Re: [PATCH v5 2/3] media: atmel: introduce microchip csi2dc driver
From: Eugen.Hristev
Date: Wed Nov 18 2020 - 04:48:29 EST
On 18.11.2020 11:11, Jacopo Mondi wrote:
> Hello Eugen,
>
> On Tue, Nov 17, 2020 at 05:24:30PM +0000, Eugen.Hristev@xxxxxxxxxxxxx wrote:
>> On 17.11.2020 14:09, Laurent Pinchart wrote:
>>> Hello everybody,
>>>
>>> On Tue, Nov 17, 2020 at 12:59:02PM +0100, Jacopo Mondi wrote:
>>>> On Thu, Nov 12, 2020 at 03:34:36PM +0200, Eugen Hristev wrote:
>>>>> Microchip CSI2DC (CSI2 Demultiplexer Controller) is a misc bridge device
>>>>> that converts a byte stream in IDI Synopsys format (coming from a CSI2HOST)
>>>>> to a pixel stream that can be captured by a sensor controller.
>>>>>
>>>>> Signed-off-by: Eugen Hristev <eugen.hristev@xxxxxxxxxxxxx>
>>>>> ---
>>>>> Hello,
>>>>>
>>>>> There still are some open questions regarding the last reviews on the ML.
>>>>>
>>>>> Regarding the format list which is not const, I cannot have it const because
>>>>> I reference elements from this list into a dynamic list which is non-const.
>>>>>
>>>>> Regarding the presence of the v4l2_dev, without this I cannot add a notifier
>>>>> for this device to have it async completed , when the underlying subdevice
>>>>> finishes probing.
>>>>
>>>> I see you have a discussion with Sakari on-going on this, and I have
>>>> the very same comment he had.
>>>>
>>>> Should a sub-notifier be used and this driver should not register any
>>>> v4l2_dev but integrate in the media graph of the receiver ?
>>>>
>>>> For reference, rcar-csi2 uses sub-notifiers and register in the
>>>> rcar-vin media graph, it doesn't need to set up a v4l2_dev.
>>>>
>>>> rcar-vin (the DMA engine driver) that register video nodes and the v4l2-dev
>>>> also registers the device nodes for all the connected subdevices.
>>>
>>> Sub-notifiers were invited for this purpose, so I think they're the
>>> right tool.
>>>
>>>>> Regarding the callbacks for set frame interval, set frame size, etc, I cannot
>>>>> remove them because losing functionality to the underlying subdevice, as
>>>>> explained in :
>>>>> https://lkml.org/lkml/2020/10/12/427
>>>>
>>>> Using the frame sizes as in the example you reported, if this device has
>>>> limitations on the supported sizes, (a min and a max most
>>>> probably) that's the only thing that should be exposed from this
>>>> driver subdevice. The sensor subdevice instead will report the
>>>> available discrete sizes and userspace configures the sensor's source
>>>> pad and the bridge's sink pad with the same sizes to properly set up the
>>>> capture pipeline. video0 should not report the sizes exposed by the sensor
>>>> nor any information that depends on what's connected to it.
>>>
>>> Agreed. Subdev drivers should remain simple and cover only the subdev
>>> they manage. Bundling it together is the job of userspace. In the legacy
>>> model it uses to be the job of the top-level V4L2 driver (the one
>>> implementing v4l2_device), but that's legacy, especially for embedded
>>> camera use cases. In any case, the top-level logic must not be spread
>>> across subdevices.
>>>
>>>> I know what your next question is: what about existing application
>>>> that only know about video0 and want to setup everything from there. I
>>>> understand vendors not wanting to wrap gstreamer or what is used in
>>>> the wild in custom scripts that sets up the pipeline opportunely
>>>> before streaming, and I'm pretty sure you know what is the solution I
>>>> would propose for this :)
>>>>
>>>> I've just gone through the same path with RPi's Unicam, that does the
>>>> same thing as your one does: it's an MC driver, but wants to control
>>>> the sensor through the v4l2-subdev kAPI to support legacy use cases
>>>> not covered by libcamera. The fact is that we're stuck in this limbo
>>>> where MC pipeline tries to decouple components one from each other and
>>>> delegate their configuration to userspace, but at the same time
>>>> vendors want to be able to use the simple plain video node centric
>>>> interface because of their existing user base, so I understand your
>>>> discomfort, but the 'simple plain' approach is what's preventing Linux
>>>> platforms to be in-par with proprietary and custom solutions for any
>>>> use case that is not a simple frame capture, so I think it's fair for
>>>> mainline to push for this evolution to happen. Just my2c of course.
>>>
>>> Compatibility with the userspace of downstream kernels is important, but
>>> that's a downstream kernel issue. In mainline we want to go the MC way,
>>> and downstream can then, if desired, implement the configuration of the
>>> pipeline in the top-level V4L2 driver in the kernel to support the
>>> existing userbase. For new drivers I wouldn't recommend that, for
>>> existing downstream drivers it may be required.
>>
>> Hello,
>>
>> First of all thank you for the review and explanations.
>>
>> I am still confused about how does the userspace know which elements are
>> in the pipeline and how to connect them ?
>
> Using the media-controller uAPI to inspect, explore and configure the
> media graph (it's mostly about enabling and disabling media links),
> and using the v4l2 subdev uAPI to configure formats, sizes and such on
> each subdevice device node associated with a media entity. Finally,
> using the v4l2 uAPI to stream from the top-level driver that expose a
> video device node.
Okay but which application in userspace uses this userAPI ?
So I could use it to test my pipeline.
>
> It's really (conceptually) not different than doing the same using
> media-ctl and v4l2-ctl.
>
>> And if this is the case, what is the purpose of the endpoints in the DT ?
>>
>
> DTS describe the hardware. Usually a driver parses the firmware graph
> to create and model the media graph and the available connections
> between media entities but there's no requirement to have a 1-to-1
> match (as far I'm aware).
So the top level v4l2 driver should parse the whole graph ?
>
> Media links on a media graph then need to be enabled opportunely
> depending on the desired use case, the platform topology etc
>
>> At this moment, I just use the /dev/video0, because the top v4l2 driver
>> configures the subdevice, and this subdevice configures it's own
>> subdevice and so on, until the sensor itself.
>>
>> My top v4l2 driver will not 'complete' unless a subdevice registers
>> itself , and if this subdevice does not provide any information
>> regarding its formats, the probe of the top v4l2 driver will fail,
>> because it will not find compatibility between it's supported input
>> formats and what the subdevice provides.
>
> ouch, I see.. Which driver is that ?
The atmel image sensor controller driver.
All sensor drivers that we use (for example : ov5640, ov7740, ov7670)
register as subdevices, and the atmel-isc will 'complete' when the
sensor is probed and registered.
To have the csi2dc module compatible with what we have on atmel-isc,
it's natural to have the csi2dc register itself as a subdevice such that
the atmel-isc can 'complete'.
>
> If that's an MC driver, and that seems the case, it should not bother
> validating the subdevice formats, but only care about listing what its
> capabilities are.
if this atmel-isc does not offer format capabilities for the userspace
via the /dev/video0, then all applications that we use (gstreamer
v4l2src, ffmpeg ) fail.
