[PATCH v8 0/2] i2c: designware: fix WACF2200 touchscreen probe ordering

[PATCH v8 0/2] i2c: designware: fix WACF2200 touchscreen probe ordering

[Hardik Prakash]

This series fixes the non-functional WACF2200 touchscreen on Lenovo
Yoga 7 14AGP11 by:

1. Reverting the earlier DMI-specific GPIO IRQ quirk (patch 1 of the
   original series, already in Torvalds' tree) which is no longer
   needed.

2. Adding a generic probe deferral in i2c-designware that walks ACPI
   child devices, finds GpioInt resources, and defers probe until the
   referenced GPIO controllers are fully bound.

Testing confirms patch 2 alone is sufficient — the existing
amd_gpio_irq_enable() flow works correctly once probe ordering is
fixed, making the manual IRQ restoration in the reverted patch
unnecessary.

The probe ordering race was confirmed via dynamic debug tracing:

  0.285952  amd_gpio_probe: registering gpiochip  <- GPIO chip visible
  0.287121  amd_gpio_probe: requesting parent IRQ <- probe still running
  0.301454  AMDI0010:02 dw_i2c_plat_probe: start  <- races here
  2.348157  lost arbitration

Tested on Lenovo Yoga 7 14AGP11 (83TD), Fedora 44, kernel 7.1.0-rc5+.

v8:
 - Add revert of now-unnecessary patch 1 (Mario Limonciello)
 - Collect Acked-by (Bartosz Golaszewski) and Reviewed-by (Mario Limonciello)

v7:
 - Fix unused 'ret' variable in check_child_gpioint() (kernel test robot)
 - Replace BugLink: with Link: (Bartosz Golaszewski)
 - Add Reported-by and Closes tags for kernel test robot warning

v6:
 - Replace DMI-specific deferral with generic GpioInt dependency check
   walking ACPI child devices (suggested by Mario Limonciello)

v5:
 - Add blank line before #include <linux/acpi.h> (Bartosz Golaszewski)
 - Use scoped_guard(device, gpio_dev) (Bartosz Golaszewski)

v4:
 - Rebase onto Linus Walleij's tree
 - Use --base so series is correctly 1/1 (Andy Shevchenko)

v3:
 - Fix variable declaration style (Andy Shevchenko)
 - Add BugLink tag (Andy Shevchenko)
 - CC AMD engineers (Andy Shevchenko)

v2:
 - Replace custom HID/UID lookup with acpi_dev_get_first_match_dev()
 - Use device_is_bound() under device_lock() with explanatory comments
 - Add Assisted-by tags per coding-assistants.rst

Kernel bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=221494
Related: https://bugzilla.kernel.org/show_bug.cgi?id=221454

Hardik Prakash (2):
  Revert "pinctrl-amd: enable IRQ for WACF2200 touchscreen on Lenovo
    Yoga 7 14AGP11"
  i2c: designware: defer probe if child GpioInt controllers are not
    bound

 drivers/i2c/busses/i2c-designware-platdrv.c | 156 ++++++++++++++++++++
 drivers/pinctrl/pinctrl-amd.c               |  35 -----
 2 files changed, 156 insertions(+), 35 deletions(-)

base-commit: 8fde5d1d47f6

[PATCH v8 1/2] Revert "pinctrl-amd: enable IRQ for WACF2200 touchscreen on Lenovo Yoga 7 14AGP11"

[Hardik Prakash]

This reverts commit 3812a9e84265a5cdd90d29fe8d97a023e91fb945.

The probe ordering fix in the following patch ensures amd_gpio_probe()
completes before i2c-designware probes AMDI0010:02, allowing the
existing amd_gpio_irq_enable() flow to work correctly. The manual IRQ
restoration added by this patch is therefore no longer needed.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=221494
Signed-off-by: Hardik Prakash <hardikprakash.official@gmail.com>
---
 drivers/pinctrl/pinctrl-amd.c | 35 -----------------------------------
 1 file changed, 35 deletions(-)

diff --git a/drivers/pinctrl/pinctrl-amd.c b/drivers/pinctrl/pinctrl-amd.c
index 64315b0edf2a..e3128b0045d2 100644
--- a/drivers/pinctrl/pinctrl-amd.c
+++ b/drivers/pinctrl/pinctrl-amd.c
@@ -26,7 +26,6 @@
 #include <linux/interrupt.h>
 #include <linux/bitops.h>
 #include <linux/pinctrl/pinconf.h>
-#include <linux/dmi.h>
 #include <linux/pinctrl/pinconf-generic.h>
 #include <linux/pinctrl/pinmux.h>
 #include <linux/string_choices.h>
@@ -40,39 +39,6 @@
 static struct amd_gpio *pinctrl_dev;
 #endif
 
-static const struct dmi_system_id amd_gpio_quirk_yoga7_14agp11[] = {
-	{
-		.matches = {
-			DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-			DMI_MATCH(DMI_PRODUCT_NAME, "83TD"),
-			DMI_MATCH(DMI_BOARD_NAME, "LNVNB161216"),
-		},
-	},
-	{ }
-};
-
-static void amd_gpio_apply_quirks(struct amd_gpio *gpio_dev)
-{
-	const unsigned int pin = 157; /* WACF2200 GpioInt per ACPI _CRS */
-	unsigned long flags;
-	u32 reg;
-
-	if (!dmi_check_system(amd_gpio_quirk_yoga7_14agp11))
-		return;
-	if (pin >= gpio_dev->gc.ngpio)
-		return;
-
-	raw_spin_lock_irqsave(&gpio_dev->lock, flags);
-	reg = readl(gpio_dev->base + pin * 4);
-	reg |= BIT(INTERRUPT_ENABLE_OFF) | BIT(INTERRUPT_MASK_OFF);
-	writel(reg, gpio_dev->base + pin * 4);
-	raw_spin_unlock_irqrestore(&gpio_dev->lock, flags);
-
-	dev_info(&gpio_dev->pdev->dev,
-		 "Enabled IRQ for GPIO %u (Yoga 7 14AGP11 touchscreen)\n",
-		 pin);
-}
-
 static int amd_gpio_get_direction(struct gpio_chip *gc, unsigned offset)
 {
 	unsigned long flags;
@@ -1253,7 +1219,6 @@ static int amd_gpio_probe(struct platform_device *pdev)
 
 	/* Disable and mask interrupts */
 	amd_gpio_irq_init(gpio_dev);
-	amd_gpio_apply_quirks(gpio_dev);
 
 	girq = &gpio_dev->gc.irq;
 	gpio_irq_chip_set_chip(girq, &amd_gpio_irqchip);
-- 
2.54.0

[PATCH v8 2/2] i2c: designware: defer probe if child GpioInt controllers are not bound

[Hardik Prakash]

I2C controllers may have child devices with GpioInt resources that
depend on GPIO controllers to be fully initialized. If the I2C
controller probes and enumerates children before the referenced GPIO
controller has completed probe, GPIO interrupts may not be properly
configured, leading to device failures.

On Lenovo Yoga 7 14AGP11, the WACF2200 touchscreen (child of
AMDI0010:02) has a GpioInt resource pointing to GPIO 157 on the
pinctrl-amd controller (AMDI0030:00). When i2c-designware probes
AMDI0010:02 before pinctrl-amd finishes initializing, I2C transactions
occur before the GPIO IRQ quirk in amd_gpio_probe() has run, causing:

  i2c_designware AMDI0010:02: i2c_dw_handle_tx_abort: lost arbitration

Add a generic dependency check in i2c-designware that walks ACPI child
devices, identifies any GpioInt resources, resolves the referenced GPIO
controllers, and defers probe if those controllers are not yet bound.

This ensures GPIO controllers complete initialization (including IRQ
setup and quirks) before I2C child enumeration begins, fixing the race
without device-specific quirks or DMI matching.

The probe ordering race was confirmed via dynamic debug tracing:

  0.285952  amd_gpio_probe: registering gpiochip  <- GPIO chip visible
  0.287121  amd_gpio_probe: requesting parent IRQ <- probe still running
  0.301454  AMDI0010:02 dw_i2c_plat_probe: start  <- races here
  2.348157  lost arbitration

Suggested-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Hardik Prakash <hardikprakash.official@gmail.com>
Acked-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
Assisted-by: Claude:claude-sonnet-4-6
Assisted-by: GPT-Codex:gpt-5.2-codex
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202605240959.Kcf1lIg4-lkp@intel.com/
Link: https://bugzilla.kernel.org/show_bug.cgi?id=221494
---
 drivers/i2c/busses/i2c-designware-platdrv.c | 156 ++++++++++++++++++++
 1 file changed, 156 insertions(+)

diff --git a/drivers/i2c/busses/i2c-designware-platdrv.c b/drivers/i2c/busses/i2c-designware-platdrv.c
index 3351c4a9ef11..1c01b0460385 100644
--- a/drivers/i2c/busses/i2c-designware-platdrv.c
+++ b/drivers/i2c/busses/i2c-designware-platdrv.c
@@ -8,6 +8,8 @@
  * Copyright (C) 2007 MontaVista Software Inc.
  * Copyright (C) 2009 Provigent Ltd.
  */
+
+#include <linux/acpi.h>
 #include <linux/clk-provider.h>
 #include <linux/clk.h>
 #include <linux/delay.h>
@@ -130,6 +132,152 @@ static int i2c_dw_probe_lock_support(struct dw_i2c_dev *dev)
 	return 0;
 }
 
+#ifdef CONFIG_ACPI
+struct gpio_dep_ctx {
+	struct list_head gpio_controllers;
+	int ret;
+};
+
+struct gpio_controller_ref {
+	struct list_head node;
+	char *path;
+};
+
+static int check_gpioint_resource(struct acpi_resource *ares, void *data)
+{
+	struct gpio_dep_ctx *ctx = data;
+	struct acpi_resource_gpio *agpio;
+	struct gpio_controller_ref *ref, *tmp;
+	bool found = false;
+
+	if (ares->type != ACPI_RESOURCE_TYPE_GPIO)
+		return 1;
+
+	agpio = &ares->data.gpio;
+	if (agpio->connection_type != ACPI_RESOURCE_GPIO_TYPE_INT)
+		return 1;
+
+	/* Check if we've already tracked this GPIO controller */
+	list_for_each_entry(tmp, &ctx->gpio_controllers, node) {
+		if (!strcmp(tmp->path, agpio->resource_source.string_ptr)) {
+			found = true;
+			break;
+		}
+	}
+
+	if (!found) {
+		ref = kzalloc(sizeof(*ref), GFP_KERNEL);
+		if (!ref) {
+			ctx->ret = -ENOMEM;
+			return 0;
+		}
+
+		ref->path = kstrdup(agpio->resource_source.string_ptr, GFP_KERNEL);
+		if (!ref->path) {
+			kfree(ref);
+			ctx->ret = -ENOMEM;
+			return 0;
+		}
+
+		list_add_tail(&ref->node, &ctx->gpio_controllers);
+	}
+
+	return 1;
+}
+
+static int check_child_gpioint(struct acpi_device *adev, void *data)
+{
+	struct gpio_dep_ctx *ctx = data;
+	struct list_head res_list;
+
+	INIT_LIST_HEAD(&res_list);
+
+	acpi_dev_get_resources(adev, &res_list, check_gpioint_resource, ctx);
+	acpi_dev_free_resource_list(&res_list);
+
+	if (ctx->ret < 0)
+		return ctx->ret;
+
+	return 0;
+}
+
+static int i2c_dw_check_gpio_dependencies(struct device *dev)
+{
+	struct acpi_device *adev = ACPI_COMPANION(dev);
+	struct gpio_dep_ctx ctx = { .ret = 0 };
+	struct gpio_controller_ref *ref, *tmp;
+	int ret = 0;
+
+	if (!adev)
+		return 0;
+
+	INIT_LIST_HEAD(&ctx.gpio_controllers);
+
+	/* Walk all child devices and collect GpioInt controller references */
+	ret = acpi_dev_for_each_child(adev, check_child_gpioint, &ctx);
+	if (ret < 0 || ctx.ret < 0) {
+		ret = ctx.ret ?: ret;
+		goto cleanup;
+	}
+
+	/* For each GPIO controller, check if its parent device is bound */
+	list_for_each_entry(ref, &ctx.gpio_controllers, node) {
+		acpi_handle handle;
+		acpi_status status;
+		struct acpi_device *gpio_adev;
+		struct device *gpio_dev;
+		bool bound;
+
+		status = acpi_get_handle(NULL, ref->path, &handle);
+		if (ACPI_FAILURE(status))
+			continue;
+
+		gpio_adev = acpi_fetch_acpi_dev(handle);
+		if (!gpio_adev)
+			continue;
+
+		gpio_dev = acpi_get_first_physical_node(gpio_adev);
+		acpi_dev_put(gpio_adev);
+
+		if (!gpio_dev) {
+			ret = -EPROBE_DEFER;
+			goto cleanup;
+		}
+
+		/*
+		 * Check if the GPIO controller's device is bound. If not,
+		 * defer probe to ensure GPIO initialization (including IRQ
+		 * setup and quirks) is complete before we enumerate I2C
+		 * child devices.
+		 */
+		scoped_guard(device, gpio_dev) {
+			bound = device_is_bound(gpio_dev);
+		}
+		if (!bound) {
+			put_device(gpio_dev);
+			ret = -EPROBE_DEFER;
+			goto cleanup;
+		}
+
+		put_device(gpio_dev);
+	}
+
+cleanup:
+	list_for_each_entry_safe(ref, tmp, &ctx.gpio_controllers, node) {
+		list_del(&ref->node);
+		kfree(ref->path);
+		kfree(ref);
+	}
+
+	return ret;
+}
+#else
+static int i2c_dw_check_gpio_dependencies(struct device *dev)
+{
+	return 0;
+}
+#endif /* CONFIG_ACPI */
+
 static int dw_i2c_plat_probe(struct platform_device *pdev)
 {
 	u32 flags = (uintptr_t)device_get_match_data(&pdev->dev);
@@ -138,6 +286,14 @@ static int dw_i2c_plat_probe(struct platform_device *pdev)
 	struct dw_i2c_dev *dev;
 	int irq, ret;
 
+	/*
+	 * Check if any child devices have GpioInt resources, and if so,
+	 * defer probe until those GPIO controllers are fully bound.
+	 */
+	ret = i2c_dw_check_gpio_dependencies(device);
+	if (ret)
+		return ret;
+
 	irq = platform_get_irq_optional(pdev, 0);
 	if (irq == -ENXIO)
 		flags |= ACCESS_POLLING;
-- 
2.54.0

Re: [PATCH v8 1/2] Revert "pinctrl-amd: enable IRQ for WACF2200 touchscreen on Lenovo Yoga 7 14AGP11"

[Mario Limonciello]

On 5/29/26 12:08, Hardik Prakash wrote:
> This reverts commit 3812a9e84265a5cdd90d29fe8d97a023e91fb945.
> 
> The probe ordering fix in the following patch ensures amd_gpio_probe()
> completes before i2c-designware probes AMDI0010:02, allowing the
> existing amd_gpio_irq_enable() flow to work correctly. The manual IRQ
> restoration added by this patch is therefore no longer needed.
> 
> Link: https://bugzilla.kernel.org/show_bug.cgi?id=221494
> Signed-off-by: Hardik Prakash <hardikprakash.official@gmail.com>

Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>

So that this reverts in older kernels probably want this:

Fixes: 3812a9e84265a ("pinctrl-amd: enable IRQ for WACF2200 touchscreen 
on Lenovo Yoga 7 14AGP11")

> ---
>   drivers/pinctrl/pinctrl-amd.c | 35 -----------------------------------
>   1 file changed, 35 deletions(-)
> 
> diff --git a/drivers/pinctrl/pinctrl-amd.c b/drivers/pinctrl/pinctrl-amd.c
> index 64315b0edf2a..e3128b0045d2 100644
> --- a/drivers/pinctrl/pinctrl-amd.c
> +++ b/drivers/pinctrl/pinctrl-amd.c
> @@ -26,7 +26,6 @@
>   #include <linux/interrupt.h>
>   #include <linux/bitops.h>
>   #include <linux/pinctrl/pinconf.h>
> -#include <linux/dmi.h>
>   #include <linux/pinctrl/pinconf-generic.h>
>   #include <linux/pinctrl/pinmux.h>
>   #include <linux/string_choices.h>
> @@ -40,39 +39,6 @@
>   static struct amd_gpio *pinctrl_dev;
>   #endif
>   
> -static const struct dmi_system_id amd_gpio_quirk_yoga7_14agp11[] = {
> -	{
> -		.matches = {
> -			DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
> -			DMI_MATCH(DMI_PRODUCT_NAME, "83TD"),
> -			DMI_MATCH(DMI_BOARD_NAME, "LNVNB161216"),
> -		},
> -	},
> -	{ }
> -};
> -
> -static void amd_gpio_apply_quirks(struct amd_gpio *gpio_dev)
> -{
> -	const unsigned int pin = 157; /* WACF2200 GpioInt per ACPI _CRS */
> -	unsigned long flags;
> -	u32 reg;
> -
> -	if (!dmi_check_system(amd_gpio_quirk_yoga7_14agp11))
> -		return;
> -	if (pin >= gpio_dev->gc.ngpio)
> -		return;
> -
> -	raw_spin_lock_irqsave(&gpio_dev->lock, flags);
> -	reg = readl(gpio_dev->base + pin * 4);
> -	reg |= BIT(INTERRUPT_ENABLE_OFF) | BIT(INTERRUPT_MASK_OFF);
> -	writel(reg, gpio_dev->base + pin * 4);
> -	raw_spin_unlock_irqrestore(&gpio_dev->lock, flags);
> -
> -	dev_info(&gpio_dev->pdev->dev,
> -		 "Enabled IRQ for GPIO %u (Yoga 7 14AGP11 touchscreen)\n",
> -		 pin);
> -}
> -
>   static int amd_gpio_get_direction(struct gpio_chip *gc, unsigned offset)
>   {
>   	unsigned long flags;
> @@ -1253,7 +1219,6 @@ static int amd_gpio_probe(struct platform_device *pdev)
>   
>   	/* Disable and mask interrupts */
>   	amd_gpio_irq_init(gpio_dev);
> -	amd_gpio_apply_quirks(gpio_dev);
>   
>   	girq = &gpio_dev->gc.irq;
>   	gpio_irq_chip_set_chip(girq, &amd_gpio_irqchip);

Re: [PATCH v8 2/2] i2c: designware: defer probe if child GpioInt controllers are not bound

[Mario Limonciello]

On 5/29/26 12:08, Hardik Prakash wrote:
> I2C controllers may have child devices with GpioInt resources that
> depend on GPIO controllers to be fully initialized. If the I2C
> controller probes and enumerates children before the referenced GPIO
> controller has completed probe, GPIO interrupts may not be properly
> configured, leading to device failures.
> 
> On Lenovo Yoga 7 14AGP11, the WACF2200 touchscreen (child of
> AMDI0010:02) has a GpioInt resource pointing to GPIO 157 on the
> pinctrl-amd controller (AMDI0030:00). When i2c-designware probes
> AMDI0010:02 before pinctrl-amd finishes initializing, I2C transactions
> occur before the GPIO IRQ quirk in amd_gpio_probe() has run, causing:
> 
>    i2c_designware AMDI0010:02: i2c_dw_handle_tx_abort: lost arbitration
> 
> Add a generic dependency check in i2c-designware that walks ACPI child
> devices, identifies any GpioInt resources, resolves the referenced GPIO
> controllers, and defers probe if those controllers are not yet bound.
> 
> This ensures GPIO controllers complete initialization (including IRQ
> setup and quirks) before I2C child enumeration begins, fixing the race
> without device-specific quirks or DMI matching.
> 
> The probe ordering race was confirmed via dynamic debug tracing:
> 
>    0.285952  amd_gpio_probe: registering gpiochip  <- GPIO chip visible
>    0.287121  amd_gpio_probe: requesting parent IRQ <- probe still running
>    0.301454  AMDI0010:02 dw_i2c_plat_probe: start  <- races here
>    2.348157  lost arbitration
> 
> Suggested-by: Mario Limonciello <mario.limonciello@amd.com>
> Signed-off-by: Hardik Prakash <hardikprakash.official@gmail.com>
> Acked-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
> Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
> Assisted-by: Claude:claude-sonnet-4-6
> Assisted-by: GPT-Codex:gpt-5.2-codex
> Reported-by: kernel test robot <lkp@intel.com>
> Closes: https://lore.kernel.org/oe-kbuild-all/202605240959.Kcf1lIg4-lkp@intel.com/
> Link: https://bugzilla.kernel.org/show_bug.cgi?id=221494

Fixes: 3812a9e84265a ("pinctrl-amd: enable IRQ for WACF2200 touchscreen 
on Lenovo Yoga 7 14AGP11")

> ---
>   drivers/i2c/busses/i2c-designware-platdrv.c | 156 ++++++++++++++++++++
>   1 file changed, 156 insertions(+)
> 
> diff --git a/drivers/i2c/busses/i2c-designware-platdrv.c b/drivers/i2c/busses/i2c-designware-platdrv.c
> index 3351c4a9ef11..1c01b0460385 100644
> --- a/drivers/i2c/busses/i2c-designware-platdrv.c
> +++ b/drivers/i2c/busses/i2c-designware-platdrv.c
> @@ -8,6 +8,8 @@
>    * Copyright (C) 2007 MontaVista Software Inc.
>    * Copyright (C) 2009 Provigent Ltd.
>    */
> +
> +#include <linux/acpi.h>
>   #include <linux/clk-provider.h>
>   #include <linux/clk.h>
>   #include <linux/delay.h>
> @@ -130,6 +132,152 @@ static int i2c_dw_probe_lock_support(struct dw_i2c_dev *dev)
>   	return 0;
>   }
>   
> +#ifdef CONFIG_ACPI
> +struct gpio_dep_ctx {
> +	struct list_head gpio_controllers;
> +	int ret;
> +};
> +
> +struct gpio_controller_ref {
> +	struct list_head node;
> +	char *path;
> +};
> +
> +static int check_gpioint_resource(struct acpi_resource *ares, void *data)
> +{
> +	struct gpio_dep_ctx *ctx = data;
> +	struct acpi_resource_gpio *agpio;
> +	struct gpio_controller_ref *ref, *tmp;
> +	bool found = false;
> +
> +	if (ares->type != ACPI_RESOURCE_TYPE_GPIO)
> +		return 1;
> +
> +	agpio = &ares->data.gpio;
> +	if (agpio->connection_type != ACPI_RESOURCE_GPIO_TYPE_INT)
> +		return 1;
> +
> +	/* Check if we've already tracked this GPIO controller */
> +	list_for_each_entry(tmp, &ctx->gpio_controllers, node) {
> +		if (!strcmp(tmp->path, agpio->resource_source.string_ptr)) {
> +			found = true;
> +			break;
> +		}
> +	}
> +
> +	if (!found) {
> +		ref = kzalloc(sizeof(*ref), GFP_KERNEL);
> +		if (!ref) {
> +			ctx->ret = -ENOMEM;
> +			return 0;
> +		}
> +
> +		ref->path = kstrdup(agpio->resource_source.string_ptr, GFP_KERNEL);
> +		if (!ref->path) {
> +			kfree(ref);
> +			ctx->ret = -ENOMEM;
> +			return 0;
> +		}
> +
> +		list_add_tail(&ref->node, &ctx->gpio_controllers);
> +	}
> +
> +	return 1;
> +}
> +
> +static int check_child_gpioint(struct acpi_device *adev, void *data)
> +{
> +	struct gpio_dep_ctx *ctx = data;
> +	struct list_head res_list;
> +
> +	INIT_LIST_HEAD(&res_list);
> +
> +	acpi_dev_get_resources(adev, &res_list, check_gpioint_resource, ctx);
> +	acpi_dev_free_resource_list(&res_list);
> +
> +	if (ctx->ret < 0)
> +		return ctx->ret;
> +
> +	return 0;
> +}
> +
> +static int i2c_dw_check_gpio_dependencies(struct device *dev)
> +{
> +	struct acpi_device *adev = ACPI_COMPANION(dev);
> +	struct gpio_dep_ctx ctx = { .ret = 0 };
> +	struct gpio_controller_ref *ref, *tmp;
> +	int ret = 0;
> +
> +	if (!adev)
> +		return 0;
> +
> +	INIT_LIST_HEAD(&ctx.gpio_controllers);
> +
> +	/* Walk all child devices and collect GpioInt controller references */
> +	ret = acpi_dev_for_each_child(adev, check_child_gpioint, &ctx);
> +	if (ret < 0 || ctx.ret < 0) {
> +		ret = ctx.ret ?: ret;
> +		goto cleanup;
> +	}
> +
> +	/* For each GPIO controller, check if its parent device is bound */
> +	list_for_each_entry(ref, &ctx.gpio_controllers, node) {
> +		acpi_handle handle;
> +		acpi_status status;
> +		struct acpi_device *gpio_adev;
> +		struct device *gpio_dev;
> +		bool bound;
> +
> +		status = acpi_get_handle(NULL, ref->path, &handle);
> +		if (ACPI_FAILURE(status))
> +			continue;
> +
> +		gpio_adev = acpi_fetch_acpi_dev(handle);
> +		if (!gpio_adev)
> +			continue;
> +
> +		gpio_dev = acpi_get_first_physical_node(gpio_adev);
> +		acpi_dev_put(gpio_adev);
> +
> +		if (!gpio_dev) {
> +			ret = -EPROBE_DEFER;
> +			goto cleanup;
> +		}
> +
> +		/*
> +		 * Check if the GPIO controller's device is bound. If not,
> +		 * defer probe to ensure GPIO initialization (including IRQ
> +		 * setup and quirks) is complete before we enumerate I2C
> +		 * child devices.
> +		 */
> +		scoped_guard(device, gpio_dev) {
> +			bound = device_is_bound(gpio_dev);
> +		}
> +		if (!bound) {
> +			put_device(gpio_dev);
> +			ret = -EPROBE_DEFER;
> +			goto cleanup;
> +		}
> +
> +		put_device(gpio_dev);
> +	}
> +
> +cleanup:
> +	list_for_each_entry_safe(ref, tmp, &ctx.gpio_controllers, node) {
> +		list_del(&ref->node);
> +		kfree(ref->path);
> +		kfree(ref);
> +	}
> +
> +	return ret;
> +}
> +#else
> +static int i2c_dw_check_gpio_dependencies(struct device *dev)
> +{
> +	return 0;
> +}
> +#endif /* CONFIG_ACPI */
> +
>   static int dw_i2c_plat_probe(struct platform_device *pdev)
>   {
>   	u32 flags = (uintptr_t)device_get_match_data(&pdev->dev);
> @@ -138,6 +286,14 @@ static int dw_i2c_plat_probe(struct platform_device *pdev)
>   	struct dw_i2c_dev *dev;
>   	int irq, ret;
>   
> +	/*
> +	 * Check if any child devices have GpioInt resources, and if so,
> +	 * defer probe until those GPIO controllers are fully bound.
> +	 */
> +	ret = i2c_dw_check_gpio_dependencies(device);
> +	if (ret)
> +		return ret;
> +
>   	irq = platform_get_irq_optional(pdev, 0);
>   	if (irq == -ENXIO)
>   		flags |= ACCESS_POLLING;

Re: [PATCH v8 0/2] i2c: designware: fix WACF2200 touchscreen probe ordering

[Mario Limonciello]

On 5/29/26 12:08, Hardik Prakash wrote:
> This series fixes the non-functional WACF2200 touchscreen on Lenovo
> Yoga 7 14AGP11 by:
> 
> 1. Reverting the earlier DMI-specific GPIO IRQ quirk (patch 1 of the
>     original series, already in Torvalds' tree) which is no longer
>     needed.
> 
> 2. Adding a generic probe deferral in i2c-designware that walks ACPI
>     child devices, finds GpioInt resources, and defers probe until the
>     referenced GPIO controllers are fully bound.
> 
> Testing confirms patch 2 alone is sufficient — the existing
> amd_gpio_irq_enable() flow works correctly once probe ordering is
> fixed, making the manual IRQ restoration in the reverted patch
> unnecessary.
> 
> The probe ordering race was confirmed via dynamic debug tracing:
> 
>    0.285952  amd_gpio_probe: registering gpiochip  <- GPIO chip visible
>    0.287121  amd_gpio_probe: requesting parent IRQ <- probe still running
>    0.301454  AMDI0010:02 dw_i2c_plat_probe: start  <- races here
>    2.348157  lost arbitration
> 
> Tested on Lenovo Yoga 7 14AGP11 (83TD), Fedora 44, kernel 7.1.0-rc5+.
> 
> v8:
>   - Add revert of now-unnecessary patch 1 (Mario Limonciello)
>   - Collect Acked-by (Bartosz Golaszewski) and Reviewed-by (Mario Limonciello)
> 
> v7:
>   - Fix unused 'ret' variable in check_child_gpioint() (kernel test robot)
>   - Replace BugLink: with Link: (Bartosz Golaszewski)
>   - Add Reported-by and Closes tags for kernel test robot warning
> 
> v6:
>   - Replace DMI-specific deferral with generic GpioInt dependency check
>     walking ACPI child devices (suggested by Mario Limonciello)
> 
> v5:
>   - Add blank line before #include <linux/acpi.h> (Bartosz Golaszewski)
>   - Use scoped_guard(device, gpio_dev) (Bartosz Golaszewski)
> 
> v4:
>   - Rebase onto Linus Walleij's tree
>   - Use --base so series is correctly 1/1 (Andy Shevchenko)
> 
> v3:
>   - Fix variable declaration style (Andy Shevchenko)
>   - Add BugLink tag (Andy Shevchenko)
>   - CC AMD engineers (Andy Shevchenko)
> 
> v2:
>   - Replace custom HID/UID lookup with acpi_dev_get_first_match_dev()
>   - Use device_is_bound() under device_lock() with explanatory comments
>   - Add Assisted-by tags per coding-assistants.rst
> 
> Kernel bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=221494
> Related: https://bugzilla.kernel.org/show_bug.cgi?id=221454
> 
> Hardik Prakash (2):
>    Revert "pinctrl-amd: enable IRQ for WACF2200 touchscreen on Lenovo
>      Yoga 7 14AGP11"
>    i2c: designware: defer probe if child GpioInt controllers are not
>      bound
> 
>   drivers/i2c/busses/i2c-designware-platdrv.c | 156 ++++++++++++++++++++
>   drivers/pinctrl/pinctrl-amd.c               |  35 -----
>   2 files changed, 156 insertions(+), 35 deletions(-)
> 
> base-commit: 8fde5d1d47f6

I've left some tags on your series, no need to resend for them though if 
no other feedback.

They should be able to be picked up by the maintainers with b4.

Re: [PATCH v8 0/2] i2c: designware: fix WACF2200 touchscreen probe ordering

[Wolfram Sang]

[-- Attachment #1: Type: text/plain, Size: 1170 bytes --]

On Fri, May 29, 2026 at 03:38:35PM +0530, Hardik Prakash wrote:
> This series fixes the non-functional WACF2200 touchscreen on Lenovo
> Yoga 7 14AGP11 by:
> 
> 1. Reverting the earlier DMI-specific GPIO IRQ quirk (patch 1 of the
>    original series, already in Torvalds' tree) which is no longer
>    needed.
> 
> 2. Adding a generic probe deferral in i2c-designware that walks ACPI
>    child devices, finds GpioInt resources, and defers probe until the
>    referenced GPIO controllers are fully bound.
> 
> Testing confirms patch 2 alone is sufficient — the existing
> amd_gpio_irq_enable() flow works correctly once probe ordering is
> fixed, making the manual IRQ restoration in the reverted patch
> unnecessary.
> 
> The probe ordering race was confirmed via dynamic debug tracing:
> 
>   0.285952  amd_gpio_probe: registering gpiochip  <- GPIO chip visible
>   0.287121  amd_gpio_probe: requesting parent IRQ <- probe still running
>   0.301454  AMDI0010:02 dw_i2c_plat_probe: start  <- races here
>   2.348157  lost arbitration
> 
> Tested on Lenovo Yoga 7 14AGP11 (83TD), Fedora 44, kernel 7.1.0-rc5+.

Shall this go in via i2c?


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [PATCH v8 2/2] i2c: designware: defer probe if child GpioInt controllers are not bound

[Wolfram Sang]

[-- Attachment #1: Type: text/plain, Size: 2388 bytes --]

On Fri, May 29, 2026 at 12:12:31PM +0200, Mario Limonciello wrote:
> 
> 
> On 5/29/26 12:08, Hardik Prakash wrote:
> > I2C controllers may have child devices with GpioInt resources that
> > depend on GPIO controllers to be fully initialized. If the I2C
> > controller probes and enumerates children before the referenced GPIO
> > controller has completed probe, GPIO interrupts may not be properly
> > configured, leading to device failures.
> > 
> > On Lenovo Yoga 7 14AGP11, the WACF2200 touchscreen (child of
> > AMDI0010:02) has a GpioInt resource pointing to GPIO 157 on the
> > pinctrl-amd controller (AMDI0030:00). When i2c-designware probes
> > AMDI0010:02 before pinctrl-amd finishes initializing, I2C transactions
> > occur before the GPIO IRQ quirk in amd_gpio_probe() has run, causing:
> > 
> >    i2c_designware AMDI0010:02: i2c_dw_handle_tx_abort: lost arbitration
> > 
> > Add a generic dependency check in i2c-designware that walks ACPI child
> > devices, identifies any GpioInt resources, resolves the referenced GPIO
> > controllers, and defers probe if those controllers are not yet bound.
> > 
> > This ensures GPIO controllers complete initialization (including IRQ
> > setup and quirks) before I2C child enumeration begins, fixing the race
> > without device-specific quirks or DMI matching.
> > 
> > The probe ordering race was confirmed via dynamic debug tracing:
> > 
> >    0.285952  amd_gpio_probe: registering gpiochip  <- GPIO chip visible
> >    0.287121  amd_gpio_probe: requesting parent IRQ <- probe still running
> >    0.301454  AMDI0010:02 dw_i2c_plat_probe: start  <- races here
> >    2.348157  lost arbitration
> > 
> > Suggested-by: Mario Limonciello <mario.limonciello@amd.com>
> > Signed-off-by: Hardik Prakash <hardikprakash.official@gmail.com>
> > Acked-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
> > Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
> > Assisted-by: Claude:claude-sonnet-4-6
> > Assisted-by: GPT-Codex:gpt-5.2-codex
> > Reported-by: kernel test robot <lkp@intel.com>
> > Closes: https://lore.kernel.org/oe-kbuild-all/202605240959.Kcf1lIg4-lkp@intel.com/
> > Link: https://bugzilla.kernel.org/show_bug.cgi?id=221494
> 
> Fixes: 3812a9e84265a ("pinctrl-amd: enable IRQ for WACF2200 touchscreen on
> Lenovo Yoga 7 14AGP11")

No Rev-by from you, Mario?


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [PATCH v8 2/2] i2c: designware: defer probe if child GpioInt controllers are not bound

[Wolfram Sang]

[-- Attachment #1: Type: text/plain, Size: 2220 bytes --]

On Fri, May 29, 2026 at 03:38:37PM +0530, Hardik Prakash wrote:
> I2C controllers may have child devices with GpioInt resources that
> depend on GPIO controllers to be fully initialized. If the I2C
> controller probes and enumerates children before the referenced GPIO
> controller has completed probe, GPIO interrupts may not be properly
> configured, leading to device failures.
> 
> On Lenovo Yoga 7 14AGP11, the WACF2200 touchscreen (child of
> AMDI0010:02) has a GpioInt resource pointing to GPIO 157 on the
> pinctrl-amd controller (AMDI0030:00). When i2c-designware probes
> AMDI0010:02 before pinctrl-amd finishes initializing, I2C transactions
> occur before the GPIO IRQ quirk in amd_gpio_probe() has run, causing:
> 
>   i2c_designware AMDI0010:02: i2c_dw_handle_tx_abort: lost arbitration
> 
> Add a generic dependency check in i2c-designware that walks ACPI child
> devices, identifies any GpioInt resources, resolves the referenced GPIO
> controllers, and defers probe if those controllers are not yet bound.
> 
> This ensures GPIO controllers complete initialization (including IRQ
> setup and quirks) before I2C child enumeration begins, fixing the race
> without device-specific quirks or DMI matching.
> 
> The probe ordering race was confirmed via dynamic debug tracing:
> 
>   0.285952  amd_gpio_probe: registering gpiochip  <- GPIO chip visible
>   0.287121  amd_gpio_probe: requesting parent IRQ <- probe still running
>   0.301454  AMDI0010:02 dw_i2c_plat_probe: start  <- races here
>   2.348157  lost arbitration
> 
> Suggested-by: Mario Limonciello <mario.limonciello@amd.com>
> Signed-off-by: Hardik Prakash <hardikprakash.official@gmail.com>
> Acked-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
> Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
> Assisted-by: Claude:claude-sonnet-4-6
> Assisted-by: GPT-Codex:gpt-5.2-codex
> Reported-by: kernel test robot <lkp@intel.com>
> Closes: https://lore.kernel.org/oe-kbuild-all/202605240959.Kcf1lIg4-lkp@intel.com/
> Link: https://bugzilla.kernel.org/show_bug.cgi?id=221494

In case this goes in via some other tree:

Acked-by: Wolfram Sang <wsa+renesas@sang-engineering.com>


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [PATCH v8 2/2] i2c: designware: defer probe if child GpioInt controllers are not bound

[Mario Limonciello]

On 5/29/26 16:42, Wolfram Sang wrote:
> On Fri, May 29, 2026 at 12:12:31PM +0200, Mario Limonciello wrote:
>>
>>
>> On 5/29/26 12:08, Hardik Prakash wrote:
>>> I2C controllers may have child devices with GpioInt resources that
>>> depend on GPIO controllers to be fully initialized. If the I2C
>>> controller probes and enumerates children before the referenced GPIO
>>> controller has completed probe, GPIO interrupts may not be properly
>>> configured, leading to device failures.
>>>
>>> On Lenovo Yoga 7 14AGP11, the WACF2200 touchscreen (child of
>>> AMDI0010:02) has a GpioInt resource pointing to GPIO 157 on the
>>> pinctrl-amd controller (AMDI0030:00). When i2c-designware probes
>>> AMDI0010:02 before pinctrl-amd finishes initializing, I2C transactions
>>> occur before the GPIO IRQ quirk in amd_gpio_probe() has run, causing:
>>>
>>>     i2c_designware AMDI0010:02: i2c_dw_handle_tx_abort: lost arbitration
>>>
>>> Add a generic dependency check in i2c-designware that walks ACPI child
>>> devices, identifies any GpioInt resources, resolves the referenced GPIO
>>> controllers, and defers probe if those controllers are not yet bound.
>>>
>>> This ensures GPIO controllers complete initialization (including IRQ
>>> setup and quirks) before I2C child enumeration begins, fixing the race
>>> without device-specific quirks or DMI matching.
>>>
>>> The probe ordering race was confirmed via dynamic debug tracing:
>>>
>>>     0.285952  amd_gpio_probe: registering gpiochip  <- GPIO chip visible
>>>     0.287121  amd_gpio_probe: requesting parent IRQ <- probe still running
>>>     0.301454  AMDI0010:02 dw_i2c_plat_probe: start  <- races here
>>>     2.348157  lost arbitration
>>>
>>> Suggested-by: Mario Limonciello <mario.limonciello@amd.com>
>>> Signed-off-by: Hardik Prakash <hardikprakash.official@gmail.com>
>>> Acked-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
>>> Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
>>> Assisted-by: Claude:claude-sonnet-4-6
>>> Assisted-by: GPT-Codex:gpt-5.2-codex
>>> Reported-by: kernel test robot <lkp@intel.com>
>>> Closes: https://lore.kernel.org/oe-kbuild-all/202605240959.Kcf1lIg4-lkp@intel.com/
>>> Link: https://bugzilla.kernel.org/show_bug.cgi?id=221494
>>
>> Fixes: 3812a9e84265a ("pinctrl-amd: enable IRQ for WACF2200 touchscreen on
>> Lenovo Yoga 7 14AGP11")
> 
> No Rev-by from you, Mario?
> 

I added to the individual patches instead, all LGTM.

