* [PATCH v5 0/4] snvs_pwrkey - code improvements and add report event
@ 2026-07-15 9:33 joy.zou
2026-07-15 9:33 ` [PATCH v5 1/4] Input: snvs_pwrkey - make use of dev_err_probe() joy.zou
` (3 more replies)
0 siblings, 4 replies; 9+ messages in thread
From: joy.zou @ 2026-07-15 9:33 UTC (permalink / raw)
To: Dmitry Torokhov, Frank Li, Bough Chen, Peng Fan, Jacky Bai, Ye Li
Cc: Joy Zou, imx, linux-input, linux-kernel, Joy Zou
This patch series improves the snvs_pwrkey driver with better code quality
and add report press event.
The main improvements include:
1. Clean up the code by using local device pointers and dev_err_probe() for
better readability and easier debugging.
2. Fix potential event loss during system suspend by reporting key press events
in the timer callback.
Signed-off-by: Joy Zou <joy.zou@oss.nxp.com>
---
Changes in v5:
- Replace SIMPLE_DEV_PM_OPS with DEFINE_SIMPLE_DEV_PM_OPS and remove
__maybe_unused from suspend/resume callbacks for patch #4.
- Use pm_ptr() to wrap pm_ops pointer in platform_driver for patch #4.
- Replace suspended flag check in interrupt handler with a pending_press
latch: set pending_press in hardirq context, consume and report the
press event from the timer callback in softirq context for patch #4.
- Link to v4: https://patch.msgid.link/20260618-b4-pwrkey-v4-0-4bfda105bdf3@oss.nxp.com
Changes in v4:
- Use dev_err_probe() for devm_input_allocate_device() to simplify error handling
for patch #1.
- Add Reviewed-by tag for patch #2 and #3.
- Modify the subject description for patch #2.
- Link to v3: https://patch.msgid.link/20260615-b4-pwrkey-v3-0-9510b1173f6e@oss.nxp.com
Changes in v3:
- Add spinlock for pdata->keystate and pdata->suspended per AI review comments.
- Replace hardcode value 1 with local variable keystate in input_report_key()
under suspended.
- Link to v2: https://lore.kernel.org/r/20260604-pwrkey-cleanup-v2-0-59fa0ddf0d7c@nxp.com
Changes in v2:
- Drop dev_err_probe() change for platform_get_irq() per AI review comments.
- Add new patch #2 replace hardcoding -EINVAL with original error code per
AI review comments.
- Use dev instead of &pdev->dev for devm_input_allocate_device(), which was
missed in patch v1 per AI review comments.
- Add a boolean variable suspended and PM callback functions to replace
the use of the is_suspended field per AI review comments.
- Move event report handle to else branch in suspended state, since the
pdata->minor_rev == 0 branch has no debounce detection per AI review
comments.
- Modify patch #3 and #4 commit message.
- Add Reviewed-by tag for patch #1.
- Link to v1: https://lore.kernel.org/r/20260326-pwrkey-cleanup-v1-0-d85d7c0bf275@nxp.com
To: Dmitry Torokhov <dmitry.torokhov@gmail.com>
To: Frank Li <Frank.Li@nxp.com>
To: Bough Chen <haibo.chen@nxp.com>
To: Peng Fan <peng.fan@nxp.com>
To: Jacky Bai <ping.bai@nxp.com>
To: Ye Li <ye.li@nxp.com>
Cc: imx@lists.linux.dev
Cc: linux-input@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
---
Joy Zou (4):
Input: snvs_pwrkey - make use of dev_err_probe()
Input: snvs_pwrkey - propagate error code of platform_get_irq()
Input: snvs_pwrkey - use local device pointer to simple code
Input: snvs_pwrkey - add press event reporting to avoid event loss during suspend
drivers/input/keyboard/snvs_pwrkey.c | 143 ++++++++++++++++++++++++-----------
1 file changed, 98 insertions(+), 45 deletions(-)
---
base-commit: 34cf6dafc47441dfb6b356a095b89c3585a93714
change-id: 20260615-b4-pwrkey-a60b2043cd29
Best regards,
--
Joy Zou <joy.zou@oss.nxp.com>
^ permalink raw reply [flat|nested] 9+ messages in thread
* [PATCH v5 1/4] Input: snvs_pwrkey - make use of dev_err_probe()
2026-07-15 9:33 [PATCH v5 0/4] snvs_pwrkey - code improvements and add report event joy.zou
@ 2026-07-15 9:33 ` joy.zou
2026-07-15 9:33 ` [PATCH v5 2/4] Input: snvs_pwrkey - propagate error code of platform_get_irq() joy.zou
` (2 subsequent siblings)
3 siblings, 0 replies; 9+ messages in thread
From: joy.zou @ 2026-07-15 9:33 UTC (permalink / raw)
To: Dmitry Torokhov, Frank Li, Bough Chen, Peng Fan, Jacky Bai, Ye Li
Cc: Joy Zou, imx, linux-input, linux-kernel, Joy Zou
From: Joy Zou <joy.zou@nxp.com>
Add dev_err_probe() at return path of probe() to support users to
identify issues easier.
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Joy Zou <joy.zou@nxp.com>
---
Changes in v4:
1. Use dev_err_probe() for devm_input_allocate_device() to simplify
error handling.
