c..859dee167b91 100644
@@ -11,9 +11,13 @@ Required properties:
 
 - reg : Specifies base physical address and size of the registers.
 
+Optional properties:
+- timeout-sec: contains the watchdog timeout in seconds.
+
 Example:
 
 wdt: watchdog at 10000000 {
 	compatible = "mediatek,mt6589-wdt";
 	reg = <0x10000000 0x18>;
+	timeout-sec = <10>;
 };
drivers/watchdog/mtk_wdt.c +1 -1
index 7ed417a765c7..fcdc10ec28a3 100644
@@ -57,7 +57,7 @@
 #define DRV_VERSION		"1.0"
 
 static bool nowayout = WATCHDOG_NOWAYOUT;
-static unsigned int timeout = WDT_MAX_TIMEOUT;
+static unsigned int timeout;
 
 struct mtk_wdt_dev {
 	struct watchdog_device wdt_dev;
Signature · 1 line
2.15.1
maintainer
#10 lore raw

Re: [PATCH v3 5/7] watchdog: mtk: allow setting timeout in devicetree

Quoted 6 lines · Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

By following best practice described in Documentation/watchdog/watchdog-kernel-api.txt, it also let us to set timout-sec property in devicetree.

Same problems in this one.

Quoted 7 lines
---
 Documentation/devicetree/bindings/watchdog/mtk-wdt.txt | 4 ++++
 drivers/watchdog/mtk_wdt.c                             | 2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)
maintainer
#11 lore raw

Re: [PATCH v3 5/7] watchdog: mtk: allow setting timeout in devicetree

Quoted 10 lines · Rob Herring
Quoted 6 lines · Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

By following best practice described in Documentation/watchdog/watchdog-kernel-api.txt, it also let us to set timout-sec property in devicetree.

Same problems in this one.

Same answer. In the existing driver;

         mtk_wdt->wdt_dev.timeout = WDT_MAX_TIMEOUT;

Guenter

Quoted 8 lines
Quoted 7 lines
---
  Documentation/devicetree/bindings/watchdog/mtk-wdt.txt | 4 ++++
  drivers/watchdog/mtk_wdt.c                             | 2 +-
  2 files changed, 5 insertions(+), 1 deletion(-)
maintainer Acked-by on v3 5/7
#12 lore raw

Re: [PATCH v3 5/7] watchdog: mtk: allow setting timeout in devicetree

In reply to Marcus Folkesson
Quoted 38 lines · Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

By following best practice described in Documentation/watchdog/watchdog-kernel-api.txt, it also let us to set timout-sec property in devicetree.

---
 Documentation/devicetree/bindings/watchdog/mtk-wdt.txt | 4 ++++
 drivers/watchdog/mtk_wdt.c                             | 2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)
Documentation/devicetree/bindings/watchdog/mtk-wdt.txt +4 -0
index 5b38a30e608c..859dee167b91 100644
@@ -11,9 +11,13 @@ Required properties:
 
 - reg : Specifies base physical address and size of the registers.
 
+Optional properties:
+- timeout-sec: contains the watchdog timeout in seconds.
+
 Example:
 
 wdt: watchdog at 10000000 {
 	compatible = "mediatek,mt6589-wdt";
 	reg = <0x10000000 0x18>;
+	timeout-sec = <10>;
 };
drivers/watchdog/mtk_wdt.c +0 -0
index 7ed417a765c7..fcdc10ec28a3 100644

drivers/watchdog/mtk_wdt.c line 63

Quoted 8 lines
drivers/watchdog/mtk_wdt.c +1 -1
@@ -57,7 +57,7 @@
 #define DRV_VERSION		"1.0"
 
 static bool nowayout = WATCHDOG_NOWAYOUT;
-static unsigned int timeout = WDT_MAX_TIMEOUT;
+static unsigned int timeout;
 
 struct mtk_wdt_dev {
 	struct watchdog_device wdt_dev;
#13 lore raw

PATCH v3 6/7 watchdog: meson: allow setting timeout in devicetree

In reply to Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

By following best practice described in Documentation/watchdog/watchdog-kernel-api.txt, it also let us to set timout-sec property in devicetree.

