Re: [V9, 2/2] media: i2c: ov02a10: Add OV02A10 image sensor driver

[Dongchun Zhu <dongchun.zhu@mediatek.com>]

Hi Tomasz,

On Wed, 2020-06-10 at 19:44 +0000, Tomasz Figa wrote:
> Hi Dongchun,
> 
> On Sat, May 23, 2020 at 04:41:03PM +0800, Dongchun Zhu wrote:
> > Add a V4L2 sub-device driver for OV02A10 image sensor.
> > 
> > Signed-off-by: Dongchun Zhu <dongchun.zhu@mediatek.com>
> > ---
> >  MAINTAINERS                 |    1 +
> >  drivers/media/i2c/Kconfig   |   13 +
> >  drivers/media/i2c/Makefile  |    1 +
> >  drivers/media/i2c/ov02a10.c | 1025 +++++++++++++++++++++++++++++++++++++++++++
> >  4 files changed, 1040 insertions(+)
> >  create mode 100644 drivers/media/i2c/ov02a10.c
> > 
> 
> Thank you for the patch. Please see my comments inline.
> 
> [snip]
> > diff --git a/drivers/media/i2c/ov02a10.c b/drivers/media/i2c/ov02a10.c
> > new file mode 100644
> > index 0000000..160a0b5
> > --- /dev/null
> > +++ b/drivers/media/i2c/ov02a10.c
> [snip]
> > +static const char * const ov02a10_test_pattern_menu[] = {
> > +	"Disabled",
> > +	"Color Bar",
> 
> nit: We should normalize this to one of the standard names. What is the
> pattern on this sensor? Is it perhaps "Eight Vertical Colour Bars"?
> 

Yes. It is one kind of 'Eight Vertical Colour Bars'.
This pattern is called as 'MIPI color bar' per the datasheet.
Can we here use 'Vertical Color Bar' or 'MIPI Color Bar'?

> > +};
> [snip]
> > +static int ov02a10_set_fmt(struct v4l2_subdev *sd,
> > +			   struct v4l2_subdev_pad_config *cfg,
> > +			   struct v4l2_subdev_format *fmt)
> > +{
> > +	struct ov02a10 *ov02a10 = to_ov02a10(sd);
> > +	struct v4l2_mbus_framefmt *mbus_fmt = &fmt->format;
> > +
> > +	mutex_lock(&ov02a10->mutex);
> > +
> 
> 
> Don't we need to handle the case when fmt->which is V4L2_SUBDEV_FORMAT_TRY,
> which is used for trying the format, but not applying it to the hardware?
> 

Got it :-)

> > +	if (ov02a10->streaming) {
> > +		mutex_unlock(&ov02a10->mutex);
> > +		return -EBUSY;
> > +	}
> > +
> > +	/* Only one sensor mode supported */
> > +	mbus_fmt->code = ov02a10->fmt.code;
> > +	ov02a10_fill_fmt(ov02a10->cur_mode, mbus_fmt);
> > +	ov02a10->fmt = fmt->format;
> > +
> > +	mutex_unlock(&ov02a10->mutex);
> > +
> > +	return 0;
> > +}
> > +
> > +static int ov02a10_get_fmt(struct v4l2_subdev *sd,
> > +			   struct v4l2_subdev_pad_config *cfg,
> > +			   struct v4l2_subdev_format *fmt)
> > +{
> > +	struct ov02a10 *ov02a10 = to_ov02a10(sd);
> > +	struct v4l2_mbus_framefmt *mbus_fmt = &fmt->format;
> > +
> > +	mutex_lock(&ov02a10->mutex);
> > +
> > +	fmt->format = ov02a10->fmt;
> 
> Ditto.
> 
> > +	mbus_fmt->code = ov02a10->fmt.code;
> > +	ov02a10_fill_fmt(ov02a10->cur_mode, mbus_fmt);
> > +
> > +	mutex_unlock(&ov02a10->mutex);
> > +
> > +	return 0;
> > +}
> > +
> > +static int ov02a10_enum_mbus_code(struct v4l2_subdev *sd,
> > +				  struct v4l2_subdev_pad_config *cfg,
> > +				  struct v4l2_subdev_mbus_code_enum *code)
> > +{
> > +	struct ov02a10 *ov02a10 = to_ov02a10(sd);
> > +
> > +	if (code->index >= ARRAY_SIZE(supported_modes))
> > +		return -EINVAL;
> 
> Hmm, supported_modes[] doesn't seem to hold the information about mbus
> codes. Should this just perhaps be "!= 0"?
> 

Understood.

> > +
> > +	code->code = ov02a10->fmt.code;
> > +
> > +	return 0;
> > +}
> [snip]
> > +static int ov02a10_entity_init_cfg(struct v4l2_subdev *sd,
> > +				   struct v4l2_subdev_pad_config *cfg)
> > +{
> > +	struct v4l2_subdev_format fmt = {
> > +		.which = cfg ? V4L2_SUBDEV_FORMAT_TRY : V4L2_SUBDEV_FORMAT_ACTIVE,
> > +		.format = {
> > +			.width = 1600,
> > +			.height = 1200,
> > +		}
> > +	};
> > +
> > +	ov02a10_set_fmt(sd, cfg, &fmt);
> > +
> > +	return 0;
> > +}
> > +
> 
> I'm not familiar with this init_cfg operation and the documentation is very
> sparse about it. Sakari, is this a correct implementation?
> 
> [snip]
> > +static int ov02a10_set_test_pattern(struct ov02a10 *ov02a10, int pattern)
> > +{
> > +	struct i2c_client *client = v4l2_get_subdevdata(&ov02a10->subdev);
> > +	int ret;
> > +
> > +	ret = i2c_smbus_write_byte_data(client, REG_PAGE_SWITCH, REG_ENABLE);
> > +	if (ret < 0)
> > +		return ret;
> > +
> > +	ret = i2c_smbus_write_byte_data(client, OV02A10_REG_TEST_PATTERN,
> > +					pattern);
> > +	if (ret < 0)
> > +		return ret;
> > +
> > +	ret = i2c_smbus_write_byte_data(client, REG_GLOBAL_EFFECTIVE,
> > +					REG_ENABLE);
> > +	if (ret < 0)
> > +		return ret;
> > +
> > +	return i2c_smbus_write_byte_data(client, REG_SC_CTRL_MODE,
> > +					 SC_CTRL_MODE_STREAMING);
> 
> Why is this needed? Does writing the test pattern register stop streaming?
> 

