Skip to contentSkip to conversation

watchdog: npcm: Add reset status detection support

PATCH Superseded Tomer Maimon 3 participants 9 messages Jul 5 – Jul 6

Subsystems (4)

v3 of 4 Superseded Latest is v4 →

lore
Share & tools
lore

Tomer Maimon

author maintainer
#0 lore raw

v3 cover watchdog: npcm: Add reset status detection support

This series documents and implements reset indication reporting for the NPCM watchdog driver on NPCM7xx and NPCM8xx systems, and documents the optional GCR syscon property used by that support.

Patch 1 updates the watchdog binding to allow the optional ``nuvoton,sysgcr`` property used for reset-cause reporting. Patch 2 adds watchdog documentation that describes the bootstatus mapping.
Patch 3 reads the SoC reset indication bits and maps them into the existing watchdog bootstatus flags for NPCM750 and NPCM845, while leaving WPCM450 unchanged.

Addressed comments from:
- Krzysztof Kozlowski: https://patchwork.ozlabs.org/project/openbmc/patch/[email protected]/

Changes since version 2:
- reword patch 1 to explain the hardware reason for the GCR phandle and why it is needed for the reset-status support, per Krzysztof's review
- cache NPCM750 reset status in SCRPAD2 so later watchdog probes can report the same boot-time state

Changes since version 1:
- Modify reset detection handle in the watchodg.
- reword patch 1 to use the GCR syscon-property wording from the applied NPCM reset binding update and drop the optional property from the binding example.
- reword the patch subjects and commit message bodies to match current kernel dt-bindings, docs, and watchdog style.

Tomer Maimon (3):
dt-bindings: watchdog: npcm: add GCR syscon property
docs: watchdog: npcm: Add reset status description
watchdog: npcm: add bootstatus support

 .../watchdog/nuvoton,npcm750-wdt.yaml         |   6 +
 Documentation/watchdog/index.rst              |   1 +
 Documentation/watchdog/npcm_wdt.rst           |  71 ++++++
 drivers/watchdog/npcm_wdt.c                   | 217 +++++++++++++++++-
 4 files changed, 293 insertions(+), 2 deletions(-)
 create mode 100644 Documentation/watchdog/npcm_wdt.rst
Signature · 1 line
2.34.1
author maintainer
#1 lore raw

PATCH v3 1/3 dt-bindings: watchdog: npcm: add GCR syscon property

NPCM750 and NPCM845 latch watchdog reset indications in the SoC GCR block rather than in the watchdog block itself.

Add the optional nuvoton,sysgcr phandle so watchdog nodes can reference the shared GCR reset-status registers that hold those latched watchdog reset indications.

This is needed by the following reset-status support, which reads those latches and reports watchdog-caused resets through bootstatus.

---
 .../devicetree/bindings/watchdog/nuvoton,npcm750-wdt.yaml   | 6 ++++++
 1 file changed, 6 insertions(+)
Documentation/devicetree/bindings/watchdog/nuvoton,npcm750-wdt.yaml +6 -0
index 7aa30f5b5c49..99c225f166db 100644
@@ -40,6 +40,12 @@ properties:
   clock-frequency:
     description: Frequency in Hz of the clock that drives the NPCM timer.
 
+  nuvoton,sysgcr:
+    $ref: /schemas/types.yaml#/definitions/phandle
+    description:
+      a phandle to access the GCR reset-status registers that latch
+      watchdog reset indications on NPCM750 and NPCM845.
+
 required:
   - compatible
   - reg
Signature · 1 line
2.34.1
author maintainer
#3 lore raw

PATCH v3 2/3 docs: watchdog: npcm: Add reset status description

In reply to Tomer Maimon

Add documentation describing how the NPCM watchdog driver reports reset causes through bootstatus on NPCM750 and NPCM845 systems.

Document the reset flag mapping, the watchdog instance mapping for WDIOF_CARDRESET, and the NPCM750/NPCM845 latch handling. Also mention sysfs bootstatus reporting when watchdog sysfs support is enabled.

---
 Documentation/watchdog/index.rst    |  1 +
 Documentation/watchdog/npcm_wdt.rst | 71 +++++++++++++++++++++++++++++
 2 files changed, 72 insertions(+)
 create mode 100644 Documentation/watchdog/npcm_wdt.rst
Documentation/watchdog/index.rst +1 -0
index dbc702b31a43..eb9e76275fa0 100644
@@ -25,4 +25,5 @@ Driver specific
     hpwdt
     ics-wdt
     mlx-wdt
+    npcm_wdt
     pcwd-watchdog
Documentation/watchdog/npcm_wdt.rst new file +71 -0

71 lines not on this page · raw message

Signature · 1 line
2.34.1
author maintainer
#6 lore raw

PATCH v3 3/3 watchdog: npcm: add bootstatus support

In reply to Tomer Maimon

The NPCM750 uses RESSR and the NPCM845 uses INTCR2 to latch reset indications. Read those bits during probe and map them into watchdog bootstatus flags.

