Re: [PATCH v5 04/11] media: bcm2835-unicam: Add support for CCP2/CSI2 camera interface

[Laurent Pinchart]

Hi Sakari,

On Mon, Jul 03, 2023 at 12:45:07AM +0300, Laurent Pinchart wrote:
> On Sun, Jul 02, 2023 at 06:18:21PM +0000, Sakari Ailus wrote:
> > On Sun, Jul 02, 2023 at 06:23:56PM +0300, Laurent Pinchart wrote:
> > > On Fri, Feb 25, 2022 at 11:29:18AM +0200, Sakari Ailus wrote:
> > > > On Tue, Feb 08, 2022 at 04:50:20PM +0100, Jean-Michel Hautbois wrote:
> > > > > Add driver for the Unicam camera receiver block on BCM283x processors.
> > > > > It is represented as two video device nodes: unicam-image and
> > > > > unicam-embedded which are connected to an internal subdev (named
> > > > > unicam-subdev) in order to manage streams routing.
> > > > > 
> > > > > Signed-off-by: Dave Stevenson <dave.stevenson@xxxxxxxxxxxxxxx>
> > > > > Signed-off-by: Naushir Patuck <naush@xxxxxxxxxxxxxxx>
> > > > > Signed-off-by: Jean-Michel Hautbois <jeanmichel.hautbois@xxxxxxxxxxxxxxxx>
> > > > > 
> > > > > ---
> > > > > v4:
> > > > >   - Add the vendor prefox for DT name
> > > > >   - Use the reg-names in DT parsing
> > > > >   - Remove MAINTAINERS entry
> > > > > 
> > > > > v3 main changes:
> > > > >   - Change code organization
> > > > >   - Remove unused variables
> > > > >   - Correct the fmt_meta functions
> > > > >   - Rewrite the start/stop streaming
> > > > >     - You can now start the image node alone, but not the metadata one
> > > > >     - The buffers are allocated per-node
> > > > >     - only the required stream is started, if the route exists and is
> > > > >       enabled
> > > > >   - Prefix the macros with UNICAM_ to not have too generic names
> > > > >   - Drop colorspace support
> > > > >     -> This is causing issues in the try-fmt v4l2-compliance test
> > > > >   test VIDIOC_G_FMT: OK
> > > > > 	fail: v4l2-test-formats.cpp(363): colorspace >= 0xff
> > > > > 	fail: v4l2-test-formats.cpp(465): testColorspace(!node->is_io_mc, pix.pixelformat, pix.colorspace, pix.ycbcr_enc, pix.quantization)
> > > > >   test VIDIOC_TRY_FMT: FAIL
> > > > > 	fail: v4l2-test-formats.cpp(363): colorspace >= 0xff
> > > > > 	fail: v4l2-test-formats.cpp(465): testColorspace(!node->is_io_mc, pix.pixelformat, pix.colorspace, pix.ycbcr_enc, pix.quantization)
> > > > >   test VIDIOC_S_FMT: FAIL
> > > > > 
> > > > > v2: Remove the unicam_{info,debug,error} macros and use
> > > > > dev_dbg/dev_err instead.
> > > > > ---
> > > > >  drivers/media/platform/Kconfig                |    1 +
> > > > >  drivers/media/platform/Makefile               |    2 +
> > > > >  drivers/media/platform/bcm2835/Kconfig        |   21 +
> > > > >  drivers/media/platform/bcm2835/Makefile       |    3 +
> > > > >  .../platform/bcm2835/bcm2835-unicam-regs.h    |  253 ++
> > > > >  .../media/platform/bcm2835/bcm2835-unicam.c   | 2570 +++++++++++++++++
> > > > >  6 files changed, 2850 insertions(+)
> > > > >  create mode 100644 drivers/media/platform/bcm2835/Kconfig
> > > > >  create mode 100644 drivers/media/platform/bcm2835/Makefile
> > > > >  create mode 100644 drivers/media/platform/bcm2835/bcm2835-unicam-regs.h
> > > > >  create mode 100644 drivers/media/platform/bcm2835/bcm2835-unicam.c
> > > 
> > > [snip]
> > > 
> > > > > diff --git a/drivers/media/platform/bcm2835/bcm2835-unicam-regs.h b/drivers/media/platform/bcm2835/bcm2835-unicam-regs.h
> > > > > new file mode 100644
> > > > > index 000000000000..b8d297076a02
> > > > > --- /dev/null
> > > > > +++ b/drivers/media/platform/bcm2835/bcm2835-unicam-regs.h
> > > 
> > > [snip]
> > > 
> > > > > +static int unicam_connect_of_subdevs(struct unicam_device *unicam)
> > > > > +{
> > > > > +	struct v4l2_fwnode_endpoint ep = { };
> > > > > +	struct fwnode_handle *ep_handle;
> > > > > +	struct v4l2_async_subdev *asd;
> > > > > +	unsigned int lane;
> > > > > +	int ret = -EINVAL;
> > > > > +
> > > > > +	if (of_property_read_u32(unicam->dev->of_node, "brcm,num-data-lanes",
> > > > > +				 &unicam->max_data_lanes) < 0) {
> > > > 
> > > > As you're already using fwnode API below, you could use
> > > > device_property_read_u32() here.
> > > > 
> > > > You can then replace of_device.h by mod_devicetable.h. Up to you.
> > > > 
> > > > > +		dev_err(unicam->dev, "DT property %s not set\n",
> > > > > +			"brcm,num-data-lanes");
> > > > > +		return -EINVAL;
> > > > > +	}
> > > > > +
> > > > > +	/* Get the local endpoint and remote device. */
> > > > > +	ep_handle = fwnode_graph_get_endpoint_by_id(dev_fwnode(unicam->dev),
> > > > > +						    0, 0,
> > > > > +						    FWNODE_GRAPH_ENDPOINT_NEXT);
> > > > > +	if (!ep_handle) {
> > > > > +		dev_err(unicam->dev, "No endpoint\n");
> > > > > +		return -ENODEV;
> > > > > +	}
> > > > > +
> > > > > +	/* Parse the local endpoint and validate its configuration. */
> > > > > +	if (v4l2_fwnode_endpoint_alloc_parse(ep_handle, &ep)) {
> > > > 
> > > > As you don't need link-frequencies property parsing, you should use
> > > > v4l2_fwnode_endpoint_parse(). That avoids having to call
> > > > v4l2_fwnode_endpoint_free().
> > > > 
> > > > > +		dev_err(unicam->dev, "could not parse endpoint\n");
> > > > > +		goto cleanup_exit;
> > > > > +	}
> > > > > +
> > > > > +	dev_dbg(unicam->dev, "parsed local endpoint, bus_type %u\n",
> > > > > +		ep.bus_type);
> > > > > +
> > > > > +	unicam->bus_type = ep.bus_type;
> > > > > +
> > > > > +	switch (ep.bus_type) {
> > > > > +	case V4L2_MBUS_CSI2_DPHY:
> > > > > +		switch (ep.bus.mipi_csi2.num_data_lanes) {
> > > > > +		case 1:
> > > > > +		case 2:
> > > > > +		case 4:
> > > > > +			break;
> > > > > +
> > > > > +		default:
> > > > > +			dev_err(unicam->dev, "%u data lanes not supported\n",
> > > > > +				ep.bus.mipi_csi2.num_data_lanes);
> > > > > +			goto cleanup_exit;
> > > > > +		}
> > > > > +
> > > > > +		for (lane = 0; lane < ep.bus.mipi_csi2.num_data_lanes; lane++) {
> > > > > +			if (ep.bus.mipi_csi2.data_lanes[lane] != lane + 1) {
> > > > > +				dev_err(unicam->dev, "data lanes reordering not supported\n");
> > > > > +				goto cleanup_exit;
> > > > > +			}
> > > > > +		}
> > > > > +
> > > > > +		if (ep.bus.mipi_csi2.num_data_lanes > unicam->max_data_lanes) {
> > > > > +			dev_err(unicam->dev, "endpoint requires %u data lanes when %u are supported\n",
> > > > > +				ep.bus.mipi_csi2.num_data_lanes,
> > > > > +				unicam->max_data_lanes);
> > > > > +		}
> > > > > +
> > > > > +		unicam->active_data_lanes = ep.bus.mipi_csi2.num_data_lanes;
> > > > > +		unicam->bus_flags = ep.bus.mipi_csi2.flags;
> > > > > +
> > > > > +		break;
> > > > > +
> > > > > +	case V4L2_MBUS_CCP2:
> > > > > +		if (ep.bus.mipi_csi1.clock_lane != 0 ||
> > > > > +		    ep.bus.mipi_csi1.data_lane != 1) {
> > > > > +			dev_err(unicam->dev, "unsupported lanes configuration\n");
> > > > 
> > > > If the hardware doesn't support lane remapping for CCP2, then that should
> > > > be reflected in DT bindings, i.e. data-lanes isn't relevant. There's no
> > > > need to check that here.
> > > 
> > > Should the above check for CSI-2 be dropped as well then ?
> > 
> > Same for CSI-2, too: if there's nothing to configure there (lane remapping)
> > there's no need to validate that part of the DT either.
> 
> OK, I'll drop that.