What can we use then, if we cannot use these applications ?
>
>>
>> So what to do in this case ? Libcamera will configure this whole
>> pipeline for me , if the drivers just probe ( and start/stop streaming
>> by selecting the respective registers) ? or how to do this whole
>> pipeline configuration ?
>>
>
> Well, libcamera offers a framework to enable specialized code (called
> "pipeline handler") to operate under an high-level API towards application
> and implement image tuning algorithms on top of that common API.
>
> The code to configure the pipeline is device specific (unless your
> pipeline is very simple). So it's not that libcamera just magically
> knows how to operate on every platforms it runs on, it needs platform
> specific code to do so. What you get in exchange is an high level API
> to offer to application developers, a framework to implement 3A and
> tuning algorithms, integration with a growing set of other userspace
> frameworks (android, gstreamer, pipewire and a legacy v4l2-compat
> layer)
I tried a buildroot with libcamera and gstreamer libcamerasrc , to see
how it goes. Libcamera does not recognize the /dev/video0 device that we
have. The v4l2 compatibility layer should not perform exactly this ?
# gst-launch-1.0 libcamerasrc ! fakesink
Setting pipeline to PAUSED ...
[0:00:28.448687200] [216] WARN IPAManager ipa_manager.cpp:147 No IPA
found in '/usr/lib/libcamera'
[0:00:28.448890800] [216] INFO Camera camera_manager.cpp:287 libcamera
v0.0.0+54328-2020.11-rc2-dirty (2020-11-17T20:58:37+02:00)
ERROR: from element
/GstPipeline:pipeline0/GstLibcameraSrc:libcamerasrc0: Could not find any
supported camera on this system.
Additional debug info:
../src/gstreamer/gstlibcamerasrc.cpp(236): gst_libcamera_src_open ():
/GstPipeline:pipeline0/GstLibcameraSrc:libcamerasrc0:
libcamera::CameraMananger::cameras() is empty
ERROR: pipeline doesn't want to preroll.
Failed to set pipeline to PAUSED.
Setting pipeline to NULL ...
Freeing pipeline ...
Thanks,
Eugen
>
> We have a lengthy pipeline handler writers guide that might help
> better explain this:
>
> https://git.linuxtv.org/libcamera.git/tree/Documentation/guides/pipeline-handler.rst
> (you might want to compile libcamera to be able to render it in a browser)
>
> Hope this helps to clarify things a bit
>
>>
>>>
>>>> Some more comments on the implementation below
>>>>
>>>>> Changes in v5:
>>>>> - only in bindings
>>>>>
>>>>> Changes in v4:
>>>>> - now using get_mbus_config ops to get data from the subdevice, like the
>>>>> virtual channel id, and the clock type.
>>>>> - now having possibility to select any of the RAW10 data modes
>>>>> - at completion time, select which formats are also available in the subdevice,
>>>>> and move to the dynamic list accordingly
>>>>> - changed the pipeline integration, do not advertise subdev ready at probe time.
>>>>> wait until completion is done, and then start a workqueue that will register
>>>>> this device as a subdevice for the next element in pipeline.
>>>>> - moved the s_power code into a different function called now csi2dc_power
>>>>> that is called with CONFIG_PM functions. This is also called at completion,
>>>>> to have the device ready in case CONFIG_PM is not selected on the platform.
>>>>> - merged try_fmt into set_fmt
>>>>> - driver cleanup, wrapped lines over 80 characters
>>>>>
>>>>> Changes in v2:
>>>>> - moved driver to platform/atmel
>>>>> - fixed minor things as per Sakari's review
>>>>> - still some things from v2 review are not yet addressed, to be followed up
>>>>>
>>>>> drivers/media/platform/atmel/Kconfig | 13 +
>>>>> drivers/media/platform/atmel/Makefile | 1 +
>>>>> .../media/platform/atmel/microchip-csi2dc.c | 785 ++++++++++++++++++
>>>>> 3 files changed, 799 insertions(+)
>>>>> create mode 100644 drivers/media/platform/atmel/microchip-csi2dc.c
>>>>>
>>>>> diff --git a/drivers/media/platform/atmel/Kconfig b/drivers/media/platform/atmel/Kconfig
>>>>> index 99b51213f871..147e1c14129b 100644
>>>>> --- a/drivers/media/platform/atmel/Kconfig
>>>>> +++ b/drivers/media/platform/atmel/Kconfig
>>>>> @@ -32,3 +32,16 @@ config VIDEO_ATMEL_ISI
>>>>> help
>>>>> This module makes the ATMEL Image Sensor Interface available
>>>>> as a v4l2 device.
>>>>> +
>>>>> +config VIDEO_MICROCHIP_CSI2DC
>>>>> + tristate "Microchip CSI2 Demux Controller"
>>>>> + depends on VIDEO_V4L2 && COMMON_CLK && OF
>>>>> + depends on ARCH_AT91 || COMPILE_TEST
>>>>> + select MEDIA_CONTROLLER
>>>>> + select VIDEO_V4L2_SUBDEV_API
>>>>> + select V4L2_FWNODE
>>>>> + help
>>>>> + CSI2 Demux Controller driver. CSI2DC is a helper chip
>>>>> + that converts IDI interface byte stream to a parallel pixel stream.
>>>>> + It supports various RAW formats as input.
>>>>> + Performs clock domain crossing between hardware blocks.
>>>>
>>>> Is the last phrase necessary ? Also missing the usual
>>>>
>>>> To compile this driver as a module, choose M here: the
>>>> module will be called ...
>>>>
>>>>> diff --git a/drivers/media/platform/atmel/Makefile b/drivers/media/platform/atmel/Makefile
>>>>> index c5c01556c653..8e80af500bf5 100644
>>>>> --- a/drivers/media/platform/atmel/Makefile
>>>>> +++ b/drivers/media/platform/atmel/Makefile
>>>>> @@ -5,3 +5,4 @@ atmel-xisc-objs = atmel-sama7g5-isc.o atmel-isc-base.o
>>>>> obj-$(CONFIG_VIDEO_ATMEL_ISI) += atmel-isi.o
>>>>> obj-$(CONFIG_VIDEO_ATMEL_ISC) += atmel-isc.o
>>>>> obj-$(CONFIG_VIDEO_ATMEL_XISC) += atmel-xisc.o
>>>>> +obj-$(CONFIG_VIDEO_MICROCHIP_CSI2DC) += microchip-csi2dc.o
>>>>> diff --git a/drivers/media/platform/atmel/microchip-csi2dc.c b/drivers/media/platform/atmel/microchip-csi2dc.c
>>>>> new file mode 100644
>>>>> index 000000000000..fda1d5882dbb
>>>>> --- /dev/null
>>>>> +++ b/drivers/media/platform/atmel/microchip-csi2dc.c
>>>>> @@ -0,0 +1,785 @@
>>>>> +// SPDX-License-Identifier: GPL-2.0
>>>>
>>>> Nit: I see GPL-2.0 listed as deprecated by spdx.org
>>>> Should you use one of:
>>>> GPL-2.0-only
>>>> GPL-2.0-or-later
>>>>
>>>>> +/*
>>>>> + * Microchip CSI2 Demux Controller (CSI2DC) driver
>>>>> + *
>>>>> + * Copyright (C) 2018-2020 Microchip Technology, Inc.
>>>>> + *
>>>>> + * Author: Eugen Hristev <eugen.hristev@xxxxxxxxxxxxx>
>>>>> + *
>>>>> + */
>>>>> +
>>>>> +#include <linux/clk.h>
>>>>> +#include <linux/module.h>
>>>>> +#include <linux/of.h>
>>>>> +#include <linux/of_graph.h>
>>>>> +#include <linux/platform_device.h>
>>>>> +#include <linux/pm_runtime.h>
>>>>> +#include <linux/videodev2.h>
>>>>> +
>>>>> +#include <media/v4l2-device.h>
>>>>> +#include <media/v4l2-fwnode.h>
>>>>> +#include <media/v4l2-subdev.h>
>>>>> +#include <media/videobuf2-dma-contig.h>
>>>>> +
>>>>> +/* Global configuration register */
>>>>> +#define CSI2DC_GCFG 0x0
>>>>> +
>>>>> +/* MIPI sensor pixel clock is free running */
>>>>> +#define CSI2DC_GCFG_MIPIFRN BIT(0)
>>>>> +/* Output waveform inter-line minimum delay */
>>>>> +#define CSI2DC_GCFG_HLC(v) ((v) << 4)
>>>>> +#define CSI2DC_GCFG_HLC_MASK GENMASK(7, 4)
>>>>> +/* SAMA7G5 requires a HLC delay of 15 */
>>>>> +#define SAMA7G5_HLC (15)
>>>>> +
>>>>> +/* Global control register */
>>>>> +#define CSI2DC_GCTLR 0x04
>>>>> +#define CSI2DC_GCTLR_SWRST BIT(0)
>>>>> +
>>>>> +/* Global status register */
>>>>> +#define CSI2DC_GS 0x08
>>>>> +
>>>>> +/* SSP interrupt status register */
>>>>> +#define CSI2DC_SSPIS 0x28
>>>>> +/* Pipe update register */
>>>>> +#define CSI2DC_PU 0xC0
>>>>> +/* Video pipe attributes update */
>>>>> +#define CSI2DC_PU_VP BIT(0)
>>>>> +
>>>>> +/* Pipe update status register */
>>>>> +#define CSI2DC_PUS 0xC4
>>>>> +
>>>>> +/* Video pipeline enable register */
>>>>> +#define CSI2DC_VPE 0xF8
>>>>> +#define CSI2DC_VPE_ENABLE BIT(0)
>>>>> +
>>>>> +/* Video pipeline configuration register */
>>>>> +#define CSI2DC_VPCFG 0xFC
>>>>> +/* Data type */
>>>>> +#define CSI2DC_VPCFG_DT(v) ((v) << 0)
>>>>> +#define CSI2DC_VPCFG_DT_MASK GENMASK(5, 0)
>>>>> +/* Virtual channel identifier */
>>>>> +#define CSI2DC_VPCFG_VC(v) ((v) << 6)
>>>>> +#define CSI2DC_VPCFG_VC_MASK GENMASK(7, 6)
>>>>> +/* Decompression enable */
>>>>> +#define CSI2DC_VPCFG_DE BIT(8)
>>>>> +/* Decoder mode */
>>>>> +#define CSI2DC_VPCFG_DM(v) ((v) << 9)
>>>>> +#define CSI2DC_VPCFG_DM_DECODER8TO12 0
>>>>> +/* Decoder predictor 2 selection */
>>>>> +#define CSI2DC_VPCFG_DP2 BIT(12)
>>>>> +/* Recommended memory storage */
>>>>> +#define CSI2DC_VPCFG_RMS BIT(13)
>>>>> +/* Post adjustment */
>>>>> +#define CSI2DC_VPCFG_PA BIT(14)
>>>>> +
>>>>> +/* Video pipeline column register */
>>>>> +#define CSI2DC_VPCOL 0x100
>>>>> +/* Column number */
>>>>> +#define CSI2DC_VPCOL_COL(v) ((v) << 0)
>>>>> +#define CSI2DC_VPCOL_COL_MASK GENMASK(15, 0)
>>>>> +
>>>>> +/* Video pipeline row register */
>>>>> +#define CSI2DC_VPROW 0x104
>>>>> +/* Row number */
>>>>> +#define CSI2DC_VPROW_ROW(v) ((v) << 0)
>>>>> +#define CSI2DC_VPROW_ROW_MASK GENMASK(15, 0)
>>>>> +
>>>>> +/* Version register */
>>>>> +#define CSI2DC_VERSION 0x1FC
>>>>> +
>>>>> +/* register read/write helpers */
>>>>> +#define csi2dc_readl(st, reg) readl_relaxed((st)->base + (reg))
>>>>> +#define csi2dc_writel(st, reg, val) writel_relaxed((val), \
>>>>> + (st)->base + (reg))
>>>>> +
>>>>> +/* supported RAW data types */
>>>>> +#define CSI2DC_DT_RAW6 0x28
>>>>> +#define CSI2DC_DT_RAW7 0x29
>>>>> +#define CSI2DC_DT_RAW8 0x2A
>>>>> +#define CSI2DC_DT_RAW10 0x2B
>>>>> +#define CSI2DC_DT_RAW12 0x2C
>>>>> +#define CSI2DC_DT_RAW14 0x2D
>>>>> +
>>>>> +struct csi2dc_format {
>>>>> + u32 mbus_code;
>>>>> + u32 dt;
>>>>> +};
>>>>> +
>>>>> +static struct csi2dc_format csi2dc_formats_list[] = {
>>>>> + {
>>>>> + .mbus_code = MEDIA_BUS_FMT_SRGGB10_1X10,
>>>>> + .dt = CSI2DC_DT_RAW10,
>>>>> + }, {
>>>>> + .mbus_code = MEDIA_BUS_FMT_SBGGR10_1X10,
>>>>> + .dt = CSI2DC_DT_RAW10,
>>>>> + }, {
>>>>> + .mbus_code = MEDIA_BUS_FMT_SGRBG10_1X10,
>>>>> + .dt = CSI2DC_DT_RAW10,
>>>>> + }, {
>>>>> + .mbus_code = MEDIA_BUS_FMT_SGBRG10_1X10,
>>>>> + .dt = CSI2DC_DT_RAW10,
>>>>> + },
>>>>> +};
>>>>> +
>>>>> +enum mipi_csi_pads {
>>>>> + CSI2DC_PAD_SINK = 0,
>>>>> + CSI2DC_PAD_SOURCE = 1,
>>>>> + CSI2DC_PADS_NUM = 2,
>>>>> +};
>>>>> +
>>>>> +struct csi2dc_device {
>>>>> + void __iomem *base;
>>>>> + struct v4l2_subdev csi2dc_sd;
>>>>> + struct device *dev;
>>>>> + struct v4l2_device v4l2_dev;
>>>>> + struct clk *pclk;
>>>>> + struct clk *scck;
>>>>> +
>>>>> + bool video_pipe;
>>>>> +
>>>>> + u32 num_fmts;
>>>>> + struct csi2dc_format **formats;
>>>>> +
>>>>> + struct csi2dc_format *cur_fmt;
>>>>> + struct csi2dc_format *try_fmt;
>>>>> +
>>>>> + struct media_pad pads[CSI2DC_PADS_NUM];
>>>>> +
>>>>> + bool clk_gated;
>>>>> + u32 vc;
>>>>> +
>>>>> + struct v4l2_async_subdev *asd;
>>>>> + struct v4l2_async_notifier notifier;