For NPCM845, cache the sampled INTCR2 state in SCRPAD10 after the reset status bits are cleared so later probes can report the same boot-time state. Also report WDIOF_CARDRESET for the watchdog instance whose reset bit is latched, while leaving WPCM450 behavior unchanged.

---
 drivers/watchdog/npcm_wdt.c | 217 +++++++++++++++++++++++++++++++++++-
 1 file changed, 215 insertions(+), 2 deletions(-)
drivers/watchdog/npcm_wdt.c +215 -2
index e62ea054bc61..cecfd6e574d8 100644
@@ -7,14 +7,52 @@
 #include <linux/delay.h>
 #include <linux/interrupt.h>
 #include <linux/kernel.h>
+#include <linux/mfd/syscon.h>
 #include <linux/module.h>
 #include <linux/of_irq.h>
 #include <linux/platform_device.h>
+#include <linux/regmap.h>
 #include <linux/slab.h>
 #include <linux/watchdog.h>
 
 #define NPCM_WTCR	0x1C
 
+/* NPCM GCR module */
+#define NPCM_RESSR_OFFSET		0x6C
+#define NPCM_INTCR2_OFFSET		0x60
+#define NPCM7XX_SCRPAD2_OFFSET		0x84
+#define NPCM8XX_SCRPAD10_OFFSET		0xE28
+
+#define NPCM_PORST			BIT(31)
+#define NPCM_CORST			BIT(30)
+#define NPCM_WD0RST			BIT(29)
+#define NPCM_SWR1RST			BIT(28)
+#define NPCM_SWR2RST			BIT(27)
+#define NPCM_SWR3RST			BIT(26)
+#define NPCM_SWR4RST			BIT(25)
+#define NPCM_WD1RST			BIT(24)
+#define NPCM_WD2RST			BIT(23)
+#define NPCM_RST			GENMASK(31, 23)
+#define NPCM8XX_TIP_RESET		BIT(25) /* Replaces SWRST4 on NPCM8xx */
+
+/* Per-instance mapping of MMIO base address to its RESSR/INTCR2 reset bit. */
+struct npcm_wdt_rst_map {
+	phys_addr_t	base;
+	u32		rst_bit;
+};
+
+struct npcm_wdt_status_map {
+	u32	rst_bit;
+	u32	wdiof_flag;
+};
+
+struct npcm_wdt_data {
+	const struct npcm_wdt_rst_map		*rst_map;
+	unsigned int				rst_map_size;
+	const struct npcm_wdt_status_map	*status_map;
+	unsigned int				status_map_size;
+};
+
 #define NPCM_WTCLK	(BIT(10) | BIT(11))	/* Clock divider */
 #define NPCM_WTE	BIT(7)			/* Enable */
 #define NPCM_WTIE	BIT(6)			/* Enable irq */
@@ -47,6 +85,50 @@ struct npcm_wdt {
 	struct clk		*clk;
 };
 
+static const struct npcm_wdt_rst_map npcm750_rst_map[] = {
+	{ 0xf000801c, NPCM_WD0RST },
+	{ 0xf000901c, NPCM_WD1RST },
+	{ 0xf000a01c, NPCM_WD2RST },
+};
+
+static const struct npcm_wdt_status_map npcm750_status_map[] = {
+	{ NPCM_PORST, WDIOF_OVERHEAT },
+	{ NPCM_CORST, WDIOF_FANFAULT },
+	{ NPCM_SWR1RST, WDIOF_EXTERN1 },
+	{ NPCM_SWR2RST, WDIOF_EXTERN2 },
+	{ NPCM_SWR3RST, WDIOF_POWERUNDER },
+	{ NPCM_SWR4RST, WDIOF_POWEROVER },
+};
+
+static const struct npcm_wdt_data npcm750_data = {
+	.rst_map = npcm750_rst_map,
+	.rst_map_size = ARRAY_SIZE(npcm750_rst_map),
+	.status_map = npcm750_status_map,
+	.status_map_size = ARRAY_SIZE(npcm750_status_map),
+};
+
+static const struct npcm_wdt_rst_map npcm845_rst_map[] = {
+	{ 0xf000801c, NPCM_WD0RST },
+	{ 0xf000901c, NPCM_WD1RST },
+	{ 0xf000a01c, NPCM_WD2RST },
+};
+
+static const struct npcm_wdt_status_map npcm845_status_map[] = {
+	{ NPCM_PORST, WDIOF_OVERHEAT },
+	{ NPCM_CORST, WDIOF_FANFAULT },
+	{ NPCM_SWR1RST, WDIOF_EXTERN1 },
+	{ NPCM_SWR2RST, WDIOF_EXTERN2 },
+	{ NPCM_SWR3RST, WDIOF_POWERUNDER },
+	{ NPCM8XX_TIP_RESET, WDIOF_POWEROVER },
+};
+
+static const struct npcm_wdt_data npcm845_data = {
+	.rst_map = npcm845_rst_map,
+	.rst_map_size = ARRAY_SIZE(npcm845_rst_map),
+	.status_map = npcm845_status_map,
+	.status_map_size = ARRAY_SIZE(npcm845_status_map),
+};
+
 static inline struct npcm_wdt *to_npcm_wdt(struct watchdog_device *wdd)
 {
 	return container_of(wdd, struct npcm_wdt, wdd);
@@ -169,6 +251,111 @@ static bool npcm_is_running(struct watchdog_device *wdd)
 	return readl(wdt->reg) & NPCM_WTE;
 }
 