Changes in v2:
1. Drop dev_err_probe() change for platform_get_irq() per AI review comments.
2. Add Reviewed-by tag.
---
drivers/input/keyboard/snvs_pwrkey.c | 44 +++++++++++++-----------------------
1 file changed, 16 insertions(+), 28 deletions(-)
diff --git a/drivers/input/keyboard/snvs_pwrkey.c b/drivers/input/keyboard/snvs_pwrkey.c
index 954055aaf6e2..8cc6863d26ed 100644
--- a/drivers/input/keyboard/snvs_pwrkey.c
+++ b/drivers/input/keyboard/snvs_pwrkey.c
@@ -124,17 +124,15 @@ static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
/* Get SNVS register Page */
np = pdev->dev.of_node;
if (!np)
- return -ENODEV;
+ return dev_err_probe(&pdev->dev, -ENODEV, "Device tree node not found\n");
pdata = devm_kzalloc(&pdev->dev, sizeof(*pdata), GFP_KERNEL);
if (!pdata)
return -ENOMEM;
pdata->snvs = syscon_regmap_lookup_by_phandle(np, "regmap");
- if (IS_ERR(pdata->snvs)) {
- dev_err(&pdev->dev, "Can't get snvs syscon\n");
- return PTR_ERR(pdata->snvs);
- }
+ if (IS_ERR(pdata->snvs))
+ return dev_err_probe(&pdev->dev, PTR_ERR(pdata->snvs), "Can't get snvs syscon\n");
if (of_property_read_u32(np, "linux,keycode", &pdata->keycode)) {
pdata->keycode = KEY_POWER;
@@ -142,10 +140,9 @@ static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
}
clk = devm_clk_get_optional_enabled(&pdev->dev, NULL);
- if (IS_ERR(clk)) {
- dev_err(&pdev->dev, "Failed to get snvs clock (%pe)\n", clk);
- return PTR_ERR(clk);
- }
+ if (IS_ERR(clk))
+ return dev_err_probe(&pdev->dev, PTR_ERR(clk),
+ "Failed to get snvs clock (%pe)\n", clk);
pdata->wakeup = of_property_read_bool(np, "wakeup-source");
@@ -165,9 +162,8 @@ static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
bpt = (val / 5) - 1;
break;
default:
- dev_err(&pdev->dev,
- "power-off-time-sec %d out of range\n", val);
- return -EINVAL;
+ return dev_err_probe(&pdev->dev, -EINVAL,
+ "power-off-time-sec %d out of range\n", val);
}
regmap_update_bits(pdata->snvs, SNVS_LPCR_REG, SNVS_LPCR_BPT_MASK,
@@ -185,10 +181,8 @@ static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
timer_setup(&pdata->check_timer, imx_imx_snvs_check_for_events, 0);
input = devm_input_allocate_device(&pdev->dev);
- if (!input) {
- dev_err(&pdev->dev, "failed to allocate the input device\n");
- return -ENOMEM;
- }
+ if (!input)
+ return dev_err_probe(&pdev->dev, -ENOMEM, "failed to allocate the input device\n");
input->name = pdev->name;
input->phys = "snvs-pwrkey/input0";
@@ -198,10 +192,8 @@ static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
/* input customer action to cancel release timer */
error = devm_add_action(&pdev->dev, imx_snvs_pwrkey_act, pdata);
- if (error) {
- dev_err(&pdev->dev, "failed to register remove action\n");
- return error;
- }
+ if (error)
+ return dev_err_probe(&pdev->dev, error, "failed to register remove action\n");
pdata->input = input;
platform_set_drvdata(pdev, pdata);
@@ -209,16 +201,12 @@ static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
error = devm_request_irq(&pdev->dev, pdata->irq,
imx_snvs_pwrkey_interrupt,
0, pdev->name, pdev);
- if (error) {
- dev_err(&pdev->dev, "interrupt not available.\n");
- return error;
- }
+ if (error)
+ return dev_err_probe(&pdev->dev, error, "interrupt not available.\n");
error = input_register_device(input);
- if (error < 0) {
- dev_err(&pdev->dev, "failed to register input device\n");
- return error;
- }
+ if (error < 0)
+ return dev_err_probe(&pdev->dev, error, "failed to register input device\n");
device_init_wakeup(&pdev->dev, pdata->wakeup);
error = dev_pm_set_wake_irq(&pdev->dev, pdata->irq);
--
2.34.1
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH v5 2/4] Input: snvs_pwrkey - propagate error code of platform_get_irq()
2026-07-15 9:33 [PATCH v5 0/4] snvs_pwrkey - code improvements and add report event joy.zou
2026-07-15 9:33 ` [PATCH v5 1/4] Input: snvs_pwrkey - make use of dev_err_probe() joy.zou
@ 2026-07-15 9:33 ` joy.zou
2026-07-15 9:42 ` sashiko-bot
2026-07-15 9:33 ` [PATCH v5 3/4] Input: snvs_pwrkey - use local device pointer to simple code joy.zou
2026-07-15 9:33 ` [PATCH v5 4/4] Input: snvs_pwrkey - add press event reporting to avoid event loss during suspend joy.zou
3 siblings, 1 reply; 9+ messages in thread
From: joy.zou @ 2026-07-15 9:33 UTC (permalink / raw)
To: Dmitry Torokhov, Frank Li, Bough Chen, Peng Fan, Jacky Bai, Ye Li
Cc: Joy Zou, imx, linux-input, linux-kernel, Joy Zou
From: Joy Zou <joy.zou@nxp.com>
Hardcoding -EINVAL discards the actual error code, which breaks probe
deferral (-EPROBE_DEFER) and loses critical diagnostic information
needed for proper kernel error handling.
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Joy Zou <joy.zou@nxp.com>
---
Changes in v4:
1. modify the subject description.
2. add Reviewed-by tag.
---
drivers/input/keyboard/snvs_pwrkey.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/input/keyboard/snvs_pwrkey.c b/drivers/input/keyboard/snvs_pwrkey.c
index 8cc6863d26ed..d58bbbe9fd58 100644
--- a/drivers/input/keyboard/snvs_pwrkey.c
+++ b/drivers/input/keyboard/snvs_pwrkey.c
@@ -148,7 +148,7 @@ static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
pdata->irq = platform_get_irq(pdev, 0);
if (pdata->irq < 0)
- return -EINVAL;
+ return pdata->irq;
error = of_property_read_u32(np, "power-off-time-sec", &val);
if (!error) {
--
2.34.1
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH v5 3/4] Input: snvs_pwrkey - use local device pointer to simple code
2026-07-15 9:33 [PATCH v5 0/4] snvs_pwrkey - code improvements and add report event joy.zou
2026-07-15 9:33 ` [PATCH v5 1/4] Input: snvs_pwrkey - make use of dev_err_probe() joy.zou
2026-07-15 9:33 ` [PATCH v5 2/4] Input: snvs_pwrkey - propagate error code of platform_get_irq() joy.zou
@ 2026-07-15 9:33 ` joy.zou
2026-07-15 9:38 ` sashiko-bot
2026-07-15 9:33 ` [PATCH v5 4/4] Input: snvs_pwrkey - add press event reporting to avoid event loss during suspend joy.zou
3 siblings, 1 reply; 9+ messages in thread
From: joy.zou @ 2026-07-15 9:33 UTC (permalink / raw)
To: Dmitry Torokhov, Frank Li, Bough Chen, Peng Fan, Jacky Bai, Ye Li
Cc: Joy Zou, imx, linux-input, linux-kernel, Joy Zou
From: Joy Zou <joy.zou@nxp.com>
Use local struct device pointer to avoid reference the platform_device
pointer every time.
No functional change.
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Joy Zou <joy.zou@nxp.com>
---
Changes for v4:
1. Add Reviewed-by tag.
Changes for v2:
1. Use dev instead of &pdev->dev for devm_input_allocate_device(),
which was missed in patch v1 per AI review comments.
2. Modify commit message.
---
drivers/input/keyboard/snvs_pwrkey.c | 41 ++++++++++++++++++------------------
1 file changed, 21 insertions(+), 20 deletions(-)
diff --git a/drivers/input/keyboard/snvs_pwrkey.c b/drivers/input/keyboard/snvs_pwrkey.c
index d58bbbe9fd58..cbe44a38d2b3 100644
--- a/drivers/input/keyboard/snvs_pwrkey.c
+++ b/drivers/input/keyboard/snvs_pwrkey.c
@@ -112,6 +112,7 @@ static void imx_snvs_pwrkey_act(void *pdata)
static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
{
+ struct device *dev = &pdev->dev;
struct pwrkey_drv_data *pdata;
struct input_dev *input;
struct device_node *np;
@@ -122,26 +123,26 @@ static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
u32 vid;
/* Get SNVS register Page */
- np = pdev->dev.of_node;
+ np = dev->of_node;
if (!np)
- return dev_err_probe(&pdev->dev, -ENODEV, "Device tree node not found\n");
+ return dev_err_probe(dev, -ENODEV, "Device tree node not found\n");
- pdata = devm_kzalloc(&pdev->dev, sizeof(*pdata), GFP_KERNEL);
+ pdata = devm_kzalloc(dev, sizeof(*pdata), GFP_KERNEL);
if (!pdata)
return -ENOMEM;
pdata->snvs = syscon_regmap_lookup_by_phandle(np, "regmap");
if (IS_ERR(pdata->snvs))
- return dev_err_probe(&pdev->dev, PTR_ERR(pdata->snvs), "Can't get snvs syscon\n");
+ return dev_err_probe(dev, PTR_ERR(pdata->snvs), "Can't get snvs syscon\n");
if (of_property_read_u32(np, "linux,keycode", &pdata->keycode)) {
pdata->keycode = KEY_POWER;
- dev_warn(&pdev->dev, "KEY_POWER without setting in dts\n");
+ dev_warn(dev, "KEY_POWER without setting in dts\n");
}
- clk = devm_clk_get_optional_enabled(&pdev->dev, NULL);
+ clk = devm_clk_get_optional_enabled(dev, NULL);
if (IS_ERR(clk))
- return dev_err_probe(&pdev->dev, PTR_ERR(clk),
+ return dev_err_probe(dev, PTR_ERR(clk),
"Failed to get snvs clock (%pe)\n", clk);
pdata->wakeup = of_property_read_bool(np, "wakeup-source");
@@ -162,7 +163,7 @@ static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
bpt = (val / 5) - 1;
break;
default:
- return dev_err_probe(&pdev->dev, -EINVAL,
+ return dev_err_probe(dev, -EINVAL,
"power-off-time-sec %d out of range\n", val);
}
@@ -180,9 +181,9 @@ static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
timer_setup(&pdata->check_timer, imx_imx_snvs_check_for_events, 0);
- input = devm_input_allocate_device(&pdev->dev);
+ input = devm_input_allocate_device(dev);
if (!input)
- return dev_err_probe(&pdev->dev, -ENOMEM, "failed to allocate the input device\n");
+ return dev_err_probe(dev, -ENOMEM, "failed to allocate the input device\n");
input->name = pdev->name;
input->phys = "snvs-pwrkey/input0";
@@ -191,27 +192,27 @@ static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
input_set_capability(input, EV_KEY, pdata->keycode);
/* input customer action to cancel release timer */
- error = devm_add_action(&pdev->dev, imx_snvs_pwrkey_act, pdata);
+ error = devm_add_action(dev, imx_snvs_pwrkey_act, pdata);
if (error)
- return dev_err_probe(&pdev->dev, error, "failed to register remove action\n");
+ return dev_err_probe(dev, error, "failed to register remove action\n");
pdata->input = input;
platform_set_drvdata(pdev, pdata);
- error = devm_request_irq(&pdev->dev, pdata->irq,
- imx_snvs_pwrkey_interrupt,
- 0, pdev->name, pdev);
+ error = devm_request_irq(dev, pdata->irq,