Re: [PATCH v8 1/2] Revert "pinctrl-amd: enable IRQ for WACF2200 touchscreen on Lenovo Yoga 7 14AGP11"

[Linus Walleij]

On Fri, May 29, 2026 at 12:09 PM Hardik Prakash
<hardikprakash.official@gmail.com> wrote:

> This reverts commit 3812a9e84265a5cdd90d29fe8d97a023e91fb945.
>
> The probe ordering fix in the following patch ensures amd_gpio_probe()
> completes before i2c-designware probes AMDI0010:02, allowing the
> existing amd_gpio_irq_enable() flow to work correctly. The manual IRQ
> restoration added by this patch is therefore no longer needed.
>
> Link: https://bugzilla.kernel.org/show_bug.cgi?id=221494
> Signed-off-by: Hardik Prakash <hardikprakash.official@gmail.com>

Patch applied for fixes.

Yours,
Linus Walleij

Re: [PATCH v8 2/2] i2c: designware: defer probe if child GpioInt controllers are not bound

[Linus Walleij]

On Fri, May 29, 2026 at 4:43 PM Wolfram Sang
<wsa+renesas@sang-engineering.com> wrote:

> In case this goes in via some other tree:
>
> Acked-by: Wolfram Sang <wsa+renesas@sang-engineering.com>

I queued the revert in the pinctrl fixes, you can take this one (2/2).

Yours,
Linus Walleij

Re: [PATCH v8 2/2] i2c: designware: defer probe if child GpioInt controllers are not bound

[Wolfram Sang]

[-- Attachment #1: Type: text/plain, Size: 288 bytes --]


> > In case this goes in via some other tree:
> >
> > Acked-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
> 
> I queued the revert in the pinctrl fixes, you can take this one (2/2).

Is there really no dependency on patch 1? I'd feel safer if you'd take
this as well...


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [PATCH v8 2/2] i2c: designware: defer probe if child GpioInt controllers are not bound

[Linus Walleij]

On Fri, May 29, 2026 at 10:55 PM Wolfram Sang
<wsa+renesas@sang-engineering.com> wrote:
> > > In case this goes in via some other tree:
> > >
> > > Acked-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
> >
> > I queued the revert in the pinctrl fixes, you can take this one (2/2).
>
> Is there really no dependency on patch 1? I'd feel safer if you'd take
> this as well...

OK then, you're right. Let's keep it together.

Patch applied!

Yours,
Linus Walleij

Re: [PATCH v8 2/2] i2c: designware: defer probe if child GpioInt controllers are not bound

[Wolfram Sang]

[-- Attachment #1: Type: text/plain, Size: 60 bytes --]


> OK then, you're right. Let's keep it together.

Thanks!


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [PATCH v8 2/2] i2c: designware: defer probe if child GpioInt controllers are not bound

[Nathan Chancellor]

Hi Hardik,

On Fri, May 29, 2026 at 03:38:37PM +0530, Hardik Prakash wrote:
> I2C controllers may have child devices with GpioInt resources that
> depend on GPIO controllers to be fully initialized. If the I2C
> controller probes and enumerates children before the referenced GPIO
> controller has completed probe, GPIO interrupts may not be properly
> configured, leading to device failures.
> 
> On Lenovo Yoga 7 14AGP11, the WACF2200 touchscreen (child of
> AMDI0010:02) has a GpioInt resource pointing to GPIO 157 on the
> pinctrl-amd controller (AMDI0030:00). When i2c-designware probes
> AMDI0010:02 before pinctrl-amd finishes initializing, I2C transactions
> occur before the GPIO IRQ quirk in amd_gpio_probe() has run, causing:
> 
>   i2c_designware AMDI0010:02: i2c_dw_handle_tx_abort: lost arbitration
> 
> Add a generic dependency check in i2c-designware that walks ACPI child
> devices, identifies any GpioInt resources, resolves the referenced GPIO
> controllers, and defers probe if those controllers are not yet bound.
> 
> This ensures GPIO controllers complete initialization (including IRQ
> setup and quirks) before I2C child enumeration begins, fixing the race
> without device-specific quirks or DMI matching.
> 
> The probe ordering race was confirmed via dynamic debug tracing:
> 
>   0.285952  amd_gpio_probe: registering gpiochip  <- GPIO chip visible
>   0.287121  amd_gpio_probe: requesting parent IRQ <- probe still running
>   0.301454  AMDI0010:02 dw_i2c_plat_probe: start  <- races here
>   2.348157  lost arbitration
> 
> Suggested-by: Mario Limonciello <mario.limonciello@amd.com>
> Signed-off-by: Hardik Prakash <hardikprakash.official@gmail.com>
> Acked-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
> Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
> Assisted-by: Claude:claude-sonnet-4-6
> Assisted-by: GPT-Codex:gpt-5.2-codex
> Reported-by: kernel test robot <lkp@intel.com>
> Closes: https://lore.kernel.org/oe-kbuild-all/202605240959.Kcf1lIg4-lkp@intel.com/
> Link: https://bugzilla.kernel.org/show_bug.cgi?id=221494

I bisected boot issues with a few of my test machines to this change in -next
as commit ef76a3a28c79 ("i2c: designware: defer probe if child GpioInt
controllers are not bound"). I consistently get no display output and I do not
have serial access to these machines, so I do not have much to go on here at
the moment. Occasionally, the network controller will be initialized so I can
remote in but I have not been able to do it recently to try and gather logs.
What information would be useful for trying to figure out what is going on
here? I seem to recall a crash in strcmp() in one log but I cannot be too sure
since it was late at night when I was doing my initial triage.

# bad: [08484c504b55a98bd100527fbe10a3caf55ff3ff] Add linux-next specific files for 20260601
# good: [e43ffb69e0438cddd72aaa30898b4dc446f664f8] Linux 7.1-rc6
git bisect start '08484c504b55a98bd100527fbe10a3caf55ff3ff' 'e43ffb69e0438cddd72aaa30898b4dc446f664f8'
# good: [57fb910db1b6051476b91a4d4ae18bc027a7f36d] Merge branch 'master' of https://git.kernel.org/pub/scm/linux/kernel/git/herbert/cryptodev-2.6.git
git bisect good 57fb910db1b6051476b91a4d4ae18bc027a7f36d
# good: [18ea7e3ca91e7a9eb9e43c11154350cec160830b] Merge branch 'for-next' of https://git.kernel.org/pub/scm/linux/kernel/git/jassibrar/mailbox.git
git bisect good 18ea7e3ca91e7a9eb9e43c11154350cec160830b
# good: [a5696dd9c080352e25b576acf24450cef255b6eb] Merge branch 'tty-next' of https://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty.git
git bisect good a5696dd9c080352e25b576acf24450cef255b6eb
# good: [c158564f40f0d2a68833610ac136f9714fc51b16] Merge branch 'staging-next' of https://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging.git
git bisect good c158564f40f0d2a68833610ac136f9714fc51b16
# bad: [4240c928a5edeb13fb3953248dfc45e5d9f42f00] Merge branch 'for-next' of https://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl.git
git bisect bad 4240c928a5edeb13fb3953248dfc45e5d9f42f00
# good: [2e24f5964b72ef213ecdb959a0b0d60b2dc00a3b] Merge branch 'for-next' of https://git.kernel.org/pub/scm/linux/kernel/git/remoteproc/linux.git
git bisect good 2e24f5964b72ef213ecdb959a0b0d60b2dc00a3b
# good: [6a350ccc4dc3f46ed2ee2592e3050956e415ef64] pinctrl: add new generic groups/function creation function for pinmux
git bisect good 6a350ccc4dc3f46ed2ee2592e3050956e415ef64
# good: [b12e12ee4138e30d786eda02223e87044c989bb1] gpiolib: Mark gpio_devt, gpiolib_initialized and gpio_stub_drv as __ro_after_init
git bisect good b12e12ee4138e30d786eda02223e87044c989bb1
# good: [446fa334d186316e76cbdc4e94e42af7d040a79c] pinctrl: qcom: Replace open coded eoi call with irq_chip_eoi_parent()
git bisect good 446fa334d186316e76cbdc4e94e42af7d040a79c
# good: [a8754838f83a9905af516f38dd2633744a94f71a] gpio: max77620: Unify usage of space and comma in platform_device_id array
git bisect good a8754838f83a9905af516f38dd2633744a94f71a
# bad: [ef76a3a28c79b628890431aa344af633e892035b] i2c: designware: defer probe if child GpioInt controllers are not bound
git bisect bad ef76a3a28c79b628890431aa344af633e892035b
# good: [b0c13ec17438577f90b379d448dfed1233e2c0a4] pinctrl: mcp23s08: Read spi-present-mask as u8 not u32
git bisect good b0c13ec17438577f90b379d448dfed1233e2c0a4
# good: [3f786abd23951f3f600a62fef42469d9200d5f52] Revert "pinctrl-amd: enable IRQ for WACF2200 touchscreen on Lenovo Yoga 7 14AGP11"
git bisect good 3f786abd23951f3f600a62fef42469d9200d5f52
# first bad commit: [ef76a3a28c79b628890431aa344af633e892035b] i2c: designware: defer probe if child GpioInt controllers are not bound

-- 
Cheers,
Nathan

Re: [PATCH v8 2/2] i2c: designware: defer probe if child GpioInt controllers are not bound

[Hardik Prakash]

On Wed, Jun 03, 2026 at 00:23, Nathan Chancellor wrote:
> I bisected boot issues with a few of my test machines to this change
> in -next as commit ef76a3a28c79. I seem to recall a crash in strcmp()
> in one log but I cannot be too sure.

