iio: adc: qcom-spmi-iadc: balance enable_irq_wake() on driver unbind
Subsystems (2)
PATCH 1/1 iio: adc: qcom-spmi-iadc: balance enable_irq_wake() on driver unbind
iadc_probe() calls enable_irq_wake() after a successful devm_request_irq(), but the driver has no remove callback or matching disable_irq_wake(), so the wake reference count on the IRQ is leaked on module unload or driver unbind.
Check the IRQ request error first, then register a devm action that calls disable_irq_wake() so the wake reference is released in the same scope as the enable. While here, drop the inverted "if (!ret) ... else return ret" in favour of the standard "if (ret) return ret;" pattern.
--- drivers/iio/adc/qcom-spmi-iadc.c | 18 +++++++++++++++--- 1 file changed, 15 insertions(+), 3 deletions(-)
drivers/iio/adc/qcom-spmi-iadc.c +15 -3
@@ -481,6 +481,11 @@ static const struct iio_chan_spec iadc_channels[] = { }, }; +static void iadc_disable_irq_wake (void *data) +{ +disable_irq_wake ((unsigned long )data); +} +static int iadc_probe (struct platform_device *pdev) {struct device_node *node = pdev->dev.of_node;
@@ -538,9 +543,16 @@ static int iadc_probe(struct platform_device *pdev)if (!iadc->poll_eoc) { ret =devm_request_irq (dev, irq_eoc, iadc_isr,0 ,"spmi-iadc" , iadc); -if (!ret) -enable_irq_wake (irq_eoc); -else +if (ret) +return ret; + + ret =enable_irq_wake (irq_eoc); +if (ret) +return ret; + + ret =devm_add_action_or_reset (dev, iadc_disable_irq_wake, + (void *)(unsigned long )irq_eoc); +if (ret)return ret; }else { ret =devm_device_init_wakeup (iadc->dev);
Signature · 1 line
2.43.0
Jonathan Cameron
maintainerQuoted 5 lines · On Thu, 21 May 2026 00:09:24 +0500 Stepan Ionichev <[email protected]> wrote:
iadc_probe()callsenable_irq_wake()after a successfuldevm_request_irq(), but the driver has no remove callback or matchingdisable_irq_wake(), so the wake reference count on the IRQ is leaked on module unload or driver unbind.
Quoted 7 lines
Check the IRQ request error first, then register a devm action that calls
disable_irq_wake()so the wake reference is released in the same scope as the enable. While here, drop the inverted "if (!ret) ... else return ret" in favour of the standard "if (ret) return ret;" pattern.Signed-off-by Stepan Ionichev <[email protected]>
Looks mostly good - one trivial question inline.
Quoted 9 lines
--- drivers/iio/adc/qcom-spmi-iadc.c | 18 +++++++++++++++--- 1 file changed, 15 insertions(+), 3 deletions(-)drivers/iio/adc/qcom-spmi-iadc.c +0
-0
drivers/iio/adc/qcom-spmi-iadc.c static const struct iio_chan_spec iadc_channels[] = { · line 486
Quoted 6 lines
drivers/iio/adc/qcom-spmi-iadc.c +3
-0@@ -481,6 +481,11 @@ static const struct iio_chan_spec iadc_channels[] = { }, }; +static void iadc_disable_irq_wake (void *data) +{ +disable_irq_wake ((unsigned long )data);
Why not cast it to an int given that's what disable_irq_wake() takes
Quoted 11 lines
drivers/iio/adc/qcom-spmi-iadc.c +2
-2+} +static int iadc_probe (struct platform_device *pdev) {struct device_node *node = pdev->dev.of_node;@@ -538,9 +543,16 @@ static int iadc_probe(struct platform_device *pdev)if (!iadc->poll_eoc) { ret =devm_request_irq (dev, irq_eoc, iadc_isr,0 ,"spmi-iadc" , iadc); -if (!ret) -enable_irq_wake (irq_eoc);
drivers/iio/adc/qcom-spmi-iadc.c static int iadc_probe(struct platform_device *pdev) · line 554
Quoted 10 lines
drivers/iio/adc/qcom-spmi-iadc.c +9
-1-else +if (ret) +return ret; + + ret =enable_irq_wake (irq_eoc); +if (ret) +return ret; + + ret =devm_add_action_or_reset (dev, iadc_disable_irq_wake, + (void *)(unsigned long )irq_eoc);
What is the unsigned long bit of this cast giving us? Just casting to void * should be enough I think.
Quoted 4 lines
drivers/iio/adc/qcom-spmi-iadc.c +1
-0+if (ret)return ret; }else { ret =devm_device_init_wakeup (iadc->dev);
Jonathan Cameron
maintainerQuoted 51 lines · On Fri, 22 May 2026 15:58:49 +0100 Jonathan Cameron <[email protected]> wrote:
Quoted 5 lines · On Thu, 21 May 2026 00:09:24 +0500 Stepan Ionichev <[email protected]> wrote:
iadc_probe()callsenable_irq_wake()after a successfuldevm_request_irq(), but the driver has no remove callback or matchingdisable_irq_wake(), so the wake reference count on the IRQ is leaked on module unload or driver unbind.Quoted 7 lines
Check the IRQ request error first, then register a devm action that calls
disable_irq_wake()so the wake reference is released in the same scope as the enable. While here, drop the inverted "if (!ret) ... else return ret" in favour of the standard "if (ret) return ret;" pattern.Signed-off-by Stepan Ionichev <[email protected]>Looks mostly good - one trivial question inline.
Quoted 9 lines
--- drivers/iio/adc/qcom-spmi-iadc.c | 18 +++++++++++++++--- 1 file changed, 15 insertions(+), 3 deletions(-)drivers/iio/adc/qcom-spmi-iadc.c +0
-0drivers/iio/adc/qcom-spmi-iadc.c static const struct iio_chan_spec iadc_channels[] = { · line 486
Quoted 6 lines
drivers/iio/adc/qcom-spmi-iadc.c +3
-0@@ -481,6 +481,11 @@ static const struct iio_chan_spec iadc_channels[] = { }, }; +static void iadc_disable_irq_wake (void *data) +{ +disable_irq_wake ((unsigned long )data);Why not cast it to an int given that's what
disable_irq_wake()takesQuoted 16 lines
drivers/iio/adc/qcom-spmi-iadc.c +6
-3+} +static int iadc_probe (struct platform_device *pdev) {struct device_node *node = pdev->dev.of_node;@@ -538,9 +543,16 @@ static int iadc_probe(struct platform_device *pdev)if (!iadc->poll_eoc) { ret =devm_request_irq (dev, irq_eoc, iadc_isr,0 ,"spmi-iadc" , iadc); -if (!ret) -enable_irq_wake (irq_eoc); -else +if (ret) +return ret; + + ret =enable_irq_wake (irq_eoc);
Quoted 8 lines
drivers/iio/adc/qcom-spmi-iadc.c static int iadc_probe(struct platform_device *pdev) · line 554
Quoted 5 lines
drivers/iio/adc/qcom-spmi-iadc.c +5
-0+if (ret) +return ret; + + ret =devm_add_action_or_reset (dev, iadc_disable_irq_wake, + (void *)(unsigned long )irq_eoc);What is the unsigned long bit of this cast giving us? Just casting to void * should be enough I think.
The answer to this was size mismatch - so I was completely wrong ;(
So as I just replied to v2 this was correct, but could do with a Fixes tag.
Thanks
Jonathan
Quoted 8 lines
Quoted 6 lines
drivers/iio/adc/qcom-spmi-iadc.c +1
-0+if (ret)return ret; }else { ret =devm_device_init_wakeup (iadc->dev);