Re: [PATCH v5 00/10] media: subdev: Improve stream enable/disable machinery

[Umang Jain]

Hi Tomi,

On 16/04/24 7:25 pm, Tomi Valkeinen wrote:
> This series works on the .s_stream, .enable_streams, .disable_streams
> related code. The main feature introduced here, in the last two patchs,
> is adding support for .enable/disable_streams() for subdevs that do not
> implement full streams support.
>
> Additionally we add support for the privacy led when
> v4l2_subdev_enable/disable_streams() is used.
>
> I have tested this on RPi5.

I have also tested the series on i.MX8MP with IMX283 hence,

Tested-by: Umang Jain <umang.jain@xxxxxxxxxxxxxxxx>
>   Tomi
>
> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@xxxxxxxxxxxxxxxx>
> ---
> Changes in v5:
> - Fix issues with non-routing case:
>    - Always set 'found_streams' variable instead of leaving it
>      uninitialized in "Support single-stream case in
>      v4l2_subdev_enable/disable_streams()"
>    - Fix the "implicit" stream from bit 1 to bit 0 in "Support
>      non-routing subdevs in v4l2_subdev_s_stream_helper()"
> - Link to v4: https://lore.kernel.org/r/20240416-enable-streams-impro-v4-0-1d370c9c2b6d@xxxxxxxxxxxxxxxx
>
> Changes in v4:
> - Added Rb tags
> - Rename 'streaming_enabled' to 's_stream_enabled'
> - Cosmetic changes (comments / patch descs)
> - Added new patch "media: subdev: Support non-routing subdevs in  v4l2_subdev_s_stream_helper()".
> - Link to v3: https://lore.kernel.org/r/20240410-enable-streams-impro-v3-0-e5e7a5da7420@xxxxxxxxxxxxxxxx
>
> Changes in v3:
> - Rebased on top of "[PATCH v2 1/1] media: v4l: Don't turn on privacy LED if streamon fails"
> - Drop extra "!!" in "media: subdev: Fix use of sd->enabled_streams in  call_s_stream()"
> - Enable privacy LED after a succesfull stream enable in  "media: subdev: Support privacy led in v4l2_subdev_enable/disable_streams()"
> - Init 'cfg' variable when declaring in "media: subdev: Refactor v4l2_subdev_enable/disable_streams()"
> - Link to v2: https://lore.kernel.org/r/20240405-enable-streams-impro-v2-0-22bca967665d@xxxxxxxxxxxxxxxx
>
> Changes in v2:
> - New patches for privacy led
> - Use v4l2_subdev_has_op()
> - Use BITS_PER_BYTE instead of 8
> - Fix 80+ line issues
> - Fix typos
> - Check for pad < 64 also in the non-routing .enable/disable_streams code path
> - Dropped "media: subdev: Support enable/disable_streams for non-streams
>    subdevs", which is implemented in a different way in new patches in this series
> - Link to v1: https://lore.kernel.org/r/20240404-enable-streams-impro-v1-0-1017a35bbe07@xxxxxxxxxxxxxxxx
>
> ---
> Tomi Valkeinen (10):
>        media: subdev: Add privacy led helpers
>        media: subdev: Use v4l2_subdev_has_op() in v4l2_subdev_enable/disable_streams()
>        media: subdev: Add checks for subdev features
>        media: subdev: Fix use of sd->enabled_streams in call_s_stream()
>        media: subdev: Improve v4l2_subdev_enable/disable_streams_fallback
>        media: subdev: Add v4l2_subdev_is_streaming()
>        media: subdev: Support privacy led in v4l2_subdev_enable/disable_streams()
>        media: subdev: Refactor v4l2_subdev_enable/disable_streams()
>        media: subdev: Support single-stream case in v4l2_subdev_enable/disable_streams()
>        media: subdev: Support non-routing subdevs in v4l2_subdev_s_stream_helper()
>
>   drivers/media/v4l2-core/v4l2-subdev.c | 378 ++++++++++++++++++++--------------
>   include/media/v4l2-subdev.h           |  25 ++-
>   2 files changed, 245 insertions(+), 158 deletions(-)
> ---
> base-commit: 6547a87b1ffc9b3c3a17f20f71016de98c169bbb
> change-id: 20240404-enable-streams-impro-db8bcd898471
>
> Best regards,