+static void npcm_get_reset_status(struct npcm_wdt *wdt, struct device *dev,
+				  const struct npcm_wdt_data *data,
+				  resource_size_t start)
+{
+	struct regmap *gcr_regmap;
+	u32 rstval = 0;
+	unsigned int i;
+	int ret;
+
+	if (!data)
+		return;
+
+	gcr_regmap = syscon_regmap_lookup_by_phandle(dev->of_node,
+						     "nuvoton,sysgcr");
+	if (IS_ERR(gcr_regmap)) {
+		dev_warn(dev,
+			 "Failed to find nuvoton,sysgcr, WD reset status not supported\n");
+		return;
+	}
+
+	if (of_device_is_compatible(dev->of_node, "nuvoton,npcm845-wdt")) {
+		ret = regmap_read(gcr_regmap, NPCM_INTCR2_OFFSET, &rstval);
+		if (ret) {
+			dev_warn(dev, "Failed to read INTCR2 reset status: %d\n",
+				 ret);
+			return;
+		}
+
+		if (rstval & NPCM_RST) {
+			ret = regmap_write(gcr_regmap, NPCM_INTCR2_OFFSET,
+					   rstval & ~NPCM_RST);
+			if (ret) {
+				dev_warn(dev,
+					 "Failed to clear INTCR2 reset status: %d\n",
+					 ret);
+				return;
+			}
+
+			ret = regmap_write(gcr_regmap, NPCM8XX_SCRPAD10_OFFSET,
+					   rstval);
+			if (ret) {
+				dev_warn(dev,
+					 "Failed to cache reset status in SCRPAD10: %d\n",
+					 ret);
+				return;
+			}
+		} else {
+			ret = regmap_read(gcr_regmap, NPCM8XX_SCRPAD10_OFFSET,
+					  &rstval);
+			if (ret) {
+				dev_warn(dev,
+					 "Failed to read cached reset status from SCRPAD10: %d\n",
+					 ret);
+				return;
+			}
+		}
+	} else if (of_device_is_compatible(dev->of_node, "nuvoton,npcm750-wdt")) {
+		ret = regmap_read(gcr_regmap, NPCM_RESSR_OFFSET, &rstval);
+		if (ret) {
+			dev_warn(dev, "Failed to read RESSR reset status: %d\n",
+				 ret);
+			return;
+		}
+
+		if (rstval & NPCM_RST) {
+			ret = regmap_write(gcr_regmap, NPCM_RESSR_OFFSET, 
+					   rstval & ~NPCM_RST);
+			if (ret) {
+				dev_warn(dev, "Failed to clear RESSR reset status: %d\n", ret);
+				return;
+			}
+
+			ret = regmap_write(gcr_regmap, NPCM7XX_SCRPAD2_OFFSET,
+					   rstval);
+			if (ret) {
+				dev_warn(dev,
+					 "Failed to cache reset status in SCRPAD2: %d\n", ret);
+				return;
+			}
+		} else {
+			ret = regmap_read(gcr_regmap, NPCM7XX_SCRPAD2_OFFSET,
+					  &rstval);
+			if (ret) {
+				dev_warn(dev,
+					 "Failed to read cached reset status from SCRPAD2: %d\n",
+					 ret);
+				return;
+			}
+		}
+	}
+
+	for (i = 0; i < data->status_map_size; i++) {
+		if (rstval & data->status_map[i].rst_bit)
+			wdt->wdd.bootstatus |= data->status_map[i].wdiof_flag;
+	}
+
+	for (i = 0; i < data->rst_map_size; i++) {
+		if (data->rst_map[i].base == start &&
+		    rstval & data->rst_map[i].rst_bit) {
+			wdt->wdd.bootstatus |= WDIOF_CARDRESET;
+			break;
+		}
+	}
+}
+
 static const struct watchdog_info npcm_wdt_info = {
 	.identity	= KBUILD_MODNAME,
 	.options	= WDIOF_SETTIMEOUT
@@ -176,6 +363,20 @@ static const struct watchdog_info npcm_wdt_info = {
 			| WDIOF_MAGICCLOSE,
 };
 