>>>>> +
>>>>> + struct v4l2_subdev *input_sd;
>>>>> +
>>>>> + u32 remote_pad;
>>>>> +
>>>>> + bool completed;
>>>>> + bool powered_on;
>>>>> +
>>>>> + struct work_struct workq;
>>>>> +};
>>>>> +
>>>>> +static void csi2dc_vp_update(struct csi2dc_device *csi2dc)
>>>>> +{
>>>>> + u32 vp;
>>>>> +
>>>>> + vp = CSI2DC_VPCFG_DT(csi2dc->cur_fmt->dt) & CSI2DC_VPCFG_DT_MASK;
>>>>> + vp |= CSI2DC_VPCFG_VC(csi2dc->vc) & CSI2DC_VPCFG_VC_MASK;
>>>>> + vp &= ~CSI2DC_VPCFG_DE;
>>>>> + vp |= CSI2DC_VPCFG_DM(CSI2DC_VPCFG_DM_DECODER8TO12);
>>>>> + vp &= ~CSI2DC_VPCFG_DP2;
>>>>> + vp &= ~CSI2DC_VPCFG_RMS;
>>>>> + vp |= CSI2DC_VPCFG_PA;
>>>>> +
>>>>> + csi2dc_writel(csi2dc, CSI2DC_VPCFG, vp);
>>>>> + csi2dc_writel(csi2dc, CSI2DC_VPE, CSI2DC_VPE_ENABLE);
>>>>> + csi2dc_writel(csi2dc, CSI2DC_PU, CSI2DC_PU_VP);
>>>>> +}
>>>>> +
>>>>> +static inline struct csi2dc_device *
>>>>> +csi2dc_sd_to_csi2dc_device(struct v4l2_subdev *csi2dc_sd)
>>>>> +{
>>>>> + return container_of(csi2dc_sd, struct csi2dc_device, csi2dc_sd);
>>>>> +}
>>>>> +
>>>>> +static int csi2dc_enum_mbus_code(struct v4l2_subdev *csi2dc_sd,
>>>>> + struct v4l2_subdev_pad_config *cfg,
>>>>> + struct v4l2_subdev_mbus_code_enum *code)
>>>>> +{
>>>>> + struct csi2dc_device *csi2dc = csi2dc_sd_to_csi2dc_device(csi2dc_sd);
>>>>> +
>>>>> + if (code->index >= csi2dc->num_fmts)
>>>>> + return -EINVAL;
>>>>> +
>>>>> + code->code = csi2dc->formats[code->index]->mbus_code;
>>>>> + return 0;
>>>>> +}
>>>>> +
>>>>> +static int csi2dc_set_fmt(struct v4l2_subdev *csi2dc_sd,
>>>>> + struct v4l2_subdev_pad_config *cfg,
>>>>> + struct v4l2_subdev_format *req_fmt)
>>>>> +{
>>>>> + struct csi2dc_device *csi2dc = csi2dc_sd_to_csi2dc_device(csi2dc_sd);
>>>>> + struct csi2dc_format *fmt;
>>>>> + int ret, i;
>>>>> +
>>>>> + if (!csi2dc->completed) {
>>>>> + dev_dbg((csi2dc)->dev, "subdev not registered yet\n");
>>>>> + return 0;
>>>>> + }
>>>>> +
>>>>> + for (fmt = csi2dc->formats[0], i = 0; i < csi2dc->num_fmts; fmt++, i++)
>>>>> + if (req_fmt->format.code == fmt->mbus_code)
>>>>> + csi2dc->try_fmt = fmt;
>>>>> +
>>>>> + /* in case we could not find the desired format, default to something */
>>>>> + if (!csi2dc->try_fmt ||
>>>>> + req_fmt->format.code != csi2dc->try_fmt->mbus_code) {
>>>>> + csi2dc->try_fmt = csi2dc->formats[0];
>>>>> +
>>>>> + dev_dbg(csi2dc->dev,
>>>>> + "CSI2DC unsupported format 0x%x, defaulting to 0x%x\n",
>>>>> + req_fmt->format.code, csi2dc->formats[0]->mbus_code);
>>>>> +
>>>>> + req_fmt->format.code = csi2dc->formats[0]->mbus_code;
>>>>> + }
>>>>> +
>>>>> + /* if we are just trying, we are done */
>>>>> + if (req_fmt->which == V4L2_SUBDEV_FORMAT_TRY)
>>>>> + return 0;
>>>>> +
>>>>> + ret = v4l2_subdev_call(csi2dc->input_sd, pad, set_fmt, cfg, req_fmt);
>>>>> + if (ret) {
>>>>> + dev_err(csi2dc->dev, "input subdev failed %d\n", ret);
>>>>> + return ret;
>>>>> + }
>>>>> +
>>>>> + csi2dc->cur_fmt = csi2dc->try_fmt;
>>>>> + /* update video pipe */
>>>>> + csi2dc_vp_update(csi2dc);
>>>>> +
>>>>> + dev_dbg(csi2dc->dev, "CSI2DC new format: 0x%x\n", req_fmt->format.code);
>>>>> + return 0;
>>>>> +}
>>>>> +
>>>>> +static int csi2dc_formats_init(struct csi2dc_device *csi2dc)
>>>>> +{
>>>>> + struct csi2dc_format *fmt;
>>>>> + struct v4l2_subdev *subdev = csi2dc->input_sd;
>>>>> + unsigned int num_fmts, i;
>>>>> +
>>>>> + struct v4l2_subdev_mbus_code_enum mbus_code = {
>>>>> + .which = V4L2_SUBDEV_FORMAT_ACTIVE,
>>>>> + .index = 0,
>>>>> + };
>>>>> +
>>>>> + num_fmts = 0;
>>>>> +
>>>>> + csi2dc->formats = devm_kcalloc(csi2dc->dev,
>>>>> + ARRAY_SIZE(csi2dc_formats_list),
>>>>> + sizeof(*csi2dc->formats), GFP_KERNEL);
>>>>> + if (!csi2dc->formats)
>>>>> + return -ENOMEM;
>>>>> +
>>>>> + while (!v4l2_subdev_call(subdev, pad, enum_mbus_code, NULL,
>>>>> + &mbus_code)) {
>>>>> + mbus_code.index++;
>>>>> + for (fmt = &csi2dc_formats_list[0], i = 0;
>>>>> + i < ARRAY_SIZE(csi2dc_formats_list); i++, fmt++)
>>>>> + if (fmt->mbus_code == mbus_code.code) {
>>>>> + csi2dc->formats[num_fmts] = fmt;
>>>>> + num_fmts++;
>>>>> + }
>>>>> + }
>>>>
>>>> For the reasons explained above and by Sakari, I think this should be dropped
>>>>
>>>>> +
>>>>> + if (!num_fmts)
>>>>> + return -ENXIO;
>>>>> +
>>>>> + csi2dc->num_fmts = num_fmts;
>>>>> +
>>>>> + return 0;
>>>>> +}
>>>>> +
>>>>> +static int csi2dc_power(struct csi2dc_device *csi2dc, int on)
>>>>> +{
>>>>> + int ret = 0;
>>>>> +
>>>>> + if (!csi2dc->completed) {
>>>>> + dev_dbg((csi2dc)->dev, "subdev not registered yet\n");
>>>>> + return 0;
>>>>> + }
>>>>
>>>> Can this happen ?