Actually, I'm wondering if it would make sense to tell the parsing
functions whether lane reordering is supported or not. The checks could
then be moved to the framework. What do you think ?

> > > > > +			goto cleanup_exit;
> > > > > +		}
> > > > > +
> > > > > +		unicam->max_data_lanes = 1;
> > > > > +		unicam->active_data_lanes = 1;
> > > > > +		unicam->bus_flags = ep.bus.mipi_csi1.strobe;
> > > > > +		break;
> > > > > +
> > > > > +	default:
> > > > > +		/* Unsupported bus type */
> > > > > +		dev_err(unicam->dev, "unsupported bus type %u\n",
> > > > > +			ep.bus_type);
> > > > > +		goto cleanup_exit;
> > > > > +	}
> > > > > +
> > > > > +	dev_dbg(unicam->dev, "%s bus, %u data lanes, flags=0x%08x\n",
> > > > > +		unicam->bus_type == V4L2_MBUS_CSI2_DPHY ? "CSI-2" : "CCP2",
> > > > > +		unicam->active_data_lanes, unicam->bus_flags);
> > > > 
> > > > V4l2-fwnode already prints this information I believe.
> > > 
> > > True. It does so with pr_debug() though, it would be nice to use
> > > dev_dbg(). That's a candidate for a separate fix of course.
> > 
> > The reason for using pr_debug() is that the device isn't used by the fwnode
> > framework. Would you add that just for debug prints? Note that the device
> > nodes themselves are already being printed so it adds little to the
> > usefulness of the messages.
> 
> I'll send patches, we can discuss their usefulness there.

-- 
Regards,

Laurent Pinchart