+static const struct watchdog_info npcm_wdt_rst_info = {
+	.identity	= KBUILD_MODNAME,
+	.options	= WDIOF_SETTIMEOUT
+			| WDIOF_KEEPALIVEPING
+			| WDIOF_MAGICCLOSE
+			| WDIOF_CARDRESET
+			| WDIOF_OVERHEAT
+			| WDIOF_FANFAULT
+			| WDIOF_EXTERN1
+			| WDIOF_EXTERN2
+			| WDIOF_POWERUNDER
+			| WDIOF_POWEROVER,
+};
+
 static const struct watchdog_ops npcm_wdt_ops = {
 	.owner = THIS_MODULE,
 	.start = npcm_wdt_start,
@@ -188,7 +389,10 @@ static const struct watchdog_ops npcm_wdt_ops = {
 static int npcm_wdt_probe(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
+	const struct npcm_wdt_data *data = device_get_match_data(dev);
+	struct resource *res;
 	struct npcm_wdt *wdt;
+	resource_size_t start;
 	int irq;
 	int ret;
 
@@ -196,10 +400,16 @@ static int npcm_wdt_probe(struct platform_device *pdev)
 	if (!wdt)
 		return -ENOMEM;
 
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res)
+		return -EINVAL;
+
 	wdt->reg = devm_platform_ioremap_resource(pdev, 0);
 	if (IS_ERR(wdt->reg))
 		return PTR_ERR(wdt->reg);
 
+	start = res->start;
+
 	wdt->clk = devm_clk_get_optional(&pdev->dev, NULL);
 	if (IS_ERR(wdt->clk))
 		return PTR_ERR(wdt->clk);
@@ -208,7 +418,7 @@ static int npcm_wdt_probe(struct platform_device *pdev)
 	if (irq < 0)
 		return irq;
 
-	wdt->wdd.info = &npcm_wdt_info;
+	wdt->wdd.info = data ? &npcm_wdt_rst_info : &npcm_wdt_info;
 	wdt->wdd.ops = &npcm_wdt_ops;
 	wdt->wdd.min_timeout = 1;
 	wdt->wdd.max_timeout = 2750;
@@ -220,6 +430,8 @@ static int npcm_wdt_probe(struct platform_device *pdev)
 	/* Ensure timeout is able to be represented by the hardware */
 	npcm_wdt_set_timeout(&wdt->wdd, wdt->wdd.timeout);
 
+	npcm_get_reset_status(wdt, dev, data, start);
+
 	if (npcm_is_running(&wdt->wdd)) {
 		/* Restart with the default or device-tree specified timeout */
 		npcm_wdt_start(&wdt->wdd);
@@ -243,7 +455,8 @@ static int npcm_wdt_probe(struct platform_device *pdev)
 #ifdef CONFIG_OF
 static const struct of_device_id npcm_wdt_match[] = {
 	{.compatible = "nuvoton,wpcm450-wdt"},
-	{.compatible = "nuvoton,npcm750-wdt"},
+	{.compatible = "nuvoton,npcm750-wdt", .data = &npcm750_data},
+	{.compatible = "nuvoton,npcm845-wdt", .data = &npcm845_data},
 	{},
 };
 MODULE_DEVICE_TABLE(of, npcm_wdt_match);
Signature · 1 line
2.34.1
#4 lore raw

Re: [PATCH v3 2/3] docs: watchdog: npcm: Add reset status description

In reply to Tomer Maimon
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [High] The watchdog driver abuses standard Watchdog UAPI constants to report non-watchdog platform reset…

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] The watchdog driver abuses standard Watchdog UAPI constants to report non-watchdog platform reset causes, violating the userspace ABI contract. --

commit 72fd6c8b87d96376853f3dcd5ef1597495cf9817
Author: Tomer Maimon <[email protected]>

docs: watchdog: npcm: Add reset status description

Add documentation describing how the NPCM watchdog driver reports reset causes through bootstatus on NPCM750 and NPCM845 systems.