>>>>
>>>>> +
>>>>> + if (csi2dc->powered_on == on)
>>>>> + return 0;
>>>>
>>>> Same question
>>>>
>>>>> +
>>>>> + if (on)
>>>>> + ret = clk_prepare_enable(csi2dc->scck);
>>>>> + else
>>>>> + clk_disable_unprepare(csi2dc->scck);
>>>>> + if (ret)
>>>>> + dev_err(csi2dc->dev, "failed to enable scck: %d\n", ret);
>>>>> +
>>>>> + /* if powering up, deassert reset line */
>>>>> + if (on)
>>>>> + csi2dc_writel(csi2dc, CSI2DC_GCTLR, CSI2DC_GCTLR_SWRST);
>>>>> +
>>>>> + /* if powering down, assert reset line */
>>>>> + if (!on)
>>>>> + csi2dc_writel(csi2dc, CSI2DC_GCTLR, !CSI2DC_GCTLR_SWRST);
>>>>> + if (!ret)
>>>>> + csi2dc->powered_on = on;
>>>>
>>>> Wouldn't this be better coded as
>>>>
>>>> if (on) {
>>>> clk_prepare_enable
>>>> csi2dc_writel(SWRST)
>>>> powered_on = true
>>>> } else {
>>>> clk_disable_unrepare
>>>> csi2dc_writel(!SWRST)
>>>> powered_on = false
>>>> }
>>>>> +
>>>>> + return ret;
>>>>> +}
>>>>> +
>>>>> +static int csi2dc_s_stream(struct v4l2_subdev *csi2dc_sd, int enable)
>>>>> +{
>>>>> + struct csi2dc_device *csi2dc = csi2dc_sd_to_csi2dc_device(csi2dc_sd);
>>>>> +
>>>>> + if (!csi2dc->completed) {
>>>>> + dev_dbg((csi2dc)->dev, "subdev not registered yet\n");
>>>>> + return 0;
>>>>> + }
>>>>> +
>>>>> + return v4l2_subdev_call(csi2dc->input_sd, video, s_stream, enable);
>>>>> +}
>>>>> +
>>>>> +static int csi2dc_g_frame_interval(struct v4l2_subdev *csi2dc_sd,
>>>>> + struct v4l2_subdev_frame_interval *interval)
>>>>> +{
>>>>> + struct csi2dc_device *csi2dc = csi2dc_sd_to_csi2dc_device(csi2dc_sd);
>>>>> +
>>>>> + if (!csi2dc->completed) {
>>>>> + dev_dbg((csi2dc)->dev, "subdev not registered yet\n");
>>>>> + return 0;
>>>>> + }
>>>>> +
>>>>> + return v4l2_subdev_call(csi2dc->input_sd, video, g_frame_interval,
>>>>> + interval);
>>>>> +}
>>>>> +
>>>>> +static int csi2dc_s_frame_interval(struct v4l2_subdev *csi2dc_sd,
>>>>> + struct v4l2_subdev_frame_interval *interval)
>>>>> +{
>>>>> + struct csi2dc_device *csi2dc = csi2dc_sd_to_csi2dc_device(csi2dc_sd);
>>>>> +
>>>>> + if (!csi2dc->completed) {
>>>>> + dev_dbg((csi2dc)->dev, "subdev not registered yet\n");
>>>>> + return 0;
>>>>> + }
>>>>> +
>>>>> + return v4l2_subdev_call(csi2dc->input_sd, video, s_frame_interval,
>>>>> + interval);
>>>>> +}
>>>>> +
>>>>> +static int csi2dc_enum_frame_size(struct v4l2_subdev *csi2dc_sd,
>>>>> + struct v4l2_subdev_pad_config *cfg,
>>>>> + struct v4l2_subdev_frame_size_enum *fse)
>>>>> +{
>>>>> + struct csi2dc_device *csi2dc = csi2dc_sd_to_csi2dc_device(csi2dc_sd);
>>>>> +
>>>>> + if (!csi2dc->completed) {
>>>>> + dev_dbg((csi2dc)->dev, "subdev not registered yet\n");
>>>>> + return 0;
>>>>> + }
>>>>> +
>>>>> + return v4l2_subdev_call(csi2dc->input_sd, pad, enum_frame_size, cfg,
>>>>> + fse);
>>>>> +}
>>>>> +
>>>>> +static int csi2dc_enum_frame_interval(struct v4l2_subdev *csi2dc_sd,
>>>>> + struct v4l2_subdev_pad_config *cfg,
>>>>> + struct v4l2_subdev_frame_interval_enum *fie)
>>>>> +{
>>>>> + struct csi2dc_device *csi2dc = csi2dc_sd_to_csi2dc_device(csi2dc_sd);
>>>>> +
>>>>> + if (!csi2dc->completed) {
>>>>> + dev_dbg((csi2dc)->dev, "subdev not registered yet\n");
>>>>> + return 0;
>>>>> + }
>>>>> +
>>>>> + return v4l2_subdev_call(csi2dc->input_sd, pad, enum_frame_interval, cfg,
>>>>> + fie);
>>>>> +}
>>>>> +
>>>>> +static const struct v4l2_subdev_pad_ops csi2dc_pad_ops = {
>>>>> + .enum_mbus_code = csi2dc_enum_mbus_code,
>>>>> + .set_fmt = csi2dc_set_fmt,
>>>>> + .enum_frame_size = csi2dc_enum_frame_size,
>>>>> + .enum_frame_interval = csi2dc_enum_frame_interval,
>>>>> +};
>>>>> +
>>>>> +static const struct v4l2_subdev_video_ops csi2dc_video_ops = {
>>>>> + .s_stream = csi2dc_s_stream,
>>>>> + .g_frame_interval = csi2dc_g_frame_interval,
>>>>> + .s_frame_interval = csi2dc_s_frame_interval,
>>>>> +};
>>>>> +
>>>>> +static const struct v4l2_subdev_ops csi2dc_subdev_ops = {
>>>>> + .pad = &csi2dc_pad_ops,
>>>>> + .video = &csi2dc_video_ops,
>>>>> +};
>>>>> +
>>>>> +static int csi2dc_get_mbus_config(struct csi2dc_device *csi2dc)
>>>>> +{
>>>>> + struct v4l2_mbus_config mbus_config = { 0 };
>>>>> + int ret;
>>>>> +
>>>>> + ret = v4l2_subdev_call(csi2dc->input_sd, pad, get_mbus_config,
>>>>> + csi2dc->remote_pad, &mbus_config);
>>>>> + if (ret == -ENOIOCTLCMD) {
>>>>> + dev_dbg(csi2dc->dev,
>>>>> + "no remote mbus configuration available\n");
>>>>> + goto csi2dc_get_mbus_config_defaults;
>>>>> + }
>>>>> +
>>>>> + if (ret) {
>>>>> + dev_err(csi2dc->dev,
>>>>> + "failed to get remote mbus configuration\n");
>>>>> + goto csi2dc_get_mbus_config_defaults;
>>>>
>>>> This should probably be an error reported to the caller. But this
>>>> function return value is not checked, so it's probably fine.