Thank you for bisecting this.

The strcmp() crash is likely the issue. In
check_gpioint_resource(), we access agpio->resource_source.string_ptr
without checking whether it is NULL first:

    if (!strcmp(tmp->path, agpio->resource_source.string_ptr))
    ref->path = kstrdup(agpio->resource_source.string_ptr, ...)

The acpi_resource_source struct has a string_length field -- when it is
0, string_ptr will be NULL. On your machines, likely some GPIO resource
apparently has no resource source string, which we did not account for.

I am preparing a fix that skips GPIO resources with no resource source
string (string_length == 0 || string_ptr == NULL). I will test it on
my hardware first and send a patch shortly.

Sorry for the regression.

Thanks,
Hardik

On Wed, 3 Jun 2026 at 00:23, Nathan Chancellor <nathan@kernel.org> wrote:
>
> Hi Hardik,
>
> On Fri, May 29, 2026 at 03:38:37PM +0530, Hardik Prakash wrote:
> > I2C controllers may have child devices with GpioInt resources that
> > depend on GPIO controllers to be fully initialized. If the I2C
> > controller probes and enumerates children before the referenced GPIO
> > controller has completed probe, GPIO interrupts may not be properly
> > configured, leading to device failures.
> >
> > On Lenovo Yoga 7 14AGP11, the WACF2200 touchscreen (child of
> > AMDI0010:02) has a GpioInt resource pointing to GPIO 157 on the
> > pinctrl-amd controller (AMDI0030:00). When i2c-designware probes
> > AMDI0010:02 before pinctrl-amd finishes initializing, I2C transactions
> > occur before the GPIO IRQ quirk in amd_gpio_probe() has run, causing:
> >
> >   i2c_designware AMDI0010:02: i2c_dw_handle_tx_abort: lost arbitration
> >
> > Add a generic dependency check in i2c-designware that walks ACPI child
> > devices, identifies any GpioInt resources, resolves the referenced GPIO
> > controllers, and defers probe if those controllers are not yet bound.
> >
> > This ensures GPIO controllers complete initialization (including IRQ
> > setup and quirks) before I2C child enumeration begins, fixing the race
> > without device-specific quirks or DMI matching.
> >
> > The probe ordering race was confirmed via dynamic debug tracing:
> >
> >   0.285952  amd_gpio_probe: registering gpiochip  <- GPIO chip visible
> >   0.287121  amd_gpio_probe: requesting parent IRQ <- probe still running
> >   0.301454  AMDI0010:02 dw_i2c_plat_probe: start  <- races here
> >   2.348157  lost arbitration
> >
> > Suggested-by: Mario Limonciello <mario.limonciello@amd.com>
> > Signed-off-by: Hardik Prakash <hardikprakash.official@gmail.com>
> > Acked-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
> > Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
> > Assisted-by: Claude:claude-sonnet-4-6
> > Assisted-by: GPT-Codex:gpt-5.2-codex
> > Reported-by: kernel test robot <lkp@intel.com>
> > Closes: https://lore.kernel.org/oe-kbuild-all/202605240959.Kcf1lIg4-lkp@intel.com/
> > Link: https://bugzilla.kernel.org/show_bug.cgi?id=221494
>
> I bisected boot issues with a few of my test machines to this change in -next
> as commit ef76a3a28c79 ("i2c: designware: defer probe if child GpioInt
> controllers are not bound"). I consistently get no display output and I do not
> have serial access to these machines, so I do not have much to go on here at
> the moment. Occasionally, the network controller will be initialized so I can
> remote in but I have not been able to do it recently to try and gather logs.
> What information would be useful for trying to figure out what is going on
> here? I seem to recall a crash in strcmp() in one log but I cannot be too sure
> since it was late at night when I was doing my initial triage.
>
> # bad: [08484c504b55a98bd100527fbe10a3caf55ff3ff] Add linux-next specific files for 20260601
> # good: [e43ffb69e0438cddd72aaa30898b4dc446f664f8] Linux 7.1-rc6
> git bisect start '08484c504b55a98bd100527fbe10a3caf55ff3ff' 'e43ffb69e0438cddd72aaa30898b4dc446f664f8'
> # good: [57fb910db1b6051476b91a4d4ae18bc027a7f36d] Merge branch 'master' of https://git.kernel.org/pub/scm/linux/kernel/git/herbert/cryptodev-2.6.git
> git bisect good 57fb910db1b6051476b91a4d4ae18bc027a7f36d
> # good: [18ea7e3ca91e7a9eb9e43c11154350cec160830b] Merge branch 'for-next' of https://git.kernel.org/pub/scm/linux/kernel/git/jassibrar/mailbox.git
> git bisect good 18ea7e3ca91e7a9eb9e43c11154350cec160830b
> # good: [a5696dd9c080352e25b576acf24450cef255b6eb] Merge branch 'tty-next' of https://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty.git
> git bisect good a5696dd9c080352e25b576acf24450cef255b6eb
> # good: [c158564f40f0d2a68833610ac136f9714fc51b16] Merge branch 'staging-next' of https://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging.git
> git bisect good c158564f40f0d2a68833610ac136f9714fc51b16
> # bad: [4240c928a5edeb13fb3953248dfc45e5d9f42f00] Merge branch 'for-next' of https://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl.git
> git bisect bad 4240c928a5edeb13fb3953248dfc45e5d9f42f00
> # good: [2e24f5964b72ef213ecdb959a0b0d60b2dc00a3b] Merge branch 'for-next' of https://git.kernel.org/pub/scm/linux/kernel/git/remoteproc/linux.git
> git bisect good 2e24f5964b72ef213ecdb959a0b0d60b2dc00a3b
> # good: [6a350ccc4dc3f46ed2ee2592e3050956e415ef64] pinctrl: add new generic groups/function creation function for pinmux
> git bisect good 6a350ccc4dc3f46ed2ee2592e3050956e415ef64
> # good: [b12e12ee4138e30d786eda02223e87044c989bb1] gpiolib: Mark gpio_devt, gpiolib_initialized and gpio_stub_drv as __ro_after_init
> git bisect good b12e12ee4138e30d786eda02223e87044c989bb1
> # good: [446fa334d186316e76cbdc4e94e42af7d040a79c] pinctrl: qcom: Replace open coded eoi call with irq_chip_eoi_parent()
> git bisect good 446fa334d186316e76cbdc4e94e42af7d040a79c
> # good: [a8754838f83a9905af516f38dd2633744a94f71a] gpio: max77620: Unify usage of space and comma in platform_device_id array
> git bisect good a8754838f83a9905af516f38dd2633744a94f71a
> # bad: [ef76a3a28c79b628890431aa344af633e892035b] i2c: designware: defer probe if child GpioInt controllers are not bound
> git bisect bad ef76a3a28c79b628890431aa344af633e892035b
> # good: [b0c13ec17438577f90b379d448dfed1233e2c0a4] pinctrl: mcp23s08: Read spi-present-mask as u8 not u32
> git bisect good b0c13ec17438577f90b379d448dfed1233e2c0a4
> # good: [3f786abd23951f3f600a62fef42469d9200d5f52] Revert "pinctrl-amd: enable IRQ for WACF2200 touchscreen on Lenovo Yoga 7 14AGP11"
> git bisect good 3f786abd23951f3f600a62fef42469d9200d5f52
> # first bad commit: [ef76a3a28c79b628890431aa344af633e892035b] i2c: designware: defer probe if child GpioInt controllers are not bound
>
> --
> Cheers,
> Nathan

Re: [PATCH v8 2/2] i2c: designware: defer probe if child GpioInt controllers are not bound

[Andy Shevchenko]

On Fri, May 29, 2026 at 12:12:31PM +0200, Mario Limonciello wrote:
> On 5/29/26 12:08, Hardik Prakash wrote:
> > I2C controllers may have child devices with GpioInt resources that
> > depend on GPIO controllers to be fully initialized. If the I2C
> > controller probes and enumerates children before the referenced GPIO
> > controller has completed probe, GPIO interrupts may not be properly
> > configured, leading to device failures.
> > 
> > On Lenovo Yoga 7 14AGP11, the WACF2200 touchscreen (child of
> > AMDI0010:02) has a GpioInt resource pointing to GPIO 157 on the
> > pinctrl-amd controller (AMDI0030:00). When i2c-designware probes
> > AMDI0010:02 before pinctrl-amd finishes initializing, I2C transactions
> > occur before the GPIO IRQ quirk in amd_gpio_probe() has run, causing:
> > 
> >    i2c_designware AMDI0010:02: i2c_dw_handle_tx_abort: lost arbitration
> > 
> > Add a generic dependency check in i2c-designware that walks ACPI child
> > devices, identifies any GpioInt resources, resolves the referenced GPIO
> > controllers, and defers probe if those controllers are not yet bound.
> > 
> > This ensures GPIO controllers complete initialization (including IRQ
> > setup and quirks) before I2C child enumeration begins, fixing the race
> > without device-specific quirks or DMI matching.
> > 
> > The probe ordering race was confirmed via dynamic debug tracing:
> > 
> >    0.285952  amd_gpio_probe: registering gpiochip  <- GPIO chip visible
> >    0.287121  amd_gpio_probe: requesting parent IRQ <- probe still running
> >    0.301454  AMDI0010:02 dw_i2c_plat_probe: start  <- races here
> >    2.348157  lost arbitration
> > 
> > Suggested-by: Mario Limonciello <mario.limonciello@amd.com>
> > Signed-off-by: Hardik Prakash <hardikprakash.official@gmail.com>
> > Acked-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
> > Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
> > Assisted-by: Claude:claude-sonnet-4-6
> > Assisted-by: GPT-Codex:gpt-5.2-codex

> > Reported-by: kernel test robot <lkp@intel.com>
> > Closes: https://lore.kernel.org/oe-kbuild-all/202605240959.Kcf1lIg4-lkp@intel.com/

This patch definitely wasn't the result of the above report. These two tags are
misused.

> > Link: https://bugzilla.kernel.org/show_bug.cgi?id=221494

> Fixes: 3812a9e84265a ("pinctrl-amd: enable IRQ for WACF2200 touchscreen on
> Lenovo Yoga 7 14AGP11")

Must be oneline.


-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH v8 2/2] i2c: designware: defer probe if child GpioInt controllers are not bound

[Andy Shevchenko]

On Fri, May 29, 2026 at 11:59:24PM +0200, Linus Walleij wrote:
> On Fri, May 29, 2026 at 10:55 PM Wolfram Sang
> <wsa+renesas@sang-engineering.com> wrote:
> > > > In case this goes in via some other tree:
> > > >
> > > > Acked-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
> > >
> > > I queued the revert in the pinctrl fixes, you can take this one (2/2).
> >
> > Is there really no dependency on patch 1? I'd feel safer if you'd take
> > this as well...
> 
> OK then, you're right. Let's keep it together.
> 
> Patch applied!

Can we drop this patch or revert? I would like (as I said earlier) to review it
closer, I just returned from conferences and vacation and got a ~16 hundreds of
new emails to go through.

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH v8 2/2] i2c: designware: defer probe if child GpioInt controllers are not bound

[Andy Shevchenko]

On Fri, May 29, 2026 at 03:38:37PM +0530, Hardik Prakash wrote:
> I2C controllers may have child devices with GpioInt resources that
> depend on GPIO controllers to be fully initialized. If the I2C
> controller probes and enumerates children before the referenced GPIO
> controller has completed probe, GPIO interrupts may not be properly
> configured, leading to device failures.
> 
> On Lenovo Yoga 7 14AGP11, the WACF2200 touchscreen (child of
> AMDI0010:02) has a GpioInt resource pointing to GPIO 157 on the
> pinctrl-amd controller (AMDI0030:00). When i2c-designware probes
> AMDI0010:02 before pinctrl-amd finishes initializing, I2C transactions
> occur before the GPIO IRQ quirk in amd_gpio_probe() has run, causing:
> 
>   i2c_designware AMDI0010:02: i2c_dw_handle_tx_abort: lost arbitration
> 
> Add a generic dependency check in i2c-designware that walks ACPI child
> devices, identifies any GpioInt resources, resolves the referenced GPIO
> controllers, and defers probe if those controllers are not yet bound.
> 
> This ensures GPIO controllers complete initialization (including IRQ
> setup and quirks) before I2C child enumeration begins, fixing the race
> without device-specific quirks or DMI matching.
> 
> The probe ordering race was confirmed via dynamic debug tracing:
> 
>   0.285952  amd_gpio_probe: registering gpiochip  <- GPIO chip visible
>   0.287121  amd_gpio_probe: requesting parent IRQ <- probe still running
>   0.301454  AMDI0010:02 dw_i2c_plat_probe: start  <- races here
>   2.348157  lost arbitration
> 
> Suggested-by: Mario Limonciello <mario.limonciello@amd.com>
> Signed-off-by: Hardik Prakash <hardikprakash.official@gmail.com>
> Acked-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
> Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
> Assisted-by: Claude:claude-sonnet-4-6
> Assisted-by: GPT-Codex:gpt-5.2-codex

> Reported-by: kernel test robot <lkp@intel.com>
> Closes: https://lore.kernel.org/oe-kbuild-all/202605240959.Kcf1lIg4-lkp@intel.com/

These two tags are misused.

> Link: https://bugzilla.kernel.org/show_bug.cgi?id=221494

...

> +#ifdef CONFIG_ACPI
> +struct gpio_dep_ctx {
> +	struct list_head gpio_controllers;
> +	int ret;
> +};

Useless data type. You can use list_head directly as a context. See below.

> +struct gpio_controller_ref {
> +	struct list_head node;
> +	char *path;

No const?

> +};
> +
> +static int check_gpioint_resource(struct acpi_resource *ares, void *data)
> +{
> +	struct gpio_dep_ctx *ctx = data;
> +	struct acpi_resource_gpio *agpio;
> +	struct gpio_controller_ref *ref, *tmp;
> +	bool found = false;

> +	if (ares->type != ACPI_RESOURCE_TYPE_GPIO)
> +		return 1;
> +
> +	agpio = &ares->data.gpio;
> +	if (agpio->connection_type != ACPI_RESOURCE_GPIO_TYPE_INT)
> +		return 1;

The idea is not to use the above outside of the (GPIO) core parts.

> +	/* Check if we've already tracked this GPIO controller */
> +	list_for_each_entry(tmp, &ctx->gpio_controllers, node) {
> +		if (!strcmp(tmp->path, agpio->resource_source.string_ptr)) {
> +			found = true;
> +			break;
> +		}
> +	}
> +
> +	if (!found) {
> +		ref = kzalloc(sizeof(*ref), GFP_KERNEL);
> +		if (!ref) {
> +			ctx->ret = -ENOMEM;
> +			return 0;
> +		}
> +
> +		ref->path = kstrdup(agpio->resource_source.string_ptr, GFP_KERNEL);
> +		if (!ref->path) {
> +			kfree(ref);
> +			ctx->ret = -ENOMEM;
> +			return 0;
> +		}
> +
> +		list_add_tail(&ref->node, &ctx->gpio_controllers);
> +	}
> +
> +	return 1;

This is a duplication with gpiolib-acpi-core.c.

> +}
> +
> +static int check_child_gpioint(struct acpi_device *adev, void *data)
> +{
> +	struct gpio_dep_ctx *ctx = data;
> +	struct list_head res_list;
> +
> +	INIT_LIST_HEAD(&res_list);
> +
> +	acpi_dev_get_resources(adev, &res_list, check_gpioint_resource, ctx);
> +	acpi_dev_free_resource_list(&res_list);

> +	if (ctx->ret < 0)

How do we clear partially filled gpio_controllers list?

> +		return ctx->ret;

Huh?! Use returned value from acpi_dev_get_resources().

> +	return 0;
> +}
> +
> +static int i2c_dw_check_gpio_dependencies(struct device *dev)
> +{
> +	struct acpi_device *adev = ACPI_COMPANION(dev);
> +	struct gpio_dep_ctx ctx = { .ret = 0 };

Unneeded assignment, {} will suffice.

> +	struct gpio_controller_ref *ref, *tmp;
> +	int ret = 0;


Split assignment and move it here

	adev = ACPI_COMPANION(...);
> +	if (!adev)
> +		return 0;
> +
> +	INIT_LIST_HEAD(&ctx.gpio_controllers);
> +
> +	/* Walk all child devices and collect GpioInt controller references */
> +	ret = acpi_dev_for_each_child(adev, check_child_gpioint, &ctx);
> +	if (ret < 0 || ctx.ret < 0) {
> +		ret = ctx.ret ?: ret;
> +		goto cleanup;
> +	}
> +
> +	/* For each GPIO controller, check if its parent device is bound */
> +	list_for_each_entry(ref, &ctx.gpio_controllers, node) {
> +		acpi_handle handle;
> +		acpi_status status;
> +		struct acpi_device *gpio_adev;
> +		struct device *gpio_dev;
> +		bool bound;
> +
> +		status = acpi_get_handle(NULL, ref->path, &handle);
> +		if (ACPI_FAILURE(status))
> +			continue;
> +
> +		gpio_adev = acpi_fetch_acpi_dev(handle);
> +		if (!gpio_adev)
> +			continue;

Again, a lot of duplication with gpiolib-acpi-core.c. Please, find a way how to
reuse the code.

> +		gpio_dev = acpi_get_first_physical_node(gpio_adev);
> +		acpi_dev_put(gpio_adev);

> +

Unneeded blank line.

> +		if (!gpio_dev) {
> +			ret = -EPROBE_DEFER;
> +			goto cleanup;
> +		}
> +
> +		/*
> +		 * Check if the GPIO controller's device is bound. If not,
> +		 * defer probe to ensure GPIO initialization (including IRQ
> +		 * setup and quirks) is complete before we enumerate I2C
> +		 * child devices.
> +		 */
> +		scoped_guard(device, gpio_dev) {
> +			bound = device_is_bound(gpio_dev);
> +		}

Redundant parentheses.

> +		if (!bound) {
> +			put_device(gpio_dev);
> +			ret = -EPROBE_DEFER;
> +			goto cleanup;
> +		}
> +
> +		put_device(gpio_dev);
> +	}
> +
> +cleanup:
> +	list_for_each_entry_safe(ref, tmp, &ctx.gpio_controllers, node) {
> +		list_del(&ref->node);
> +		kfree(ref->path);
> +		kfree(ref);
> +	}

Yes, this needs to be present on error path above.

> +	return ret;
> +}
> +#else
> +static int i2c_dw_check_gpio_dependencies(struct device *dev)
> +{
> +	return 0;
> +}
> +#endif /* CONFIG_ACPI */

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH v8 2/2] i2c: designware: defer probe if child GpioInt controllers are not bound

[Andy Shevchenko]

On Wed, Jun 03, 2026 at 01:20:24AM +0530, Hardik Prakash wrote:
> On Wed, Jun 03, 2026 at 00:23, Nathan Chancellor wrote:
> > I bisected boot issues with a few of my test machines to this change
> > in -next as commit ef76a3a28c79. I seem to recall a crash in strcmp()
> > in one log but I cannot be too sure.
> 
> Thank you for bisecting this.
> 
> The strcmp() crash is likely the issue. In
> check_gpioint_resource(), we access agpio->resource_source.string_ptr
> without checking whether it is NULL first:
> 
>     if (!strcmp(tmp->path, agpio->resource_source.string_ptr))
>     ref->path = kstrdup(agpio->resource_source.string_ptr, ...)
> 
> The acpi_resource_source struct has a string_length field -- when it is
> 0, string_ptr will be NULL. On your machines, likely some GPIO resource
> apparently has no resource source string, which we did not account for.
> 
> I am preparing a fix that skips GPIO resources with no resource source
> string (string_length == 0 || string_ptr == NULL). I will test it on
> my hardware first and send a patch shortly.
> 
> Sorry for the regression.

Linus, if you applied this patch, please drop it. There are problems with patch
code wise and functional wise. It needs more love and work.

> On Wed, 3 Jun 2026 at 00:23, Nathan Chancellor <nathan@kernel.org> wrote:
> > On Fri, May 29, 2026 at 03:38:37PM +0530, Hardik Prakash wrote:
> > > I2C controllers may have child devices with GpioInt resources that
> > > depend on GPIO controllers to be fully initialized. If the I2C
> > > controller probes and enumerates children before the referenced GPIO
> > > controller has completed probe, GPIO interrupts may not be properly
> > > configured, leading to device failures.
> > >
> > > On Lenovo Yoga 7 14AGP11, the WACF2200 touchscreen (child of
> > > AMDI0010:02) has a GpioInt resource pointing to GPIO 157 on the
> > > pinctrl-amd controller (AMDI0030:00). When i2c-designware probes
> > > AMDI0010:02 before pinctrl-amd finishes initializing, I2C transactions
> > > occur before the GPIO IRQ quirk in amd_gpio_probe() has run, causing:
> > >
> > >   i2c_designware AMDI0010:02: i2c_dw_handle_tx_abort: lost arbitration
> > >
> > > Add a generic dependency check in i2c-designware that walks ACPI child
> > > devices, identifies any GpioInt resources, resolves the referenced GPIO
> > > controllers, and defers probe if those controllers are not yet bound.
> > >
> > > This ensures GPIO controllers complete initialization (including IRQ
> > > setup and quirks) before I2C child enumeration begins, fixing the race
> > > without device-specific quirks or DMI matching.
> > >
> > > The probe ordering race was confirmed via dynamic debug tracing:
> > >
> > >   0.285952  amd_gpio_probe: registering gpiochip  <- GPIO chip visible
> > >   0.287121  amd_gpio_probe: requesting parent IRQ <- probe still running
> > >   0.301454  AMDI0010:02 dw_i2c_plat_probe: start  <- races here
> > >   2.348157  lost arbitration

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH v8 2/2] i2c: designware: defer probe if child GpioInt controllers are not bound

[Hardik Prakash]

On Wed, Jun 03, 2026 at 04:26, Andy Shevchenko wrote:
> Linus, if you applied this patch, please drop it. There are problems
> with patch code wise and functional wise. It needs more love and work.

Understood. I agree that the patch should be dropped. The NULL pointer
dereference reported by Nathan and the code issues you've identified
both need to be addressed properly before resubmission.

I will rework the patch addressing all your feedback and the NULL check
for resource_source.string_ptr. I will send a new version once all of the
issues are rectified.

Thanks,
Hardik

On Wed, 3 Jun 2026 at 04:26, Andy Shevchenko
<andriy.shevchenko@intel.com> wrote:
>
> On Wed, Jun 03, 2026 at 01:20:24AM +0530, Hardik Prakash wrote:
> > On Wed, Jun 03, 2026 at 00:23, Nathan Chancellor wrote:
> > > I bisected boot issues with a few of my test machines to this change
> > > in -next as commit ef76a3a28c79. I seem to recall a crash in strcmp()
> > > in one log but I cannot be too sure.
> >
> > Thank you for bisecting this.
> >
> > The strcmp() crash is likely the issue. In
> > check_gpioint_resource(), we access agpio->resource_source.string_ptr
> > without checking whether it is NULL first:
> >
> >     if (!strcmp(tmp->path, agpio->resource_source.string_ptr))
> >     ref->path = kstrdup(agpio->resource_source.string_ptr, ...)
> >
> > The acpi_resource_source struct has a string_length field -- when it is
> > 0, string_ptr will be NULL. On your machines, likely some GPIO resource
> > apparently has no resource source string, which we did not account for.
> >
> > I am preparing a fix that skips GPIO resources with no resource source
> > string (string_length == 0 || string_ptr == NULL). I will test it on
> > my hardware first and send a patch shortly.
> >
> > Sorry for the regression.
>
> Linus, if you applied this patch, please drop it. There are problems with patch
> code wise and functional wise. It needs more love and work.
>
> > On Wed, 3 Jun 2026 at 00:23, Nathan Chancellor <nathan@kernel.org> wrote:
> > > On Fri, May 29, 2026 at 03:38:37PM +0530, Hardik Prakash wrote:
> > > > I2C controllers may have child devices with GpioInt resources that
> > > > depend on GPIO controllers to be fully initialized. If the I2C
> > > > controller probes and enumerates children before the referenced GPIO
> > > > controller has completed probe, GPIO interrupts may not be properly
> > > > configured, leading to device failures.
> > > >
> > > > On Lenovo Yoga 7 14AGP11, the WACF2200 touchscreen (child of
> > > > AMDI0010:02) has a GpioInt resource pointing to GPIO 157 on the
> > > > pinctrl-amd controller (AMDI0030:00). When i2c-designware probes
> > > > AMDI0010:02 before pinctrl-amd finishes initializing, I2C transactions
> > > > occur before the GPIO IRQ quirk in amd_gpio_probe() has run, causing:
> > > >
> > > >   i2c_designware AMDI0010:02: i2c_dw_handle_tx_abort: lost arbitration
> > > >
> > > > Add a generic dependency check in i2c-designware that walks ACPI child
> > > > devices, identifies any GpioInt resources, resolves the referenced GPIO
> > > > controllers, and defers probe if those controllers are not yet bound.
> > > >
> > > > This ensures GPIO controllers complete initialization (including IRQ
> > > > setup and quirks) before I2C child enumeration begins, fixing the race
> > > > without device-specific quirks or DMI matching.
> > > >
> > > > The probe ordering race was confirmed via dynamic debug tracing:
> > > >
> > > >   0.285952  amd_gpio_probe: registering gpiochip  <- GPIO chip visible
> > > >   0.287121  amd_gpio_probe: requesting parent IRQ <- probe still running
> > > >   0.301454  AMDI0010:02 dw_i2c_plat_probe: start  <- races here
> > > >   2.348157  lost arbitration
>
> --
> With Best Regards,
> Andy Shevchenko
>
>

Re: [PATCH v8 2/2] i2c: designware: defer probe if child GpioInt controllers are not bound

[Hardik Prakash]

On Wed, Jun 03, 2026 at 04:25, Andy Shevchenko wrote:
> Useless data type. You can use list_head directly as a context.
> ...
> Again, a lot of duplication with gpiolib-acpi-core.c. Please, find a
> way how to reuse the code.

1. Remove gpio_dep_ctx, pass struct list_head * directly as void *data
2. const char *path in gpio_controller_ref
3. Use acpi_gpio_get_irq_resource() instead of open-coding the type
   checks -- this removes the duplication with gpiolib-acpi-core.c and
   also requires changing #ifdef CONFIG_ACPI to
   #if defined(CONFIG_ACPI) && defined(CONFIG_GPIOLIB)
4. Add NULL check for resource_source.string_ptr
5. Use return value of acpi_dev_get_resources() in check_child_gpioint()
   instead of ctx->ret
6. Ensure cleanup of partially filled list on all error paths
7. Remove Reported-by and Closes tags (carried over from a previous patch)
8. Fixes: tag on one line
9. {} instead of { .ret = 0 } for zero initialisation
10. Split adev declaration and assignment
11. Remove unneeded blank line after acpi_dev_put()
12. For "Redundant parentheses" on scoped_guard -- I believe you mean
    the bool bound pattern is unnecessary and we should return/goto
    directly from inside the guard instead?

Please correct me if I've misunderstood any point.

Thanks,
Hardik

Re: [PATCH v8 2/2] i2c: designware: defer probe if child GpioInt controllers are not bound

[Andy Shevchenko]

On Wed, Jun 03, 2026 at 12:53:26PM +0530, Hardik Prakash wrote:
> On Wed, Jun 03, 2026 at 04:25, Andy Shevchenko wrote:
> > Useless data type. You can use list_head directly as a context.
> > ...
> > Again, a lot of duplication with gpiolib-acpi-core.c. Please, find a
> > way how to reuse the code.
> 
> 1. Remove gpio_dep_ctx, pass struct list_head * directly as void *data
> 2. const char *path in gpio_controller_ref
> 3. Use acpi_gpio_get_irq_resource() instead of open-coding the type
>    checks -- this removes the duplication with gpiolib-acpi-core.c and
>    also requires changing #ifdef CONFIG_ACPI to
>    #if defined(CONFIG_ACPI) && defined(CONFIG_GPIOLIB)
> 4. Add NULL check for resource_source.string_ptr
> 5. Use return value of acpi_dev_get_resources() in check_child_gpioint()
>    instead of ctx->ret
> 6. Ensure cleanup of partially filled list on all error paths
> 7. Remove Reported-by and Closes tags (carried over from a previous patch)
> 8. Fixes: tag on one line
> 9. {} instead of { .ret = 0 } for zero initialisation
> 10. Split adev declaration and assignment
> 11. Remove unneeded blank line after acpi_dev_put()

> 12. For "Redundant parentheses" on scoped_guard -- I believe you mean
>     the bool bound pattern is unnecessary and we should return/goto
>     directly from inside the guard instead?

Nope, the scoped_guard() for a single statement should be written without {}.

	scoped_guard(...)
		do_something(...);

> Please correct me if I've misunderstood any point.

Otherwise the rest sounds like a good plan.

-- 
With Best Regards,
Andy Shevchenko