+ imx_snvs_pwrkey_interrupt,
+ 0, pdev->name, pdev);
if (error)
- return dev_err_probe(&pdev->dev, error, "interrupt not available.\n");
+ return dev_err_probe(dev, error, "interrupt not available.\n");
error = input_register_device(input);
if (error < 0)
- return dev_err_probe(&pdev->dev, error, "failed to register input device\n");
+ return dev_err_probe(dev, error, "failed to register input device\n");
- device_init_wakeup(&pdev->dev, pdata->wakeup);
- error = dev_pm_set_wake_irq(&pdev->dev, pdata->irq);
+ device_init_wakeup(dev, pdata->wakeup);
+ error = dev_pm_set_wake_irq(dev, pdata->irq);
if (error)
- dev_err(&pdev->dev, "irq wake enable failed.\n");
+ dev_err(dev, "irq wake enable failed.\n");
return 0;
}
--
2.34.1
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH v5 4/4] Input: snvs_pwrkey - add press event reporting to avoid event loss during suspend
2026-07-15 9:33 [PATCH v5 0/4] snvs_pwrkey - code improvements and add report event joy.zou
` (2 preceding siblings ...)
2026-07-15 9:33 ` [PATCH v5 3/4] Input: snvs_pwrkey - use local device pointer to simple code joy.zou
@ 2026-07-15 9:33 ` joy.zou
2026-07-15 9:46 ` sashiko-bot
2026-07-15 16:01 ` Frank Li
3 siblings, 2 replies; 9+ messages in thread
From: joy.zou @ 2026-07-15 9:33 UTC (permalink / raw)
To: Dmitry Torokhov, Frank Li, Bough Chen, Peng Fan, Jacky Bai, Ye Li
Cc: Joy Zou, imx, linux-input, linux-kernel, Joy Zou
From: Joy Zou <joy.zou@nxp.com>
The driver implements debounce protection using a timer-based mechanism:
when a key interrupt occurs, a timer is scheduled to verify the key state
after DEBOUNCE_TIME before reporting the event. This works well during
normal operation.
However, key press events can be lost during system resume on platforms
like i.MX8MQ-EVK because:
1. During the no_irq resume phase, PCIe driver restoration can take up to
200ms with IRQs disabled.
2. The power key interrupt remains pending during the no_irq phase.
3. If the key is released before IRQs are re-enabled, the timer eventually
runs but sees the key as released and skips reporting the event.
To prevent event loss during system suspend, set a pending_press flag in
the interrupt handler and report the press event from the timer callback
when the flag is set. This avoids out-of-order event delivery and keeps
the existing timer-based debounce mechanism for normal operation.
Signed-off-by: Joy Zou <joy.zou@nxp.com>
---
Changes for v5:
1. Replace SIMPLE_DEV_PM_OPS with DEFINE_SIMPLE_DEV_PM_OPS and remove
__maybe_unused from suspend/resume callbacks.
2. Use pm_ptr() to wrap pm_ops pointer in platform_driver.
3. Replace suspended flag check in interrupt handler with a pending_press
latch: set pending_press in hardirq context, consume and report the
press event from the timer callback in softirq context.
Changes for v3:
1. Add spinlock for pdata->keystate and pdata->suspended per AI review
comments.
2. Replace hardcode value 1 with local variable keystate in input_report_key()
under suspended.
Changes for v2:
1. Add a boolean variable suspended and PM callback functions to replace
the use of the is_suspended field per AI review comments.
2. Move event report handle to else branch in suspended state, since the
pdata->minor_rev == 0 branch has no debounce detection per AI review
comments.
3. Modify the commit message.
---
drivers/input/keyboard/snvs_pwrkey.c | 72 ++++++++++++++++++++++++++++++++++--
1 file changed, 68 insertions(+), 4 deletions(-)
diff --git a/drivers/input/keyboard/snvs_pwrkey.c b/drivers/input/keyboard/snvs_pwrkey.c
index cbe44a38d2b3..970c7d65bd57 100644
--- a/drivers/input/keyboard/snvs_pwrkey.c
+++ b/drivers/input/keyboard/snvs_pwrkey.c
@@ -39,6 +39,9 @@ struct pwrkey_drv_data {
int keycode;
int keystate; /* 1:pressed */
int wakeup;
+ bool suspended; /* Track suspend state */
+ bool pending_press; /* Key pressed during suspend, report from timer callback */
+ spinlock_t lock; /* Protects keystate, suspended and pending_press */
struct timer_list check_timer;
struct input_dev *input;
u8 minor_rev;
@@ -49,14 +52,38 @@ static void imx_imx_snvs_check_for_events(struct timer_list *t)
struct pwrkey_drv_data *pdata = timer_container_of(pdata, t,
check_timer);
struct input_dev *input = pdata->input;
+ bool state_changed = false;
+ bool pending_press;
u32 state;
regmap_read(pdata->snvs, SNVS_HPSR_REG, &state);
state = state & SNVS_HPSR_BTN ? 1 : 0;
- /* only report new event if status changed */
- if (state ^ pdata->keystate) {
- pdata->keystate = state;
+ scoped_guard(spinlock_irqsave, &pdata->lock) {
+ pending_press = pdata->pending_press;
+ if (pending_press) {
+ pdata->pending_press = false;
+ pdata->keystate = 1;
+ }
+ /* only report new event if status changed */
+ if (state ^ pdata->keystate) {
+ pdata->keystate = state;
+ state_changed = true;
+ }
+ }
+
+ /*
+ * Report a press event latched during suspend. If the key is still
+ * held, state_changed will be 0 (keystate already set to 1 above),
+ * so no duplicate press is reported. If already released,
+ * state_changed will fire next to report the release.