>>>>
>>>>> + }
>>>>> +
>>>>> + if (mbus_config.flags & V4L2_MBUS_CSI2_CHANNEL_0)
>>>>> + csi2dc->vc = 0;
>>>>> + else if (mbus_config.flags & V4L2_MBUS_CSI2_CHANNEL_1)
>>>>> + csi2dc->vc = 1;
>>>>> + else if (mbus_config.flags & V4L2_MBUS_CSI2_CHANNEL_2)
>>>>> + csi2dc->vc = 2;
>>>>> + else if (mbus_config.flags & V4L2_MBUS_CSI2_CHANNEL_3)
>>>>> + csi2dc->vc = 3;
>>>>> +
>>>>> + dev_dbg(csi2dc->dev, "subdev sending on channel %d\n", csi2dc->vc);
>>>>> +
>>>>> + csi2dc->clk_gated = mbus_config.flags &
>>>>> + V4L2_MBUS_CSI2_NONCONTINUOUS_CLOCK;
>>>>> +
>>>>> + dev_dbg(csi2dc->dev, "%s clock\n",
>>>>> + csi2dc->clk_gated ? "gated" : "free running");
>>>>> +
>>>>> + return 0;
>>>>> +
>>>>> +csi2dc_get_mbus_config_defaults:
>>>>> + csi2dc->vc = 0; /* Virtual ID 0 by default */
>>>>> + csi2dc->clk_gated = false; /* Free running clock by default */
>>>>> +
>>>>> + return 0;
>>>>> +}
>>>>> +
>>>>> +static int csi2dc_async_bound(struct v4l2_async_notifier *notifier,
>>>>> + struct v4l2_subdev *subdev,
>>>>> + struct v4l2_async_subdev *asd)
>>>>> +{
>>>>> + struct csi2dc_device *csi2dc = container_of(notifier->v4l2_dev,
>>>>> + struct csi2dc_device, v4l2_dev);
>>>>> + int pad;
>>>>> +
>>>>> + csi2dc->input_sd = subdev;
>>>>> +
>>>>> + pad = media_entity_get_fwnode_pad(&subdev->entity,
>>>>> + asd->match.fwnode,
>>>>> + MEDIA_PAD_FL_SOURCE);
>>>>> + if (pad < 0) {
>>>>> + dev_err(csi2dc->dev, "Failed to find pad for %s\n",
>>>>> + subdev->name);
>>>>> + return pad;
>>>>> + }
>>>>> +
>>>>> + csi2dc->remote_pad = pad;
>>>>> +
>>>>> + csi2dc_get_mbus_config(csi2dc);
>>>>
>>>> The closer to s_stream you call get_mbus_config, the less the chances
>>>> of receiving a stale configuration, if this can happen.
>>>>
>>>>> +
>>>>> + return 0;
>>>>> +}
>>>>> +
>>>>> +static int csi2dc_async_complete(struct v4l2_async_notifier *notifier)
>>>>> +{
>>>>> + struct csi2dc_device *csi2dc =
>>>>> + container_of(notifier->v4l2_dev, struct csi2dc_device,
>>>>> + v4l2_dev);
>>>>> + int ret;
>>>>> +
>>>>> + ret = csi2dc_formats_init(csi2dc);
>>>>> + if (ret)
>>>>> + return ret;
>>>>> +
>>>>> + ret = v4l2_device_register_subdev_nodes(&csi2dc->v4l2_dev);
>>>>> + if (ret < 0) {
>>>>> + v4l2_err(&csi2dc->v4l2_dev,
>>>>> + "failed to register subdev nodes\n");
>>>>> + return ret;
>>>>> + }
>>>>> +
>>>>> + csi2dc->completed = true;
>>>>> +
>>>>> + csi2dc_writel(csi2dc, CSI2DC_GCFG,
>>>>> + (SAMA7G5_HLC & CSI2DC_GCFG_HLC_MASK) |
>>>>> + (csi2dc->clk_gated ? 0 : CSI2DC_GCFG_MIPIFRN));
>>>>> +
>>>>> + csi2dc_writel(csi2dc, CSI2DC_VPCOL,
>>>>> + CSI2DC_VPCOL_COL(0xFFF) & CSI2DC_VPCOL_COL_MASK);
>>>>> + csi2dc_writel(csi2dc, CSI2DC_VPROW,
>>>>> + CSI2DC_VPROW_ROW(0xFFF) & CSI2DC_VPROW_ROW_MASK);
>>>>> +
>>>>> + /* once we are completed, we can register ourselves in the pipeline */
>>>>> + schedule_work(&csi2dc->workq);
>>>>> +
>>>>> + return ret;
>>>>> +}
>>>>> +
>>>>> +static const struct v4l2_async_notifier_operations csi2dc_async_ops = {
>>>>> + .bound = csi2dc_async_bound,
>>>>> + .complete = csi2dc_async_complete,
>>>>> +};
>>>>> +
>>>>> +static void csi2dc_cleanup_notifier(struct csi2dc_device *csi2dc)
>>>>> +{
>>>>> + v4l2_async_notifier_unregister(&csi2dc->notifier);
>>>>> + v4l2_async_notifier_cleanup(&csi2dc->notifier);
>>>>> +}
>>>>> +
>>>>> +static int csi2dc_prepare_notifier(struct csi2dc_device *csi2dc,
>>>>> + struct device_node *input_parent)
>>>>> +{
>>>>> + int ret;
>>>>> +
>>>>> + v4l2_async_notifier_init(&csi2dc->notifier);
>>>>> +
>>>>> + csi2dc->asd = kzalloc(sizeof(*csi2dc->asd), GFP_KERNEL);
>>>>> + if (!csi2dc->asd)
>>>>> + return -ENOMEM;
>>>>> +
>>>>> + csi2dc->asd->match_type = V4L2_ASYNC_MATCH_FWNODE;
>>>>> + csi2dc->asd->match.fwnode = of_fwnode_handle(input_parent);
>>>>
>>>> If you use v4l2_async_notifier_add_fwnode_subdev() the framework does
>>>> this for you.
