Re: [PATCH v12 00/17] media: cadence,ti: CSI2RX Multistream Support

[Tomi Valkeinen]

Hi,

On 13/03/2026 11:06, Rishikesh Donadkar wrote:
> This series adds multi-stream support and PM support for Cadence CSI2RX
> and TI CSI2RX SHIM drivers.
> 

Looks fine to me. Tested on AM62A SK + V3Link board.

 Tomi

> PM patches are picked from:
> https://lore.kernel.org/all/20250902-ti_csi_pm-v2-0-59a3be199940@xxxxxxxxxxxxxxxx/
> 
> PATCH 01 :    Remove word size alignment restriction on frame width
> PATCH 02-08:  Support multiple DMA contexts/video nodes in TI CSI2RX
> PATCH 09-10:  Use get_frame_desc to propagate virtual channel
>               information across Cadence and TI CSI-RX subdevs
> PATCH 11-12:  Use new multi-stream APIs across the drivers to support
>               multiplexed cameras from sources like UB960 (FPDLink)
> PATCH 13:     Optimize stream on by submitting all queued buffers to DMA
> PATCH 14:     Change the drain architecture to support multi-stream,
>               implement completion barriers for last drain
> PATCH 15-17:  Runtime PM and System PM support for CSI-RX.
> 
> Testing for this series has been done on top of media tree with 4x IMX219
> camera modules connected to TI's AM62A using V3 Link fusion mini board.
> 
> Runtime PM and System PM has been tested with IMX219 camera module
> connect to TI's AM62.
> 
> Follwing formats and resolutions have been tested:
> 
> Format    | Resolution | Sensor | 
> ----------|------------|--------|
> SRGGB10   | 1640x1232  | IMX219 | 
> SRGGB10   | 1920x1080  | IMX219 |
> SRGGB8    | 1640x1232  | IMX219 | 
> SRGGB8    | 1920x1080  | IMX219 |
> 
> Overlay and defconfig changes for the same can be found below:
> https://github.com/RISHI27-dot/linux/commits/u/multistream_v12
> 
> v4l2-compliance results:
> https://gist.github.com/Rishikesh-D/85d2f1ceab61fdc4e3340349c12823b4
> 
> ---
> Changes in v12:
> [PATCH v12 06/17] media: ti: j721e-csi2rx: add a subdev for the core
> device
> - Report the format not being present as error and use dev_err instead of dev_dbg
> - Add .has_pad_interdep
> [PATCH v12 11/17] media: cadence: csi2rx: add multistream support
> - Add .has_pad_interdep
> [PATCH v12 14/17] media: ti: j721e-csi2rx: Change the drain architecture for multistream
> - Remove debug print when wait_for_completion() timeouts
> [PATCH v12 16/17] media: ti: j721e-csi2rx: Support runtime suspend
> - Init dma and request channels at probe time
> 
> Link to (v11):
> https://lore.kernel.org/all/20260209183456.1906327-1-r-donadkar@xxxxxx/
> ---
> Changes in v11:
> [PATCH v11 04/18] media: ti: j721e-csi2rx: prepare SHIM code for multiple contexts
> - Remove register write that caused bogus image for SRGGB10
> [PATCH v11 10/18] media: ti: j721e-csi2rx: add support for processing virtual channels
> - Update commit message to mention about DT being handled
> [PATCH v11 12/18] media: ti: j721e-csi2rx: add multistream support
> - s/form/from
> - Move code for falling back to default VC and DT inside
>   ti_csi2rx_get_vc_and_dt()
> 
> Link to (v10):
> https://lore.kernel.org/all/20260121135424.1185710-1-r-donadkar@xxxxxx/
> ---
> Changes in v10:
> 
> # New patches in v10:
> [PATCH v10 12/18] media: cadence: csi2rx: Use the stream from route
> 
> #Changes
> [PATCH v10 06/18] media: ti: j721e-csi2rx: add a subdev for the core
> - Remove unwanted mutex
> [PATCH v10 10/18] media: ti: j721e-csi2rx: add support for processing
> - Improve commit message to remove info about get_frame_desc() usage
>   in single/multi stream usecases
> - Add break when stream found
> - Report error when no stream found
> [PATCH v10 11/18] media: cadence: csi2rx: add multistream support
> - Move the code to set default VC0 inside if statement
