On Thu, 2023-12-07 at 09:18 +0100, Krzysztof Kozlowski wrote:
>
> External email : Please do not click links or open attachments until
> you have verified the sender or the content.
> On 07/12/2023 06:20, Zhi Mao wrote:
> > Add a V4L2 sub-device driver for Galaxycore GC08A3 image sensor.
> >
> > Reviewed-By: [email protected]
>
> I don't see review given here:
>
>
https://lore.kernel.org/linux-media/[email protected]/
>
> This does not look like real review. Where was it performed? How
> thorough was it? How many review iterations did it include? Why
> there
> is no name but anonymous review?
>
[mtk]: Sorry, I removed this "Reviewed-By:" messge. please review
patch:v3
>
>
> > Signed-off-by: Zhi Mao <[email protected]>
> > ---
> > drivers/media/i2c/Kconfig | 14 +
> > drivers/media/i2c/Makefile | 1 +
> > drivers/media/i2c/gc08a3.c | 1888
> ++++++++++++++++++++++++++++++++++++
> > 3 files changed, 1903 insertions(+)
> > create mode 100644 drivers/media/i2c/gc08a3.c
> >
>
> ...
>
> > +static inline struct gc08a3 *to_gc08a3(struct v4l2_subdev *sd)
> > +{
> > +return container_of(sd, struct gc08a3, sd);
> > +}
> > +
> > +static int gc08a3_power_on(struct device *dev)
> > +{
> > +struct i2c_client *client = to_i2c_client(dev);
> > +struct v4l2_subdev *sd = i2c_get_clientdata(client);
> > +struct gc08a3 *gc08a3 = to_gc08a3(sd);
> > +int ret;
> > +
> > +gpiod_set_value_cansleep(gc08a3->enable_gpio, 0);
>
> Why do you put make enable GPIO low? That's not how enable GPIO is
> supposed to work...
>
[mtk]: fixed in patch:v3

> > +usleep_range(GC08A3_MIN_SLEEP_US, GC08A3_MAX_SLEEP_US);
> > +
> > +ret = regulator_bulk_enable(GC08A3_NUM_SUPPLIES, gc08a3-
> >supplies);
> > +if (ret < 0) {
> > +dev_err(gc08a3->dev, "failed to enable regulators: %d\n", ret);
> > +return ret;
> > +}
> > +
> > +ret = clk_prepare_enable(gc08a3->xclk);
> > +if (ret < 0) {
> > +regulator_bulk_disable(GC08A3_NUM_SUPPLIES, gc08a3->supplies);
> > +dev_err(gc08a3->dev, "clk prepare enable failed\n");
> > +return ret;
> > +}
> > +
> > +usleep_range(GC08A3_MIN_SLEEP_US, GC08A3_MAX_SLEEP_US);
> > +
> > +gpiod_set_value_cansleep(gc08a3->enable_gpio, 1);
> > +usleep_range(GC08A3_MIN_SLEEP_US, GC08A3_MAX_SLEEP_US);
> > +
> > +return 0;
> > +}
> > +
> > +static int gc08a3_power_off(struct device *dev)
> > +{
> > +struct i2c_client *client = to_i2c_client(dev);
> > +struct v4l2_subdev *sd = i2c_get_clientdata(client);
> > +struct gc08a3 *gc08a3 = to_gc08a3(sd);
> > +
> > +gpiod_set_value_cansleep(gc08a3->enable_gpio, 0);
>
> How making enable GPIO low is related to power off? Enable means you
> turn on some feature, not shutdown. Look at common GPIO consumer
> bindings in the kernel.
>
[mtk]:Here, we rename it to "reset-pin", according to sensor power off
sequence in sepc, we should pull low this pin.
>
> ...
>
> > +static int gc08a3_probe(struct i2c_client *client)
> > +{
> > +struct device *dev = &client->dev;
> > +struct gc08a3 *gc08a3;
> > +int ret;
> > +
> > +ret = gc08a3_parse_fwnode(dev);
> > +if (ret)
> > +return ret;
> > +
> > +gc08a3 = devm_kzalloc(dev, sizeof(*gc08a3), GFP_KERNEL);
> > +if (!gc08a3)
> > +return -ENOMEM;
> > +
> > +gc08a3->dev = dev;
> > +
> > +gc08a3->xclk = devm_clk_get(dev, NULL);
> > +if (IS_ERR(gc08a3->xclk))
> > +return dev_err_probe(dev, PTR_ERR(gc08a3->xclk),
> > + "failed to get xclk\n");
>
> Misaligned indentation
>
[mtk]: fixed in patch:v3

> > +
> > +ret = clk_set_rate(gc08a3->xclk, GC08A3_DEFAULT_CLK_FREQ);
> > +if (ret)
> > +return dev_err_probe(dev, ret,
> > + "failed to set xclk frequency\n");
>
> Misaligned indentation
[mtk]: fixed in patch:v3

>
> > +
> > +ret = gc08a3_get_regulators(dev, gc08a3);
> > +if (ret < 0)
> > +return dev_err_probe(dev, ret,
> > + "failed to get regulators\n");
>
> Misaligned indentation
>
[mtk]: fixed in patch:v3

> > +
> > +gc08a3->enable_gpio = devm_gpiod_get(dev, "enable",
> GPIOD_OUT_LOW);
> > +if (IS_ERR(gc08a3->enable_gpio))
> > +return dev_err_probe(dev, PTR_ERR(gc08a3->enable_gpio),
> > + "failed to get gpio\n");
>
> Misaligned indentation... probably entire code is misaligned.
>
[mtk]: fixed in patch:v3

>
> Best regards,
> Krzysztof
>

2024-01-09 11:25:36

by Sakari Ailus

[permalink] [raw]

Subject: Re: [PATCH 1/2] media: i2c: Add GC08A3 image sensor driver

Hi Zhi,

On Tue, Jan 09, 2024 at 10:41:15AM +0000, Zhi Mao (毛智) wrote:
> > > +static const char *const gc08a3_supply_name[] = {
> > > +"avdd",
> > > +"dvdd",
> > > +"dovdd",
> > > +};
> > > +
> > > +#define GC08A3_NUM_SUPPLIES ARRAY_SIZE(gc08a3_supply_name)
> >
> > Please use ARRAY_SIZE(...) directly.
> >
> [mtk]: About "ARRAY_SIZE", creating a macro with a descriptive name can
> improve readability of code, especially when it is used in multiple
> locations in codes. and it seems a common usage in sensor drivers. Can
> we keep this usage in gc08a3 driver?

It improves readability even more if you use ARRAY_SIZE() directly as then
it's easy to see you're dealing with a single array. GC08A3_NUM_SUPPLIES is
thus a useless definition.

..

> > > +static int gc08a3_g_mbus_config(struct v4l2_subdev *sd, unsigned
> > int pad,
> > > +struct v4l2_mbus_config *config)
> > > +{
> > > +config->type = V4L2_MBUS_CSI2_DPHY;
> > > +config->bus.mipi_csi2.num_data_lanes = 4;
> > > +config->bus.mipi_csi2.flags = 0;
> > > +return 0;
> > > +}
> >
> > As you return a static configuration, there's no need to implement
> > g_mbus_config().
> >
> [mtk]: we can not remove this function, because meidatek ISP driver
> will use this interface to get some information.

Please fix the Mediatek ISP driver in that case.

I'm also open to adding a V4L2 framework function to obtain the number of
lanes (and other configuration) for an upstream sub-device, either using
the local endpoint or g_mbus_config if the sub-device driver implements
that.

--
Regards,

Sakari Ailus