Quoted 5 lines
Documentation/watchdog/npcm_wdt.rst new file +0 -0
new file mode 100644
index 0000000000000..cdc45ade9e0d9

[ ... ]

Quoted 5 lines

+.. list-table:: Reset-cause mapping

Documentation/watchdog/npcm_wdt.rst new file +4 -0
+   :header-rows: 1
+
+   * - Platform
+     - Reset indication

Documentation/watchdog/npcm_wdt.rst line 32

Quoted 10 lines
Documentation/watchdog/npcm_wdt.rst new file +10 -0
+     - Bootstatus flag
+     - Reported meaning
+   * - NPCM750 and NPCM845
+     - ``PORST``
+     - ``WDIOF_OVERHEAT``
+     - power-on reset
+   * - NPCM750 and NPCM845
+     - ``CORST``
+     - ``WDIOF_FANFAULT``
+     - core reset

[Severity: High]
Is it correct to map generic Watchdog UAPI thermal and fan failure flags to standard platform reset indications?

The generic Watchdog API defines strict semantics for its status flags. For instance, WDIOF_OVERHEAT means a reset due to CPU overheat, and WDIOF_FANFAULT means a fan failed. By mapping standard SoC reset causes like Power-On Reset (PORST) to WDIOF_OVERHEAT and Core Reset (CORST) to WDIOF_FANFAULT, might this break the userspace ABI contract?

During a normal cold boot, reading the SoC latch would unconditionally expose this to userspace via the WDIOC_GETBOOTSTATUS ioctl or sysfs. Standard userspace monitoring daemons (like systemd or wdctl) reading bootstatus might erroneously conclude the system suffered a thermal event or fan failure, potentially triggering unintended emergency actions or false alarms.

Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=2
#7 lore raw

Re: [PATCH v3 3/3] watchdog: npcm: add bootstatus support

In reply to Tomer Maimon
Thank you for your contribution! Sashiko AI review found 4 potential issue(s) to consider:

Thank you for your contribution! Sashiko AI review found 4 potential issue(s) to consider:

New issues:
- [High] Non-atomic read-modify-write on shared syscon registers corrupts concurrent updates or incorrectly clears W1C bits.
- [High] The driver hijacks standardized watchdog bootstatus flags to report unrelated SoC reset reasons.
- [Low] The driver hardcodes physical MMIO base addresses to identify driver instances.

Pre-existing issues:
- [High] The driver leaks a clock enable reference on every timeout change and fails to clean up on probe errors. --

commit 3d82c9ef067a4770a83258e13cbb0ed566d4998b
Author: Tomer Maimon <[email protected]>

watchdog: npcm: add bootstatus support

The NPCM750 uses RESSR and the NPCM845 uses INTCR2 to latch reset indications. Read those bits during probe and map them into watchdog bootstatus flags. For NPCM845, cache the sampled INTCR2 state in SCRPAD10 after the reset status bits are cleared.

Quoted 4 lines
drivers/watchdog/npcm_wdt.c +0 -0
index e62ea054bc618..82ff3956dcf21 100644

[ ... ]

drivers/watchdog/npcm_wdt.c struct npcm_wdt { · line 92

Quoted 5 lines
drivers/watchdog/npcm_wdt.c +5 -0
+static const struct npcm_wdt_rst_map npcm750_rst_map[] = {
+	{ 0xf000801c, NPCM_WD0RST },
+	{ 0xf000901c, NPCM_WD1RST },
+	{ 0xf000a01c, NPCM_WD2RST },
+};

[Severity: Low]
Does this code hardcode physical MMIO base addresses to identify driver instances?

Matching the resource start address against hardcoded values in npcm_get_reset_status() can be brittle if the memory map changes in future SoC revisions or alternative device tree mappings. Is there a way to use device tree aliases or instance IDs to identify the hardware block instead?

drivers/watchdog/npcm_wdt.c struct npcm_wdt { · line 101

Quoted 9 lines
drivers/watchdog/npcm_wdt.c +9 -0
+
+static const struct npcm_wdt_status_map npcm750_status_map[] = {
+	{ NPCM_PORST, WDIOF_OVERHEAT },
+	{ NPCM_CORST, WDIOF_FANFAULT },
+	{ NPCM_SWR1RST, WDIOF_EXTERN1 },
+	{ NPCM_SWR2RST, WDIOF_EXTERN2 },
+	{ NPCM_SWR3RST, WDIOF_POWERUNDER },
+	{ NPCM_SWR4RST, WDIOF_POWEROVER },
+};

[Severity: High]
Does this mapping hijack standardized watchdog bootstatus flags to report unrelated SoC reset reasons?

Translating NPCM SoC-specific reset reasons into standard UAPI flags might cause issues in userspace. For example, WDIOF_OVERHEAT and WDIOF_FANFAULT are strictly defined in the UAPI to indicate thermal overheating and fan failures.

Could system monitoring tools misinterpret a normal cold boot (NPCM_CORST) as a severe hardware failure and trigger unintended alerts or shutdowns?

