* [PATCH 01/27] drm/encoder: Add of_graph port to struct drm_encoder
From: Sascha Hauer @ 2022-01-26 14:55 UTC (permalink / raw)
To: dri-devel
Cc: linux-arm-kernel, linux-rockchip, devicetree, kernel, Andy Yan,
Benjamin Gaignard, Michael Riesch, Sandy Huang,
Heiko Stübner, Peter Geis, Sascha Hauer
In-Reply-To: <20220126145549.617165-1-s.hauer@pengutronix.de>
Add a device node to drm_encoder which corresponds with the port node
in the DT description of the encoder. This allows drivers to find the
of_graph link between a crtc and an encoder.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
---
include/drm/drm_encoder.h | 2 ++
1 file changed, 2 insertions(+)
diff --git a/include/drm/drm_encoder.h b/include/drm/drm_encoder.h
index 6e91a0280f31..3acd054b1eb3 100644
--- a/include/drm/drm_encoder.h
+++ b/include/drm/drm_encoder.h
@@ -99,6 +99,8 @@ struct drm_encoder {
struct drm_device *dev;
struct list_head head;
+ struct device_node *port;
+
struct drm_mode_object base;
char *name;
/**
--
2.30.2
^ permalink raw reply related
* Re: [PATCH] arm64: dts: renesas: ulcb-kf: add 9-asix sensor device
From: Geert Uytterhoeven @ 2022-01-26 14:55 UTC (permalink / raw)
To: Nikita Yushchenko
Cc: Magnus Damm, Rob Herring, Linux-Renesas,
open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
Linux Kernel Mailing List
In-Reply-To: <20220112205205.4082026-1-nikita.yoush@cogentembedded.com>
Hi Nikita,
Thanks for your patch!
On Wed, Jan 12, 2022 at 9:52 PM Nikita Yushchenko
<nikita.yoush@cogentembedded.com> wrote:
> This adds nodes for lsm9ds0 sensor installed on the KF board.
>
> With this patch, the sensor data becomes available over iio sysfs
> interface.
>
> Interrupt definition is not added yet, because the interrupt lines of
> lsm9ds0 are pulled to VCC on the board, which implies need for
> active-low configuration. But st_sensors drivers currently can't work
> with active-low interrupts on this chip.
That's unfortunate, as DT describes hardware, not limitations of the
software stack.
> Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
> --- a/arch/arm64/boot/dts/renesas/ulcb-kf.dtsi
> +++ b/arch/arm64/boot/dts/renesas/ulcb-kf.dtsi
> @@ -66,6 +66,13 @@ hdmi_3v3: regulator-hdmi-3v3 {
> regulator-max-microvolt = <3300000>;
> };
>
> + accel_3v3: regulator-acc-3v3 {
> + compatible = "regulator-fixed";
> + regulator-name = "accel-3v3";
> + regulator-min-microvolt = <3300000>;
> + regulator-max-microvolt = <3300000>;
> + };
> +
> hdmi1-out {
> compatible = "hdmi-connector";
> type = "a";
> @@ -208,6 +215,22 @@ pcm3168a_endpoint_c: endpoint {
> };
> };
> };
> +
> + lsm9ds0_acc_mag@1d {
Please use standard node names: accelerometer@1d?
> + compatible = "st,lsm9ds0-imu";
> + reg = <0x1d>;
> +
> + vdd-supply = <&accel_3v3>;
> + vddio-supply = <&accel_3v3>;
According to the bindings, the supplies are not required, so you can
leave them out? Or are the bindings wrong?
(The bindings also say "interrupts: maxItems 2", while the "interrupts:
description" says up to three interrupts, doh...)
> + };
> +
> + lsm9ds0_gyro@6b {
gyroscope@6b?
> + compatible = "st,lsm9ds0-gyro";
> + reg = <0x6b>;
> +
> + vdd-supply = <&accel_3v3>;
> + vddio-supply = <&accel_3v3>;
> + };
> };
> };
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
^ permalink raw reply
* Re: [RFC PATCH v2 4/7] media: bcm2835-unicam: Add support for for CCP2/CSI2 camera interface
From: Jean-Michel Hautbois @ 2022-01-26 14:54 UTC (permalink / raw)
To: Laurent Pinchart
Cc: dave.stevenson, devicetree, kernel-list, linux-arm-kernel,
linux-kernel, linux-media, linux-rpi-kernel, lukasz, mchehab,
naush, robh, tomi.valkeinen, Maxime Ripard
In-Reply-To: <0212af6d-a5de-05bb-161b-4271692dee59@ideasonboard.com>
+CC Maxime :-)
On 26/01/2022 15:21, Jean-Michel Hautbois wrote:
> Hi Laurent, thanks for the review !
>
> A question for Dave below :-)
>
> On 23/01/2022 00:26, Laurent Pinchart wrote:
>> Hi Jean-Michel,
>>
>> Thank you for the patch.
>>
>> Dave, Naush, there are a few questions for you below.
>>
>> On Fri, Jan 21, 2022 at 09:18:07AM +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@raspberrypi.com>
>>> Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
>>> Signed-off-by: Jean-Michel Hautbois
>>> <jeanmichel.hautbois@ideasonboard.com>
>>>
>>> ---
>>> in v2: Remove the unicam_{info,debug,error} macros and use
>>> dev_dbg/dev_err instead.
>>> ---
>>> MAINTAINERS | 1 +
>>> drivers/media/platform/Kconfig | 1 +
>>> drivers/media/platform/Makefile | 2 +
>>> drivers/media/platform/bcm2835/Kconfig | 21 +
>>> drivers/media/platform/bcm2835/Makefile | 3 +
>>> .../media/platform/bcm2835/bcm2835-unicam.c | 2678 +++++++++++++++++
>>> .../media/platform/bcm2835/vc4-regs-unicam.h | 253 ++
>>> 7 files changed, 2959 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.c
>>> create mode 100644 drivers/media/platform/bcm2835/vc4-regs-unicam.h
>>>
>>> diff --git a/MAINTAINERS b/MAINTAINERS
>>> index 07f238fd5ff9..b17bb533e007 100644
>>> --- a/MAINTAINERS
>>> +++ b/MAINTAINERS
>>> @@ -3684,6 +3684,7 @@ M: Raspberry Pi Kernel Maintenance
>>> <kernel-list@raspberrypi.com>
>>> L: linux-media@vger.kernel.org
>>> S: Maintained
>>> F: Documentation/devicetree/bindings/media/brcm,bcm2835-unicam.yaml
>>> +F: drivers/media/platform/bcm2835/
>>> BROADCOM BCM47XX MIPS ARCHITECTURE
>>> M: Hauke Mehrtens <hauke@hauke-m.de>
>>> diff --git a/drivers/media/platform/Kconfig
>>> b/drivers/media/platform/Kconfig
>>> index 9fbdba0fd1e7..033b0358fbb8 100644
>>> --- a/drivers/media/platform/Kconfig
>>> +++ b/drivers/media/platform/Kconfig
>>> @@ -170,6 +170,7 @@ source "drivers/media/platform/am437x/Kconfig"
>>> source "drivers/media/platform/xilinx/Kconfig"
>>> source "drivers/media/platform/rcar-vin/Kconfig"
>>> source "drivers/media/platform/atmel/Kconfig"
>>> +source "drivers/media/platform/bcm2835/Kconfig"
>>> source "drivers/media/platform/sunxi/Kconfig"
>>> config VIDEO_TI_CAL
>>> diff --git a/drivers/media/platform/Makefile
>>> b/drivers/media/platform/Makefile
>>> index 19bcbced7382..689e64857db1 100644
>>> --- a/drivers/media/platform/Makefile
>>> +++ b/drivers/media/platform/Makefile
>>> @@ -85,6 +85,8 @@ obj-$(CONFIG_VIDEO_QCOM_CAMSS) += qcom/camss/
>>> obj-$(CONFIG_VIDEO_QCOM_VENUS) += qcom/venus/
>>> +obj-y += bcm2835/
>>
>> One day it would be nice to clean the media Kconfig and Makefile
>> infrastructure to use a consistent style for all drivers. One can always
>> dream :-)
>>
>>> +
>>> obj-y += sunxi/
>>> obj-$(CONFIG_VIDEO_MESON_GE2D) += meson/ge2d/
>>> diff --git a/drivers/media/platform/bcm2835/Kconfig
>>> b/drivers/media/platform/bcm2835/Kconfig
>>> new file mode 100644
>>> index 000000000000..bd1370199650
>>> --- /dev/null
>>> +++ b/drivers/media/platform/bcm2835/Kconfig
>>> @@ -0,0 +1,21 @@
>>> +# Broadcom VideoCore4 V4L2 camera support
>>> +
>>> +config VIDEO_BCM2835_UNICAM
>>> + tristate "Broadcom BCM283x/BCM271x Unicam video capture driver"
>>> + depends on VIDEO_V4L2
>>> + depends on ARCH_BCM2835 || COMPILE_TEST
>>
>> I would put the arch dependency first, but that may be just me.
>>
>>> + select VIDEO_V4L2_SUBDEV_API
>>> + select MEDIA_CONTROLLER
>>> + select VIDEOBUF2_DMA_CONTIG
>>> + select V4L2_FWNODE
>>
>> And sort there alphabetically.
>>
>>> + help
>>> + Say Y here to enable support for the BCM283x/BCM271x CSI-2
>>> receiver.
>>> + This is a V4L2 driver that controls the CSI-2 receiver directly,
>>> + independently from the VC4 firmware.
>>> + This driver is mutually exclusive with the use of
>>> bcm2835-camera. The
>>> + firmware will disable all access to the peripheral from within
>>> the
>>> + firmware if it finds a DT node using it, and bcm2835-camera will
>>> + therefore fail to probe.
>>
>> Dave, Naush, what should we do with the bcm2835-camera driver in staging
>> ? Do we need to keep it to support older Pi models ?
>>
>>> +
>>> + To compile this driver as a module, choose M here. The module
>>> will be
>>> + called bcm2835-unicam.
>>> diff --git a/drivers/media/platform/bcm2835/Makefile
>>> b/drivers/media/platform/bcm2835/Makefile
>>> new file mode 100644
>>> index 000000000000..a98aba03598a
>>> --- /dev/null
>>> +++ b/drivers/media/platform/bcm2835/Makefile
>>> @@ -0,0 +1,3 @@
>>> +# Makefile for BCM2835 Unicam driver
>>> +
>>> +obj-$(CONFIG_VIDEO_BCM2835_UNICAM) += bcm2835-unicam.o
>>> diff --git a/drivers/media/platform/bcm2835/bcm2835-unicam.c
>>> b/drivers/media/platform/bcm2835/bcm2835-unicam.c
>>> new file mode 100644
>>> index 000000000000..7636496be00a
>>> --- /dev/null
>>> +++ b/drivers/media/platform/bcm2835/bcm2835-unicam.c
>>> @@ -0,0 +1,2678 @@
>>> +// SPDX-License-Identifier: GPL-2.0-only
>>> +/*
>>> + * BCM283x / BCM271x Unicam Capture Driver
>>> + *
>>> + * Copyright (C) 2017-2020 - Raspberry Pi (Trading) Ltd.
>>> + *
>>> + * Dave Stevenson <dave.stevenson@raspberrypi.com>
>>> + *
>>> + * Based on TI am437x driver by
>>> + * Benoit Parrot <bparrot@ti.com>
>>> + * Lad, Prabhakar <prabhakar.csengg@gmail.com>
>>> + *
>>> + * and TI CAL camera interface driver by
>>> + * Benoit Parrot <bparrot@ti.com>
>>> + *
>>> + *
>>> + * There are two camera drivers in the kernel for BCM283x - this one
>>> + * and bcm2835-camera (currently in staging).
>>> + *
>>> + * This driver directly controls the Unicam peripheral - there is no
>>> + * involvement with the VideoCore firmware. Unicam receives CSI-2 or
>>> + * CCP2 data and writes it into SDRAM.
>>> + * The only potential processing options are to repack Bayer data
>>> into an
>>> + * alternate format, and applying windowing.
>>> + * The repacking does not shift the data, so can repack
>>> V4L2_PIX_FMT_Sxxxx10P
>>> + * to V4L2_PIX_FMT_Sxxxx10, or V4L2_PIX_FMT_Sxxxx12P to
>>> V4L2_PIX_FMT_Sxxxx12,
>>> + * but not generically up to V4L2_PIX_FMT_Sxxxx16. The driver will
>>> add both
>>> + * formats where the relevant formats are defined, and will
>>> automatically
>>> + * configure the repacking as required.
>>> + * Support for windowing may be added later.
>>> + *
>>> + * It should be possible to connect this driver to any sensor with a
>>> + * suitable output interface and V4L2 subdevice driver.
>>> + *
>>> + * bcm2835-camera uses the VideoCore firmware to control the sensor,
>>> + * Unicam, ISP, and all tuner control loops. Fully processed frames are
>>> + * delivered to the driver by the firmware. It only has sensor drivers
>>> + * for Omnivision OV5647, and Sony IMX219 sensors.
>>> + *
>>> + * The two drivers are mutually exclusive for the same Unicam instance.
>>> + * The VideoCore firmware checks the device tree configuration
>>> during boot.
>>> + * If it finds device tree nodes called csi0 or csi1 it will block the
>>> + * firmware from accessing the peripheral, and bcm2835-camera will
>>> + * not be able to stream data.
>>> + */
>>> +
>>> +#include <linux/clk.h>
>>> +#include <linux/delay.h>
>>> +#include <linux/device.h>
>>> +#include <linux/dma-mapping.h>
>>> +#include <linux/err.h>
>>> +#include <linux/init.h>
>>> +#include <linux/interrupt.h>
>>> +#include <linux/io.h>
>>> +#include <linux/module.h>
>>> +#include <linux/of_device.h>
>>> +#include <linux/of_graph.h>
>>> +#include <linux/pinctrl/consumer.h>
>>> +#include <linux/platform_device.h>
>>> +#include <linux/pm_runtime.h>
>>> +#include <linux/slab.h>
>>> +#include <linux/uaccess.h>
>>> +#include <linux/videodev2.h>
>>> +
>>> +#include <media/v4l2-common.h>
>>> +#include <media/v4l2-ctrls.h>
>>> +#include <media/v4l2-dev.h>
>>> +#include <media/v4l2-device.h>
>>> +#include <media/v4l2-dv-timings.h>
>>> +#include <media/v4l2-event.h>
>>> +#include <media/v4l2-ioctl.h>
>>> +#include <media/v4l2-fwnode.h>
>>> +#include <media/videobuf2-dma-contig.h>
>>> +
>>> +#include <media/v4l2-async.h>
>>> +#define v4l2_async_nf_add_subdev __v4l2_async_nf_add_subdev
>>
>> Don't do that ! It's the kind of pattern that a malicious actor would
>> use to obfuscate code and get vulnerabilities merged.
>>
>>> +
>>> +#include "vc4-regs-unicam.h"
>>> +
>>> +#define UNICAM_MODULE_NAME "unicam"
>>> +#define UNICAM_VERSION "0.1.0"
>>
>> I doubt the version would ever be incremented, so let's just drop it.
>>
>>> +
>>> +static int debug;
>>> +module_param(debug, int, 0644);
>>> +MODULE_PARM_DESC(debug, "Debug level 0-3");
>>
>> Not used.
>>
>>> +
>>> +/*
>>> + * Unicam must request a minimum of 250Mhz from the VPU clock.
>>> + * Otherwise the input FIFOs overrun and cause image corruption.
>>> + */
>>> +#define MIN_VPU_CLOCK_RATE (250 * 1000 * 1000)
>>> +/*
>>> + * To protect against a dodgy sensor driver never returning an error
>>> from
>>> + * enum_mbus_code, set a maximum index value to be used.
>>> + */
>>> +#define MAX_ENUM_MBUS_CODE 128
>>> +
>>> +/*
>>> + * Stride is a 16 bit register, but also has to be a multiple of 32.
>>> + */
>>> +#define BPL_ALIGNMENT 32
>>> +#define MAX_BYTESPERLINE ((1 << 16) - BPL_ALIGNMENT)
>>> +/*
>>> + * Max width is therefore determined by the max stride divided by
>>> + * the number of bits per pixel. Take 32bpp as a
>>> + * worst case.
>>> + * No imposed limit on the height, so adopt a square image for want
>>> + * of anything better.
>>> + */
>>> +#define MAX_WIDTH (MAX_BYTESPERLINE / 4)
>>> +#define MAX_HEIGHT MAX_WIDTH
>>> +/* Define a nominal minimum image size */
>>> +#define MIN_WIDTH 16
>>> +#define MIN_HEIGHT 16
>>> +/* Default size of the embedded buffer */
>>> +#define UNICAM_EMBEDDED_SIZE 16384
>>> +
>>> +/*
>>> + * Size of the dummy buffer. Can be any size really, but the DMA
>>> + * allocation works in units of page sizes.
>>> + */
>>> +#define DUMMY_BUF_SIZE (PAGE_SIZE)
>>
>> No need for parentheses.
>>
>> Some of those macros have toogeneric names and may result in namespace
>> clashes. Please prefix them all with UNICAM_.
>>
>>> +
>>> +#define UNICAM_SD_PAD_SINK 0
>>> +#define UNICAM_SD_PAD_FIRST_SOURCE 1
>>
>> I'd name the source pads explicitly.
>>
>> #define UNICAM_SD_PAD_SOURCE_IMAGE 1
>> #define UNICAM_SD_PAD_SOURCE_METADATA 2
>>
>> (or META instead of METADATA, up to you, as long as we use consistent
>> naming through the driver).
>>
>>> +#define UNICAM_SD_NUM_SOURCE_PADS 2
>>> +#define UNICAM_SD_NUM_PADS (1 + UNICAM_SD_NUM_SOURCE_PADS)
>>> +
>>> +static inline bool unicam_sd_pad_is_sink(u32 pad)
>>> +{
>>> + /* Camera RX has 1 sink pad, and N source pads */
>>> + return pad == 0;
>>> +}
>>> +
>>> +static inline bool unicam_sd_pad_is_source(u32 pad)
>>> +{
>>> + /* Camera RX has 1 sink pad, and N source pads */
>>> + return pad >= UNICAM_SD_PAD_FIRST_SOURCE &&
>>> + pad <= UNICAM_SD_NUM_SOURCE_PADS;
>>
>> return pad != UNICAM_SD_PAD_SINK;
>>
>> and drop UNICAM_SD_PAD_FIRST_SOURCE and UNICAM_SD_NUM_SOURCE_PADS.
>>
>>> +}
>>> +
>>> +enum node_types {
>>
>> enum unicam_node_type {
>>
>>> + IMAGE_NODE,
>>> + METADATA_NODE,
>>> + MAX_NODES
>>> +};
>>
>> UNICAM_ prefixes please.
>>
>>> +
>>> +#define MASK_CS_DEFAULT BIT(V4L2_COLORSPACE_DEFAULT)
>>
>> This and several of the macros below are not used.
>>
>>> +#define MASK_CS_SMPTE170M BIT(V4L2_COLORSPACE_SMPTE170M)
>>> +#define MASK_CS_SMPTE240M BIT(V4L2_COLORSPACE_SMPTE240M)
>>> +#define MASK_CS_REC709 BIT(V4L2_COLORSPACE_REC709)
>>> +#define MASK_CS_BT878 BIT(V4L2_COLORSPACE_BT878)
>>> +#define MASK_CS_470_M BIT(V4L2_COLORSPACE_470_SYSTEM_M)
>>> +#define MASK_CS_470_BG BIT(V4L2_COLORSPACE_470_SYSTEM_BG)
>>> +#define MASK_CS_JPEG BIT(V4L2_COLORSPACE_JPEG)
>>> +#define MASK_CS_SRGB BIT(V4L2_COLORSPACE_SRGB)
>>> +#define MASK_CS_OPRGB BIT(V4L2_COLORSPACE_OPRGB)
>>> +#define MASK_CS_BT2020 BIT(V4L2_COLORSPACE_BT2020)
>>> +#define MASK_CS_RAW BIT(V4L2_COLORSPACE_RAW)
>>> +#define MASK_CS_DCI_P3 BIT(V4L2_COLORSPACE_DCI_P3)
>>> +
>>> +#define MAX_COLORSPACE 32
>>> +
>>> +/*
>>> + * struct unicam_fmt - Unicam media bus format information
>>> + * @pixelformat: V4L2 pixel format FCC identifier. 0 if n/a.
>>> + * @repacked_fourcc: V4L2 pixel format FCC identifier if the data is
>>> expanded
>>> + * out to 16bpp. 0 if n/a.
>>> + * @code: V4L2 media bus format code.
>>> + * @depth: Bits per pixel as delivered from the source.
>>> + * @csi_dt: CSI data type.
>>> + * @valid_colorspaces: Bitmask of valid colorspaces so that the
>>> Media Controller
>>> + * centric try_fmt can validate the colorspace and pass
>>> + * v4l2-compliance.
>>
>> I'd actually drop colorspace support completely. Unicam doesn't affect
>> the colorspace. It will output whatever it receives from the sensor. You
>> can accept any colorspace on the sink pad of the subdev, and just
>> propagate that to the source pad.
>>
>> If that causes v4l2-compliance failures, then we have a problem either
>> in the sensor drivers, or in v4l2-compliance.
>>
>>> + * @check_variants: Flag to denote that there are multiple mediabus
>>> formats
>>> + * still in the list that could match this V4L2 format.
>>> + * @mc_skip: Media Controller shouldn't list this format via
>>> ENUM_FMT as it is
>>> + * a duplicate of an earlier format.
>>> + * @metadata_fmt: This format only applies to the metadata pad.
>>> + */
>>> +struct unicam_fmt {
>>> + u32 fourcc;
>>> + u32 repacked_fourcc;
>>> + u32 code;
>>> + u8 depth;
>>> + u8 csi_dt;
>>> + u32 valid_colorspaces;
>>> + u8 check_variants:1;
>>> + u8 mc_skip:1;
>>> + u8 metadata_fmt:1;
>>> +};
>>> +
>>> +static const struct unicam_fmt formats[] = {
>>> + /* YUV Formats */
>>> + {
>>> + .fourcc = V4L2_PIX_FMT_YUYV,
>>> + .code = MEDIA_BUS_FMT_YUYV8_2X8,
>>
>> The convention for CSI-2 buses is to used the _1X16 media bus codes for
>> YUV. You can drop the first four entries, and drop the check_variants
>> and mc_skip fields.