---
 Documentation/devicetree/bindings/watchdog/meson-wdt.txt | 4 ++++
 drivers/watchdog/meson_wdt.c                             | 2 +-
 2 files changed, 5 insertions(+), 1 deletion(-)
Documentation/devicetree/bindings/watchdog/meson-wdt.txt +4 -0
index 8a6d84cb36c9..7588cc3971bf 100644
@@ -9,9 +9,13 @@ Required properties:
 	"amlogic,meson8m2-wdt" and "amlogic,meson8b-wdt" on Meson8m2 SoCs
 - reg : Specifies base physical address and size of the registers.
 
+Optional properties:
+- timeout-sec: contains the watchdog timeout in seconds.
+
 Example:
 
 wdt: watchdog at c1109900 {
 	compatible = "amlogic,meson6-wdt";
 	reg = <0xc1109900 0x8>;
+	timeout-sec = <10>;
 };
drivers/watchdog/meson_wdt.c +1 -1
index 304274c67735..cd0275a6cdac 100644
@@ -36,7 +36,7 @@
 #define MESON_SEC_TO_TC(s, c)	((s) * (c))
 
 static bool nowayout = WATCHDOG_NOWAYOUT;
-static unsigned int timeout = MESON_WDT_TIMEOUT;
+static unsigned int timeout;
 
 struct meson_wdt_data {
 	unsigned int enable;
Signature · 1 line
2.15.1
#14 lore raw

PATCH v3 7/7 watchdog: coh901327: make use of timeout-secs provided in devicetree

In reply to Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

Following best practice described in Documentation/watchdog/watchdog-kernel-api.txt to make use of the parameter logic.

---

Notes:
v3:
- Reformat and use coh901327_wdt.timeout instead of margin when print out timout in probe function.

v2:
- Set .timeout in coh901327_wdt structure declaration.
- Set .min_timeout to 1 instead of 0. I could not find a datasheet for coh901327, so I'm not sure if 0 is valid. However, 0 seems wrong to me and most driver has 1 as min value. If it should be 0, please let me know and I have to set another initial value for margin.

 drivers/watchdog/coh901327_wdt.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)
drivers/watchdog/coh901327_wdt.c +8 -6
index 4410337f4f7f..500af8a7ec5a 100644
@@ -67,7 +67,9 @@
 #define U300_WDOG_IFR_WILL_BARK_IRQ_FORCE_ENABLE			0x0001U
 
 /* Default timeout in seconds = 1 minute */
-static unsigned int margin = 60;
+#define U300_WDOG_DEFAULT_TIMEOUT					60
+
+static unsigned int margin;
 static int irq;
 static void __iomem *virtbase;
 static struct device *parent;
@@ -235,8 +237,9 @@ static struct watchdog_device coh901327_wdt = {
 	 * timeout register is max
 	 * 0x7FFF = 327670ms ~= 327s.
 	 */
-	.min_timeout = 0,
+	.min_timeout = 1,
 	.max_timeout = 327,
+	.timeout = U300_WDOG_DEFAULT_TIMEOUT,
 };
 
 static int __exit coh901327_remove(struct platform_device *pdev)
@@ -315,16 +318,15 @@ static int __init coh901327_probe(struct platform_device *pdev)
 		goto out_no_irq;
 	}
 
-	ret = watchdog_init_timeout(&coh901327_wdt, margin, dev);
-	if (ret < 0)
-		coh901327_wdt.timeout = 60;
+	watchdog_init_timeout(&coh901327_wdt, margin, dev);
 
 	coh901327_wdt.parent = dev;
 	ret = watchdog_register_device(&coh901327_wdt);
 	if (ret)
 		goto out_no_wdog;
 
-	dev_info(dev, "initialized. timer margin=%d sec\n", margin);
+	dev_info(dev, "initialized. (timeout=%d sec)\n",
+			coh901327_wdt.timeout);
 	return 0;
 
 out_no_wdog:
Signature · 1 line
2.15.1
maintainer Reviewed-by on v3 7/7
#15 lore raw

Re: [v3, 7/7] watchdog: coh901327: make use of timeout-secs provided in devicetree

Quoted 8 lines · Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

Following best practice described in Documentation/watchdog/watchdog-kernel-api.txt to make use of the parameter logic.