+ */
+ if (pending_press) {
+ input_report_key(input, pdata->keycode, 1);
+ input_sync(input);
+ }
+
+ if (state_changed) {
input_event(input, EV_KEY, pdata->keycode, state);
input_sync(input);
pm_relax(pdata->input->dev.parent);
@@ -92,8 +119,17 @@ static irqreturn_t imx_snvs_pwrkey_interrupt(int irq, void *dev_id)
input_sync(input);
pm_relax(input->dev.parent);
} else {
+ /*
+ * If the key is pressed during suspend, latch it so
+ * the timer callback can report the press event in
+ * softirq context, avoiding out-of-order events.
+ */
+ scoped_guard(spinlock_irqsave, &pdata->lock) {
+ if (pdata->suspended)
+ pdata->pending_press = true;
+ }
mod_timer(&pdata->check_timer,
- jiffies + msecs_to_jiffies(DEBOUNCE_TIME));
+ jiffies + msecs_to_jiffies(DEBOUNCE_TIME));
}
}
@@ -151,6 +187,7 @@ static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
if (pdata->irq < 0)
return pdata->irq;
+ spin_lock_init(&pdata->lock);
error = of_property_read_u32(np, "power-off-time-sec", &val);
if (!error) {
switch (val) {
@@ -217,6 +254,32 @@ static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
return 0;
}
+static int imx_snvs_pwrkey_suspend(struct device *dev)
+{
+ struct platform_device *pdev = to_platform_device(dev);
+ struct pwrkey_drv_data *pdata = platform_get_drvdata(pdev);
+
+ scoped_guard(spinlock_irqsave, &pdata->lock)
+ pdata->suspended = true;
+
+ return 0;
+}
+
+static int imx_snvs_pwrkey_resume(struct device *dev)
+{
+ struct platform_device *pdev = to_platform_device(dev);
+ struct pwrkey_drv_data *pdata = platform_get_drvdata(pdev);
+
+ scoped_guard(spinlock_irqsave, &pdata->lock)
+ pdata->suspended = false;
+
+ return 0;
+}
+
+static DEFINE_SIMPLE_DEV_PM_OPS(imx_snvs_pwrkey_pm_ops,
+ imx_snvs_pwrkey_suspend,
+ imx_snvs_pwrkey_resume);
+
static const struct of_device_id imx_snvs_pwrkey_ids[] = {
{ .compatible = "fsl,sec-v4.0-pwrkey" },
{ /* sentinel */ }
@@ -227,6 +290,7 @@ static struct platform_driver imx_snvs_pwrkey_driver = {
.driver = {
.name = "snvs_pwrkey",
.of_match_table = imx_snvs_pwrkey_ids,
+ .pm = pm_ptr(&imx_snvs_pwrkey_pm_ops),
},
.probe = imx_snvs_pwrkey_probe,
};
--
2.34.1
^ permalink raw reply related [flat|nested] 9+ messages in thread
* Re: [PATCH v5 3/4] Input: snvs_pwrkey - use local device pointer to simple code
2026-07-15 9:33 ` [PATCH v5 3/4] Input: snvs_pwrkey - use local device pointer to simple code joy.zou
@ 2026-07-15 9:38 ` sashiko-bot
0 siblings, 0 replies; 9+ messages in thread
From: sashiko-bot @ 2026-07-15 9:38 UTC (permalink / raw)
To: joy.zou; +Cc: imx, dmitry.torokhov, linux-input, Frank.Li
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
Pre-existing issues:
- [High] Resource Leak / Use-After-Free due to unmanaged dev_pm_set_wake_irq()
--
--- Patch [3]: [PATCH v5 3/4] Input: snvs_pwrkey - use local device pointer to simple code ---
commit a69cd5d0da1072f6e3d82c6e22dfe1b937e29159
Author: Joy Zou <joy.zou@nxp.com>
Input: snvs_pwrkey - use local device pointer to simple code
This commit introduces a local device pointer in imx_snvs_pwrkey_probe() to
simplify the code. It replaces repeated references to &pdev->dev with a
single local variable.
> diff --git a/drivers/input/keyboard/snvs_pwrkey.c b/drivers/input/keyboard/snvs_pwrkey.c
> index d58bbbe9fd58f..cbe44a38d2b3d 100644
> --- a/drivers/input/keyboard/snvs_pwrkey.c
> +++ b/drivers/input/keyboard/snvs_pwrkey.c
[ ... ]
> @@ -191,27 +192,27 @@ static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
[ ... ]
> - device_init_wakeup(&pdev->dev, pdata->wakeup);
> - error = dev_pm_set_wake_irq(&pdev->dev, pdata->irq);
> + device_init_wakeup(dev, pdata->wakeup);
> + error = dev_pm_set_wake_irq(dev, pdata->irq);
[Severity: High]
This is a pre-existing issue, but does this code leak the wakeirq structure
when the driver is unbound?
Because the driver provides no remove callback or devm action to call
dev_pm_clear_wake_irq(), the allocated wake_irq structure appears to be left
attached to dev->power.wakeirq on unbind. If the driver is re-probed, could
this cause dev_pm_set_wake_irq() to fail with -EEXIST?
Also, since the underlying IRQ is managed by devm_request_irq() and freed on
unbind, if the system enters suspend after unbinding, could the PM core
attempt to call enable_irq_wake() using the stale, freed IRQ number and
cause an oops or use-after-free?
> if (error)
> - dev_err(&pdev->dev, "irq wake enable failed.\n");
> + dev_err(dev, "irq wake enable failed.\n");
>
> return 0;
> }
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260715-b4-pwrkey-v5-0-07e7353c319e@oss.nxp.com?part=3
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH v5 2/4] Input: snvs_pwrkey - propagate error code of platform_get_irq()
2026-07-15 9:33 ` [PATCH v5 2/4] Input: snvs_pwrkey - propagate error code of platform_get_irq() joy.zou
@ 2026-07-15 9:42 ` sashiko-bot
0 siblings, 0 replies; 9+ messages in thread
From: sashiko-bot @ 2026-07-15 9:42 UTC (permalink / raw)
To: joy.zou; +Cc: linux-input, Frank.Li, dmitry.torokhov, imx
Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
Pre-existing issues:
- [High] The driver uses `dev_pm_set_wake_irq()` and `device_init_wakeup()` during probe but provides no `.remove` callback to clean them up, causing a resource leak and permanently breaking driver rebind.
- [High] The driver ignores the return value of `regmap_read()`, leading to the use of uninitialized stack variables which can trigger spurious system shutdowns.
--
--- Patch [2]: [PATCH v5 2/4] Input: snvs_pwrkey - propagate error code of platform_get_irq() ---
commit 353345395cd42cd8e38b32a7bbc24dfad430ba99
Author: Joy Zou <joy.zou@nxp.com>
Input: snvs_pwrkey - propagate error code of platform_get_irq()
This commit updates the driver to propagate the actual error code from
platform_get_irq() instead of hardcoding -EINVAL. This ensures probe
deferral and other diagnostic information are preserved.
> diff --git a/drivers/input/keyboard/snvs_pwrkey.c b/drivers/input/keyboard/snvs_pwrkey.c
> index 8cc6863d26ed8..d58bbbe9fd58f 100644
> --- a/drivers/input/keyboard/snvs_pwrkey.c
> +++ b/drivers/input/keyboard/snvs_pwrkey.c
> @@ -148,7 +148,7 @@ static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
>
> pdata->irq = platform_get_irq(pdev, 0);
> if (pdata->irq < 0)
> - return -EINVAL;
> + return pdata->irq;
>
> error = of_property_read_u32(np, "power-off-time-sec", &val);
> if (!error) {
[Severity: High]
This is a pre-existing issue, but does this driver leak the wake IRQ upon
unbind?
Looking at imx_snvs_pwrkey_probe() in drivers/input/keyboard/snvs_pwrkey.c,
the driver sets up a wake IRQ:
error = dev_pm_set_wake_irq(dev, pdata->irq);
if (error)
dev_err(dev, "irq wake enable failed.\n");
However, there is no corresponding .remove callback in the platform_driver
struct (imx_snvs_pwrkey_driver) to call dev_pm_clear_wake_irq() or
device_init_wakeup(..., false):
static struct platform_driver imx_snvs_pwrkey_driver = {
.driver = {
.name = "snvs_pwrkey",
.of_match_table = imx_snvs_pwrkey_ids,
},
.probe = imx_snvs_pwrkey_probe,
};
If the driver is unbound via sysfs and rebound, does this leave a dangling
wakeup source and cause the subsequent bind to fail with -EEXIST?
[Severity: High]
This is a pre-existing issue, but can unhandled regmap_read() errors cause
spurious system shutdowns?
In drivers/input/keyboard/snvs_pwrkey.c, imx_snvs_pwrkey_interrupt() reads
the status register without checking the return value:
regmap_read(pdata->snvs, SNVS_LPSR_REG, &lp_status);
if (lp_status & SNVS_LPSR_SPO) {
A similar pattern exists in imx_imx_snvs_check_for_events():
regmap_read(pdata->snvs, SNVS_HPSR_REG, &state);
state = state & SNVS_HPSR_BTN ? 1 : 0;
And also in imx_snvs_pwrkey_probe():
regmap_read(pdata->snvs, SNVS_HPVIDR1_REG, &vid);
pdata->minor_rev = vid & 0xff;
If regmap_read() fails due to a clock glitch or bus fault, won't these
stack variables remain uninitialized? Could an uninitialized variable
randomly evaluate to true and be interpreted as a button press,
immediately propagating a spurious power key event?