[ ... ]

Quoted 4 lines
drivers/watchdog/npcm_wdt.c +4 -0
+static void npcm_get_reset_status(struct npcm_wdt *wdt, struct device *dev,
+				  const struct npcm_wdt_data *data,
+				  resource_size_t start)
+{

[ ... ]

drivers/watchdog/npcm_wdt.c static bool npcm_is_running(struct watchdog_device *wdd) · line 284

Quoted 11 lines
drivers/watchdog/npcm_wdt.c +11 -0
+	if (of_device_is_compatible(dev->of_node, "nuvoton,npcm845-wdt")) {
+		ret = regmap_read(gcr_regmap, NPCM_INTCR2_OFFSET, &rstval);
+		if (ret) {
+			dev_warn(dev, "Failed to read INTCR2 reset status: %d\n",
+				 ret);
+			return;
+		}
+
+		if (rstval & NPCM_RST) {
+			ret = regmap_write(gcr_regmap, NPCM_INTCR2_OFFSET,
+					   rstval & ~NPCM_RST);

[Severity: High]
Can this non-atomic read-modify-write sequence on the shared syscon register corrupt concurrent updates?

If INTCR2 (and RESSR below) are standard read-write registers, this sequence could silently overwrite modifications made by other drivers sharing the syscon.

Alternatively, if these are Write-1-to-Clear (W1C) registers, writing rstval & ~NPCM_RST will write 1s to all other active status bits, inadvertently clearing them, while writing 0 to the watchdog bits, failing to clear the intended reset flags.

Would it be safer to use regmap_update_bits() for atomic updates, or if W1C, write only the target bits without a read-modify-write?

[ ... ]

Quoted 9 lines
drivers/watchdog/npcm_wdt.c +2 -0
@@ -220,6 +430,8 @@ static int npcm_wdt_probe(struct platform_device *pdev)
 	/* Ensure timeout is able to be represented by the hardware */
 	npcm_wdt_set_timeout(&wdt->wdd, wdt->wdd.timeout);
 
+	npcm_get_reset_status(wdt, dev, data, start);
+
 	if (npcm_is_running(&wdt->wdd)) {
 		/* Restart with the default or device-tree specified timeout */
 		npcm_wdt_start(&wdt->wdd);

drivers/watchdog/npcm_wdt.c

Quoted 7 lines
drivers/watchdog/npcm_wdt.c +0 -0
 		set_bit(WDOG_HW_RUNNING, &wdt->wdd.status);
 	}
 
 	ret = devm_request_irq(dev, irq, npcm_wdt_interrupt, 0, "watchdog",
 			       wdt);
 	if (ret)
 		return ret;

[Severity: High]
This is a pre-existing issue, but does this code leak a clock enable reference?

In npcm_wdt_probe(), if the hardware is running, npcm_wdt_start() is called:

npcm_wdt_start()

    clk_prepare_enable(wdt->clk);

If subsequent probe steps fail (e.g., devm_request_irq() returns an error), probe returns without unwinding the clock enable.

Additionally, npcm_wdt_set_timeout() calls npcm_wdt_start() when the watchdog is active. Since there is no matching disable on timeout change, every WDIOC_SETTIMEOUT ioctl increments the clock's refcount. This prevents the clock from ever being disabled when the watchdog is stopped, leading to a resource leak.

Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=3
#2 lore raw

Re: [PATCH v3 1/3] dt-bindings: watchdog: npcm: add GCR syscon property

In reply to Tomer Maimon
Quoted 11 lines · Tomer Maimon

NPCM750 and NPCM845 latch watchdog reset indications in the SoC GCR block rather than in the watchdog block itself.

Add the optional nuvoton,sysgcr phandle so watchdog nodes can reference the shared GCR reset-status registers that hold those latched watchdog reset indications.

This is needed by the following reset-status support, which reads those latches and reports watchdog-caused resets through bootstatus.

Since the 450 doesn't support this, please make it conditional.
pw-bot: changes-requested

Thanks,
Conor.

Quoted 24 lines
---
 .../devicetree/bindings/watchdog/nuvoton,npcm750-wdt.yaml   | 6 ++++++
 1 file changed, 6 insertions(+)
Documentation/devicetree/bindings/watchdog/nuvoton,npcm750-wdt.yaml +6 -0
index 7aa30f5b5c49..99c225f166db 100644
@@ -40,6 +40,12 @@ properties:
   clock-frequency:
     description: Frequency in Hz of the clock that drives the NPCM timer.
 
+  nuvoton,sysgcr:
+    $ref: /schemas/types.yaml#/definitions/phandle
+    description:
+      a phandle to access the GCR reset-status registers that latch
+      watchdog reset indications on NPCM750 and NPCM845.
+
 required:
   - compatible
   - reg
Signature · 1 line
2.34.1
author maintainer
#5 lore raw

Re: [PATCH v3 2/3] docs: watchdog: npcm: Add reset status description

In reply to [email protected]

Hi,

Quoted 37 lines · On Sun, 5 Jul 2026 at 17:06, <[email protected]> wrote:

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] The watchdog driver abuses standard Watchdog UAPI constants to report non-watchdog platform reset causes, violating the userspace ABI contract. --

commit 72fd6c8b87d96376853f3dcd5ef1597495cf9817
Author: Tomer Maimon <[email protected]>

docs: watchdog: npcm: Add reset status description

Add documentation describing how the NPCM watchdog driver reports reset causes through bootstatus on NPCM750 and NPCM845 systems.

Quoted 5 lines
Documentation/watchdog/npcm_wdt.rst new file +0 -0
new file mode 100644
index 0000000000000..cdc45ade9e0d9

[ ... ]

Quoted 5 lines

+.. list-table:: Reset-cause mapping

Documentation/watchdog/npcm_wdt.rst new file +4 -0
+   :header-rows: 1
+
+   * - Platform
+     - Reset indication

Documentation/watchdog/npcm_wdt.rst line 32

Quoted 10 lines
Documentation/watchdog/npcm_wdt.rst new file +10 -0
+     - Bootstatus flag
+     - Reported meaning
+   * - NPCM750 and NPCM845
+     - ``PORST``
+     - ``WDIOF_OVERHEAT``
+     - power-on reset
+   * - NPCM750 and NPCM845
+     - ``CORST``
+     - ``WDIOF_FANFAULT``
+     - core reset
Quoted 4 lines

[Severity: High]
Is it correct to map generic Watchdog UAPI thermal and fan failure flags to standard platform reset indications?

Yes, it is something that was already discussed with the WD maintainer; This is why it is mentioned in the rst document

Quoted 6 lines

The generic Watchdog API defines strict semantics for its status flags. For instance, WDIOF_OVERHEAT means a reset due to CPU overheat, and WDIOF_FANFAULT means a fan failed. By mapping standard SoC reset causes like Power-On Reset (PORST) to WDIOF_OVERHEAT and Core Reset (CORST) to WDIOF_FANFAULT, might this break the userspace ABI contract?

Quoted 8 lines

During a normal cold boot, reading the SoC latch would unconditionally expose this to userspace via the WDIOC_GETBOOTSTATUS ioctl or sysfs. Standard userspace monitoring daemons (like systemd or wdctl) reading bootstatus might erroneously conclude the system suffered a thermal event or fan failure, potentially triggering unintended emergency actions or false alarms.

--
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=2

Thanks,

Tomer

author maintainer
#8 lore raw

Re: [PATCH v3 3/3] watchdog: npcm: add bootstatus support

In reply to [email protected]

Hi,

Quoted 5 lines · On Sun, 5 Jul 2026 at 17:10, <[email protected]> wrote:

Thank you for your contribution! Sashiko AI review found 4 potential issue(s) to consider:

New issues:
- [High] Non-atomic read-modify-write on shared syscon registers corrupts concurrent updates or incorrectly clears W1C bits.

Not needed, only the WD using this register.

Quoted 1 line

- [High] The driver hijacks standardized watchdog bootstatus flags to report unrelated SoC reset reasons.

Approved by the maintainer.

Quoted 23 lines

- [Low] The driver hardcodes physical MMIO base addresses to identify driver instances.

Pre-existing issues:
- [High] The driver leaks a clock enable reference on every timeout change and fails to clean up on probe errors. --

commit 3d82c9ef067a4770a83258e13cbb0ed566d4998b
Author: Tomer Maimon <[email protected]>

watchdog: npcm: add bootstatus support

The NPCM750 uses RESSR and the NPCM845 uses INTCR2 to latch reset indications. Read those bits during probe and map them into watchdog bootstatus flags. For NPCM845, cache the sampled INTCR2 state in SCRPAD10 after the reset status bits are cleared.

Quoted 4 lines
drivers/watchdog/npcm_wdt.c +0 -0
index e62ea054bc618..82ff3956dcf21 100644

[ ... ]

Quoted 9 lines

drivers/watchdog/npcm_wdt.c struct npcm_wdt { · line 92

Quoted 5 lines
drivers/watchdog/npcm_wdt.c +5 -0
+static const struct npcm_wdt_rst_map npcm750_rst_map[] = {
+     { 0xf000801c, NPCM_WD0RST },
+     { 0xf000901c, NPCM_WD1RST },
+     { 0xf000a01c, NPCM_WD2RST },
+};

[Severity: Low]
Does this code hardcode physical MMIO base addresses to identify driver instances?

It is hardcoded.

Quoted 96 lines

Matching the resource start address against hardcoded values in npcm_get_reset_status() can be brittle if the memory map changes in future SoC revisions or alternative device tree mappings. Is there a way to use device tree aliases or instance IDs to identify the hardware block instead?

drivers/watchdog/npcm_wdt.c struct npcm_wdt { · line 101

Quoted 9 lines
drivers/watchdog/npcm_wdt.c +9 -0
+
+static const struct npcm_wdt_status_map npcm750_status_map[] = {
+     { NPCM_PORST, WDIOF_OVERHEAT },
+     { NPCM_CORST, WDIOF_FANFAULT },
+     { NPCM_SWR1RST, WDIOF_EXTERN1 },
+     { NPCM_SWR2RST, WDIOF_EXTERN2 },
+     { NPCM_SWR3RST, WDIOF_POWERUNDER },
+     { NPCM_SWR4RST, WDIOF_POWEROVER },
+};

[Severity: High]
Does this mapping hijack standardized watchdog bootstatus flags to report unrelated SoC reset reasons?

Translating NPCM SoC-specific reset reasons into standard UAPI flags might cause issues in userspace. For example, WDIOF_OVERHEAT and WDIOF_FANFAULT are strictly defined in the UAPI to indicate thermal overheating and fan failures.

Could system monitoring tools misinterpret a normal cold boot (NPCM_CORST) as a severe hardware failure and trigger unintended alerts or shutdowns?

[ ... ]

Quoted 4 lines
drivers/watchdog/npcm_wdt.c +4 -0
+static void npcm_get_reset_status(struct npcm_wdt *wdt, struct device *dev,
+                               const struct npcm_wdt_data *data,
+                               resource_size_t start)
+{

[ ... ]

drivers/watchdog/npcm_wdt.c static bool npcm_is_running(struct watchdog_device *wdd) · line 284

Quoted 11 lines
drivers/watchdog/npcm_wdt.c +11 -0
+     if (of_device_is_compatible(dev->of_node, "nuvoton,npcm845-wdt")) {
+             ret = regmap_read(gcr_regmap, NPCM_INTCR2_OFFSET, &rstval);
+             if (ret) {
+                     dev_warn(dev, "Failed to read INTCR2 reset status: %d\n",
+                              ret);
+                     return;
+             }
+
+             if (rstval & NPCM_RST) {
+                     ret = regmap_write(gcr_regmap, NPCM_INTCR2_OFFSET,
+                                        rstval & ~NPCM_RST);

[Severity: High]
Can this non-atomic read-modify-write sequence on the shared syscon register corrupt concurrent updates?

If INTCR2 (and RESSR below) are standard read-write registers, this sequence could silently overwrite modifications made by other drivers sharing the syscon.

Alternatively, if these are Write-1-to-Clear (W1C) registers, writing rstval & ~NPCM_RST will write 1s to all other active status bits, inadvertently clearing them, while writing 0 to the watchdog bits, failing to clear the intended reset flags.

Would it be safer to use regmap_update_bits() for atomic updates, or if W1C, write only the target bits without a read-modify-write?

[ ... ]

Quoted 9 lines
drivers/watchdog/npcm_wdt.c +2 -0
@@ -220,6 +430,8 @@ static int npcm_wdt_probe(struct platform_device *pdev)
      /* Ensure timeout is able to be represented by the hardware */
      npcm_wdt_set_timeout(&wdt->wdd, wdt->wdd.timeout);
 
+     npcm_get_reset_status(wdt, dev, data, start);
+
      if (npcm_is_running(&wdt->wdd)) {
              /* Restart with the default or device-tree specified timeout */
              npcm_wdt_start(&wdt->wdd);
Quoted 7 lines
              set_bit(WDOG_HW_RUNNING, &wdt->wdd.status);
      }

      ret = devm_request_irq(dev, irq, npcm_wdt_interrupt, 0, "watchdog",
                             wdt);
      if (ret)
              return ret;

[Severity: High]
This is a pre-existing issue, but does this code leak a clock enable reference?

In npcm_wdt_probe(), if the hardware is running, npcm_wdt_start() is called:

npcm_wdt_start()

    clk_prepare_enable(wdt->clk);

If subsequent probe steps fail (e.g., devm_request_irq() returns an error), probe returns without unwinding the clock enable.

Quoted 8 lines

Additionally, npcm_wdt_set_timeout() calls npcm_wdt_start() when the watchdog is active. Since there is no matching disable on timeout change, every WDIOC_SETTIMEOUT ioctl increments the clock's refcount. This prevents the clock from ever being disabled when the watchdog is stopped, leading to a resource leak.

--
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=3

Thanks,

Tomer