Add a v4l2 subdevice driver for the Sony IMX283 image sensor.
  
The IMX283 is a 20MP Diagonal 15.86 mm (Type 1) CMOS Image Sensor with
Square Pixel for Color Cameras.
    
The following features are supported:
- Manual exposure an gain control support
- vblank/hblank/link freq control support
- Test pattern support control
- Arbitrary horizontal and vertical cropping
- Supported resolution:
   - 5472x3648 @ 20fps (SRGGB12)
   - 5472x3648 @ 25fps (SRGGB10)
   - 2736x1824 @ 50fps (SRGGB12)

The driver is tested on mainline branch v6.8-rc2 on IMX8MP Debix-SOM-A.
Additional testing has been done on RPi5 with the downstream BSP.

Changes in v4:
- fix 32-bit build error around u64 divisions (use do_div)
- Fix hmax default and minimum values

Changes in v3:
- fix headers includes
- Improve #define(s) readability
- Drop __func__ from error logs
- Use HZ_PER_MHZ instead of MEGA
- mdsel* variables should be u8
- Use container_of_const() instead of container_of()
- Use clamp() used of clamp_t variant
- Use streams API imx283_{enable|disable}_streams (**NOTE**)
- Properly fix PM runtime handling
  (pm_ptr(), DEFINE_RUNTIME_DEV_PM_OPS,
   imx283_runtime_suspend, imx283_runtime_resume)
- Fix format modifiers, signed-ness at various places

changes in v2 (summary):
- Use u32 wherever possible
- Use MEGA macro instead of self defined MHZ() macro
- Properly refine regs using CCI
- Drop tracking of current mode. Shifted to infer from active state directly.
  (Laurent's review)
- Cont. from above: Pass the struct imx283_mode to functions whereever required.
- Remove unused comments
- Remove custom mutex. Use control handler one instead.
- Drop imx283_reset_colorspace() and inline
- Set colorspace field properly (drop _DEFAULTS)
- Use __maybe_unused for imx283_power_on() and imx283_power_off()
- Store controls  v4l2_ctrl handles for those required, not all.
- Drop imx283_free_controls(). Use v4l2_ctrl_handler_free
- fix reset-gpios handling and add it to DT schema
- fix data-lanes property in DT schema
- fix IMX283 Kconfig
- Remove unused macros
- Alphabetical case consistency

Signed-off-by: Umang Jain <umang.jain@ideasonboard.com>
---
Kieran Bingham (1):
      media: i2c: Add imx283 camera sensor driver

Umang Jain (2):
      media: dt-bindings: media: Add bindings for IMX283
      fixups

 .../devicetree/bindings/media/i2c/sony,imx283.yaml |  107 ++
 MAINTAINERS                                        |    9 +
 drivers/media/i2c/Kconfig                          |   10 +
 drivers/media/i2c/Makefile                         |    1 +
 drivers/media/i2c/imx283.c                         | 1605 ++++++++++++++++++++
 5 files changed, 1732 insertions(+)
---
base-commit: 54ee11761885407056f4ca60309739e2db6b02dc
change-id: 20240402-kernel-name-extraversion-2b08d441e08c

Best regards,
-- 
Umang Jain <umang.jain@ideasonboard.com>

Hi all,

PLease ignore the series, I was testing/learning the b4 tool.

I did pass --offline-mode but it has sent the patches anyway :-//

On 02/04/24 1:59 pm, Umang Jain wrote:
> Add a v4l2 subdevice driver for the Sony IMX283 image sensor.
>    
> The IMX283 is a 20MP Diagonal 15.86 mm (Type 1) CMOS Image Sensor with
> Square Pixel for Color Cameras.
>      
> The following features are supported:
> - Manual exposure an gain control support
> - vblank/hblank/link freq control support
> - Test pattern support control
> - Arbitrary horizontal and vertical cropping
> - Supported resolution:
>     - 5472x3648 @ 20fps (SRGGB12)
>     - 5472x3648 @ 25fps (SRGGB10)
>     - 2736x1824 @ 50fps (SRGGB12)
>
> The driver is tested on mainline branch v6.8-rc2 on IMX8MP Debix-SOM-A.
> Additional testing has been done on RPi5 with the downstream BSP.
>
> Changes in v4:
> - fix 32-bit build error around u64 divisions (use do_div)
> - Fix hmax default and minimum values
>
> Changes in v3:
> - fix headers includes
> - Improve #define(s) readability
> - Drop __func__ from error logs
> - Use HZ_PER_MHZ instead of MEGA
> - mdsel* variables should be u8
> - Use container_of_const() instead of container_of()
> - Use clamp() used of clamp_t variant
> - Use streams API imx283_{enable|disable}_streams (**NOTE**)
> - Properly fix PM runtime handling
>    (pm_ptr(), DEFINE_RUNTIME_DEV_PM_OPS,
>     imx283_runtime_suspend, imx283_runtime_resume)
> - Fix format modifiers, signed-ness at various places
>
> changes in v2 (summary):
> - Use u32 wherever possible
> - Use MEGA macro instead of self defined MHZ() macro
> - Properly refine regs using CCI
> - Drop tracking of current mode. Shifted to infer from active state directly.
>    (Laurent's review)
> - Cont. from above: Pass the struct imx283_mode to functions whereever required.
> - Remove unused comments
> - Remove custom mutex. Use control handler one instead.
> - Drop imx283_reset_colorspace() and inline
> - Set colorspace field properly (drop _DEFAULTS)
> - Use __maybe_unused for imx283_power_on() and imx283_power_off()
> - Store controls  v4l2_ctrl handles for those required, not all.
> - Drop imx283_free_controls(). Use v4l2_ctrl_handler_free
> - fix reset-gpios handling and add it to DT schema
> - fix data-lanes property in DT schema
> - fix IMX283 Kconfig
> - Remove unused macros
> - Alphabetical case consistency
>
> Signed-off-by: Umang Jain <umang.jain@ideasonboard.com>
> ---
> Kieran Bingham (1):
>        media: i2c: Add imx283 camera sensor driver
>
> Umang Jain (2):
>        media: dt-bindings: media: Add bindings for IMX283
>        fixups
>
>   .../devicetree/bindings/media/i2c/sony,imx283.yaml |  107 ++
>   MAINTAINERS                                        |    9 +
>   drivers/media/i2c/Kconfig                          |   10 +
>   drivers/media/i2c/Makefile                         |    1 +
>   drivers/media/i2c/imx283.c                         | 1605 ++++++++++++++++++++
>   5 files changed, 1732 insertions(+)
> ---
> base-commit: 54ee11761885407056f4ca60309739e2db6b02dc
> change-id: 20240402-kernel-name-extraversion-2b08d441e08c
>
> Best regards,

On Tue, Apr 02, 2024 at 02:01:19PM +0530, Umang Jain wrote:
> Hi all,
> 
> PLease ignore the series, I was testing/learning the b4 tool.
> 
> I did pass --offline-mode but it has sent the patches anyway :-//

Hm... I see why this would be confusing and I will try to fix this.

(the --offline-mode flag is to prevent b4 from querying lore.kernel.org, but
it should also bubble up to other functionality that calls remote services,
like b4 send).

-K