> [PATCH v10 13/18] media: ti: j721e-csi2rx: add multistream support
> - Make the pixel reset assert and de-assert more symmetric across
>   the start and stop path
> - Remove the vc_cached variable and associated checks
> - Move enable_count == 0 check at start of
>   ti_csi2rx_sd_disable_streams() with a WARN_ON
> - Update the commit message to s/4/32 for number of possible streams
> [PATCH v10 15/18] media: ti: j721e-csi2rx: Change the drain
>  architecture for multistream
> - Change the commit message to talk about streams only and not
>   VC/DT
> - Improve commit message to give better description regarding continuous
>   drain
> - Squash [PATCH v9 16/19] into this patch
> [PATCH v10 18/18] media: ti: j721e-csi2rx: Support system suspend
>  using pm_notifier
> - Add comment before register_pm_notifier() call
> 
> Link to (v9):
> https://lore.kernel.org/all/20251230083220.2405247-1-r-donadkar@xxxxxx/
> ---
> Changes in v9:
> 
> # New patches in v9:
> [PATCH v9 08/18] media: staging: starfive: Move to .enable/disable_streams API
> - Remove the s_stream API and use the enable/disable_streams API in stf
>   drivers
> [PATCH v9 16/19] media: ti: j721e-csi2rx: Return the partial frame as error
> - Return the next frame after drain to usersapce as VB2_BUF_STATE_ERROR
> 
> # Changes:
> [PATCH v9 01/18] media: ti: j721e-csi2rx: Remove word size alignment on frame width
> - Ad R-by Tomi
> [PATCH v9 06/18] media: ti: j721e-csi2rx: add a subdev for the core device
> - Squash [PATCH v8 08/18] into this patch so that we dont have to use
>   s_stream
> - Use BIT_U64() instead of BIT()
> [PATCH v9 07/18] media: cadence: csi2rx: Move to .enable/disable_streams API
> - Use BIT_U64() instead of BIT()
> [PATCH v9 11/19] media: ti: j721e-csi2rx: add support for processing virtual channels
> - Update to commit message to s/hardcoded/VC0/
> - Remove the mention of caching VC data from the commit message as that
>   is done in 13/19
> - Update the commit message to s/per stream/single stream/
> [PATCH v9 12/19] media: cadence: csi2rx: add multistream support
> - Remove the s_stream fallback
> [PATCH v9 13/19] media: ti: j721e-csi2rx: add multistream support
> - Move the code that is accessing the state from .start_streaming() call
>   to .enable_streams() call
> - Remove the check for V4L2_SUBDEV_ROUTE_FL_ACTIVE
> - Change the logic of for_each_active_route() usage to break the loop
>   when the route is found
> - Change the format variable to static const
> - Update the commit message to s/4/32 for number of possible streams
> [PATCH v9 17/19] media: cadence: csi2rx: Support runtime PM
> - Shift to goto based error handing in the .enable_streams() function
> [PATCH v9 18/19] media: ti: j721e-csi2rx: Support runtime suspend
> - Make variable i and ret unsigned
> - Move runtime_enable() call before notifier_register() in the .probe()
> - In the .remove() function if the runtime PM state is active, set it to
>   suspended and remove the pm_runtime_set_suspended() call at the end.
> - Drop the call to init_dma() from init/cleanup_ctx()
> - Remove pm_runtime_set_active() and pm_request_idle() from .probe()
> 
> Link to (v8):
> https://lore.kernel.org/all/20251112115459.2479225-1-r-donadkar@xxxxxx/#t
> ---
> Changes in v8:
> 
> # New patches in v8:
> [PATCH v8 07/18] media: cadence: csi2rx: Move to .enable/disable_streams API
> [PATCH v8 08/18] media: ti: j721e-csi2rx: Move to .enable/disable_streams API
> - Remove the s_stream API and use the enable/disable_streams API right
>   after adding the subdev for core device
> 
> Changes:
> [PATCH v8 01/18] media: ti: j721e-csi2rx: Remove word size alignment on frame width
> - Remove define for PSIL_WORD_SIZE_BYTES.
> [PATCH v8 06/18] media: ti: j721e-csi2rx: add a subdev for the core device
> - Use v4l2_subdev_state_get_format() instead of using the get_fmt from the 
>   subdev pad ops in the ti_csi2rx_link_validate() function.
> [PATCH v8 11/18] media: ti: j721e-csi2rx: add support for processing virtual channels
> - Move routing support form this patch to PATCH 13/18.
> [PATCH v8 12/18] media: cadence: csi2rx: add multistream support
> - Move support for enable/disable_streams() API that was added in this patch to
>   PATCH 07/18 earlier in this series.
> - Remove the mutex from the driver
> [PATCH v8 13/18] media: ti: j721e-csi2rx: add multistream support
> - Move support for enable/disable_streams() API that was added in this patch to
>   PATCH 08/18 earlier in this series.
> - Remove csi->mutex protection inside enable/disable_streams() function calls to
>   solve circular locking dependency warning.
> - Mention that DT filtering is also supported in the commit message.
> 
> Link to (v7):
> https://lore.kernel.org/all/20250911102832.1583440-1-r-donadkar@xxxxxx/#t
> 
> ---
> Changes in v7:
> 
> [PATCH v6 03/16] media: ti: j721e-csi2rx: separate out device and context
> - Resolve bisect test fail because of undefined 'csi' variable
> - Resolve bisect test fail because of 'vdev' not being in scope when used
> [PATCH v6 07/16] media: ti: j721e-csi2rx: get number of contexts from device tree
> - Resolve compiler warning regarding missing error code 'ret'
> [PATCH v6 12/16] media: ti: j721e-csi2rx: Submit all available buffers
> - Remove unused variable
> 
> Link to (v6):
> https://lore.kernel.org/all/20250908134729.3940366-1-r-donadkar@xxxxxx/
> 
> ---
> Changes in v6:
> 
> # New patches in v6:
> 
> [PATCH v6 14/16] media: cadence: csi2rx: Support runtime PM
> [PATCH v6 15/16] media: ti: j721e-csi2rx: Support runtime suspend
> [PATCH v6 16/16] media: ti: j721e-csi2rx: Support system suspend
> using pm_notifier
> 
> # Changes in patches from v5:
> [PATCH v5 01/16] media: ti: j721e-csi2rx: Remove word size alignment
> on frame width
> - Remove restrictions on minimum frame width in ti_csi2rx_fill_fmt()
> - Add Reviewed-by from Abhilash
> [PATCH v5 02/16] dt-bindings: media: ti,j721e-csi2rx-shim: Support 32 dma chans
> - No change
> [PATCH v5 03/16] media: ti: j721e-csi2rx: separate out device and context
> - No change
> [PATCH v5 04/16] media: ti: j721e-csi2rx: prepare SHIM code for multiple contexts
> - No change
> [PATCH v5 05/16] media: ti: j721e-csi2rx: allocate DMA channel based on context index
> - No change
> [PATCH v5 06/16] media: ti: j721e-csi2rx: add a subdev for the core device
> - No change
> [PATCH v5 07/16] media: ti: j721e-csi2rx: get number of contexts from device tree
> - Solve compilation warnings
> [PATCH v5 08/16] media: cadence: csi2rx: add get_frame_desc wrapper
> - No change
> [PATCH v5 09/16] media: ti: j721e-csi2rx: add support for processing virtual channels
> - Remove unnecessary call to v4l2_unlock_state()
> [PATCH v5 10/16] media: cadence: csi2rx: add multistream support
> - No change
> [PATCH v5 11/16] media: ti: j721e-csi2rx: add multistream support
> - No change
> [PATCH v5 12/16] media: ti: j721e-csi2rx: Submit all available buffers
> - No change
> [PATCH v5 13/16] media: ti: j721e-csi2rx: Change the drain architecture for multistream
> - Squash implementation of completion barriers for last drain cycle
> - Make struct completion per ctx
> 
> Link to (v5):
> https://lore.kernel.org/all/20250825142522.1826188-1-r-donadkar@xxxxxx/
> 
> ---
> Changes in v5:
> 
> # New patches in v5:
> 
> [PATCH v5 01/14] media: ti: j721e-csi2rx: Remove word size alignment
> [PATCH v5 14/14] media: ti: j721e-csi2rx: Wait for the last drain
> 
> # Changes in patches from v4:
> 
> [PATCH v4 01/12] dt-bindings: media: ti,j721e-csi2rx-shim: Support 32 dma chans
> - No change
> [PATCH v4 02/12] media: ti: j721e-csi2rx: separate out device and context
> - No change
> [PATCH v4 03/12] media: ti: j721e-csi2rx: prepare SHIM code for multiple contexts
> - No change
> [PATCH v4 04/12] media: ti: j721e-csi2rx: allocate DMA channel based on context index
> - No change
> [PATCH v4 05/12] media: ti: j721e-csi2rx: add a subdev for the core device
> - No change
> [PATCH v4 06/12] media: ti: j721e-csi2rx: get number of contexts from device tree
> - No change
> [PATCH v4 07/12] media: cadence: csi2rx: add get_frame_desc wrapper
> - No change
> [PATCH v4 08/12] media: ti: j721e-csi2rx: add support for processing virtual channels
> - No change
> [PATCH v4 09/12] media: cadence: csi2rx: add multistream support
> - No change
> [PATCH v4 10/12] media: ti: j721e-csi2rx: add multistream support
> - Serialize stream stop
> - Remove the break statement to avoid early return in the loop, as
>   reported by Sjoerd
> [PATCH v4 11/12] media: ti: j721e-csi2rx: Submit all available buffers
> - Delete the list node on DMA error to avoid kernel panic
> [PATCH v4 12/12] media: ti: j721e-csi2rx: Change the drain architecture for multistream
> - Mention about next frame after drain being bogus
> 
> Link to (v4):
>   https://lore.kernel.org/all/20250514112527.1983068-1-r-donadkar@xxxxxx/
> 
> Changes in v4:
> 
> [PATCH 01/13] dt-bindings: media: ti,j721e-csi2rx-shim: Support 32 dma chans
>   - No change
> [PATCH 02/13] media: ti: j721e-csi2rx: separate out device and context
>   - Add ctx identifier in the dev_err() message
>   - No change
> [PATCH 03/13] media: ti: j721e-csi2rx: prepare SHIM code for multiple contexts
>   - Reduced the name string lenght from 32 chars to 5 chars
> [PATCH 04/13] media: ti: j721e-csi2rx: allocate DMA channel based on context index
>   - No change
> [PATCH 05/13] media: ti: j721e-csi2rx: add a subdev for the core device
>   - Add .enum_mbus_code callback
>   - Replace statically allocated struct with a global static const struct
>     v4l2_mbus_framefmt and used that in the _init_state() function
> [PATCH 06/13] media: ti: j721e-csi2rx: get number of contexts from device tree
>   - Fix the drain buffer being leaked
>   - If the shows more number of ctx than the TI_CSI2RX_MAX_CTX, return an error
>     instead of warning
> [PATCH 07/13] media: cadence: csi2rx: add get_frame_desc wrapper
>   - No change
> [PATCH 08/13] media: ti: j721e-csi2rx: add support for processing virtual channels
>   - Call ti_csi2rx_get_vc() only once on first stream start and cache the VC data in
>     the driver, use the corresponding VC in all subsequent stream starts.