Looking back to the setting history, I found it was suggested by OV.
I would leave your question to OV, and update their feedback.

> [snip]
> > +static int ov02a10_initialize_controls(struct ov02a10 *ov02a10)
> > +{
> > +	struct i2c_client *client = v4l2_get_subdevdata(&ov02a10->subdev);
> > +	const struct ov02a10_mode *mode;
> > +	struct v4l2_ctrl_handler *handler;
> > +	struct v4l2_ctrl *ctrl;
> > +	u64 exposure_max;
> > +	u32 pixel_rate;
> > +	int ret;
> > +
> > +	handler = &ov02a10->ctrl_handler;
> > +	mode = ov02a10->cur_mode;
> > +	ret = v4l2_ctrl_handler_init(handler, 7);
> > +	if (ret)
> > +		return ret;
> > +
> > +	handler->lock = &ov02a10->mutex;
> > +
> > +	ctrl = v4l2_ctrl_new_int_menu(handler, NULL, V4L2_CID_LINK_FREQ, 0, 0,
> > +				      link_freq_menu_items);
> > +	if (ctrl)
> > +		ctrl->flags |= V4L2_CTRL_FLAG_READ_ONLY;
> > +
> > +	pixel_rate = to_pixel_rate(0);
> > +	v4l2_ctrl_new_std(handler, NULL, V4L2_CID_PIXEL_RATE, 0, pixel_rate, 1,
> > +			  pixel_rate);
> > +
> > +	exposure_max = mode->vts_def - 4;
> > +	ov02a10->exposure = v4l2_ctrl_new_std(handler, &ov02a10_ctrl_ops,
> > +					      V4L2_CID_EXPOSURE,
> > +					      OV02A10_EXPOSURE_MIN,
> > +					      exposure_max,
> > +					      OV02A10_EXPOSURE_STEP,
> > +					      mode->exp_def);
> > +
> > +	v4l2_ctrl_new_std(handler, &ov02a10_ctrl_ops,
> > +			  V4L2_CID_ANALOGUE_GAIN,
> > +			  OV02A10_GAIN_MIN,
> > +			  OV02A10_GAIN_MAX,
> > +			  OV02A10_GAIN_STEP,
> > +			  OV02A10_GAIN_DEFAULT);
> > +
> > +	v4l2_ctrl_new_std_menu_items(handler, &ov02a10_ctrl_ops,
> > +				     V4L2_CID_TEST_PATTERN,
> > +				     ARRAY_SIZE(ov02a10_test_pattern_menu) - 1,
> > +				     0, 0, ov02a10_test_pattern_menu);
> > +
> 
> I can see that we're missing some controls here now, VBLANK and HBLANK if I
> remember correctly. Even though read-only, some userspace need those to
> get information about how the sensor operates.
> 

Yes. I made a mistake.

> > +	if (handler->error) {
> > +		ret = handler->error;
> > +		dev_err(&client->dev, "failed to init controls(%d)\n", ret);
> > +		goto err_free_handler;
> > +	}
> > +
> > +	ov02a10->subdev.ctrl_handler = handler;
> > +
> > +	return 0;
> > +
> > +err_free_handler:
> > +	v4l2_ctrl_handler_free(handler);
> > +
> > +	return ret;
> > +}
> [snip]
> > +	pm_runtime_enable(dev);
> > +	if (!pm_runtime_enabled(dev)) {
> > +		ret = ov02a10_power_on(dev);
> > +		if (ret < 0) {
> > +			dev_err(dev, "failed to power on: %d\n", ret);
> > +			goto err_free_handler;
> > +		}
> > +	}
> > +
> > +	ret = v4l2_async_register_subdev(&ov02a10->subdev);
> > +	if (ret) {
> > +		dev_err(dev, "failed to register V4L2 subdev: %d", ret);
> > +		if (!pm_runtime_enabled(dev))
> > +			ov02a10_power_off(dev);
> 
> Please don't mix inline and error-path error handling, as it makes it
> difficult to tell if it's correct. Please move this below the appropriate
> err label instead.
> 

Fixed in next release.

> > +		goto err_clean_entity;
> > +	}
> > +
> > +	return 0;
> > +
> > +err_clean_entity:
> 
> If one calls pm_runtime_enable() in the probe path, one needs to call
> pm_runtime_disable() on the error and remove paths.
> 

Yes, fixed in next release.

> > +	media_entity_cleanup(&ov02a10->subdev.entity);
> > +err_free_handler:
> > +	v4l2_ctrl_handler_free(ov02a10->subdev.ctrl_handler);
> > +err_destroy_mutex:
> > +	mutex_destroy(&ov02a10->mutex);
> > +
> > +	return ret;
> > +}
> > +
> 
> Best regards,
> Tomasz