Quoted 64 lines

---

Notes:
v3:
- Reformat and use coh901327_wdt.timeout instead of margin when print out timout in probe function.

v2:
- Set .timeout in coh901327_wdt structure declaration.
- Set .min_timeout to 1 instead of 0. I could not find a datasheet for coh901327, so I'm not sure if 0 is valid. However, 0 seems wrong to me and most driver has 1 as min value. If it should be 0, please let me know and I have to set another initial value for margin.

 drivers/watchdog/coh901327_wdt.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)
drivers/watchdog/coh901327_wdt.c +8 -6
index 4410337f4f7f..500af8a7ec5a 100644
@@ -67,7 +67,9 @@
 #define U300_WDOG_IFR_WILL_BARK_IRQ_FORCE_ENABLE			0x0001U
 
 /* Default timeout in seconds = 1 minute */
-static unsigned int margin = 60;
+#define U300_WDOG_DEFAULT_TIMEOUT					60
+
+static unsigned int margin;
 static int irq;
 static void __iomem *virtbase;
 static struct device *parent;
@@ -235,8 +237,9 @@ static struct watchdog_device coh901327_wdt = {
 	 * timeout register is max
 	 * 0x7FFF = 327670ms ~= 327s.
 	 */
-	.min_timeout = 0,
+	.min_timeout = 1,
 	.max_timeout = 327,
+	.timeout = U300_WDOG_DEFAULT_TIMEOUT,
 };
 
 static int __exit coh901327_remove(struct platform_device *pdev)
@@ -315,16 +318,15 @@ static int __init coh901327_probe(struct platform_device *pdev)
 		goto out_no_irq;
 	}
 
-	ret = watchdog_init_timeout(&coh901327_wdt, margin, dev);
-	if (ret < 0)
-		coh901327_wdt.timeout = 60;
+	watchdog_init_timeout(&coh901327_wdt, margin, dev);
 
 	coh901327_wdt.parent = dev;
 	ret = watchdog_register_device(&coh901327_wdt);
 	if (ret)
 		goto out_no_wdog;
 
-	dev_info(dev, "initialized. timer margin=%d sec\n", margin);
+	dev_info(dev, "initialized. (timeout=%d sec)\n",
+			coh901327_wdt.timeout);
 	return 0;
 
 out_no_wdog:
Reviewed-by on v3 7/7
#16 lore raw

Re: [PATCH v3 7/7] watchdog: coh901327: make use of timeout-secs provided in devicetree

In reply to Marcus Folkesson
Quoted 15 lines · Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

Following best practice described in Documentation/watchdog/watchdog-kernel-api.txt to make use of the parameter logic.

---

Notes:
v3:
- Reformat and use coh901327_wdt.timeout instead of margin when print out timout in probe function.

Quoted 7 lines

v2:
- Set .timeout in coh901327_wdt structure declaration.
- Set .min_timeout to 1 instead of 0. I could not find a datasheet for coh901327, so I'm not sure if 0 is valid. However, 0 seems wrong to me and most driver has 1 as min value. If it should be 0, please let me know and I have to set another initial value for margin.

Yours,
Linus Walleij

maintainer Reviewed-by on v3 1/7
#17 lore raw
In reply to Marcus Folkesson
Quoted 8 lines · Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

Following best practice described in Documentation/watchdog/watchdog-kernel-api.txt to make use of the parameter logic.

Quoted 41 lines

---

Notes:
v3:
- Use wdd->timeout instead of wdt_timeout when print out timout in probe function.

 drivers/watchdog/sama5d4_wdt.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/watchdog/sama5d4_wdt.c +3 -3
index 0ae947c3d7bc..255169916dbb 100644
@@ -33,7 +33,7 @@ struct sama5d4_wdt {
 	unsigned long		last_ping;
 };
 
-static int wdt_timeout = WDT_DEFAULT_TIMEOUT;
+static int wdt_timeout;
 static bool nowayout = WATCHDOG_NOWAYOUT;
 
 module_param(wdt_timeout, int, 0);
@@ -212,7 +212,7 @@ static int sama5d4_wdt_probe(struct platform_device *pdev)
 		return -ENOMEM;
 
 	wdd = &wdt->wdd;
-	wdd->timeout = wdt_timeout;
+	wdd->timeout = WDT_DEFAULT_TIMEOUT;
 	wdd->info = &sama5d4_wdt_info;
 	wdd->ops = &sama5d4_wdt_ops;
 	wdd->min_timeout = MIN_WDT_TIMEOUT;
@@ -273,7 +273,7 @@ static int sama5d4_wdt_probe(struct platform_device *pdev)
 	platform_set_drvdata(pdev, wdt);
 
 	dev_info(&pdev->dev, "initialized (timeout = %d sec, nowayout = %d)\n",
-		 wdt_timeout, nowayout);
+		 wdd->timeout, nowayout);
 
 	return 0;
 }
maintainer Acked-by on v3 1/7
#18 lore raw
In reply to Marcus Folkesson
Quoted 8 lines · Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

Following best practice described in Documentation/watchdog/watchdog-kernel-api.txt to make use of the parameter logic.

Thanks, best regards,
Nicolas

Quoted 42 lines

---

Notes:
v3:
- Use wdd->timeout instead of wdt_timeout when print out timout in probe function.

 drivers/watchdog/sama5d4_wdt.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/watchdog/sama5d4_wdt.c +3 -3
index 0ae947c3d7bc..255169916dbb 100644
@@ -33,7 +33,7 @@ struct sama5d4_wdt {
 	unsigned long		last_ping;
 };
 
-static int wdt_timeout = WDT_DEFAULT_TIMEOUT;
+static int wdt_timeout;
 static bool nowayout = WATCHDOG_NOWAYOUT;
 
 module_param(wdt_timeout, int, 0);
@@ -212,7 +212,7 @@ static int sama5d4_wdt_probe(struct platform_device *pdev)
 		return -ENOMEM;
 
 	wdd = &wdt->wdd;
-	wdd->timeout = wdt_timeout;
+	wdd->timeout = WDT_DEFAULT_TIMEOUT;
 	wdd->info = &sama5d4_wdt_info;
 	wdd->ops = &sama5d4_wdt_ops;
 	wdd->min_timeout = MIN_WDT_TIMEOUT;
@@ -273,7 +273,7 @@ static int sama5d4_wdt_probe(struct platform_device *pdev)
 	platform_set_drvdata(pdev, wdt);
 
 	dev_info(&pdev->dev, "initialized (timeout = %d sec, nowayout = %d)\n",
-		 wdt_timeout, nowayout);
+		 wdd->timeout, nowayout);
 
 	return 0;
 }
Signature · 1 line
Nicolas Ferre
#19 lore raw
In reply to Marcus Folkesson

Hi,

Am I supposed to do something more to make this patchset picked up?

Thanks,
Marcus Folkesson

Quoted 52 lines · Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

Following best practice described in Documentation/watchdog/watchdog-kernel-api.txt to make use of the parameter logic.

---

Notes:
v3:
- Use wdd->timeout instead of wdt_timeout when print out timout in probe function.

 drivers/watchdog/sama5d4_wdt.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/watchdog/sama5d4_wdt.c +3 -3
index 0ae947c3d7bc..255169916dbb 100644
@@ -33,7 +33,7 @@ struct sama5d4_wdt {
 	unsigned long		last_ping;
 };
 
-static int wdt_timeout = WDT_DEFAULT_TIMEOUT;
+static int wdt_timeout;
 static bool nowayout = WATCHDOG_NOWAYOUT;
 
 module_param(wdt_timeout, int, 0);
@@ -212,7 +212,7 @@ static int sama5d4_wdt_probe(struct platform_device *pdev)
 		return -ENOMEM;
 
 	wdd = &wdt->wdd;
-	wdd->timeout = wdt_timeout;
+	wdd->timeout = WDT_DEFAULT_TIMEOUT;
 	wdd->info = &sama5d4_wdt_info;
 	wdd->ops = &sama5d4_wdt_ops;
 	wdd->min_timeout = MIN_WDT_TIMEOUT;
@@ -273,7 +273,7 @@ static int sama5d4_wdt_probe(struct platform_device *pdev)
 	platform_set_drvdata(pdev, wdt);
 
 	dev_info(&pdev->dev, "initialized (timeout = %d sec, nowayout = %d)\n",
-		 wdt_timeout, nowayout);
+		 wdd->timeout, nowayout);
 
 	return 0;
 }
Signature · 1 line
2.15.1

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-amlogic/attachments/20180316/b80ee7cb/attachment.sig>

maintainer
#20 lore raw
Quoted 4 lines · Marcus Folkesson

Hi,

Am I supposed to do something more to make this patchset picked up?

Did you check linux-next ?

Guenter

Quoted 56 lines

Thanks,
Marcus Folkesson

Quoted 52 lines · Marcus Folkesson

watchdog_init_timeout() will allways pick timeout_param since it defaults to a valid timeout.

Following best practice described in Documentation/watchdog/watchdog-kernel-api.txt to make use of the parameter logic.

---

Notes:
v3:
- Use wdd->timeout instead of wdt_timeout when print out timout in probe function.

  drivers/watchdog/sama5d4_wdt.c | 6 +++---
  1 file changed, 3 insertions(+), 3 deletions(-)
drivers/watchdog/sama5d4_wdt.c +3 -3
index 0ae947c3d7bc..255169916dbb 100644
@@ -33,7 +33,7 @@ struct sama5d4_wdt {
  	unsigned long		last_ping;
  };
 
-static int wdt_timeout = WDT_DEFAULT_TIMEOUT;
+static int wdt_timeout;
  static bool nowayout = WATCHDOG_NOWAYOUT;
 
  module_param(wdt_timeout, int, 0);
@@ -212,7 +212,7 @@ static int sama5d4_wdt_probe(struct platform_device *pdev)
  		return -ENOMEM;
 
  	wdd = &wdt->wdd;
-	wdd->timeout = wdt_timeout;
+	wdd->timeout = WDT_DEFAULT_TIMEOUT;
  	wdd->info = &sama5d4_wdt_info;
  	wdd->ops = &sama5d4_wdt_ops;
  	wdd->min_timeout = MIN_WDT_TIMEOUT;
@@ -273,7 +273,7 @@ static int sama5d4_wdt_probe(struct platform_device *pdev)
  	platform_set_drvdata(pdev, wdt);
 
  	dev_info(&pdev->dev, "initialized (timeout = %d sec, nowayout = %d)\n",
-		 wdt_timeout, nowayout);
+		 wdd->timeout, nowayout);
 
  	return 0;
  }
Signature · 1 line
2.15.1
#21 lore raw
Quoted 7 lines · Guenter Roeck
Quoted 4 lines · Marcus Folkesson

Hi,

Am I supposed to do something more to make this patchset picked up?

Did you check linux-next ?

:-)

Thank you

/Marcus

Quoted 3 lines

Guenter

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-amlogic/attachments/20180317/8f662ce2/attachment.sig>