> [PATCH 09/13] media: cadence: csi2rx: Use new enable stream APIs
> [PATCH 10/13] media: cadence: csi2rx: Enable multi-stream support
>   - Squash the above two patches into
>     [PATCH v4 09/12] media: cadence: csi2rx: add multistream support
>   - Use already obtained csi2rx->source_pad in enable_streams() and
>     disable_streams() call
>   - Update commit message with the reason for using a custom helper for s_stream
>     instead of v4l2_subdev_s_stream_helper()
>   - Use v4l2_get_link_freq() variant that takes pad of the source as its first
>     argument instead of the one that takes v4l2_ctrl_handler
>   - Call v4l2_get_link_freq() with bpp = 0 to prevent fallback to V4L2_CID_PIXEL_RATE
>     in multi-stream case
>   - Use lock guards to simplify error handling
>   - Call csi2rx_update_vc_select() at first stream start before enabling the controller
> [PATCH 11/13] media: ti: j721e-csi2rx: add multistream support
>   - No change
> [PATCH 12/13] media: ti: j721e-csi2rx: Submit all available buffers
>   - No change
> [PATCH 13/13] media: ti: j721e-csi2rx: Change the drain architecture for multistream
>   - Fix checkpatch warning
>   - Change commit message to give a better description of the patch
> 
> Link to (v3):
>   https://lore.kernel.org/all/20250417065554.437541-1-r-donadkar@xxxxxx/
> 
> Changes in v3:
> 
> - Drop [PATCH v2 01/13] media: cadence: csi2rx: Support runtime PM from
>   v2, support for runtime PM will be added in a separate series:
>   https://lore.kernel.org/all/20250224-ti_csi_pm-v1-0-8f8c29ef646d@xxxxxxxxxxxxxxxx/
> - Change the drain architecture to prevent FIFO overflow in multistream
>   usecases.
> - With the new drain architecture, we don't need the the driver to wait
>   for userspace to start streaming on all "actively routed" video nodes
>   before starting streaming on the source. So, revert back to the capture
>   architecture where streams can be started and stopped independent
>   to each other.
> 
> Link to (v2):
>   https://lore.kernel.org/r/20240627-multistream-v2-0-6ae96c54c1c3@xxxxxx
> 
> Changes in v2:
> 
> - Change the multi-camera capture architecture to be similar to that of
>   Tomi's RPi5 FE series, where the driver will wait for userspace to
>   start streaming on all "actively routed" video nodes before starting
>   streaming on the source. This simplifies things a lot from the HW
>   perspective, which might run into deadlocks due to a shared FIFO
>   between multiple DMA channels.
> 
> - Drop a few fixes that were posted separately and are already merged
> - Fix dtschema warnings reported by Rob on [02/13]
> - Fix warnings for uninitialized `used_vc` variable in cdns-csi2rx.c
> - Return -EBUSY if someone updates routes for j721e-csi2rx subdev while
>   streaming
> - Only allow single-streams to be routed to the source pads (linked to
>   video nodes) of the j721e-csi2rx device
> - Squash the patches marked "SQUASH" in the v1 RFC series
> 
> Changhuang Liang (1):
>   media: cadence: csi2rx: Support runtime PM
> 
> Jai Luthra (9):
>   dt-bindings: media: ti,j721e-csi2rx-shim: Support 32 dma chans
>   media: ti: j721e-csi2rx: separate out device and context
>   media: ti: j721e-csi2rx: add a subdev for the core device
>   media: ti: j721e-csi2rx: add support for processing virtual channels
>   media: cadence: csi2rx: add multistream support
>   media: ti: j721e-csi2rx: add multistream support
>   media: ti: j721e-csi2rx: Submit all available buffers
>   media: ti: j721e-csi2rx: Support runtime suspend
>   media: ti: j721e-csi2rx: Support system suspend using pm_notifier
> 
> Pratyush Yadav (4):
>   media: ti: j721e-csi2rx: prepare SHIM code for multiple contexts
>   media: ti: j721e-csi2rx: allocate DMA channel based on context index
>   media: ti: j721e-csi2rx: get number of contexts from device tree
>   media: cadence: csi2rx: add get_frame_desc wrapper
> 
> Rishikesh Donadkar (3):
>   media: ti: j721e-csi2rx: Remove word size alignment on frame width
>   media: cadence: csi2rx: Move to .enable/disable_streams API
>   media: ti: j721e-csi2rx: Change the drain architecture for multistream
> 
>  .../bindings/media/ti,j721e-csi2rx-shim.yaml  |   39 +-
>  drivers/media/platform/cadence/Kconfig        |    1 +
>  drivers/media/platform/cadence/cdns-csi2rx.c  |  509 +++++--
>  drivers/media/platform/ti/Kconfig             |    1 +
>  .../platform/ti/j721e-csi2rx/j721e-csi2rx.c   | 1219 ++++++++++++-----
>  5 files changed, 1320 insertions(+), 449 deletions(-)
>