public inbox for linux-media@vger.kernel.org
 help / color / mirror / Atom feed
From: Sanjay Chitroda <sanjayembeddedse@gmail.com>
To: Hans de Goede <hansg@kernel.org>,
	sakari.ailus@linux.intel.com, mchehab@kernel.org
Cc: hverkuil+cisco@kernel.org, linux-media@vger.kernel.org,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH 2/3] media: i2c: gc0310: use cached client and device pointers
Date: Sun, 05 Apr 2026 16:31:43 +0530	[thread overview]
Message-ID: <40B63744-FD2D-49B6-AEC8-A72C9BEDF8FA@gmail.com> (raw)
In-Reply-To: <ca8e682c-f1aa-4272-8ae7-cff01a7d0327@kernel.org>



On 2 April 2026 12:38:07 am IST, Hans de Goede <hansg@kernel.org> wrote:
>Hi,
>
>On 1-Apr-26 20:16, Sanjay Chitroda wrote:
>> From: Sanjay Chitroda <sanjayembeddedse@gmail.com>
>> 
>> The driver was repeatedly retrieving the i2c_client using
>> v4l2_get_subdevdata() only to access the underlying struct device.
>> 
>> Replace this with the cached sensor->client and sensor->dev pointers,
>> which are already available in the sensor structure.
>> 
>> This simplifies the code, avoids redundant subdev lookups, and makes
>> the driver more consistent with common V4L2 sensor driver patterns.
>> No functional change intended.
>
>
>So now you've grown struct gc0310_device by 16 bytes and
>gained nothing much. And again this sound very much AI
>generated.  Which is not necessarily a bad thing if the changes
>are actually useful, but this seems like it is just code-churn
>without any benefits.
>
>Regards,
>
>Hans
>

Thanks again for the review - I agree that, taken on its own, the change does not justify the added struct size and unnecessary churn for this driver.

For context, the motivation came from looking at recent sensor cleanups, for example Laurent's series: https://lore.kernel.org/all/20250812214620.30425-1-laurent.pinchart@ideasonboard.com/

where several sensor drivers were moved toward accessing the underlying device more directly instead of repeatedly going through subdev helpers. I attempted to explore a similar direction here.

That said, you're absolutely right that in the case of gc0310 this does not provide a concrete benefit, and the existing v412_get_subdevdata() usage is already clear and optimal.

I'll drop this patch instead of revising it.

Regards, Sanjay

>
>
>> 
>> Signed-off-by: Sanjay Chitroda <sanjayembeddedse@gmail.com>
>> ---
>>  drivers/media/i2c/gc0310.c | 72 ++++++++++++++++++++------------------
>>  1 file changed, 37 insertions(+), 35 deletions(-)
>> 
>> diff --git a/drivers/media/i2c/gc0310.c b/drivers/media/i2c/gc0310.c
>> index 72a82ad4118a..e538479fee2e 100644
>> --- a/drivers/media/i2c/gc0310.c
>> +++ b/drivers/media/i2c/gc0310.c
>> @@ -84,6 +84,9 @@
>>  #define to_gc0310_sensor(x) container_of(x, struct gc0310_device, sd)
>>  
>>  struct gc0310_device {
>> +	struct device *dev;
>> +	struct i2c_client *client;
>> +
>>  	struct v4l2_subdev sd;
>>  	struct media_pad pad;
>>  
>> @@ -409,28 +412,27 @@ static int gc0310_power_on(struct device *dev)
>>  
>>  static int gc0310_detect(struct gc0310_device *sensor)
>>  {
>> -	struct i2c_client *client = v4l2_get_subdevdata(&sensor->sd);
>>  	u64 val;
>>  	int ret;
>>  
>> -	if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C))
>> +	if (!i2c_check_functionality(sensor->client->adapter, I2C_FUNC_I2C))
>>  		return -ENODEV;
>>  
>>  	ret = cci_read(sensor->regmap, GC0310_SC_CMMN_CHIP_ID_REG, &val, NULL);
>>  	if (ret < 0) {
>> -		dev_err(&client->dev, "read sensor_id failed: %d\n", ret);
>> +		dev_err(sensor->dev, "read sensor_id failed: %d\n", ret);
>>  		return -ENODEV;
>>  	}
>>  
>> -	dev_dbg(&client->dev, "sensor ID = 0x%llx\n", val);
>> +	dev_dbg(sensor->dev, "sensor ID = 0x%llx\n", val);
>>  
>>  	if (val != GC0310_ID) {
>> -		dev_err(&client->dev, "sensor ID error, read id = 0x%llx, target id = 0x%x\n",
>> +		dev_err(sensor->dev, "sensor ID error, read id = 0x%llx, target id = 0x%x\n",
>>  			val, GC0310_ID);
>>  		return -ENODEV;
>>  	}
>>  
>> -	dev_dbg(&client->dev, "detect gc0310 success\n");
>> +	dev_dbg(sensor->dev, "detect gc0310 success\n");
>>  
>>  	return 0;
>>  }
>> @@ -440,10 +442,9 @@ static int gc0310_enable_streams(struct v4l2_subdev *sd,
>>  				 u32 pad, u64 streams_mask)
>>  {
>>  	struct gc0310_device *sensor = to_gc0310_sensor(sd);
>> -	struct i2c_client *client = v4l2_get_subdevdata(sd);
>>  	int ret;
>>  
>> -	ret = pm_runtime_resume_and_get(&client->dev);
>> +	ret = pm_runtime_resume_and_get(sensor->dev);
>>  	if (ret)
>>  		return ret;
>>  
>> @@ -474,7 +475,7 @@ static int gc0310_enable_streams(struct v4l2_subdev *sd,
>>  
>>  error_power_down:
>>  	if (ret)
>> -		pm_runtime_put(&client->dev);
>> +		pm_runtime_put(sensor->dev);
>>  
>>  	return ret;
>>  }
>> @@ -484,7 +485,6 @@ static int gc0310_disable_streams(struct v4l2_subdev *sd,
>>  				  u32 pad, u64 streams_mask)
>>  {
>>  	struct gc0310_device *sensor = to_gc0310_sensor(sd);
>> -	struct i2c_client *client = v4l2_get_subdevdata(sd);
>>  	int ret = 0;
>>  
>>  	cci_write(sensor->regmap, GC0310_RESET_RELATED_REG,
>> @@ -494,7 +494,7 @@ static int gc0310_disable_streams(struct v4l2_subdev *sd,
>>  	cci_write(sensor->regmap, GC0310_RESET_RELATED_REG,
>>  		  GC0310_REGISTER_PAGE_0, &ret);
>>  
>> -	pm_runtime_put(&client->dev);
>> +	pm_runtime_put(sensor->dev);
>>  	return ret;
>>  }
>>  
>> @@ -559,7 +559,6 @@ static const struct v4l2_subdev_internal_ops gc0310_internal_ops = {
>>  
>>  static int gc0310_init_controls(struct gc0310_device *sensor)
>>  {
>> -	struct i2c_client *client = v4l2_get_subdevdata(&sensor->sd);
>>  	struct v4l2_ctrl_handler *hdl = &sensor->ctrls.handler;
>>  	struct v4l2_fwnode_device_properties props;
>>  	int exp_max, ret;
>> @@ -597,7 +596,7 @@ static int gc0310_init_controls(struct gc0310_device *sensor)
>>  				  GC0310_H_BLANK_DEFAULT, 1,
>>  				  GC0310_H_BLANK_DEFAULT);
>>  
>> -	ret = v4l2_fwnode_device_parse(&client->dev, &props);
>> +	ret = v4l2_fwnode_device_parse(sensor->dev, &props);
>>  	if (ret)
>>  		return ret;
>>  
>> @@ -621,10 +620,10 @@ static void gc0310_remove(struct i2c_client *client)
>>  	v4l2_subdev_cleanup(sd);
>>  	media_entity_cleanup(&sensor->sd.entity);
>>  	v4l2_ctrl_handler_free(&sensor->ctrls.handler);
>> -	pm_runtime_disable(&client->dev);
>> -	if (!pm_runtime_status_suspended(&client->dev)) {
>> -		gc0310_power_off(&client->dev);
>> -		pm_runtime_set_suspended(&client->dev);
>> +	pm_runtime_disable(sensor->dev);
>> +	if (!pm_runtime_status_suspended(sensor->dev)) {
>> +		gc0310_power_off(sensor->dev);
>> +		pm_runtime_set_suspended(sensor->dev);
>>  	}
>>  }
>>  
>> @@ -695,15 +694,18 @@ static int gc0310_probe(struct i2c_client *client)
>>  	if (!sensor)
>>  		return -ENOMEM;
>>  
>> -	sensor->reset = devm_gpiod_get(&client->dev, "reset", GPIOD_OUT_HIGH);
>> +	sensor->client = client;
>> +	sensor->dev = &client->dev;
>> +
>> +	sensor->reset = devm_gpiod_get(sensor->dev, "reset", GPIOD_OUT_HIGH);
>>  	if (IS_ERR(sensor->reset)) {
>> -		return dev_err_probe(&client->dev, PTR_ERR(sensor->reset),
>> +		return dev_err_probe(sensor->dev, PTR_ERR(sensor->reset),
>>  				     "getting reset GPIO\n");
>>  	}
>>  
>> -	sensor->powerdown = devm_gpiod_get(&client->dev, "powerdown", GPIOD_OUT_HIGH);
>> +	sensor->powerdown = devm_gpiod_get(sensor->dev, "powerdown", GPIOD_OUT_HIGH);
>>  	if (IS_ERR(sensor->powerdown)) {
>> -		return dev_err_probe(&client->dev, PTR_ERR(sensor->powerdown),
>> +		return dev_err_probe(sensor->dev, PTR_ERR(sensor->powerdown),
>>  				     "getting powerdown GPIO\n");
>>  	}
>>  
>> @@ -713,11 +715,11 @@ static int gc0310_probe(struct i2c_client *client)
>>  	if (IS_ERR(sensor->regmap))
>>  		return PTR_ERR(sensor->regmap);
>>  
>> -	gc0310_power_on(&client->dev);
>> +	gc0310_power_on(sensor->dev);
>>  
>> -	pm_runtime_set_active(&client->dev);
>> -	pm_runtime_get_noresume(&client->dev);
>> -	pm_runtime_enable(&client->dev);
>> +	pm_runtime_set_active(sensor->dev);
>> +	pm_runtime_get_noresume(sensor->dev);
>> +	pm_runtime_enable(sensor->dev);
>>  
>>  	ret = gc0310_detect(sensor);
>>  	if (ret)
>> @@ -734,34 +736,34 @@ static int gc0310_probe(struct i2c_client *client)
>>  
>>  	ret = media_entity_pads_init(&sensor->sd.entity, 1, &sensor->pad);
>>  	if (ret) {
>> -		dev_err_probe(&client->dev, ret, "failed to init entity pads\n");
>> +		dev_err_probe(sensor->dev, ret, "failed to init entity pads\n");
>>  		goto error_handler_free;
>>  	}
>>  
>>  	sensor->sd.state_lock = sensor->ctrls.handler.lock;
>>  	ret = v4l2_subdev_init_finalize(&sensor->sd);
>>  	if (ret) {
>> -		dev_err_probe(&client->dev, ret, "subdev init error\n");
>> +		dev_err_probe(sensor->dev, ret, "subdev init error\n");
>>  		goto error_media_entity;
>>  	}
>>  
>>  	ret = v4l2_async_register_subdev_sensor(&sensor->sd);
>>  	if (ret) {
>> -		dev_err_probe(&client->dev, ret,
>> +		dev_err_probe(sensor->dev, ret,
>>  			      "failed to register gc0310 sub-device\n");
>>  		goto error_subdev_cleanup;
>>  	}
>>  
>> -	pm_runtime_set_autosuspend_delay(&client->dev, 1000);
>> -	pm_runtime_use_autosuspend(&client->dev);
>> -	pm_runtime_put_autosuspend(&client->dev);
>> +	pm_runtime_set_autosuspend_delay(sensor->dev, 1000);
>> +	pm_runtime_use_autosuspend(sensor->dev);
>> +	pm_runtime_put_autosuspend(sensor->dev);
>>  
>>  	return 0;
>>  
>>  error_subdev_cleanup:
>>  	v4l2_subdev_cleanup(&sensor->sd);
>> -	pm_runtime_disable(&client->dev);
>> -	pm_runtime_set_suspended(&client->dev);
>> +	pm_runtime_disable(sensor->dev);
>> +	pm_runtime_set_suspended(sensor->dev);
>>  
>>  error_media_entity:
>>  	media_entity_cleanup(&sensor->sd.entity);
>> @@ -770,8 +772,8 @@ static int gc0310_probe(struct i2c_client *client)
>>  	v4l2_ctrl_handler_free(&sensor->ctrls.handler);
>>  
>>  error_power_off:
>> -	pm_runtime_put_noidle(&client->dev);
>> -	gc0310_power_off(&client->dev);
>> +	pm_runtime_put_noidle(sensor->dev);
>> +	gc0310_power_off(sensor->dev);
>>  
>>  	return ret;
>>  }
>

  reply	other threads:[~2026-04-05 11:01 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-04-01 18:16 [PATCH 0/3] media: i2c: gc0310: cleanups and sensor clock handling improvements Sanjay Chitroda
2026-04-01 18:16 ` [PATCH 1/3] media: i2c: gc0310: fix probe error handling and unwind resources properly Sanjay Chitroda
2026-04-01 19:06   ` Hans de Goede
2026-04-05 10:16     ` Sanjay Chitroda
2026-04-06 16:09       ` Hans de Goede
2026-04-07  4:32         ` Sanjay Chitroda
2026-04-01 18:16 ` [PATCH 2/3] media: i2c: gc0310: use cached client and device pointers Sanjay Chitroda
2026-04-01 19:08   ` Hans de Goede
2026-04-05 11:01     ` Sanjay Chitroda [this message]
2026-04-01 18:16 ` [PATCH 3/3] media: i2c: gc0310: Use devm_v4l2_sensor_clk_get() Sanjay Chitroda
2026-04-01 19:20   ` Hans de Goede
2026-04-19  7:38     ` Sanjay Chitroda

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=40B63744-FD2D-49B6-AEC8-A72C9BEDF8FA@gmail.com \
    --to=sanjayembeddedse@gmail.com \
    --cc=hansg@kernel.org \
    --cc=hverkuil+cisco@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-media@vger.kernel.org \
    --cc=mchehab@kernel.org \
    --cc=sakari.ailus@linux.intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox