* [PATCH] media: i2c: ov5647: Fix runtime PM refcount leak in s_ctrl
@ 2026-02-25 8:56 Xiaolei Wang
2026-02-25 10:27 ` Jacopo Mondi
2026-02-28 8:23 ` Tarang Raval
0 siblings, 2 replies; 3+ messages in thread
From: Xiaolei Wang @ 2026-02-25 8:56 UTC (permalink / raw)
To: sakari.ailus, tarang.raval, laurent.pinchart, dave.stevenson,
jacopo, mchehab, prabhakar.mahadev-lad.rj, hverkuil+cisco,
johannes.goede, hverkuil-cisco, jai.luthra, Xiaolei.Wang
Cc: linux-media, linux-kernel
Three control cases (AUTOGAIN, EXPOSURE_AUTO, ANALOGUE_GAIN) directly
return without calling pm_runtime_put(), causing runtime PM reference
count leaks.
Change these cases from 'return' to 'ret = ... break' pattern to ensure
pm_runtime_put() is always called before function exit.
Fixes: 4f66f36388d5 ("media: i2c: ov5647: Convert to CCI register access helpers")
Cc: stable@vger.kernel.org
Signed-off-by: Xiaolei Wang <xiaolei.wang@windriver.com>
---
drivers/media/i2c/ov5647.c | 12 ++++++------
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/drivers/media/i2c/ov5647.c b/drivers/media/i2c/ov5647.c
index 6a46ef7233ac..db9bd2892140 100644
--- a/drivers/media/i2c/ov5647.c
+++ b/drivers/media/i2c/ov5647.c
@@ -967,21 +967,21 @@ static int ov5647_s_ctrl(struct v4l2_ctrl *ctrl)
break;
case V4L2_CID_AUTOGAIN:
/* Non-zero turns on AGC by clearing bit 1.*/
- return cci_update_bits(sensor->regmap, OV5647_REG_AEC_AGC, BIT(1),
- ctrl->val ? 0 : BIT(1), NULL);
+ ret = cci_update_bits(sensor->regmap, OV5647_REG_AEC_AGC, BIT(1),
+ ctrl->val ? 0 : BIT(1), NULL);
break;
case V4L2_CID_EXPOSURE_AUTO:
/*
* Everything except V4L2_EXPOSURE_MANUAL turns on AEC by
* clearing bit 0.
*/
- return cci_update_bits(sensor->regmap, OV5647_REG_AEC_AGC, BIT(0),
- ctrl->val == V4L2_EXPOSURE_MANUAL ? BIT(0) : 0, NULL);
+ ret = cci_update_bits(sensor->regmap, OV5647_REG_AEC_AGC, BIT(0),
+ ctrl->val == V4L2_EXPOSURE_MANUAL ? BIT(0) : 0, NULL);
break;
case V4L2_CID_ANALOGUE_GAIN:
/* 10 bits of gain, 2 in the high register. */
- return cci_write(sensor->regmap, OV5647_REG_GAIN,
- ctrl->val & 0x3ff, NULL);
+ ret = cci_write(sensor->regmap, OV5647_REG_GAIN,
+ ctrl->val & 0x3ff, NULL);
break;
case V4L2_CID_EXPOSURE:
/*
--
2.43.0
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH] media: i2c: ov5647: Fix runtime PM refcount leak in s_ctrl
2026-02-25 8:56 [PATCH] media: i2c: ov5647: Fix runtime PM refcount leak in s_ctrl Xiaolei Wang
@ 2026-02-25 10:27 ` Jacopo Mondi
2026-02-28 8:23 ` Tarang Raval
1 sibling, 0 replies; 3+ messages in thread
From: Jacopo Mondi @ 2026-02-25 10:27 UTC (permalink / raw)
To: Xiaolei Wang
Cc: sakari.ailus, tarang.raval, laurent.pinchart, dave.stevenson,
jacopo, mchehab, prabhakar.mahadev-lad.rj, hverkuil+cisco,
johannes.goede, hverkuil-cisco, jai.luthra, linux-media,
linux-kernel
Hello Xiaolei
On Wed, Feb 25, 2026 at 04:56:21PM +0800, Xiaolei Wang wrote:
> Three control cases (AUTOGAIN, EXPOSURE_AUTO, ANALOGUE_GAIN) directly
> return without calling pm_runtime_put(), causing runtime PM reference
> count leaks.
>
> Change these cases from 'return' to 'ret = ... break' pattern to ensure
> pm_runtime_put() is always called before function exit.
>
> Fixes: 4f66f36388d5 ("media: i2c: ov5647: Convert to CCI register access helpers")
> Cc: stable@vger.kernel.org
> Signed-off-by: Xiaolei Wang <xiaolei.wang@windriver.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Thanks
j
> ---
> drivers/media/i2c/ov5647.c | 12 ++++++------
> 1 file changed, 6 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/media/i2c/ov5647.c b/drivers/media/i2c/ov5647.c
> index 6a46ef7233ac..db9bd2892140 100644
> --- a/drivers/media/i2c/ov5647.c
> +++ b/drivers/media/i2c/ov5647.c
> @@ -967,21 +967,21 @@ static int ov5647_s_ctrl(struct v4l2_ctrl *ctrl)
> break;
> case V4L2_CID_AUTOGAIN:
> /* Non-zero turns on AGC by clearing bit 1.*/
> - return cci_update_bits(sensor->regmap, OV5647_REG_AEC_AGC, BIT(1),
> - ctrl->val ? 0 : BIT(1), NULL);
> + ret = cci_update_bits(sensor->regmap, OV5647_REG_AEC_AGC, BIT(1),
> + ctrl->val ? 0 : BIT(1), NULL);
> break;
> case V4L2_CID_EXPOSURE_AUTO:
> /*
> * Everything except V4L2_EXPOSURE_MANUAL turns on AEC by
> * clearing bit 0.
> */
> - return cci_update_bits(sensor->regmap, OV5647_REG_AEC_AGC, BIT(0),
> - ctrl->val == V4L2_EXPOSURE_MANUAL ? BIT(0) : 0, NULL);
> + ret = cci_update_bits(sensor->regmap, OV5647_REG_AEC_AGC, BIT(0),
> + ctrl->val == V4L2_EXPOSURE_MANUAL ? BIT(0) : 0, NULL);
> break;
> case V4L2_CID_ANALOGUE_GAIN:
> /* 10 bits of gain, 2 in the high register. */
> - return cci_write(sensor->regmap, OV5647_REG_GAIN,
> - ctrl->val & 0x3ff, NULL);
> + ret = cci_write(sensor->regmap, OV5647_REG_GAIN,
> + ctrl->val & 0x3ff, NULL);
> break;
> case V4L2_CID_EXPOSURE:
> /*
> --
> 2.43.0
>
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] media: i2c: ov5647: Fix runtime PM refcount leak in s_ctrl
2026-02-25 8:56 [PATCH] media: i2c: ov5647: Fix runtime PM refcount leak in s_ctrl Xiaolei Wang
2026-02-25 10:27 ` Jacopo Mondi
@ 2026-02-28 8:23 ` Tarang Raval
1 sibling, 0 replies; 3+ messages in thread
From: Tarang Raval @ 2026-02-28 8:23 UTC (permalink / raw)
To: Xiaolei Wang
Cc: sakari.ailus@linux.intel.com, hverkuil-cisco@xs4all.nl,
johannes.goede@oss.qualcomm.com, jai.luthra@ideasonboard.com,
hverkuil+cisco@kernel.org,
prabhakar.mahadev-lad.rj@bp.renesas.com, mchehab@kernel.org,
jacopo@jmondi.org, dave.stevenson@raspberrypi.com,
laurent.pinchart@ideasonboard.com, linux-media@vger.kernel.org,
linux-kernel@vger.kernel.org
Hi Xiaolei,
> Three control cases (AUTOGAIN, EXPOSURE_AUTO, ANALOGUE_GAIN) directly
> return without calling pm_runtime_put(), causing runtime PM reference
> count leaks.
>
> Change these cases from 'return' to 'ret = ... break' pattern to ensure
> pm_runtime_put() is always called before function exit.
>
> Fixes: 4f66f36388d5 ("media: i2c: ov5647: Convert to CCI register access helpers")
> Cc: stable@vger.kernel.org
> Signed-off-by: Xiaolei Wang <xiaolei.wang@windriver.com>
In my last review, I missed this.
Thank you for addressing it.
Reviewed-by: Tarang Raval <tarang.raval@siliconsignals.io>
Best Regards,
Tarang
> ---
> drivers/media/i2c/ov5647.c | 12 ++++++------
> 1 file changed, 6 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/media/i2c/ov5647.c b/drivers/media/i2c/ov5647.c
> index 6a46ef7233ac..db9bd2892140 100644
> --- a/drivers/media/i2c/ov5647.c
> +++ b/drivers/media/i2c/ov5647.c
> @@ -967,21 +967,21 @@ static int ov5647_s_ctrl(struct v4l2_ctrl *ctrl)
> break;
> case V4L2_CID_AUTOGAIN:
> /* Non-zero turns on AGC by clearing bit 1.*/
> - return cci_update_bits(sensor->regmap, OV5647_REG_AEC_AGC, BIT(1),
> - ctrl->val ? 0 : BIT(1), NULL);
> + ret = cci_update_bits(sensor->regmap, OV5647_REG_AEC_AGC, BIT(1),
> + ctrl->val ? 0 : BIT(1), NULL);
> break;
> case V4L2_CID_EXPOSURE_AUTO:
> /*
> * Everything except V4L2_EXPOSURE_MANUAL turns on AEC by
> * clearing bit 0.
> */
> - return cci_update_bits(sensor->regmap, OV5647_REG_AEC_AGC, BIT(0),
> - ctrl->val == V4L2_EXPOSURE_MANUAL ? BIT(0) : 0, NULL);
> + ret = cci_update_bits(sensor->regmap, OV5647_REG_AEC_AGC, BIT(0),
> + ctrl->val == V4L2_EXPOSURE_MANUAL ? BIT(0) : 0, NULL);
> break;
> case V4L2_CID_ANALOGUE_GAIN:
> /* 10 bits of gain, 2 in the high register. */
> - return cci_write(sensor->regmap, OV5647_REG_GAIN,
> - ctrl->val & 0x3ff, NULL);
> + ret = cci_write(sensor->regmap, OV5647_REG_GAIN,
> + ctrl->val & 0x3ff, NULL);
> break;
> case V4L2_CID_EXPOSURE:
> /*
> --
> 2.43.0
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2026-02-28 8:23 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-02-25 8:56 [PATCH] media: i2c: ov5647: Fix runtime PM refcount leak in s_ctrl Xiaolei Wang
2026-02-25 10:27 ` Jacopo Mondi
2026-02-28 8:23 ` Tarang Raval
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox