Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.36 as permitted sender) client-ip=23.128.96.36;
References: <20231211175023.1680247-1-mike.rudenko@gmail.com>
 <20231211175023.1680247-8-mike.rudenko@gmail.com>
 <20231211181526.GF27535@pendragon.ideasonboard.com>
User-agent: mu4e 1.10.7; emacs 29.1
From:   Mikhail Rudenko <mike.rudenko@gmail.com>
To:     Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Cc:     linux-media@vger.kernel.org, linux-kernel@vger.kernel.org,
        Sakari Ailus <sakari.ailus@linux.intel.com>,
        Jacopo Mondi <jacopo@jmondi.org>,
        Christophe JAILLET <christophe.jaillet@wanadoo.fr>,
        Dave Stevenson <dave.stevenson@raspberrypi.com>,
        Mauro Carvalho Chehab <mchehab@kernel.org>
Subject: Re: [PATCH 07/19] media: i2c: ov4689: Use sub-device active state
Date:   Tue, 12 Dec 2023 15:27:38 +0300
In-reply-to: <20231211181526.GF27535@pendragon.ideasonboard.com>
Message-ID: <87ttonppp5.fsf@gmail.com>
MIME-Version: 1.0
Content-Type: text/plain
Precedence: bulk


On 2023-12-11 at 20:15 +02, Laurent Pinchart <laurent.pinchart@ideasonboard.com> wrote:

> Hi Mikhail,
>
> Thank you for the patch.
>
> On Mon, Dec 11, 2023 at 08:50:10PM +0300, Mikhail Rudenko wrote:
>> Use sub-device active state. Employ control handler lock to
>> synchronize access to the active state and s_stream.
>>
>> Signed-off-by: Mikhail Rudenko <mike.rudenko@gmail.com>
>> ---
>>  drivers/media/i2c/ov4689.c | 68 +++++++++++++++-----------------------
>>  1 file changed, 27 insertions(+), 41 deletions(-)
>>
>> diff --git a/drivers/media/i2c/ov4689.c b/drivers/media/i2c/ov4689.c
>> index 5392f650960c..2eef64cd0070 100644
>> --- a/drivers/media/i2c/ov4689.c
>> +++ b/drivers/media/i2c/ov4689.c
>> @@ -86,7 +86,6 @@ struct ov4689 {
>>
>>  	u32 clock_rate;
>>
>> -	struct mutex mutex; /* lock to protect ctrls and cur_mode */
>>  	struct v4l2_ctrl_handler ctrl_handler;
>>  	struct v4l2_ctrl *exposure;
>>
>> @@ -319,19 +318,6 @@ static int ov4689_set_fmt(struct v4l2_subdev *sd,
>>  	return 0;
>>  }
>>
>> -static int ov4689_get_fmt(struct v4l2_subdev *sd,
>> -			  struct v4l2_subdev_state *sd_state,
>> -			  struct v4l2_subdev_format *fmt)
>> -{
>> -	struct v4l2_mbus_framefmt *mbus_fmt = &fmt->format;
>> -	struct ov4689 *ov4689 = to_ov4689(sd);
>> -
>> -	/* only one mode supported for now */
>> -	ov4689_fill_fmt(ov4689->cur_mode, mbus_fmt);
>> -
>> -	return 0;
>> -}
>> -
>>  static int ov4689_enum_mbus_code(struct v4l2_subdev *sd,
>>  				 struct v4l2_subdev_state *sd_state,
>>  				 struct v4l2_subdev_mbus_code_enum *code)
>> @@ -405,10 +391,11 @@ static int ov4689_get_selection(struct v4l2_subdev *sd,
>>  static int ov4689_s_stream(struct v4l2_subdev *sd, int on)
>>  {
>>  	struct ov4689 *ov4689 = to_ov4689(sd);
>> +	struct v4l2_subdev_state *sd_state;
>>  	struct device *dev = ov4689->dev;
>>  	int ret = 0;
>>
>> -	mutex_lock(&ov4689->mutex);
>> +	sd_state = v4l2_subdev_lock_and_get_active_state(&ov4689->subdev);
>>
>>  	if (on) {
>>  		ret = pm_runtime_resume_and_get(dev);
>> @@ -443,7 +430,7 @@ static int ov4689_s_stream(struct v4l2_subdev *sd, int on)
>>  	}
>>
>>  unlock_and_return:
>> -	mutex_unlock(&ov4689->mutex);
>> +	v4l2_subdev_unlock_state(sd_state);
>>
>>  	return ret;
>>  }
>> @@ -506,19 +493,15 @@ static int __maybe_unused ov4689_power_off(struct device *dev)
>>  	return 0;
>>  }
>>
>> -static int ov4689_open(struct v4l2_subdev *sd, struct v4l2_subdev_fh *fh)
>> +static int ov4689_init_cfg(struct v4l2_subdev *sd,
>> +			   struct v4l2_subdev_state *sd_state)
>>  {
>> -	struct ov4689 *ov4689 = to_ov4689(sd);
>> -	struct v4l2_mbus_framefmt *try_fmt;
>> -
>> -	mutex_lock(&ov4689->mutex);
>> +	struct v4l2_mbus_framefmt *try_fmt =
>> +		v4l2_subdev_state_get_format(sd_state, 0);
>>
>> -	try_fmt = v4l2_subdev_state_get_format(fh->state, 0);
>>  	/* Initialize try_fmt */
>
> It's not necessarily a "try format" anymore, so you cand rop this
> comment, and rename the variable from try_fmt to fmt.

Ack, will rename it in v2.

>
>>  	ov4689_fill_fmt(&supported_modes[OV4689_MODE_2688_1520], try_fmt);
>>
>> -	mutex_unlock(&ov4689->mutex);
>> -
>>  	return 0;
>>  }
>>
>> @@ -526,18 +509,15 @@ static const struct dev_pm_ops ov4689_pm_ops = {
>>  	SET_RUNTIME_PM_OPS(ov4689_power_off, ov4689_power_on, NULL)
>>  };
>>
>> -static const struct v4l2_subdev_internal_ops ov4689_internal_ops = {
>> -	.open = ov4689_open,
>> -};
>> -
>>  static const struct v4l2_subdev_video_ops ov4689_video_ops = {
>>  	.s_stream = ov4689_s_stream,
>>  };
>>
>>  static const struct v4l2_subdev_pad_ops ov4689_pad_ops = {
>> +	.init_cfg = ov4689_init_cfg,
>
> Please rebase on top of the latest media tree for the next version, the
> .init_cfg() operation has been renamed to .init_state() and moved to
> v4l2_subdev_internal_ops.

Thanks for the tip, will do it in v2.

> Other than that, this patch looks good.
>
> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
>
>>  	.enum_mbus_code = ov4689_enum_mbus_code,
>>  	.enum_frame_size = ov4689_enum_frame_sizes,
>> -	.get_fmt = ov4689_get_fmt,
>> +	.get_fmt = v4l2_subdev_get_fmt,
>>  	.set_fmt = ov4689_set_fmt,
>>  	.get_selection = ov4689_get_selection,
>>  };
>> @@ -649,7 +629,6 @@ static int ov4689_initialize_controls(struct ov4689 *ov4689)
>>  	ret = v4l2_ctrl_handler_init(handler, 10);
>>  	if (ret)
>>  		return ret;
>> -	handler->lock = &ov4689->mutex;
>>
>>  	ctrl = v4l2_ctrl_new_int_menu(handler, NULL, V4L2_CID_LINK_FREQ, 0, 0,
>>  				      link_freq_menu_items);
>> @@ -862,13 +841,14 @@ static int ov4689_probe(struct i2c_client *client)
>>  		return dev_err_probe(dev, ret,
>>  				     "Failed to get power regulators\n");
>>
>> -	mutex_init(&ov4689->mutex);
>> -
>>  	sd = &ov4689->subdev;
>>  	v4l2_i2c_subdev_init(sd, client, &ov4689_subdev_ops);
>> +
>>  	ret = ov4689_initialize_controls(ov4689);
>> -	if (ret)
>> -		goto err_destroy_mutex;
>> +	if (ret) {
>> +		dev_err(dev, "Failed to initialize controls\n");
>> +		return ret;
>> +	}
>>
>>  	ret = ov4689_power_on(dev);
>>  	if (ret)
>> @@ -878,19 +858,26 @@ static int ov4689_probe(struct i2c_client *client)
>>  	if (ret)
>>  		goto err_power_off;
>>
>> -	sd->internal_ops = &ov4689_internal_ops;
>>  	sd->flags |= V4L2_SUBDEV_FL_HAS_DEVNODE;
>> +	sd->entity.function = MEDIA_ENT_F_CAM_SENSOR;
>>
>>  	ov4689->pad.flags = MEDIA_PAD_FL_SOURCE;
>> -	sd->entity.function = MEDIA_ENT_F_CAM_SENSOR;
>>  	ret = media_entity_pads_init(&sd->entity, 1, &ov4689->pad);
>>  	if (ret < 0)
>>  		goto err_power_off;
>>
>> +	sd->state_lock = ov4689->ctrl_handler.lock;
>> +	ret = v4l2_subdev_init_finalize(sd);
>> +
>> +	if (ret) {
>> +		dev_err(dev, "Could not register v4l2 device\n");
>> +		goto err_clean_entity;
>> +	}
>> +
>>  	ret = v4l2_async_register_subdev_sensor(sd);
>>  	if (ret) {
>>  		dev_err(dev, "v4l2 async register subdev failed\n");
>> -		goto err_clean_entity;
>> +		goto err_clean_subdev;
>>  	}
>>
>>  	pm_runtime_set_active(dev);
>> @@ -899,14 +886,14 @@ static int ov4689_probe(struct i2c_client *client)
>>
>>  	return 0;
>>
>> +err_clean_subdev:
>> +	v4l2_subdev_cleanup(sd);
>>  err_clean_entity:
>>  	media_entity_cleanup(&sd->entity);
>>  err_power_off:
>>  	ov4689_power_off(dev);
>>  err_free_handler:
>>  	v4l2_ctrl_handler_free(&ov4689->ctrl_handler);
>> -err_destroy_mutex:
>> -	mutex_destroy(&ov4689->mutex);
>>
>>  	return ret;
>>  }
>> @@ -918,9 +905,8 @@ static void ov4689_remove(struct i2c_client *client)
>>
>>  	v4l2_async_unregister_subdev(sd);
>>  	media_entity_cleanup(&sd->entity);
>> -
>> +	v4l2_subdev_cleanup(sd);
>>  	v4l2_ctrl_handler_free(&ov4689->ctrl_handler);
>> -	mutex_destroy(&ov4689->mutex);
>>
>>  	pm_runtime_disable(&client->dev);
>>  	if (!pm_runtime_status_suspended(&client->dev))


--
Best regards,
Mikhail Rudenko