>>>>
>>>>> +
>>>>> + ret = v4l2_async_notifier_add_subdev(&csi2dc->notifier, csi2dc->asd);
>>>>> + if (ret) {
>>>>> + dev_err(csi2dc->dev, "failed to add async notifier.\n");
>>>>> + v4l2_async_notifier_cleanup(&csi2dc->notifier);
>>>>> + goto csi2dc_prepare_notifier_err;
>>>>> + }
>>>>> +
>>>>> + csi2dc->notifier.ops = &csi2dc_async_ops;
>>>>> +
>>>>> + ret = v4l2_async_notifier_register(&csi2dc->v4l2_dev,
>>>>> + &csi2dc->notifier);
>>>>> +
>>>>> + if (ret) {
>>>>> + dev_err(csi2dc->dev, "fail to register async notifier.\n");
>>>>
>>>> Shouldn't you cleanup the notifier if registration fails ?
>>>>
>>>>> + goto csi2dc_prepare_notifier_err;
>>>>> + }
>>>>> +
>>>>> +csi2dc_prepare_notifier_err:
>>>>> + of_node_put(input_parent);
>>>>> +
>>>>> + return ret;
>>>>> +}
>>>>> +
>>>>> +static int csi2dc_of_parse(struct csi2dc_device *csi2dc,
>>>>> + struct device_node *of_node)
>>>>> +{
>>>>> + struct device_node *input_node, *sink_node, *input_parent;
>>>>> + struct v4l2_fwnode_endpoint input_endpoint = { 0 },
>>>>> + sink_endpoint = { 0 };
>>>>> + int ret;
>>>>> +
>>>>> + input_node = of_graph_get_next_endpoint(of_node, NULL);
>>>>> +
>>>>> + if (!input_node) {
>>>>> + dev_err(csi2dc->dev,
>>>>> + "missing port node at %s, input node is mandatory.\n",
>>>>
>>>> Please use %pOF to print the node name
>>>> Also, input_node should be put in the error path
>>>>
>>>>> + of_node->full_name);
>>>>> + return -EINVAL;
>>>>> + }
>>>>> +
>>>>> + ret = v4l2_fwnode_endpoint_parse(of_fwnode_handle(input_node),
>>>>> + &input_endpoint);
>>>>
>>>> Please set input_endpoint.bus_type if you know what kind of bus you're
>>>> dealing with (not having the bindings of the IDI transmitter I'm not
>>>> sure. Is this V4L2_MBUS_CSI2_DPHY ?)
>>>>
>>>>> +
>>>>> + if (ret) {
>>>>> + dev_err(csi2dc->dev, "endpoint not defined at %s\n",
>>>>> + of_node->full_name);
>>>>> + return ret;
>>>>> + }
>>>>> +
>>>>> + input_parent = of_graph_get_remote_port_parent(input_node);
>>>>> + if (!input_parent) {
>>>>
>>>> Missing of_node_put(input_node)
>>>>
>>>>> + dev_err(csi2dc->dev,
>>>>> + "could not get input node's parent node.\n");
>>>>> + return -EINVAL;
>>>>> + }
>>>>> +
>>>>> + sink_node = of_graph_get_next_endpoint(of_node, input_node);
>>>>> +
>>>>> + if (sink_node)
>>>>> + ret = v4l2_fwnode_endpoint_parse(of_fwnode_handle(sink_node),
>>>>> + &sink_endpoint);
>>>>
>>>> Same, set sink_endpoint.bus_type if you know it's MBUS_PARALLEL
>>>>
>>>>> +
>>>>> + if (!sink_node || ret) {
>>>>
>>>> If parsing fail you don't return an error but continue registering the
>>>> notifier and potentially end up registering the async subdev for this
>>>> entity even if the endpoint parsing failed.
>>>>
>>>>> + dev_info(csi2dc->dev,
>>>>> + "missing sink node at %s, data pipe available only.\n",
>>>>> + of_node->full_name);
>>>>> + } else {
>>>>> + csi2dc->video_pipe = true;
>>>>
>>>> video_pipe is unused it seems
>>>>
>>>>> +
>>>>> + dev_dbg(csi2dc->dev, "block %s %d.%d->%d.%d video pipeline\n",
>>>>> + of_node->full_name, input_endpoint.base.port,
>>>>> + input_endpoint.base.id, sink_endpoint.base.port,
>>>>> + sink_endpoint.base.id);
>>>>> + }
>>>>> +
>>>>> + of_node_put(sink_node);
>>>>> + of_node_put(input_node);
>>>>> +
>>>>> + /* prepare async notifier for subdevice completion */
>>>>> + return csi2dc_prepare_notifier(csi2dc, input_parent);
>>>>> +}
>>>>> +
>>>>> +static void csi2dc_workq_handler(struct work_struct *workq)
>>>>> +{
>>>>> + struct csi2dc_device *csi2dc = container_of(workq,
>>>>> + struct csi2dc_device, workq);
>>>>> + int ret;
>>>>> +
>>>>> + if (v4l2_async_register_subdev(&csi2dc->csi2dc_sd))
>>>>> + dev_dbg(csi2dc->dev, "failed to register the subdevice\n");
>>>>
>>>> That's peculiar, why do you have to schedule this to a workqueue
>>>> instead of doing this in the complete() callback ?
>>>>
>>>>> +
>>>>> + ret = csi2dc_power(csi2dc, true);
>>>>> + if (ret < 0)
>>>>> + v4l2_err(&csi2dc->v4l2_dev, "failed to power on\n");
>>>>
>>>> Should the device be powered on at s_stream time ? Possibly with a
>>>> pm_runtime_get_sync() call ? I see you register pm_runtime operations,
>>>> but never call get_sync() or _put().
>>>>
>>>> What if CONFIG_PM not selected ? Should this driver select it ? Do you
>>>> have use cases where that's not possible ?
>>>>
>>>>> +}
>>>>> +
>>>>> +static int csi2dc_probe(struct platform_device *pdev)
>>>>> +{
>>>>> + struct device *dev = &pdev->dev;
>>>>> + struct csi2dc_device *csi2dc;
>>>>> + struct resource *res = NULL;
>>>>> + int ret = 0;
>>>>> +
>>>>> + csi2dc = devm_kzalloc(dev, sizeof(*csi2dc), GFP_KERNEL);
>>>>> + if (!csi2dc)
>>>>> + return -ENOMEM;
>>>>> +
>>>>> + csi2dc->dev = dev;
>>>>> +
>>>>> + res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
>>>>> + if (!res)
>>>>> + return -EINVAL;
>>>>
>>>> devm_ioremap_resource() checks for the validity of 'res' so you can
>>>> drop the previous two lines
>>>>
>>>>> +
>>>>> + csi2dc->base = devm_ioremap_resource(dev, res);
>>>>> +
>>>>
>>>> Is the empy line intentional, sometimes you have one, sometimes you
>>>> don't ?
>>>>
>>>>> + if (IS_ERR(csi2dc->base)) {
>>>>> + dev_err(dev, "base address not set\n");
>>>>
>>>>
>>>>> + return PTR_ERR(csi2dc->base);
>>>>> + }
>>>>> +
>>>>> + csi2dc->pclk = devm_clk_get(dev, "pclk");
>>>>> + if (IS_ERR(csi2dc->pclk)) {
>>>>> + ret = PTR_ERR(csi2dc->pclk);
>>>>> + dev_err(dev, "failed to get pclk: %d\n", ret);
>>>>> + return ret;
>>>>
>>>> Just return PTR_ERR instead of assigning to ret
>>>>
>>>>> + }
>>>>> +
>>>>> + ret = clk_prepare_enable(csi2dc->pclk);
>>>>> + if (ret) {
>>>>> + dev_err(dev, "failed to enable pclk: %d\n", ret);
>>>>> + return ret;
>>>>> + }
>>>>
>>>> Has this clock to be enabled here or can it be done in the
>>>> pm_runtime callback ?
>>>>
>>>>> +
>>>>> + csi2dc->scck = devm_clk_get(dev, "scck");
>>>>> + if (IS_ERR(csi2dc->scck)) {
>>>>> + ret = PTR_ERR(csi2dc->scck);
>>>>> + dev_err(dev, "failed to get scck: %d\n", ret);
>>>>> + goto csi2dc_clk_fail;
>>>>> + }
>>>>> +
>>>>> + ret = v4l2_device_register(dev, &csi2dc->v4l2_dev);
>>>>> + if (ret) {
>>>>> + dev_err(dev, "unable to register v4l2 device.\n");
>>>>> + goto csi2dc_clk_fail;
>>>>> + }
>>>>> +
>>>>> + v4l2_subdev_init(&csi2dc->csi2dc_sd, &csi2dc_subdev_ops);
>>>>> +
>>>>> + csi2dc->csi2dc_sd.owner = THIS_MODULE;
>>>>> + csi2dc->csi2dc_sd.dev = dev;
>>>>> + snprintf(csi2dc->csi2dc_sd.name, sizeof(csi2dc->csi2dc_sd.name),
>>>>> + "CSI2DC.0");
>>>>> +
>>>>> + csi2dc->csi2dc_sd.flags |= V4L2_SUBDEV_FL_HAS_DEVNODE;
>>>>> + csi2dc->csi2dc_sd.entity.function = MEDIA_ENT_F_VID_IF_BRIDGE;
>>>>> + csi2dc->pads[CSI2DC_PAD_SINK].flags = MEDIA_PAD_FL_SINK;
>>>>> + csi2dc->pads[CSI2DC_PAD_SOURCE].flags = MEDIA_PAD_FL_SOURCE;
>>>>
>>>> Isn't the source pad presence conditional to the presence of
>>>> port@1 ?
>>>>
>>>>> +
>>>>> + ret = media_entity_pads_init(&csi2dc->csi2dc_sd.entity, CSI2DC_PADS_NUM,
>>>>> + csi2dc->pads);
>>>>> + if (ret < 0) {
>>>>> + dev_err(dev, "media entity init failed\n");
>>>>> + goto csi2dc_probe_entity_err;
>>>>> + }
>>>>> +
>>>>> + v4l2_set_subdevdata(&csi2dc->csi2dc_sd, pdev);
>>>>> +
>>>>> + platform_set_drvdata(pdev, &csi2dc->csi2dc_sd);
>>>>> +
>>>>> + INIT_WORK(&csi2dc->workq, csi2dc_workq_handler);
>>>>> +
>>>>> + ret = csi2dc_of_parse(csi2dc, dev->of_node);
>>>>> + if (ret)
>>>>> + goto csi2dc_probe_entity_err;
>>>>> +
>>>>> + dev_info(dev, "Microchip CSI2DC version %x\n",
>>>>> + csi2dc_readl(csi2dc, CSI2DC_VERSION));
>>>>> +
>>>>> + pm_runtime_set_active(dev);
>>>>> + pm_runtime_enable(dev);
>>>>> + pm_request_idle(dev);
>>>>> +
>>>>> + return 0;
>>>>> +
>>>>> +csi2dc_probe_entity_err:
>>>>> + media_entity_cleanup(&csi2dc->csi2dc_sd.entity);
>>>>> + v4l2_device_unregister(&csi2dc->v4l2_dev);
>>>>> +csi2dc_clk_fail:
>>>>> + clk_disable_unprepare(csi2dc->pclk);
>>>>> + return ret;
>>>>> +}
>>>>> +
>>>>> +static int csi2dc_remove(struct platform_device *pdev)
>>>>> +{
>>>>> + struct v4l2_subdev *csi2dc_sd = platform_get_drvdata(pdev);
>>>>> + struct csi2dc_device *csi2dc = csi2dc_sd_to_csi2dc_device(csi2dc_sd);
>>>>> +
>>>>> + pm_runtime_disable(&pdev->dev);
>>>>> +
>>>>> + v4l2_async_unregister_subdev(&csi2dc->csi2dc_sd);
>>>>> + csi2dc_cleanup_notifier(csi2dc);
>>>>> + media_entity_cleanup(&csi2dc->csi2dc_sd.entity);
>>>>> + v4l2_device_unregister(&csi2dc->v4l2_dev);
>>>>> + clk_disable_unprepare(csi2dc->pclk);
>>>>> +
>>>>> + return 0;
>>>>> +}
>>>>> +
>>>>> +static int __maybe_unused csi2dc_runtime_suspend(struct device *dev)
>>>>> +{
>>>>> + struct csi2dc_device *csi2dc = dev_get_drvdata(dev);
>>>>> +
>>>>> + return csi2dc_power(csi2dc, false);
>>>>> +}
>>>>> +
>>>>> +static int __maybe_unused csi2dc_runtime_resume(struct device *dev)
>>>>> +{
>>>>> + struct csi2dc_device *csi2dc = dev_get_drvdata(dev);
>>>>> +
>>>>> + return csi2dc_power(csi2dc, true);
>>>>> +}
>>>>> +
>>>>> +static const struct dev_pm_ops csi2dc_dev_pm_ops = {
>>>>> + SET_RUNTIME_PM_OPS(csi2dc_runtime_suspend, csi2dc_runtime_resume, NULL)
>>>>> +};
>>>>> +
>>>>> +static const struct of_device_id csi2dc_of_match[] = {
>>>>> + { .compatible = "microchip,sama7g5-csi2dc" },
>>>>> + { }
>>>>> +};
>>>>> +
>>>>> +MODULE_DEVICE_TABLE(of, csi2dc_of_match);
>>>>> +
>>>>> +static struct platform_driver csi2dc_driver = {
>>>>> + .probe = csi2dc_probe,
>>>>> + .remove = csi2dc_remove,
>>>>> + .driver = {
>>>>> + .name = "microchip-csi2dc",
>>>>> + .pm = &csi2dc_dev_pm_ops,
>>>>> + .of_match_table = of_match_ptr(csi2dc_of_match),
>>>>> + },
>>>>> +};
>>>>> +
>>>>> +module_platform_driver(csi2dc_driver);
>>>>> +
>>>>> +MODULE_AUTHOR("Eugen Hristev <eugen.hristev@xxxxxxxxxxxxx>");
>>>>> +MODULE_DESCRIPTION("Microchip CSI2 Demux Controller driver");
>>>>> +MODULE_LICENSE("GPL v2");
>>>>> +MODULE_SUPPORTED_DEVICE("video");
>>>>
>>>> For my education, what's MODULE_SUPPORTED_DEVICE for ?
>>>
>>> It's "not yet implemented" since 2005, I think it can be dropped :-)
>>>
>>> --
>>> Regards,
>>>
>>> Laurent Pinchart
>>>
>>