- 11 Mar, 2021 40 commits
-
-
Laurent Pinchart authored
The CSI subdev has a single source, the input video mux. There's thus a single link between the source and the CSI, which can be made immutable and enabled to simplify configuration. Signed-off-by:
Laurent Pinchart <laurent.pinchart@ideasonboard.com> Reviewed-by:
Rui Miguel Silva <rmfrfs@gmail.com> Signed-off-by:
Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by:
Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
-
Laurent Pinchart authored
Add a flags argument to the v4l2_create_fwnode_links_to_pad() function to specify the link flags. This allows drivers to create immutable links for instance. Signed-off-by:
-
Laurent Pinchart authored
The CSI driver only supports the BGGR Bayer patterns currently. The hardware supports all patterns (the only pattern-dependent hardware operation is statistics calculation, as de-bayering isn't supported), enable them in the driver too. Signed-off-by:
-
Laurent Pinchart authored
The SOF and EOF interrupts are not used. Don't enable them. Signed-off-by:
-
Laurent Pinchart authored
The DMA controller needs to be reflashed after being configured. There is however no need to do it twice, once in imx7_csi_configure() and once in imx7_csi_sw_reset(), called from imx7_csi_enable(). Remove the former. Signed-off-by:
-
Laurent Pinchart authored
The RxFIFO is enabled by imx7_csi_dmareq_rff_enable() in both imx7_csi_init() and imx7_csi_enable(). Enabling it once is enough, remove the former. Signed-off-by:
-
Laurent Pinchart authored
The CSI is enabled at the end of imx7_csi_sw_reset(), called from imx7_csi_enable(), as well as at the end of imx7_csi_enable(). The latter is enough, remove the former. Signed-off-by:
-
Laurent Pinchart authored
There's no need to read the CSICR1 register before each write in imx7_csi_rx_fifo_clear(). Simplify the function. Signed-off-by:
-
Laurent Pinchart authored
The RFF burst type is set in imx7_csi_dmareq_rff_enable(), which doesn't match the function name. Move it to imx7_csi_configure(). Signed-off-by:
-
Laurent Pinchart authored
When configuring the CSI in imx7_csi_configure(), the CSICR18 is set through a read-modify-write cycle to avoid affecting fields whose usage isn't clear. The function then sets fields depending on whether the input is interlaced or progressive, and from the parallel input or the CSI-2 receiver. Those bits are only set and never cleared. For instance, when switching from a CSI-2 source to the parallel input, the BIT_DATA_FROM_MIPI will stay set. Fix this issue by first clearing all the fields we need to configure. Add BIT_CSI_HW_ENABLE to the set of fields being cleared, as the CSI needs to start in the disabled state. This allows dropping the call to imx7_csi_hw_disable() in imx7_csi_sw_reset(). Signed-off-by:
-
Laurent Pinchart authored
Merge all the configuration steps (imx7_csi_buf_stride_set(), imx7_csi_deinterlace_enable(), imx7_csi_set_imagpara() and imx7_csi_init_default()) in the imx7_csi_configure() function. This simplifies the configuration procedure by removing most read-update-write cycles, and makes the code generally more readable. imx7_csi_init_default() is kept as it is also called from imx7_csi_deinit(). This will be addressed separately. Signed-off-by:
-
Laurent Pinchart authored
There's no need to set the buffer stride to 0 when disabling the CSI. Remove the extraneous register write. Signed-off-by:
-
Laurent Pinchart authored
The MIPI CSI-2 data type depends solely on the media bus code on the sink pad. Deriving it from the pixel format is confusing. Simplify the code. Signed-off-by:
-
Laurent Pinchart authored
The imx7_csi_hw_reset() and imx7_csi_init_interface() functions are always called together. Merge them. This allows simplifying the code by avoiding duplicated register writes. As the imx7_csi_hw_reset() function didn't perform a hardware reset by initialized the registers to reset defaults (in addition to resetting the frame counter), name the resulting function imx7_csi_init_default() as it sets default values. Signed-off-by:
-
Laurent Pinchart authored
The imx7_csi_streaming_start() function just wraps imx7_csi_enable(). Call the latter directly. Similarly, merge imx7_csi_streaming_stop() and imx7_csi_disable() as they're both called from a single location only. Signed-off-by:
-
Laurent Pinchart authored
There's no reason to delay the CSI configuration and the DMA setup after starting the source. Move it before, simplifying error handling at stream start. Signed-off-by:
-
Laurent Pinchart authored
The imx7_csi_dma_stop() function performs two operations, it waits for DMA to be complete, and then cleans up and frees buffers. Split the second part out to a new imx7_csi_dma_cleanup() function to match imx7_csi_dma_setup(), in preparation for a stream start refactoring. Signed-off-by:
-
Laurent Pinchart authored
The imx7_csi_dma_start() doesn't start DMA. Rename it to imx7_csi_dma_setup() to avoid confusion. Signed-off-by:
-
Laurent Pinchart authored
The is_csi2 field can't be accessed concurrently by imx7_csi_pad_link_validate() and imx7_csi_configure(), as the latter is called from imx7_csi_s_stream(), which is called after link validation. Drop the lock. Signed-off-by:
-
Laurent Pinchart authored
The code in imx7_csi_pad_link_validate() that checks the type of the source incorrectly handles devices that have no parallel input. In that case, the source entity of the CSI is the CSI-2 receiver, not the video mux, and the driver will proceed to check the type of the source of the CSI-2 receiver. Make the code more explicit to fix this, by handling the three cases (parallel input only, CSI-2 receiver only, and video mux) separately. Note that the driver will not correctly handle the case where only a parallel input is present, and the external entity connected to the parallel input reports a MEDIA_ENT_F_VID_IF_BRIDGE or MEDIA_ENT_F_VID_MUX function. This was broken already, and should be fixed separately. Fixes: f5ffb81f ("media: imx7: csi: Fix pad link validation") Signed-off-by:
-
Laurent Pinchart authored
Move capture format validation from imx7_csi_configure() to .link_validate(). This simplifies error handling at stream on time, as imx7_csi_configure) now becomes a void function. Signed-off-by:
-
Laurent Pinchart authored
To prepare for a major rework of the hardware initialization, and make the code easier to read, reorder functions to group them in sections. No functional change is included. Signed-off-by:
-
Laurent Pinchart authored
Describe multi-bit fields with parameterized macros where applicable, and use them to replace manual shifts and hardcoded numerical values. Add macros for the CSICR2 fields used in the driver. While at it, align the indentation for all macros, and use lower-case hex numbers. Signed-off-by:
-
Laurent Pinchart authored
The CSICR*_RESET_VAL macros only obfuscate code. Use their value directly, replacing the numerical value with the macros that describe register bits. Signed-off-by:
-
Laurent Pinchart authored
The i.MX7 camera pipeline is simple, with the CSI connected directly to the capture device without any other option. There's thus no need to allow link configurable by userspace. Make it immutable. Signed-off-by:
-
Laurent Pinchart authored
There's no need to initialize the CSI every time a link is enabled (and de-initialize it when a link is disabled). Move initialization to .s_stream() instead. Signed-off-by:
-
Laurent Pinchart authored
The control handler isn't used, drop it. Signed-off-by:
-
Laurent Pinchart authored
When the subdevice connected to the capture device has a single possible sink, there's no point in making the link mutable. Support creating immutable links. Signed-off-by:
-
Laurent Pinchart authored
Support for the MC-centric API has been tested on the i.MX7. Enable it for that platform. i.MX6 should be converted next. Signed-off-by:
-
Laurent Pinchart authored
Report to userspace that the MC-centric API is MC-centric by exposing the V4L2_CAP_IO_MC. This requires adding support for mbus code filtering in format enumeration, as required by V4L2_CAP_IO_MC. Signed-off-by:
-
Laurent Pinchart authored
When operating in MC-centric mode, the behaviour of video nodes shall not be influenced by the active configuration of the source subdev. Add a set of ioctl handlers that implement this mode, and select them when support for the legacy API is not requested. Signed-off-by:
-
Laurent Pinchart authored
The __capture_legacy_try_fmt() function returns two values through pointer arguments. One is a compose rectangle, which duplicates informationr returned through the subdev format argument, and can thus be removed. The other is the imx_media_pixfmt, which can be returned by value instead. Simplify the implementation of __capture_legacy_try_fmt() by dropping the retcc and compose arguments, and returning the imx_media_pixfmt by value. Signed-off-by:
Steve Longerbeam <slongerbeam@gmail.com> Signed-off-by:
-
Laurent Pinchart authored
capture_validate_fmt() delegates the media bus format to pixel format conversion to __capture_legacy_try_fmt(). It turns out that this can be simplified quite a lot: - The format lookup from media bus code can be performed by the recently added capture_find_format() function instead of receiving the information from a side effect of __capture_legacy_try_fmt(). - The validation of the pixel format size isn't needed, as they duplicate the validation of the compose rectangle. The pixel format size is directly derived from the size of the incoming stream, which is stored in the compose rectangle. If the compose rectangle is valid, the pixel format size will be valid too. By removing the validation of the pixel format size and using capture_find_format() to lookup the format, we can stop using __capture_legacy_try_fmt() and simplify capture_validate_fmt(). Signed-off-by:
-
Laurent Pinchart authored
Extract the format lookup code from __capture_legacy_try_fmt() to a separate function, it will be reused when validating the video node format at stream start time. Signed-off-by:
-
Laurent Pinchart authored
The correct return code to report an invalid pipeline configuration is -EPIPE. Return it instead of -EINVAL from __capture_legacy_try_fmt() when the capture format doesn't match the media bus format of the connected subdev. Signed-off-by:
-
Laurent Pinchart authored
The __capture_legacy_try_fmt() function only needs the v4l2_pix_format embedded in the v4l2_format argument it receives. Pass it the v4l2_pix_format directly, allowing on caller to allocate a smaller structure on the stack. Signed-off-by:
-
Laurent Pinchart authored
The __capture_legacy_try_fmt() function doesn't need two imx_media_pixfmt pointer variables. Remove one of them. Signed-off-by:
-
Laurent Pinchart authored
Add a parameter to the imx_media_capture_device_init() function to select between the legacy and MC-centric API. When selecting the MC-centric API, the video node doesn't inherit controls from subdevs anymore. All callers are updated to use the legacy API for now, preserving the existing behaviour. Signed-off-by:
-
Laurent Pinchart authored
The i.MX media drivers implement a legacy video node API, where the format of the video node is influenced by the active format of the connected subdev (both for enumeration and for the get, set and try format ioctls), and where controls exposed by the subdevs in the pipeline are inherited by the video node. At the same time, the drivers implement the media controller API and expose subdev video nodes to userspace. Those two modes of operation are incompatible and should not be exposed together. Furthermore, the legacy API gets in the way of proper enumeration of pixel formats on the video node, as it prevents compliance with the V4L2 specification. As a first step towards fixing this, rename all V4L2 video node ioctl handlers with a legacy prefix. This will allow implementing a new set of ioctls in parallel and gradually switching drivers. Add a task to the TODO file for the removal of the legacy API. Signed-off-by:
-
Laurent Pinchart authored
For consistency with all the other ioctl handlers, rename vidioc_querycap() to capture_querycap(). Signed-off-by:
-