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260715-b4-pwrkey-v5-0-07e7353c319e@oss.nxp.com?part=2
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH v5 4/4] Input: snvs_pwrkey - add press event reporting to avoid event loss during suspend
2026-07-15 9:33 ` [PATCH v5 4/4] Input: snvs_pwrkey - add press event reporting to avoid event loss during suspend joy.zou
@ 2026-07-15 9:46 ` sashiko-bot
2026-07-15 16:01 ` Frank Li
1 sibling, 0 replies; 9+ messages in thread
From: sashiko-bot @ 2026-07-15 9:46 UTC (permalink / raw)
To: joy.zou; +Cc: imx, dmitry.torokhov, linux-input, Frank.Li
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
Pre-existing issues:
- [High] Missing cleanup for wake IRQ leads to memory leak and device re-bind failure.
--
--- Patch [4]: [PATCH v5 4/4] Input: snvs_pwrkey - add press event reporting to avoid event loss during suspend ---
commit 0a87e611a13956d492552759681ef44d587017b8
Author: Joy Zou <joy.zou@nxp.com>
Input: snvs_pwrkey - add press event reporting to avoid event loss during suspend
This commit introduces a pending_press flag and a spinlock to handle power
key events during system resume phases where IRQs might be disabled. It
ensures that if the key is released before IRQs are re-enabled, a press
event is still reported via the timer callback.
> diff --git a/drivers/input/keyboard/snvs_pwrkey.c b/drivers/input/keyboard/snvs_pwrkey.c
> --- a/drivers/input/keyboard/snvs_pwrkey.c
> +++ b/drivers/input/keyboard/snvs_pwrkey.c
[ ... ]
> @@ -217,6 +254,32 @@ static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
> return 0;
> }
[Severity: High]
This is a pre-existing issue, but I noticed the wake IRQ setup just prior to
this return in imx_snvs_pwrkey_probe() lacks a cleanup path:
error = dev_pm_set_wake_irq(dev, pdata->irq);
Since there is no .remove callback or manual cleanup, does this leak the
struct wake_irq memory when the device is unbound?
Additionally, since dev->power.wakeirq is never cleared, would re-binding
the driver trigger the dev_WARN_ONCE ("wake irq already initialized") in
dev_pm_attach_wake_irq(), potentially causing a system panic if
panic_on_warn is enabled?
Could this be resolved by switching to the managed devm_pm_set_wake_irq()
helper?
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260715-b4-pwrkey-v5-0-07e7353c319e@oss.nxp.com?part=4
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH v5 4/4] Input: snvs_pwrkey - add press event reporting to avoid event loss during suspend
2026-07-15 9:33 ` [PATCH v5 4/4] Input: snvs_pwrkey - add press event reporting to avoid event loss during suspend joy.zou
2026-07-15 9:46 ` sashiko-bot
@ 2026-07-15 16:01 ` Frank Li
1 sibling, 0 replies; 9+ messages in thread
From: Frank Li @ 2026-07-15 16:01 UTC (permalink / raw)
To: joy.zou
Cc: Dmitry Torokhov, Frank Li, Bough Chen, Peng Fan, Jacky Bai, Ye Li,
imx, linux-input, linux-kernel, Joy Zou
On Wed, Jul 15, 2026 at 05:33:54PM +0800, joy.zou@oss.nxp.com wrote:
> From: Joy Zou <joy.zou@nxp.com>
>
> The driver implements debounce protection using a timer-based mechanism:
> when a key interrupt occurs, a timer is scheduled to verify the key state
> after DEBOUNCE_TIME before reporting the event. This works well during
> normal operation.
>
> However, key press events can be lost during system resume on platforms
> like i.MX8MQ-EVK because:
> 1. During the no_irq resume phase, PCIe driver restoration can take up to
> 200ms with IRQs disabled.
> 2. The power key interrupt remains pending during the no_irq phase.
> 3. If the key is released before IRQs are re-enabled, the timer eventually
> runs but sees the key as released and skips reporting the event.
>
> To prevent event loss during system suspend, set a pending_press flag in
> the interrupt handler and report the press event from the timer callback
> when the flag is set. This avoids out-of-order event delivery and keeps
> the existing timer-based debounce mechanism for normal operation.
>
> Signed-off-by: Joy Zou <joy.zou@nxp.com>
> ---
> Changes for v5:
> 1. Replace SIMPLE_DEV_PM_OPS with DEFINE_SIMPLE_DEV_PM_OPS and remove
> __maybe_unused from suspend/resume callbacks.
> 2. Use pm_ptr() to wrap pm_ops pointer in platform_driver.
> 3. Replace suspended flag check in interrupt handler with a pending_press
> latch: set pending_press in hardirq context, consume and report the
> press event from the timer callback in softirq context.
>
> Changes for v3:
> 1. Add spinlock for pdata->keystate and pdata->suspended per AI review
> comments.
> 2. Replace hardcode value 1 with local variable keystate in input_report_key()
> under suspended.
>
> Changes for v2:
> 1. Add a boolean variable suspended and PM callback functions to replace
> the use of the is_suspended field per AI review comments.
> 2. Move event report handle to else branch in suspended state, since the
> pdata->minor_rev == 0 branch has no debounce detection per AI review
> comments.
> 3. Modify the commit message.
> ---
Reviewed-by: Frank Li <Frank.Li@nxp.com>
> drivers/input/keyboard/snvs_pwrkey.c | 72 ++++++++++++++++++++++++++++++++++--
> 1 file changed, 68 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/input/keyboard/snvs_pwrkey.c b/drivers/input/keyboard/snvs_pwrkey.c
> index cbe44a38d2b3..970c7d65bd57 100644
> --- a/drivers/input/keyboard/snvs_pwrkey.c
> +++ b/drivers/input/keyboard/snvs_pwrkey.c
> @@ -39,6 +39,9 @@ struct pwrkey_drv_data {
> int keycode;
> int keystate; /* 1:pressed */
> int wakeup;
> + bool suspended; /* Track suspend state */
> + bool pending_press; /* Key pressed during suspend, report from timer callback */
> + spinlock_t lock; /* Protects keystate, suspended and pending_press */
> struct timer_list check_timer;
> struct input_dev *input;
> u8 minor_rev;
> @@ -49,14 +52,38 @@ static void imx_imx_snvs_check_for_events(struct timer_list *t)
> struct pwrkey_drv_data *pdata = timer_container_of(pdata, t,
> check_timer);
> struct input_dev *input = pdata->input;
> + bool state_changed = false;
> + bool pending_press;
> u32 state;
>
> regmap_read(pdata->snvs, SNVS_HPSR_REG, &state);
> state = state & SNVS_HPSR_BTN ? 1 : 0;
>
> - /* only report new event if status changed */
> - if (state ^ pdata->keystate) {
> - pdata->keystate = state;
> + scoped_guard(spinlock_irqsave, &pdata->lock) {
> + pending_press = pdata->pending_press;
> + if (pending_press) {
> + pdata->pending_press = false;
> + pdata->keystate = 1;
> + }
> + /* only report new event if status changed */
> + if (state ^ pdata->keystate) {
> + pdata->keystate = state;
> + state_changed = true;
> + }
> + }
> +
> + /*
> + * Report a press event latched during suspend. If the key is still
> + * held, state_changed will be 0 (keystate already set to 1 above),
> + * so no duplicate press is reported. If already released,
> + * state_changed will fire next to report the release.
> + */
> + if (pending_press) {
> + input_report_key(input, pdata->keycode, 1);
> + input_sync(input);
> + }
> +
> + if (state_changed) {
> input_event(input, EV_KEY, pdata->keycode, state);
> input_sync(input);
> pm_relax(pdata->input->dev.parent);
> @@ -92,8 +119,17 @@ static irqreturn_t imx_snvs_pwrkey_interrupt(int irq, void *dev_id)
> input_sync(input);
> pm_relax(input->dev.parent);
> } else {
> + /*
> + * If the key is pressed during suspend, latch it so
> + * the timer callback can report the press event in
> + * softirq context, avoiding out-of-order events.
> + */
> + scoped_guard(spinlock_irqsave, &pdata->lock) {
> + if (pdata->suspended)
> + pdata->pending_press = true;
> + }
> mod_timer(&pdata->check_timer,
> - jiffies + msecs_to_jiffies(DEBOUNCE_TIME));
> + jiffies + msecs_to_jiffies(DEBOUNCE_TIME));
> }
> }
>
> @@ -151,6 +187,7 @@ static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
> if (pdata->irq < 0)
> return pdata->irq;
>
> + spin_lock_init(&pdata->lock);
> error = of_property_read_u32(np, "power-off-time-sec", &val);
> if (!error) {
> switch (val) {
> @@ -217,6 +254,32 @@ static int imx_snvs_pwrkey_probe(struct platform_device *pdev)
> return 0;
> }
>
> +static int imx_snvs_pwrkey_suspend(struct device *dev)
> +{
> + struct platform_device *pdev = to_platform_device(dev);
> + struct pwrkey_drv_data *pdata = platform_get_drvdata(pdev);
> +
> + scoped_guard(spinlock_irqsave, &pdata->lock)
> + pdata->suspended = true;
> +
> + return 0;
> +}
> +
> +static int imx_snvs_pwrkey_resume(struct device *dev)
> +{
> + struct platform_device *pdev = to_platform_device(dev);
> + struct pwrkey_drv_data *pdata = platform_get_drvdata(pdev);
> +
> + scoped_guard(spinlock_irqsave, &pdata->lock)
> + pdata->suspended = false;
> +
> + return 0;
> +}
> +
> +static DEFINE_SIMPLE_DEV_PM_OPS(imx_snvs_pwrkey_pm_ops,
> + imx_snvs_pwrkey_suspend,
> + imx_snvs_pwrkey_resume);
> +
> static const struct of_device_id imx_snvs_pwrkey_ids[] = {
> { .compatible = "fsl,sec-v4.0-pwrkey" },
> { /* sentinel */ }
> @@ -227,6 +290,7 @@ static struct platform_driver imx_snvs_pwrkey_driver = {
> .driver = {
> .name = "snvs_pwrkey",
> .of_match_table = imx_snvs_pwrkey_ids,
> + .pm = pm_ptr(&imx_snvs_pwrkey_pm_ops),
> },
> .probe = imx_snvs_pwrkey_probe,
> };
>
> --
> 2.34.1
>
>
^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2026-07-15 16:01 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-07-15 9:33 [PATCH v5 0/4] snvs_pwrkey - code improvements and add report event joy.zou
2026-07-15 9:33 ` [PATCH v5 1/4] Input: snvs_pwrkey - make use of dev_err_probe() joy.zou
2026-07-15 9:33 ` [PATCH v5 2/4] Input: snvs_pwrkey - propagate error code of platform_get_irq() joy.zou
2026-07-15 9:42 ` sashiko-bot
2026-07-15 9:33 ` [PATCH v5 3/4] Input: snvs_pwrkey - use local device pointer to simple code joy.zou
2026-07-15 9:38 ` sashiko-bot
2026-07-15 9:33 ` [PATCH v5 4/4] Input: snvs_pwrkey - add press event reporting to avoid event loss during suspend joy.zou
2026-07-15 9:46 ` sashiko-bot
2026-07-15 16:01 ` Frank Li
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox