[PATCH v3] iio: pressure: rohm-bm1390: harden trigger handler against transient errors

Stepan Ionichev posted 1 patch 6 days, 9 hours ago
drivers/iio/pressure/rohm-bm1390.c | 23 ++++++++++++++++-------
1 file changed, 16 insertions(+), 7 deletions(-)
[PATCH v3] iio: pressure: rohm-bm1390: harden trigger handler against transient errors
Posted by Stepan Ionichev 6 days, 9 hours ago
bm1390_trigger_handler() returns from three error paths without calling
iio_trigger_notify_done(). The success path at the end does, so on a
transient regmap or read failure the trigger's use_count is never
decremented and iio_trigger_poll_chained() drops subsequent dispatches
until the trigger is reattached.

This is not a fix for a reported bug, only hardening against hardware
or bus glitches; if a glitch is persistent the device is wedged and
needs an unbind anyway, which is left to the user.

Split the function so the inner helper returns bool with the data-read
outcome, and the outer IRQ handler calls iio_trigger_notify_done()
once and reports the result via IRQ_RETVAL().

Signed-off-by: Stepan Ionichev <sozdayvek@gmail.com>
---
v3:
- Split into bm1390_handle_trigger() helper returning bool, with the IRQ
  wrapper calling iio_trigger_notify_done() once (Andy)
- Drop Fixes: and Cc: stable@; this is hardware-failure hardening, not
  a user-visible bug fix (Jonathan)
- Reframe the commit message accordingly

v1: https://lore.kernel.org/all/20260517160801.269-1-sozdayvek@gmail.com/
v2: https://lore.kernel.org/all/20260518094238.1986-1-sozdayvek@gmail.com/

 drivers/iio/pressure/rohm-bm1390.c | 23 ++++++++++++++++-------
 1 file changed, 16 insertions(+), 7 deletions(-)

diff --git a/drivers/iio/pressure/rohm-bm1390.c b/drivers/iio/pressure/rohm-bm1390.c
index 08146ca0f..007c1559f 100644
--- a/drivers/iio/pressure/rohm-bm1390.c
+++ b/drivers/iio/pressure/rohm-bm1390.c
@@ -621,17 +621,15 @@ static const struct iio_buffer_setup_ops bm1390_buffer_ops = {
 	.predisable = bm1390_buffer_predisable,
 };
 
-static irqreturn_t bm1390_trigger_handler(int irq, void *p)
+static bool bm1390_handle_trigger(struct iio_dev *idev)
 {
-	struct iio_poll_func *pf = p;
-	struct iio_dev *idev = pf->indio_dev;
 	struct bm1390_data *data = iio_priv(idev);
 	int ret, status;
 
 	/* DRDY is acked by reading status reg */
 	ret = regmap_read(data->regmap, BM1390_REG_STATUS, &status);
 	if (ret || !status)
-		return IRQ_NONE;
+		return false;
 
 	dev_dbg(data->dev, "DRDY trig status 0x%x\n", status);
 
@@ -639,7 +637,7 @@ static irqreturn_t bm1390_trigger_handler(int irq, void *p)
 		ret = bm1390_pressure_read(data, &data->buf.pressure);
 		if (ret) {
 			dev_warn(data->dev, "sample read failed %d\n", ret);
-			return IRQ_NONE;
+			return false;
 		}
 	}
 
@@ -648,15 +646,26 @@ static irqreturn_t bm1390_trigger_handler(int irq, void *p)
 				       &data->buf.temp, sizeof(data->buf.temp));
 		if (ret) {
 			dev_warn(data->dev, "temp read failed %d\n", ret);
-			return IRQ_HANDLED;
+			return true;
 		}
 	}
 
 	iio_push_to_buffers_with_ts(idev, &data->buf, sizeof(data->buf),
 				    data->timestamp);
+
+	return true;
+}
+
+static irqreturn_t bm1390_trigger_handler(int irq, void *p)
+{
+	struct iio_poll_func *pf = p;
+	struct iio_dev *idev = pf->indio_dev;
+	bool result;
+
+	result = bm1390_handle_trigger(idev);
 	iio_trigger_notify_done(idev->trig);
 
-	return IRQ_HANDLED;
+	return IRQ_RETVAL(result);
 }
 
 /* Get timestamps and wake the thread if we need to read data */
-- 
2.43.0
Re: [PATCH v3] iio: pressure: rohm-bm1390: harden trigger handler against transient errors
Posted by Matti Vaittinen 5 days, 21 hours ago
On 18/05/2026 21:11, Stepan Ionichev wrote:
> bm1390_trigger_handler() returns from three error paths without calling
> iio_trigger_notify_done(). The success path at the end does, so on a
> transient regmap or read failure the trigger's use_count is never
> decremented and iio_trigger_poll_chained() drops subsequent dispatches
> until the trigger is reattached.
> 
> This is not a fix for a reported bug, only hardening against hardware
> or bus glitches; if a glitch is persistent the device is wedged and
> needs an unbind anyway, which is left to the user.
> 
> Split the function so the inner helper returns bool with the data-read
> outcome, and the outer IRQ handler calls iio_trigger_notify_done()
> once and reports the result via IRQ_RETVAL().
> 
> Signed-off-by: Stepan Ionichev <sozdayvek@gmail.com>

Reviewed-by: Matti Vaittinen <mazziesaccount@gmail.com>

-- 
---
Matti Vaittinen
Linux kernel developer at ROHM Semiconductors
Oulu Finland

~~ When things go utterly wrong vim users can always type :help! ~~