>>
>>> + .depth = 16,
>>> + .csi_dt = 0x1e,
>>> + .check_variants = 1,
>>> + .valid_colorspaces = MASK_CS_SMPTE170M | MASK_CS_REC709 |
>>> + MASK_CS_JPEG,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_UYVY,
>>> + .code = MEDIA_BUS_FMT_UYVY8_2X8,
>>> + .depth = 16,
>>> + .csi_dt = 0x1e,
>>> + .check_variants = 1,
>>> + .valid_colorspaces = MASK_CS_SMPTE170M | MASK_CS_REC709 |
>>> + MASK_CS_JPEG,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_YVYU,
>>> + .code = MEDIA_BUS_FMT_YVYU8_2X8,
>>> + .depth = 16,
>>> + .csi_dt = 0x1e,
>>> + .check_variants = 1,
>>> + .valid_colorspaces = MASK_CS_SMPTE170M | MASK_CS_REC709 |
>>> + MASK_CS_JPEG,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_VYUY,
>>> + .code = MEDIA_BUS_FMT_VYUY8_2X8,
>>> + .depth = 16,
>>> + .csi_dt = 0x1e,
>>> + .check_variants = 1,
>>> + .valid_colorspaces = MASK_CS_SMPTE170M | MASK_CS_REC709 |
>>> + MASK_CS_JPEG,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_YUYV,
>>> + .code = MEDIA_BUS_FMT_YUYV8_1X16,
>>> + .depth = 16,
>>> + .csi_dt = 0x1e,
>>> + .mc_skip = 1,
>>> + .valid_colorspaces = MASK_CS_SMPTE170M | MASK_CS_REC709 |
>>> + MASK_CS_JPEG,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_UYVY,
>>> + .code = MEDIA_BUS_FMT_UYVY8_1X16,
>>> + .depth = 16,
>>> + .csi_dt = 0x1e,
>>> + .mc_skip = 1,
>>> + .valid_colorspaces = MASK_CS_SMPTE170M | MASK_CS_REC709 |
>>> + MASK_CS_JPEG,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_YVYU,
>>> + .code = MEDIA_BUS_FMT_YVYU8_1X16,
>>> + .depth = 16,
>>> + .csi_dt = 0x1e,
>>> + .mc_skip = 1,
>>> + .valid_colorspaces = MASK_CS_SMPTE170M | MASK_CS_REC709 |
>>> + MASK_CS_JPEG,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_VYUY,
>>> + .code = MEDIA_BUS_FMT_VYUY8_1X16,
>>> + .depth = 16,
>>> + .csi_dt = 0x1e,
>>> + .mc_skip = 1,
>>> + .valid_colorspaces = MASK_CS_SMPTE170M | MASK_CS_REC709 |
>>> + MASK_CS_JPEG,
>>> + }, {
>>> + /* RGB Formats */
>>> + .fourcc = V4L2_PIX_FMT_RGB565, /* gggbbbbb rrrrrggg */
>>> + .code = MEDIA_BUS_FMT_RGB565_2X8_LE,
>>> + .depth = 16,
>>> + .csi_dt = 0x22,
>>> + .valid_colorspaces = MASK_CS_SRGB,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_RGB565X, /* rrrrrggg gggbbbbb */
>>> + .code = MEDIA_BUS_FMT_RGB565_2X8_BE,
>>> + .depth = 16,
>>> + .csi_dt = 0x22,
>>> + .valid_colorspaces = MASK_CS_SRGB,
>>
>> CSI-2 specifies the RGB565 format unambiguously, there are no little
>> endian of big endian variants. Let's drop one of the two entries, and
>> use MEDIA_BUS_FMT_RGB565_1X15. Dave, Naush, what is the correct pixel
>> format for this, is data layed out in big endian or little endian ?
>>
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_RGB555, /* gggbbbbb arrrrrgg */
>>> + .code = MEDIA_BUS_FMT_RGB555_2X8_PADHI_LE,
>>> + .depth = 16,
>>> + .csi_dt = 0x21,
>>> + .valid_colorspaces = MASK_CS_SRGB,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_RGB555X, /* arrrrrgg gggbbbbb */
>>> + .code = MEDIA_BUS_FMT_RGB555_2X8_PADHI_BE,
>>> + .depth = 16,
>>> + .csi_dt = 0x21,
>>> + .valid_colorspaces = MASK_CS_SRGB,
>>
>> Same here, but we need to define a new MEDIA_BUS_FMT_RGB555_1X16 format.
>> The CSI-2 RGB555 data type has the padding bit between the green and
>> blue components, so I wonder if Unicam maps that to one of
>> V4L2_PIX_FMT_RGB555 or V4L2_PIX_FMT_RGB555X, or if a new pixel format is
>> needed too.
>>
>> Another option would be to drop RGB555 support until someone needs it
>> (with a comment in the table to tell the entry is missing).
>>
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_RGB24, /* rgb */
>>> + .code = MEDIA_BUS_FMT_RGB888_1X24,
>>> + .depth = 24,
>>> + .csi_dt = 0x24,
>>> + .valid_colorspaces = MASK_CS_SRGB,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_BGR24, /* bgr */
>>> + .code = MEDIA_BUS_FMT_BGR888_1X24,
>>> + .depth = 24,
>>> + .csi_dt = 0x24,
>>> + .valid_colorspaces = MASK_CS_SRGB,
>>
>> This is possibly more tricky, as CSI-2 defined RGB888 as being
>> transmitted in BGR order, but sensors could possibly also support RGB as
>> a non-standard extension (the same may CSI-2 standardizes on UYVY, but
>> many sensors can swap components). We can keep both entry I suppose.
>>
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_RGB32, /* argb */
>>> + .code = MEDIA_BUS_FMT_ARGB8888_1X32,
>>> + .depth = 32,
>>> + .csi_dt = 0x0,
>>> + .valid_colorspaces = MASK_CS_SRGB,
>>
>> There's no 32-bit RGB in CSI-2, should this be dropped ?
>>
>>> + }, {
>>> + /* Bayer Formats */
>>> + .fourcc = V4L2_PIX_FMT_SBGGR8,
>>> + .code = MEDIA_BUS_FMT_SBGGR8_1X8,
>>> + .depth = 8,
>>> + .csi_dt = 0x2a,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_SGBRG8,
>>> + .code = MEDIA_BUS_FMT_SGBRG8_1X8,
>>> + .depth = 8,
>>> + .csi_dt = 0x2a,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_SGRBG8,
>>> + .code = MEDIA_BUS_FMT_SGRBG8_1X8,
>>> + .depth = 8,
>>> + .csi_dt = 0x2a,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_SRGGB8,
>>> + .code = MEDIA_BUS_FMT_SRGGB8_1X8,
>>> + .depth = 8,
>>> + .csi_dt = 0x2a,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_SBGGR10P,
>>> + .repacked_fourcc = V4L2_PIX_FMT_SBGGR10,
>>> + .code = MEDIA_BUS_FMT_SBGGR10_1X10,
>>> + .depth = 10,
>>> + .csi_dt = 0x2b,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_SGBRG10P,
>>> + .repacked_fourcc = V4L2_PIX_FMT_SGBRG10,
>>> + .code = MEDIA_BUS_FMT_SGBRG10_1X10,
>>> + .depth = 10,
>>> + .csi_dt = 0x2b,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_SGRBG10P,
>>> + .repacked_fourcc = V4L2_PIX_FMT_SGRBG10,
>>> + .code = MEDIA_BUS_FMT_SGRBG10_1X10,
>>> + .depth = 10,
>>> + .csi_dt = 0x2b,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_SRGGB10P,
>>> + .repacked_fourcc = V4L2_PIX_FMT_SRGGB10,
>>> + .code = MEDIA_BUS_FMT_SRGGB10_1X10,
>>> + .depth = 10,
>>> + .csi_dt = 0x2b,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_SBGGR12P,
>>> + .repacked_fourcc = V4L2_PIX_FMT_SBGGR12,
>>> + .code = MEDIA_BUS_FMT_SBGGR12_1X12,
>>> + .depth = 12,
>>> + .csi_dt = 0x2c,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_SGBRG12P,
>>> + .repacked_fourcc = V4L2_PIX_FMT_SGBRG12,
>>> + .code = MEDIA_BUS_FMT_SGBRG12_1X12,
>>> + .depth = 12,
>>> + .csi_dt = 0x2c,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_SGRBG12P,
>>> + .repacked_fourcc = V4L2_PIX_FMT_SGRBG12,
>>> + .code = MEDIA_BUS_FMT_SGRBG12_1X12,
>>> + .depth = 12,
>>> + .csi_dt = 0x2c,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_SRGGB12P,
>>> + .repacked_fourcc = V4L2_PIX_FMT_SRGGB12,
>>> + .code = MEDIA_BUS_FMT_SRGGB12_1X12,
>>> + .depth = 12,
>>> + .csi_dt = 0x2c,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_SBGGR14P,
>>> + .repacked_fourcc = V4L2_PIX_FMT_SBGGR14,
>>> + .code = MEDIA_BUS_FMT_SBGGR14_1X14,
>>> + .depth = 14,
>>> + .csi_dt = 0x2d,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_SGBRG14P,
>>> + .repacked_fourcc = V4L2_PIX_FMT_SGBRG14,
>>> + .code = MEDIA_BUS_FMT_SGBRG14_1X14,
>>> + .depth = 14,
>>> + .csi_dt = 0x2d,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_SGRBG14P,
>>> + .repacked_fourcc = V4L2_PIX_FMT_SGRBG14,
>>> + .code = MEDIA_BUS_FMT_SGRBG14_1X14,
>>> + .depth = 14,
>>> + .csi_dt = 0x2d,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_SRGGB14P,
>>> + .repacked_fourcc = V4L2_PIX_FMT_SRGGB14,
>>> + .code = MEDIA_BUS_FMT_SRGGB14_1X14,
>>> + .depth = 14,
>>> + .csi_dt = 0x2d,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + /*
>>> + * 16 bit Bayer formats could be supported, but there is no CSI2
>>> + * data_type defined for raw 16, and no sensors that produce it at
>>> + * present.
>>> + */
>>
>> RAW16 is now defined in CSI-2, but that can be handled later.
>>
>>> +
>>> + /* Greyscale formats */
>>> + .fourcc = V4L2_PIX_FMT_GREY,
>>> + .code = MEDIA_BUS_FMT_Y8_1X8,
>>> + .depth = 8,
>>> + .csi_dt = 0x2a,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_Y10P,
>>> + .repacked_fourcc = V4L2_PIX_FMT_Y10,
>>> + .code = MEDIA_BUS_FMT_Y10_1X10,
>>> + .depth = 10,
>>> + .csi_dt = 0x2b,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_Y12P,
>>> + .repacked_fourcc = V4L2_PIX_FMT_Y12,
>>> + .code = MEDIA_BUS_FMT_Y12_1X12,
>>> + .depth = 12,
>>> + .csi_dt = 0x2c,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + }, {
>>> + .fourcc = V4L2_PIX_FMT_Y14P,
>>> + .repacked_fourcc = V4L2_PIX_FMT_Y14,
>>> + .code = MEDIA_BUS_FMT_Y14_1X14,
>>> + .depth = 14,
>>> + .csi_dt = 0x2d,
>>> + .valid_colorspaces = MASK_CS_RAW,
>>> + },
>>> + /* Embedded data format */
>>> + {
>>> + .fourcc = V4L2_META_FMT_8,
>>> + .code = MEDIA_BUS_FMT_METADATA_8,
>>> + .depth = 8,
>>> + .metadata_fmt = 1,
>>> + }
>>> +};
>>
>> Let's reorder the code a bit to improve readability:
>>
>> - Macro definitions
>> - Structure definitions (with the container_of wrappers following the
>> respective structure)
>> - Misc helper functions (from is_metadata_node() to reg_write_field())
>> - Format data table and helper functions (from find_format_by_code() to
>> unicam_calc_format_size_bpl())
>> - Hardware handling:
>> - From unicam_wr_dma_addr() to unicam_isr()
>> - From unicam_set_packing_config() to unicam_disable()
>> - V4L2 subdev (ops & initialization & registration)
>> - vb2 queue opq
>> - V4L2 video device (ops & initialization & registration)
>> - Probe
>>
>> I recommend adding a comment header in front of each section after the
>> structure definitions with a distinctive visual appearance, to delimit
>> code blocks very clearly. I usually use the following style:
>>
>> /*
>> -----------------------------------------------------------------------------
>>
>> * V4L2 Subdev Operations
>> */
>>
>>> +
>>> +struct unicam_buffer {
>>> + struct vb2_v4l2_buffer vb;
>>> + struct list_head list;
>>> +};
>>> +
>>> +static inline struct unicam_buffer *to_unicam_buffer(struct
>>> vb2_buffer *vb)
>>> +{
>>> + return container_of(vb, struct unicam_buffer, vb.vb2_buf);
>>> +}
>>> +
>>> +struct unicam_node {
>>> + bool registered;
>>> + int open;
>>
>> Not used. Please go through all structure fields and delete the ones
>> that are not used.
>>
>>> + bool streaming;
>>> + unsigned int node_id;
>>> + /* Source pad id on the sensor for this node */
>>> + unsigned int src_pad_id;
>>> + /* Pointer pointing to current v4l2_buffer */
>>> + struct unicam_buffer *cur_frm;
>>> + /* Pointer pointing to next v4l2_buffer */
>>> + struct unicam_buffer *next_frm;
>>> + /* video capture */
>>> + const struct unicam_fmt *fmt;
>>> + /* Used to store current pixel format */
>>> + struct v4l2_format v_fmt;
>>> + /* Used to store current mbus frame format */
>>> + struct v4l2_mbus_framefmt m_fmt;
>>> + /* Buffer queue used in video-buf */
>>> + struct vb2_queue buffer_queue;
>>> + /* Queue of filled frames */
>>> + struct list_head dma_queue;
>>> + /* IRQ lock for DMA queue */
>>> + spinlock_t dma_queue_lock;
>>> + /* lock used to access this structure */
>>> + struct mutex lock;
>>> + /* Identifies video device for this channel */
>>> + struct video_device video_dev;
>>> + /* Pointer to the parent handle */
>>> + struct unicam_device *dev;
>>> + struct media_pad pad;
>>> + unsigned int embedded_lines;
>>> + struct media_pipeline pipe;
>>> + /*
>>> + * Dummy buffer intended to be used by unicam
>>> + * if we have no other queued buffers to swap to.
>>> + */
>>> + void *dummy_buf_cpu_addr;
>>> + dma_addr_t dummy_buf_dma_addr;
>>> +};
>>> +
>>> +struct unicam_device {
>>> + struct kref kref;
>>> +
>>> + /* V4l2 specific parameters */
>>> + struct v4l2_async_subdev asd;
>>> +
>>> + /* peripheral base address */
>>> + void __iomem *base;
>>> + /* clock gating base address */
>>> + void __iomem *clk_gate_base;
>>> + /* lp clock handle */
>>> + struct clk *clock;
>>> + /* vpu clock handle */
>>> + struct clk *vpu_clock;
>>> + /* vpu clock request */
>>> + struct clk_request *vpu_req;
>>
>> Not used (and that may be a problem).
>
> In the original linux-rpi tree, there is this portion of code in
> unicam_start_streaming:
>
> dev->vpu_req = clk_request_start(dev->vpu_clock, MIN_VPU_CLOCK_RATE);
> if (!dev->vpu_req) {
> unicam_err(dev, "failed to set up VPU clock\n");
> goto error_pipeline;
> }
>
> ret = clk_prepare_enable(dev->vpu_clock);
> if (ret) {
> unicam_err(dev, "Failed to enable VPU clock: %d\n", ret);
> goto error_pipeline;
> }
>
> And this is used as this because it depends on the non-merged series
> "clk: [PATCH v2 0/3] clk: Implement a clock request API" [1]
>
> [1]: https://lore.kernel.org/all/20210914093515.260031-1-maxime@cerno.tech/
>
> That's why I modified the code and called:
> clk_set_min_rate(dev->vpu_clock, UNICAM_MIN_VPU_CLOCK_RATE);
>
> Dave, is it ok or do we need absolutely the clock request API ?
>
> JM
>
>>
>>> + /* clock status for error handling */
>>> + bool clocks_enabled;
>>> + /* V4l2 device */
>>> + struct v4l2_device v4l2_dev;
>>> + struct media_device mdev;
>>> +
>>> + /* parent device */
>>> + struct platform_device *pdev;
>>
>> The pdev field is only used to access pdev.dev, so I'd replace it with
>> the struct device.
>>
>>> + /* subdevice async Notifier */
>>> + struct v4l2_async_notifier notifier;
>>> + unsigned int sequence;
>>> +
>>> + /* ptr to sub device */
>>> + struct v4l2_subdev *sensor;
>>> + /* Pad config for the sensor */
>>> + struct v4l2_subdev_state *sensor_state;
>>> +
>>> + /* Internal subdev */
>>> + struct v4l2_subdev sd;
>>> + struct media_pad pads[UNICAM_SD_NUM_PADS];
>>> +
>>> + bool streaming;
>>
>> "streaming" is ambiguous, it's not clear it relates to the subdev. I'd
>> group all the fields related to the subdev in a structure:
>>
>> struct {
>> struct v4l2_subdev sd;
>> struct media_pad pads[UNICAM_SD_NUM_PADS];
>> bool streaming;
>> } subdev;
>>
>> (there may be more, I haven't checked)
>>
>>> +
>>> + enum v4l2_mbus_type bus_type;
>>> + /*
>>> + * Stores bus.mipi_csi2.flags for CSI2 sensors, or
>>> + * bus.mipi_csi1.strobe for CCP2.
>>> + */
>>> + unsigned int bus_flags;
>>> + unsigned int max_data_lanes;
>>> + unsigned int active_data_lanes;
>>> + bool sensor_embedded_data;
>>> +
>>> + struct unicam_node node[MAX_NODES];
>>> + struct v4l2_ctrl_handler ctrl_handler;
>>> +
>>> + bool mc_api;
>>> +};
>>> +
>>> +static inline struct unicam_device *
>>> +to_unicam_device(struct v4l2_device *v4l2_dev)
>>
>> I'd rename this to v4l2_device_to_unicam_device(), as there's another
>> container_of wrapper.
>>
>>> +{
>>> + return container_of(v4l2_dev, struct unicam_device, v4l2_dev);
>>> +}
>>> +
>>> +static inline struct unicam_device *
>>> +sd_to_unicam_device(struct v4l2_subdev *sd)
>>> +{
>>> + return container_of(sd, struct unicam_device, sd);
>>> +}
>>> +
>>> +static inline bool is_metadata_node(struct unicam_node *node)
>>> +{
>>> + return node->video_dev.device_caps & V4L2_CAP_META_CAPTURE;
>>> +}
>>> +
>>> +static inline bool is_image_node(struct unicam_node *node)
>>> +{
>>> + return node->video_dev.device_caps & V4L2_CAP_VIDEO_CAPTURE;
>>> +}
>>> +
>>> +/* Hardware access */
>>> +static inline void clk_write(struct unicam_device *dev, u32 val)
>>> +{
>>> + writel(val | 0x5a000000, dev->clk_gate_base);
>>> +}
>>
>> The name of this function and the functions below are too generic. Add a
>> unicam_ prefix.
>>
>>> +
>>> +static inline u32 reg_read(struct unicam_device *dev, u32 offset)
>>> +{
>>> + return readl(dev->base + offset);
>>> +}
>>> +
>>> +static inline void reg_write(struct unicam_device *dev, u32 offset,
>>> u32 val)
>>> +{
>>> + writel(val, dev->base + offset);
>>> +}
>>> +
>>> +static inline int get_field(u32 value, u32 mask)
>>> +{
>>> + return (value & mask) >> __ffs(mask);
>>> +}
>>> +
>>> +static inline void set_field(u32 *valp, u32 field, u32 mask)
>>> +{
>>> + u32 val = *valp;
>>> +
>>> + val &= ~mask;
>>> + val |= (field << __ffs(mask)) & mask;
>>> + *valp = val;
>>> +}
>>> +
>>> +static inline u32 reg_read_field(struct unicam_device *dev, u32 offset,
>>> + u32 mask)
>>
>> Not used.
>>
>>> +{
>>> + return get_field(reg_read(dev, offset), mask);
>>> +}
>>> +
>>> +static inline void reg_write_field(struct unicam_device *dev, u32
>>> offset,
>>> + u32 field, u32 mask)
>>> +{
>>> + u32 val = reg_read(dev, offset);
>>> +
>>> + set_field(&val, field, mask);
>>> + reg_write(dev, offset, val);
>>> +}
>>
>> I'm really not a bit fan of the read-modify-update patterns, more often
>> than not they can be replaced by direct writes. No need to fix this
>> though, it's ok.
>>
>>> +
>>> +/* Power management functions */
>>> +static inline int unicam_runtime_get(struct unicam_device *dev)
>>> +{
>>> + return pm_runtime_get_sync(&dev->pdev->dev);
>>
>> Use pm_runtime_resume_and_get() instead, as the error handling is broken
>> in the driver otherwise.
>>
>>> +}
>>> +
>>> +static inline void unicam_runtime_put(struct unicam_device *dev)
>>> +{
>>> + pm_runtime_put_sync(&dev->pdev->dev);
>>> +}
>>
>> Drop these wrappers and call the functions directly.
>>
>>> +
>>> +/* Format setup functions */
>>> +static const struct unicam_fmt *find_format_by_code(u32 code)
>>> +{
>>> + unsigned int i;
>>> +
>>> + for (i = 0; i < ARRAY_SIZE(formats); i++) {
>>> + if (formats[i].code == code)
>>> + return &formats[i];
>>> + }
>>> +
>>> + return NULL;
>>> +}
>>> +
>>> +static const struct unicam_fmt *find_format_by_fourcc(u32 fourcc)
>>> +{
>>> + unsigned int i;
>>> +
>>> + for (i = 0; i < ARRAY_SIZE(formats); ++i) {
>>> + if (formats[i].fourcc == fourcc)
>>> + return &formats[i];
>>> + }
>>> +
>>> + return NULL;
>>> +}
>>> +
>>> +static unsigned int bytes_per_line(u32 width, const struct
>>> unicam_fmt *fmt,
>>> + u32 v4l2_fourcc)
>>> +{
>>> + if (v4l2_fourcc == fmt->repacked_fourcc)
>>> + /* Repacking always goes to 16bpp */
>>> + return ALIGN(width << 1, BPL_ALIGNMENT);
>>> + else
>>> + return ALIGN((width * fmt->depth) >> 3, BPL_ALIGNMENT);
>>> +}
>>> +
>>> +static int unicam_calc_format_size_bpl(struct unicam_device *dev,
>>> + const struct unicam_fmt *fmt,
>>> + struct v4l2_format *f)
>>> +{
>>> + unsigned int min_bytesperline;
>>> +
>>> + v4l_bound_align_image(&f->fmt.pix.width, MIN_WIDTH, MAX_WIDTH, 2,
>>> + &f->fmt.pix.height, MIN_HEIGHT, MAX_HEIGHT, 0,
>>> + 0);
>>> +
>>> + min_bytesperline = bytes_per_line(f->fmt.pix.width, fmt,
>>> + f->fmt.pix.pixelformat);
>>> +
>>> + if (f->fmt.pix.bytesperline > min_bytesperline &&
>>> + f->fmt.pix.bytesperline <= MAX_BYTESPERLINE)
>>> + f->fmt.pix.bytesperline = ALIGN(f->fmt.pix.bytesperline,
>>> + BPL_ALIGNMENT);
>>> + else
>>> + f->fmt.pix.bytesperline = min_bytesperline;
>>> +
>>> + f->fmt.pix.sizeimage = f->fmt.pix.height * f->fmt.pix.bytesperline;
>>> +
>>> + dev_dbg(dev->v4l2_dev.dev, "%s: fourcc: %08X size: %dx%d bpl:%d
>>> img_size:%d\n",
>>> + __func__,
>>> + f->fmt.pix.pixelformat,
>>> + f->fmt.pix.width, f->fmt.pix.height,
>>> + f->fmt.pix.bytesperline, f->fmt.pix.sizeimage);
>>> +
>>> + return 0;
>>> +}
>>> +
>>> +static void unicam_wr_dma_addr(struct unicam_device *dev, dma_addr_t
>>> dmaaddr,
>>> + unsigned int buffer_size, int node_id)
>>> +{
>>> + dma_addr_t endaddr = dmaaddr + buffer_size;
>>> +
>>> + if (node_id == IMAGE_NODE) {
>>> + reg_write(dev, UNICAM_IBSA0, dmaaddr);
>>> + reg_write(dev, UNICAM_IBEA0, endaddr);
>>> + } else {
>>> + reg_write(dev, UNICAM_DBSA0, dmaaddr);
>>> + reg_write(dev, UNICAM_DBEA0, endaddr);
>>> + }
>>> +}
>>> +
>>> +static unsigned int unicam_get_lines_done(struct unicam_device *dev)
>>> +{
>>> + dma_addr_t start_addr, cur_addr;
>>> + unsigned int stride =
>>> dev->node[IMAGE_NODE].v_fmt.fmt.pix.bytesperline;
>>> + struct unicam_buffer *frm = dev->node[IMAGE_NODE].cur_frm;
>>> +
>>> + if (!frm)
>>> + return 0;
>>> +
>>> + start_addr = vb2_dma_contig_plane_dma_addr(&frm->vb.vb2_buf, 0);
>>> + cur_addr = reg_read(dev, UNICAM_IBWP);
>>> + return (unsigned int)(cur_addr - start_addr) / stride;
>>> +}
>>> +
>>> +static void unicam_schedule_next_buffer(struct unicam_node *node)
>>> +{
>>> + struct unicam_device *dev = node->dev;
>>> + struct unicam_buffer *buf;
>>> + unsigned int size;
>>> + dma_addr_t addr;
>>> +
>>> + buf = list_first_entry(&node->dma_queue, struct unicam_buffer,
>>> list);
>>> + node->next_frm = buf;
>>> + list_del(&buf->list);
>>> +
>>> + addr = vb2_dma_contig_plane_dma_addr(&buf->vb.vb2_buf, 0);
>>> + if (is_image_node(node)) {
>>> + size = node->v_fmt.fmt.pix.sizeimage;
>>> + unicam_wr_dma_addr(dev, addr, size, IMAGE_NODE);
>>> + } else {
>>> + size = node->v_fmt.fmt.meta.buffersize;
>>> + unicam_wr_dma_addr(dev, addr, size, METADATA_NODE);
>>> + }
>>> +}
>>> +
>>> +static void unicam_schedule_dummy_buffer(struct unicam_node *node)
>>> +{
>>> + struct unicam_device *dev = node->dev;
>>> + int node_id = is_image_node(node) ? IMAGE_NODE : METADATA_NODE;
>>> +
>>> + dev_dbg(dev->v4l2_dev.dev, "Scheduling dummy buffer for node
>>> %d\n", node_id);
>>> +
>>> + unicam_wr_dma_addr(dev, node->dummy_buf_dma_addr, DUMMY_BUF_SIZE,
>>> + node_id);
>>> + node->next_frm = NULL;
>>> +}
>>> +
>>> +static void unicam_process_buffer_complete(struct unicam_node *node,
>>> + unsigned int sequence)
>>> +{
>>> + node->cur_frm->vb.field = node->m_fmt.field;
>>> + node->cur_frm->vb.sequence = sequence;
>>> +
>>> + vb2_buffer_done(&node->cur_frm->vb.vb2_buf, VB2_BUF_STATE_DONE);
>>> +}
>>> +
>>> +static void unicam_queue_event_sof(struct unicam_device *unicam)
>>> +{
>>> + struct v4l2_event event = {
>>> + .type = V4L2_EVENT_FRAME_SYNC,
>>> + .u.frame_sync.frame_sequence = unicam->sequence,
>>> + };
>>> +
>>> + v4l2_event_queue(&unicam->node[IMAGE_NODE].video_dev, &event);
>>> +}
>>> +
>>> +/*
>>> + * unicam_isr : ISR handler for unicam capture
>>> + * @irq: irq number
>>> + * @dev_id: dev_id ptr
>>> + *
>>> + * It changes status of the captured buffer, takes next buffer from
>>> the queue
>>> + * and sets its address in unicam registers
>>> + */
>>> +static irqreturn_t unicam_isr(int irq, void *dev)
>>> +{
>>> + struct unicam_device *unicam = dev;
>>> + unsigned int lines_done = unicam_get_lines_done(dev);
>>> + unsigned int sequence = unicam->sequence;
>>> + unsigned int i;
>>> + u32 ista, sta;
>>> + bool fe;
>>> + u64 ts;
>>> +
>>> + sta = reg_read(unicam, UNICAM_STA);
>>> + /* Write value back to clear the interrupts */
>>> + reg_write(unicam, UNICAM_STA, sta);
>>> +
>>> + ista = reg_read(unicam, UNICAM_ISTA);
>>> + /* Write value back to clear the interrupts */
>>> + reg_write(unicam, UNICAM_ISTA, ista);
>>> +
>>> + dev_dbg(unicam->v4l2_dev.dev, "ISR: ISTA: 0x%X, STA: 0x%X,
>>> sequence %d, lines done %d",
>>> + ista, sta, sequence, lines_done);
>>> +
>>> + if (!(sta & (UNICAM_IS | UNICAM_PI0)))
>>> + return IRQ_HANDLED;
>>> +
>>> + /*
>>> + * Look for either the Frame End interrupt or the Packet Capture
>>> status
>>> + * to signal a frame end.
>>> + */
>>> + fe = (ista & UNICAM_FEI || sta & UNICAM_PI0);
>>
>> No need for the outer parentheses.
>>
>>> +
>>> + /*
>>> + * We must run the frame end handler first. If we have a valid
>>> next_frm
>>> + * and we get a simultaneout FE + FS interrupt, running the FS
>>> handler
>>> + * first would null out the next_frm ptr and we would have lost the
>>> + * buffer forever.
>>> + */
>>> + if (fe) {
>>> + /*
>>> + * Ensure we have swapped buffers already as we can't
>>> + * stop the peripheral. If no buffer is available, use a
>>> + * dummy buffer to dump out frames until we get a new buffer
>>> + * to use.
>>> + */
>>> + for (i = 0; i < ARRAY_SIZE(unicam->node); i++) {
>>> + if (!unicam->node[i].streaming)
>>> + continue;
>>> +
>>> + /*
>>> + * If cur_frm == next_frm, it means we have not had
>>> + * a chance to swap buffers, likely due to having
>>> + * multiple interrupts occurring simultaneously (like FE
>>> + * + FS + LS). In this case, we cannot signal the buffer
>>> + * as complete, as the HW will reuse that buffer.
>>> + */
>>> + if (unicam->node[i].cur_frm &&
>>> + unicam->node[i].cur_frm != unicam->node[i].next_frm)
>>> + unicam_process_buffer_complete(&unicam->node[i],
>>> + sequence);
>>> + unicam->node[i].cur_frm = unicam->node[i].next_frm;
>>> + }
>>> + unicam->sequence++;
>>> + }
>>> +
>>> + if (ista & UNICAM_FSI) {
>>> + /*
>>> + * Timestamp is to be when the first data byte was captured,
>>> + * aka frame start.
>>> + */
>>> + ts = ktime_get_ns();
>>> + for (i = 0; i < ARRAY_SIZE(unicam->node); i++) {
>>> + if (!unicam->node[i].streaming)
>>> + continue;
>>> +
>>> + if (unicam->node[i].cur_frm)
>>> + unicam->node[i].cur_frm->vb.vb2_buf.timestamp =
>>> + ts;
>>> + else
>>> + dev_dbg(unicam->v4l2_dev.dev, "ISR: [%d] Dropping
>>> frame, buffer not available at FS\n",
>>> + i);
>>> + /*
>>> + * Set the next frame output to go to a dummy frame
>>> + * if we have not managed to obtain another frame
>>> + * from the queue.
>>> + */
>>> + unicam_schedule_dummy_buffer(&unicam->node[i]);
>>> + }
>>> +
>>> + unicam_queue_event_sof(unicam);
>>> + }
>>> +
>>> + /*
>>> + * Cannot swap buffer at frame end, there may be a race condition
>>> + * where the HW does not actually swap it if the new frame has
>>> + * already started.
>>> + */
>>> + if (ista & (UNICAM_FSI | UNICAM_LCI) && !fe) {
>>> + for (i = 0; i < ARRAY_SIZE(unicam->node); i++) {
>>> + if (!unicam->node[i].streaming)
>>> + continue;
>>> +
>>> + spin_lock(&unicam->node[i].dma_queue_lock);
>>> + if (!list_empty(&unicam->node[i].dma_queue) &&
>>> + !unicam->node[i].next_frm)
>>> + unicam_schedule_next_buffer(&unicam->node[i]);
>>> + spin_unlock(&unicam->node[i].dma_queue_lock);
>>> + }
>>> + }
>>> +
>>> + if (reg_read(unicam, UNICAM_ICTL) & UNICAM_FCM) {
>>> + /* Switch out of trigger mode if selected */
>>> + reg_write_field(unicam, UNICAM_ICTL, 1, UNICAM_TFC);
>>> + reg_write_field(unicam, UNICAM_ICTL, 0, UNICAM_FCM);
>>> + }
>>> + return IRQ_HANDLED;
>>> +}
>>> +
>>> +/* V4L2 Common IOCTLs */
>>> +static int unicam_querycap(struct file *file, void *priv,
>>> + struct v4l2_capability *cap)
>>> +{
>>> + struct unicam_node *node = video_drvdata(file);
>>> + struct unicam_device *dev = node->dev;
>>> +
>>> + strscpy(cap->driver, UNICAM_MODULE_NAME, sizeof(cap->driver));
>>> + strscpy(cap->card, UNICAM_MODULE_NAME, sizeof(cap->card));
>>> +
>>> + snprintf(cap->bus_info, sizeof(cap->bus_info),
>>> + "platform:%s", dev_name(&dev->pdev->dev));
>>> +
>>> + cap->capabilities |= V4L2_CAP_VIDEO_CAPTURE |
>>> V4L2_CAP_META_CAPTURE;
>>> +
>>> + return 0;
>>> +}
>>> +
>>> +static int unicam_log_status(struct file *file, void *fh)
>>> +{
>>> + struct unicam_node *node = video_drvdata(file);
>>> + struct unicam_device *dev = node->dev;
>>> + u32 reg;
>>> +
>>> + /* status for sub devices */
>>> + v4l2_device_call_all(&dev->v4l2_dev, 0, core, log_status);
>>
>> I'm tempted to drop this, bit I won't insist.
>>
>>> +
>>> + dev_info(dev->v4l2_dev.dev, "-----Receiver status-----\n");
>>> + dev_info(dev->v4l2_dev.dev, "V4L2 width/height: %ux%u\n",
>>> + node->v_fmt.fmt.pix.width, node->v_fmt.fmt.pix.height);
>>> + dev_info(dev->v4l2_dev.dev, "Mediabus format: %08x\n",
>>> node->fmt->code);
>>> + dev_info(dev->v4l2_dev.dev, "V4L2 format: %08x\n",
>>> + node->v_fmt.fmt.pix.pixelformat);
>>> + reg = reg_read(dev, UNICAM_IPIPE);
>>> + dev_info(dev->v4l2_dev.dev, "Unpacking/packing: %u / %u\n",
>>> + get_field(reg, UNICAM_PUM_MASK),
>>> + get_field(reg, UNICAM_PPM_MASK));
>>> + dev_info(dev->v4l2_dev.dev, "----Live data----\n");
>>> + dev_info(dev->v4l2_dev.dev, "Programmed stride: %4u\n",
>>> + reg_read(dev, UNICAM_IBLS));
>>> + dev_info(dev->v4l2_dev.dev, "Detected resolution: %ux%u\n",
>>> + reg_read(dev, UNICAM_IHSTA),
>>> + reg_read(dev, UNICAM_IVSTA));
>>> + dev_info(dev->v4l2_dev.dev, "Write pointer: %08x\n",
>>> + reg_read(dev, UNICAM_IBWP));
>>> +
>>> + return 0;
>>> +}
>>> +
>>> +static int unicam_subscribe_event(struct v4l2_fh *fh,
>>> + const struct v4l2_event_subscription *sub)
>>> +{
>>> + switch (sub->type) {
>>> + case V4L2_EVENT_FRAME_SYNC:
>>> + return v4l2_event_subscribe(fh, sub, 2, NULL);
>>> + case V4L2_EVENT_SOURCE_CHANGE:
>>> + return v4l2_event_subscribe(fh, sub, 4, NULL);
>>
>> The drive doesn't generate V4L2_EVENT_SOURCE_CHANGE events, drop it.
>>
>>> + }
>>> +
>>> + return v4l2_ctrl_subscribe_event(fh, sub);
>>> +}
>>> +
>>> +static void unicam_notify(struct v4l2_subdev *sd,
>>> + unsigned int notification, void *arg)
>>> +{
>>> + struct unicam_device *dev = to_unicam_device(sd->v4l2_dev);
>>
>> Use sd_to_unicam_device().
>>
>>> +
>>> + switch (notification) {
>>> + case V4L2_DEVICE_NOTIFY_EVENT:
>>> + v4l2_event_queue(&dev->node[IMAGE_NODE].video_dev, arg);
>>> + break;
>>> + default:
>>> + break;
>>> + }
>>> +}
>>
>> Not needed, drop it.
>>
>>> +
>>> +/* V4L2 Media Controller Centric IOCTLs */
>>> +
>>> +static int unicam_mc_enum_fmt_vid_cap(struct file *file, void *priv,
>>> + struct v4l2_fmtdesc *f)
>>
>> Drop "mc" in all function names as that's the only option now.
>>
>>> +{
>>> + int i, j;
>>
>> Never negative, can be unsigned int.
>>
>> j is not a loop counter but the format index, I'd rename it to index to
>> make it clearer.
>>
>>> +
>>> + for (i = 0, j = 0; i < ARRAY_SIZE(formats); i++) {
>>> + if (f->mbus_code && formats[i].code != f->mbus_code)
>>> + continue;
>>> + if (formats[i].mc_skip || formats[i].metadata_fmt)
>>> + continue;
>>> +
>>> + if (formats[i].fourcc) {
>>
>> All entries have a fourcc so you can drop this check.
>>
>>> + if (j == f->index) {
>>> + f->pixelformat = formats[i].fourcc;
>>> + f->type = V4L2_BUF_TYPE_VIDEO_CAPTURE;
>>
>> Not needed, the type is an input parameter. Same below.
>>
>>> + return 0;
>>> + }
>>> + j++;
>>> + }
>>> + if (formats[i].repacked_fourcc) {
>>> + if (j == f->index) {
>>> + f->pixelformat = formats[i].repacked_fourcc;
>>> + f->type = V4L2_BUF_TYPE_VIDEO_CAPTURE;
>>> + return 0;
>>> + }
>>> + j++;
>>> + }
>>> + }
>>> +
>>> + return -EINVAL;
>>> +}
>>> +
>>> +static int unicam_mc_g_fmt_vid_cap(struct file *file, void *priv,
>>> + struct v4l2_format *f)
>>> +{
>>> + struct unicam_node *node = video_drvdata(file);
>>> +
>>> + if (!is_image_node(node))
>>> + return -EINVAL;
>>
>> Can this happen ?
>>
>>> +
>>> + *f = node->v_fmt;
>>> +
>>> + return 0;
>>> +}
>>> +
>>> +static void unicam_mc_try_fmt(struct unicam_node *node, struct
>>> v4l2_format *f,
>>> + const struct unicam_fmt **ret_fmt)
>>> +{
>>> + struct v4l2_pix_format *v4l2_format = &f->fmt.pix;
>>> + struct unicam_device *dev = node->dev;
>>> + const struct unicam_fmt *fmt;
>>> + int is_rgb;
>>
>> bool.
>>
>>> +
>>> + /*
>>> + * Default to the first format if the requested pixel format
>>> code isn't
>>> + * supported.
>>> + */
>>> + fmt = find_format_by_fourcc(v4l2_format->pixelformat);
>>> + if (!fmt) {
>>> + fmt = &formats[0];
>>> + v4l2_format->pixelformat = fmt->fourcc;
>>> + }
>>> +
>>> + unicam_calc_format_size_bpl(dev, fmt, f);
>>> +
>>> + if (v4l2_format->field == V4L2_FIELD_ANY)
>>> + v4l2_format->field = V4L2_FIELD_NONE;
>>> +
>>> + if (v4l2_format->colorspace >= MAX_COLORSPACE ||
>>> + !(fmt->valid_colorspaces & (1 << v4l2_format->colorspace))) {
>>> + v4l2_format->colorspace = __ffs(fmt->valid_colorspaces);
>>> +
>>> + v4l2_format->xfer_func =
>>> + V4L2_MAP_XFER_FUNC_DEFAULT(v4l2_format->colorspace);
>>> + v4l2_format->ycbcr_enc =
>>> + V4L2_MAP_YCBCR_ENC_DEFAULT(v4l2_format->colorspace);
>>> + is_rgb = v4l2_format->colorspace == V4L2_COLORSPACE_SRGB;
>>> + v4l2_format->quantization =
>>> + V4L2_MAP_QUANTIZATION_DEFAULT(is_rgb,
>>> + v4l2_format->colorspace,
>>> + v4l2_format->ycbcr_enc);
>>> + }
>>
>> Drop this, we can accept any colorspace produced by the sensor.
>>
>>> +
>>> + if (ret_fmt)
>>> + *ret_fmt = fmt;
>>
>> How about returning the format from the function instead ?
>>
>>> +
>>> + dev_dbg(dev->v4l2_dev.dev, "%s: %08x %ux%u (bytesperline %u
>>> sizeimage %u)\n",
>>> + __func__, v4l2_format->pixelformat,
>>> + v4l2_format->width, v4l2_format->height,
>>> + v4l2_format->bytesperline, v4l2_format->sizeimage);
>>> +}
>>> +
>>> +static int unicam_mc_try_fmt_vid_cap(struct file *file, void *priv,
>>> + struct v4l2_format *f)
>>> +{
>>> + struct unicam_node *node = video_drvdata(file);
>>> +
>>> + unicam_mc_try_fmt(node, f, NULL);
>>> + return 0;
>>> +}
>>> +
>>> +static int unicam_mc_s_fmt_vid_cap(struct file *file, void *priv,
>>> + struct v4l2_format *f)
>>> +{
>>> + struct unicam_node *node = video_drvdata(file);
>>> + struct unicam_device *dev = node->dev;
>>> + const struct unicam_fmt *fmt;
>>> +
>>> + if (vb2_is_busy(&node->buffer_queue)) {
>>> + dev_dbg(dev->v4l2_dev.dev, "%s device busy\n", __func__);
>>> + return -EBUSY;
>>> + }
>>> +
>>> + unicam_mc_try_fmt(node, f, &fmt);
>>> +
>>> + node->v_fmt = *f;
>>> + node->fmt = fmt;
>>> +
>>> + return 0;
>>> +}
>>> +
>>> +static int unicam_mc_enum_framesizes(struct file *file, void *fh,
>>> + struct v4l2_frmsizeenum *fsize)
>>> +{
>>> + struct unicam_node *node = video_drvdata(file);
>>> + struct unicam_device *dev = node->dev;
>>> +
>>> + if (fsize->index > 0)
>>> + return -EINVAL;
>>> +
>>> + if (!find_format_by_fourcc(fsize->pixel_format)) {
>>> + dev_dbg(dev->v4l2_dev.dev, "Invalid pixel format 0x%08x\n",
>>> + fsize->pixel_format);
>>
>> I'd drop this message.
>>
>>> + return -EINVAL;
>>> + }
>>> +
>>> + fsize->type = V4L2_FRMSIZE_TYPE_STEPWISE;
>>> + fsize->stepwise.min_width = MIN_WIDTH;
>>> + fsize->stepwise.max_width = MAX_WIDTH;
>>> + fsize->stepwise.step_width = 1;
>>> + fsize->stepwise.min_height = MIN_HEIGHT;
>>> + fsize->stepwise.max_height = MAX_HEIGHT;
>>> + fsize->stepwise.step_height = 1;
>>
>> Is this valid for the metadata node too ?
>>
>>> +
>>> + return 0;
>>> +}
>>> +
>>> +static int unicam_mc_enum_fmt_meta_cap(struct file *file, void *priv,
>>> + struct v4l2_fmtdesc *f)
>>> +{
>>> + struct unicam_node *node = video_drvdata(file);
>>> + int i, j;
>>> +
>>> + if (!is_metadata_node(node))
>>> + return -EINVAL;
>>
>> Same comments as for video format enumeration.
>>
>>> +
>>> + for (i = 0, j = 0; i < ARRAY_SIZE(formats); i++) {
>>> + if (f->mbus_code && formats[i].code != f->mbus_code)
>>> + continue;
>>> + if (!formats[i].metadata_fmt)
>>> + continue;
>>> +
>>> + if (formats[i].fourcc) {
>>> + if (j == f->index) {
>>> + f->pixelformat = formats[i].fourcc;
>>> + f->type = V4L2_BUF_TYPE_META_CAPTURE;
>>> + return 0;
>>> + }
>>> + j++;
>>> + }
>>> + }
>>> +
>>> + return -EINVAL;
>>> +}
>>> +
>>> +static int unicam_mc_g_fmt_meta_cap(struct file *file, void *priv,
>>> + struct v4l2_format *f)
>>> +{
>>> + struct unicam_node *node = video_drvdata(file);
>>> +
>>> + if (!is_metadata_node(node))
>>> + return -EINVAL;
>>
>> Can this happen ? Same below.
>>
>>> +
>>> + *f = node->v_fmt;
>>> +
>>> + return 0;
>>> +}
>>> +
>>> +static int unicam_mc_try_fmt_meta_cap(struct file *file, void *priv,
>>> + struct v4l2_format *f)
>>> +{
>>> + struct unicam_node *node = video_drvdata(file);
>>> +
>>> + if (!is_metadata_node(node))
>>> + return -EINVAL;
>>> +
>>> + f->fmt.meta.dataformat = V4L2_META_FMT_8;
>>> +
>>> + return 0;
>>> +}
>>> +
>>> +static int unicam_mc_s_fmt_meta_cap(struct file *file, void *priv,
>>> + struct v4l2_format *f)
>>> +{
>>> + struct unicam_node *node = video_drvdata(file);
>>> +
>>> + if (!is_metadata_node(node))
>>> + return -EINVAL;
>>> +
>>> + unicam_mc_try_fmt_meta_cap(file, priv, f);
>>> +
>>> + node->v_fmt = *f;
>>> +
>>> + return 0;
>>> +}
>>> +
>>> +static const struct v4l2_ioctl_ops unicam_mc_ioctl_ops = {
>>> + .vidioc_querycap = unicam_querycap,
>>> + .vidioc_enum_fmt_vid_cap = unicam_mc_enum_fmt_vid_cap,
>>> + .vidioc_g_fmt_vid_cap = unicam_mc_g_fmt_vid_cap,
>>> + .vidioc_try_fmt_vid_cap = unicam_mc_try_fmt_vid_cap,
>>> + .vidioc_s_fmt_vid_cap = unicam_mc_s_fmt_vid_cap,
>>> +
>>> + .vidioc_enum_fmt_meta_cap = unicam_mc_enum_fmt_meta_cap,
>>> + .vidioc_g_fmt_meta_cap = unicam_mc_g_fmt_meta_cap,
>>> + .vidioc_try_fmt_meta_cap = unicam_mc_try_fmt_meta_cap,
>>> + .vidioc_s_fmt_meta_cap = unicam_mc_s_fmt_meta_cap,
>>> +
>>> + .vidioc_enum_framesizes = unicam_mc_enum_framesizes,
>>> + .vidioc_reqbufs = vb2_ioctl_reqbufs,
>>> + .vidioc_create_bufs = vb2_ioctl_create_bufs,
>>> + .vidioc_prepare_buf = vb2_ioctl_prepare_buf,
>>> + .vidioc_querybuf = vb2_ioctl_querybuf,
>>> + .vidioc_qbuf = vb2_ioctl_qbuf,
>>> + .vidioc_dqbuf = vb2_ioctl_dqbuf,
>>> + .vidioc_expbuf = vb2_ioctl_expbuf,
>>> + .vidioc_streamon = vb2_ioctl_streamon,
>>> + .vidioc_streamoff = vb2_ioctl_streamoff,
>>> +
>>> + .vidioc_log_status = unicam_log_status,
>>> + .vidioc_subscribe_event = unicam_subscribe_event,
>>> + .vidioc_unsubscribe_event = v4l2_event_unsubscribe,
>>> +};
>>> +
>>> +static const struct media_entity_operations unicam_mc_entity_ops = {
>>> + .link_validate = v4l2_subdev_link_validate,
>>> + .has_route = v4l2_subdev_has_route,
>>> +};
>>
>> I don't think this is needed for the video nodes.
>>
>>> +
>>> +/* videobuf2 Operations */
>>> +
>>> +static int unicam_queue_setup(struct vb2_queue *vq,
>>> + unsigned int *nbuffers,
>>> + unsigned int *nplanes,
>>> + unsigned int sizes[],
>>> + struct device *alloc_devs[])
>>> +{
>>> + struct unicam_node *node = vb2_get_drv_priv(vq);
>>> + struct unicam_device *dev = node->dev;
>>> + unsigned int size = is_image_node(node) ?
>>> + node->v_fmt.fmt.pix.sizeimage :
>>> + node->v_fmt.fmt.meta.buffersize;
>>> +
>>> + if (vq->num_buffers + *nbuffers < 3)
>>> + *nbuffers = 3 - vq->num_buffers;
>>> +
>>> + if (*nplanes) {
>>> + if (sizes[0] < size) {
>>> + dev_err(dev->v4l2_dev.dev, "sizes[0] %i < size %u\n",
>>> sizes[0],
>>> + size);
>>
>> This can be a debug message, let's not give a way to unprivileged
>> userspace to flood the kernel log.
>>
>>> + return -EINVAL;
>>> + }
>>> + size = sizes[0];
>>> + }
>>> +
>>> + *nplanes = 1;
>>> + sizes[0] = size;
>>> +
>>> + return 0;
>>> +}
>>> +
>>> +static int unicam_buffer_prepare(struct vb2_buffer *vb)
>>> +{
>>> + struct unicam_node *node = vb2_get_drv_priv(vb->vb2_queue);
>>> + struct unicam_device *dev = node->dev;
>>> + struct unicam_buffer *buf = to_unicam_buffer(vb);
>>> + unsigned long size;
>>> +
>>> + if (WARN_ON(!node->fmt))
>>> + return -EINVAL;
>>> +
>>> + size = is_image_node(node) ? node->v_fmt.fmt.pix.sizeimage :
>>> + node->v_fmt.fmt.meta.buffersize;
>>> + if (vb2_plane_size(vb, 0) < size) {
>>> + dev_err(dev->v4l2_dev.dev, "data will not fit into plane
>>> (%lu < %lu)\n",
>>> + vb2_plane_size(vb, 0), size);
>>
>> Debug message here too.
>>
>>> + return -EINVAL;
>>> + }
>>> +
>>> + vb2_set_plane_payload(&buf->vb.vb2_buf, 0, size);
>>> + return 0;
>>> +}
>>> +
>>> +static void unicam_buffer_queue(struct vb2_buffer *vb)
>>> +{
>>> + struct unicam_node *node = vb2_get_drv_priv(vb->vb2_queue);
>>> + struct unicam_buffer *buf = to_unicam_buffer(vb);
>>> + unsigned long flags;
>>> +
>>> + spin_lock_irqsave(&node->dma_queue_lock, flags);
>>
>> Small optimization, as this is guaranteed to be called with interrupts
>> enabled, you can use the spin_lock_irq() variant.
>>
>>> + list_add_tail(&buf->list, &node->dma_queue);
>>> + spin_unlock_irqrestore(&node->dma_queue_lock, flags);
>>> +}
>>> +
>>> +static void unicam_set_packing_config(struct unicam_device *dev)
>>> +{
>>> + u32 pack, unpack;
>>> + u32 val;
>>> +
>>> + if (dev->node[IMAGE_NODE].v_fmt.fmt.pix.pixelformat ==
>>> + dev->node[IMAGE_NODE].fmt->fourcc) {
>>> + unpack = UNICAM_PUM_NONE;
>>> + pack = UNICAM_PPM_NONE;
>>> + } else {
>>> + switch (dev->node[IMAGE_NODE].fmt->depth) {
>>> + case 8:
>>> + unpack = UNICAM_PUM_UNPACK8;
>>> + break;
>>> + case 10:
>>> + unpack = UNICAM_PUM_UNPACK10;
>>> + break;
>>> + case 12:
>>> + unpack = UNICAM_PUM_UNPACK12;
>>> + break;
>>> + case 14:
>>> + unpack = UNICAM_PUM_UNPACK14;
>>> + break;
>>> + case 16:
>>> + unpack = UNICAM_PUM_UNPACK16;
>>> + break;
>>> + default:
>>> + unpack = UNICAM_PUM_NONE;
>>> + break;
>>> + }
>>> +
>>> + /* Repacking is always to 16bpp */
>>> + pack = UNICAM_PPM_PACK16;
>>> + }
>>> +
>>> + val = 0;
>>> + set_field(&val, unpack, UNICAM_PUM_MASK);
>>> + set_field(&val, pack, UNICAM_PPM_MASK);
>>> + reg_write(dev, UNICAM_IPIPE, val);
>>> +}
>>> +
>>> +static void unicam_cfg_image_id(struct unicam_device *dev)
>>> +{
>>> + if (dev->bus_type == V4L2_MBUS_CSI2_DPHY) {
>>> + /* CSI2 mode, hardcode VC 0 for now. */
>>> + reg_write(dev, UNICAM_IDI0,
>>> + (0 << 6) | dev->node[IMAGE_NODE].fmt->csi_dt);
>>> + } else {
>>> + /* CCP2 mode */
>>> + reg_write(dev, UNICAM_IDI0,
>>> + 0x80 | dev->node[IMAGE_NODE].fmt->csi_dt);
>>> + }
>>> +}
>>> +
>>> +static void unicam_enable_ed(struct unicam_device *dev)
>>> +{
>>> + u32 val = reg_read(dev, UNICAM_DCS);
>>> +
>>> + set_field(&val, 2, UNICAM_EDL_MASK);
>>> + /* Do not wrap at the end of the embedded data buffer */
>>> + set_field(&val, 0, UNICAM_DBOB);
>>> +
>>> + reg_write(dev, UNICAM_DCS, val);
>>> +}
>>> +
>>> +static void unicam_start_rx(struct unicam_device *dev, dma_addr_t
>>> *addr)
>>> +{
>>> + int line_int_freq = dev->node[IMAGE_NODE].v_fmt.fmt.pix.height
>>> >> 2;
>>> + unsigned int size, i;
>>> + u32 val;
>>> +
>>> + if (line_int_freq < 128)
>>> + line_int_freq = 128;
>>> +
>>> + /* Enable lane clocks */
>>> + val = 1;
>>> + for (i = 0; i < dev->active_data_lanes; i++)
>>> + val = val << 2 | 1;
>>> + clk_write(dev, val);
>>> +
>>> + /* Basic init */
>>> + reg_write(dev, UNICAM_CTRL, UNICAM_MEM);
>>> +
>>> + /* Enable analogue control, and leave in reset. */
>>> + val = UNICAM_AR;
>>> + set_field(&val, 7, UNICAM_CTATADJ_MASK);
>>> + set_field(&val, 7, UNICAM_PTATADJ_MASK);
>>> + reg_write(dev, UNICAM_ANA, val);
>>> + usleep_range(1000, 2000);
>>> +
>>> + /* Come out of reset */
>>> + reg_write_field(dev, UNICAM_ANA, 0, UNICAM_AR);
>>> +
>>> + /* Peripheral reset */
>>> + reg_write_field(dev, UNICAM_CTRL, 1, UNICAM_CPR);
>>> + reg_write_field(dev, UNICAM_CTRL, 0, UNICAM_CPR);
>>> +
>>> + reg_write_field(dev, UNICAM_CTRL, 0, UNICAM_CPE);
>>> +
>>> + /* Enable Rx control. */
>>> + val = reg_read(dev, UNICAM_CTRL);
>>> + if (dev->bus_type == V4L2_MBUS_CSI2_DPHY) {
>>> + set_field(&val, UNICAM_CPM_CSI2, UNICAM_CPM_MASK);
>>> + set_field(&val, UNICAM_DCM_STROBE, UNICAM_DCM_MASK);
>>> + } else {
>>> + set_field(&val, UNICAM_CPM_CCP2, UNICAM_CPM_MASK);
>>> + set_field(&val, dev->bus_flags, UNICAM_DCM_MASK);
>>> + }
>>> + /* Packet framer timeout */
>>> + set_field(&val, 0xf, UNICAM_PFT_MASK);
>>> + set_field(&val, 128, UNICAM_OET_MASK);
>>> + reg_write(dev, UNICAM_CTRL, val);
>>> +
>>> + reg_write(dev, UNICAM_IHWIN, 0);
>>> + reg_write(dev, UNICAM_IVWIN, 0);
>>> +
>>> + /* AXI bus access QoS setup */
>>> + val = reg_read(dev, UNICAM_PRI);
>>> + set_field(&val, 0, UNICAM_BL_MASK);
>>> + set_field(&val, 0, UNICAM_BS_MASK);
>>> + set_field(&val, 0xe, UNICAM_PP_MASK);
>>> + set_field(&val, 8, UNICAM_NP_MASK);
>>> + set_field(&val, 2, UNICAM_PT_MASK);
>>> + set_field(&val, 1, UNICAM_PE);
>>> + reg_write(dev, UNICAM_PRI, val);
>>> +
>>> + reg_write_field(dev, UNICAM_ANA, 0, UNICAM_DDL);
>>> +
>>> + /* Always start in trigger frame capture mode (UNICAM_FCM set) */
>>> + val = UNICAM_FSIE | UNICAM_FEIE | UNICAM_FCM | UNICAM_IBOB;
>>> + set_field(&val, line_int_freq, UNICAM_LCIE_MASK);
>>> + reg_write(dev, UNICAM_ICTL, val);
>>> + reg_write(dev, UNICAM_STA, UNICAM_STA_MASK_ALL);
>>> + reg_write(dev, UNICAM_ISTA, UNICAM_ISTA_MASK_ALL);
>>> +
>>> + /* tclk_term_en */
>>> + reg_write_field(dev, UNICAM_CLT, 2, UNICAM_CLT1_MASK);
>>> + /* tclk_settle */
>>> + reg_write_field(dev, UNICAM_CLT, 6, UNICAM_CLT2_MASK);
>>> + /* td_term_en */
>>> + reg_write_field(dev, UNICAM_DLT, 2, UNICAM_DLT1_MASK);
>>> + /* ths_settle */
>>> + reg_write_field(dev, UNICAM_DLT, 6, UNICAM_DLT2_MASK);
>>> + /* trx_enable */
>>> + reg_write_field(dev, UNICAM_DLT, 0, UNICAM_DLT3_MASK);
>>> +
>>> + reg_write_field(dev, UNICAM_CTRL, 0, UNICAM_SOE);
>>> +
>>> + /* Packet compare setup - required to avoid missing frame ends */
>>> + val = 0;
>>> + set_field(&val, 1, UNICAM_PCE);
>>> + set_field(&val, 1, UNICAM_GI);
>>> + set_field(&val, 1, UNICAM_CPH);
>>> + set_field(&val, 0, UNICAM_PCVC_MASK);
>>> + set_field(&val, 1, UNICAM_PCDT_MASK);
>>> + reg_write(dev, UNICAM_CMP0, val);
>>> +
>>> + /* Enable clock lane and set up terminations */
>>> + val = 0;
>>> + if (dev->bus_type == V4L2_MBUS_CSI2_DPHY) {
>>> + /* CSI2 */
>>> + set_field(&val, 1, UNICAM_CLE);
>>> + set_field(&val, 1, UNICAM_CLLPE);
>>> + if (dev->bus_flags & V4L2_MBUS_CSI2_CONTINUOUS_CLOCK) {
>>> + set_field(&val, 1, UNICAM_CLTRE);
>>> + set_field(&val, 1, UNICAM_CLHSE);
>>> + }
>>> + } else {
>>> + /* CCP2 */
>>> + set_field(&val, 1, UNICAM_CLE);
>>> + set_field(&val, 1, UNICAM_CLHSE);
>>> + set_field(&val, 1, UNICAM_CLTRE);
>>> + }
>>> + reg_write(dev, UNICAM_CLK, val);
>>> +
>>> + /*
>>> + * Enable required data lanes with appropriate terminations.
>>> + * The same value needs to be written to UNICAM_DATn registers for
>>> + * the active lanes, and 0 for inactive ones.
>>> + */
>>> + val = 0;
>>> + if (dev->bus_type == V4L2_MBUS_CSI2_DPHY) {
>>> + /* CSI2 */
>>> + set_field(&val, 1, UNICAM_DLE);
>>> + set_field(&val, 1, UNICAM_DLLPE);
>>> + if (dev->bus_flags & V4L2_MBUS_CSI2_CONTINUOUS_CLOCK) {
>>> + set_field(&val, 1, UNICAM_DLTRE);
>>> + set_field(&val, 1, UNICAM_DLHSE);
>>> + }
>>> + } else {
>>> + /* CCP2 */
>>> + set_field(&val, 1, UNICAM_DLE);
>>> + set_field(&val, 1, UNICAM_DLHSE);
>>> + set_field(&val, 1, UNICAM_DLTRE);
>>> + }
>>> + reg_write(dev, UNICAM_DAT0, val);
>>> +
>>> + if (dev->active_data_lanes == 1)
>>> + val = 0;
>>> + reg_write(dev, UNICAM_DAT1, val);
>>> +
>>> + if (dev->max_data_lanes > 2) {
>>> + /*
>>> + * Registers UNICAM_DAT2 and UNICAM_DAT3 only valid if the
>>> + * instance supports more than 2 data lanes.
>>> + */
>>> + if (dev->active_data_lanes == 2)
>>> + val = 0;
>>> + reg_write(dev, UNICAM_DAT2, val);
>>> +
>>> + if (dev->active_data_lanes == 3)
>>> + val = 0;
>>> + reg_write(dev, UNICAM_DAT3, val);
>>> + }
>>> +
>>> + reg_write(dev, UNICAM_IBLS,
>>> + dev->node[IMAGE_NODE].v_fmt.fmt.pix.bytesperline);
>>> + size = dev->node[IMAGE_NODE].v_fmt.fmt.pix.sizeimage;
>>> + unicam_wr_dma_addr(dev, addr[IMAGE_NODE], size, IMAGE_NODE);
>>> + unicam_set_packing_config(dev);
>>> + unicam_cfg_image_id(dev);
>>> +
>>> + val = reg_read(dev, UNICAM_MISC);
>>> + set_field(&val, 1, UNICAM_FL0);
>>> + set_field(&val, 1, UNICAM_FL1);
>>> + reg_write(dev, UNICAM_MISC, val);
>>> +
>>> + if (dev->node[METADATA_NODE].streaming &&
>>> dev->sensor_embedded_data) {
>>> + dev_dbg(dev->v4l2_dev.dev, "enable metadata dma\n");
>>> + size = dev->node[METADATA_NODE].v_fmt.fmt.meta.buffersize;
>>> + unicam_enable_ed(dev);
>>> + unicam_wr_dma_addr(dev, addr[METADATA_NODE], size,
>>> METADATA_NODE);
>>> + }
>>> +
>>> + /* Enable peripheral */
>>> + reg_write_field(dev, UNICAM_CTRL, 1, UNICAM_CPE);
>>> +
>>> + /* Load image pointers */
>>> + reg_write_field(dev, UNICAM_ICTL, 1, UNICAM_LIP_MASK);
>>> +
>>> + /* Load embedded data buffer pointers if needed */
>>> + if (dev->node[METADATA_NODE].streaming &&
>>> dev->sensor_embedded_data)
>>> + reg_write_field(dev, UNICAM_DCS, 1, UNICAM_LDP);
>>> +
>>> + /*
>>> + * Enable trigger only for the first frame to
>>> + * sync correctly to the FS from the source.
>>> + */
>>> + reg_write_field(dev, UNICAM_ICTL, 1, UNICAM_TFC);
>>> +}
>>> +
>>> +static void unicam_disable(struct unicam_device *dev)
>>> +{
>>> + /* Analogue lane control disable */
>>> + reg_write_field(dev, UNICAM_ANA, 1, UNICAM_DDL);
>>> +
>>> + /* Stop the output engine */
>>> + reg_write_field(dev, UNICAM_CTRL, 1, UNICAM_SOE);
>>> +
>>> + /* Disable the data lanes. */
>>> + reg_write(dev, UNICAM_DAT0, 0);
>>> + reg_write(dev, UNICAM_DAT1, 0);
>>> +
>>> + if (dev->max_data_lanes > 2) {
>>> + reg_write(dev, UNICAM_DAT2, 0);
>>> + reg_write(dev, UNICAM_DAT3, 0);
>>> + }
>>> +
>>> + /* Peripheral reset */
>>> + reg_write_field(dev, UNICAM_CTRL, 1, UNICAM_CPR);
>>> + usleep_range(50, 100);
>>> + reg_write_field(dev, UNICAM_CTRL, 0, UNICAM_CPR);
>>> +
>>> + /* Disable peripheral */
>>> + reg_write_field(dev, UNICAM_CTRL, 0, UNICAM_CPE);
>>> +
>>> + /* Clear ED setup */
>>> + reg_write(dev, UNICAM_DCS, 0);
>>> +
>>> + /* Disable all lane clocks */
>>> + clk_write(dev, 0);
>>> +}
>>> +
>>> +static void unicam_return_buffers(struct unicam_node *node,
>>> + enum vb2_buffer_state state)
>>> +{
>>> + struct unicam_buffer *buf, *tmp;
>>> + unsigned long flags;
>>> +
>>> + spin_lock_irqsave(&node->dma_queue_lock, flags);
>>
>> Same here, spin_lock_irq() would do. But isn't this function meant to be
>> called with the hardware stopped ? Shouldn't it then be safe to drop the
>> spinlock completely ?
>>
>>> + list_for_each_entry_safe(buf, tmp, &node->dma_queue, list) {
>>> + list_del(&buf->list);
>>> + vb2_buffer_done(&buf->vb.vb2_buf, state);
>>> + }
>>> +
>>> + if (node->cur_frm)
>>> + vb2_buffer_done(&node->cur_frm->vb.vb2_buf,
>>> + state);
>>> + if (node->next_frm && node->cur_frm != node->next_frm)
>>> + vb2_buffer_done(&node->next_frm->vb.vb2_buf,
>>> + state);
>>> +
>>> + node->cur_frm = NULL;
>>> + node->next_frm = NULL;
>>> + spin_unlock_irqrestore(&node->dma_queue_lock, flags);
>>> +}
>>> +
>>> +static int unicam_video_check_format(struct unicam_node *node)
>>> +{
>>> + const struct v4l2_mbus_framefmt *format;
>>> + struct media_pad *remote_pad;
>>> + struct v4l2_subdev_state *state;
>>> + int ret = 0;
>>> +
>>> + remote_pad = media_entity_remote_pad(&node->pad);
>>> + if (!remote_pad)
>>> + return -ENODEV;
>>
>> This could be done in unicam_async_complete() and cached in the
>> unicam_device structure, the same way we cache the sensor subdev
>> pointer.
>>
>>> +
>>> + state = v4l2_subdev_lock_active_state(node->dev->sensor);
>>
>> That's not correct, the video node is connected to the subdev, not to
>> the sensor.
>>
>>> +
>>> + format = v4l2_state_get_stream_format(state,
>>> + remote_pad->index, 0);
>>
>> This holds on a single line.
>>
>>> + if (!format) {
>>> + ret = -EINVAL;
>>> + goto out;
>>> + }
>>> +
>>> + if (node->fmt->code != format->code ||
>>> + node->v_fmt.fmt.pix.height != format->height ||
>>> + node->v_fmt.fmt.pix.width != format->width ||
>>> + node->v_fmt.fmt.pix.field != format->field) {
>>
>> A debug message that prints the configuration on both sides could be
>> useful for debugging here.
>>
>>> + ret = -EPIPE;
>>> + goto out;
>>> + }
>>> +
>>> +out:
>>> + v4l2_subdev_unlock_state(state);
>>> +
>>> + return ret;
>>> +}
>>> +
>>> +static int unicam_start_streaming(struct vb2_queue *vq, unsigned int
>>> count)
>>> +{
>>> + struct unicam_node *node = vb2_get_drv_priv(vq);
>>> + struct unicam_device *dev = node->dev;
>>> + dma_addr_t buffer_addr[MAX_NODES] = { 0 };
>>> + struct unicam_buffer *buf;
>>> + unsigned long flags;
>>> + unsigned int i;
>>> + int ret;
>>> +
>>> + node->streaming = true;
>>
>> Locking seems to be completely missing from the driver. I'll comment
>> more on that on v3, it will be easier after all the other cleanups.
>>
>>> +
>>> + dev->sequence = 0;
>>> + ret = unicam_runtime_get(dev);
>>> + if (ret < 0) {
>>> + dev_dbg(dev->v4l2_dev.dev, "unicam_runtime_get failed\n");
>>
>> This is really not supposed to happen, dev_err() is appropriate.
>>
>>> + goto err_streaming;
>>> + }
>>> +
>>> + ret = media_pipeline_start(node->video_dev.entity.pads,
>>> &node->pipe);
>>> + if (ret < 0) {
>>> + dev_err(dev->v4l2_dev.dev, "Failed to start media pipeline:
>>> %d\n", ret);
>>
>> dev_dbg() here too.
>>
>>> + goto err_pm_put;
>>> + }
>>> +
>>> + dev->active_data_lanes = dev->max_data_lanes;
>>
>> You need to get the number of data lanes from the sensor subdev and
>> validate it to ensure it doesn't exceed the maximum.
>>
>>> +
>>> + unicam_video_check_format(node);
>>
>> Where's the return value check ?
>>
>>> +
>>> + dev_dbg(dev->v4l2_dev.dev, "Running with %u data lanes\n",
>>> + dev->active_data_lanes);
>>> +
>>> + ret = clk_set_min_rate(dev->vpu_clock, MIN_VPU_CLOCK_RATE);
>>> + if (ret) {
>>> + dev_err(dev->v4l2_dev.dev, "failed to set up VPU clock\n");
>>> + goto error_pipeline;
>>> + }
>>> +
>>> + ret = clk_prepare_enable(dev->vpu_clock);
>>> + if (ret) {
>>> + dev_err(dev->v4l2_dev.dev, "Failed to enable VPU clock:
>>> %d\n", ret);
>>> + goto error_pipeline;
>>> + }
>>> +
>>> + ret = clk_set_rate(dev->clock, 100 * 1000 * 1000);
>>> + if (ret) {
>>> + dev_err(dev->v4l2_dev.dev, "failed to set up CSI clock\n");
>>> + goto err_vpu_clock;
>>> + }
>>> +
>>> + ret = clk_prepare_enable(dev->clock);
>>> + if (ret) {
>>> + dev_err(dev->v4l2_dev.dev, "Failed to enable CSI clock:
>>> %d\n", ret);
>>> + goto err_vpu_clock;
>>> + }
>>> +
>>> + dev_dbg(dev->v4l2_dev.dev, "node %s\n", node->video_dev.name);
>>> +
>>> + spin_lock_irqsave(&node->dma_queue_lock, flags);
>>> + buf = list_first_entry(&node->dma_queue,
>>> + struct unicam_buffer, list);
>>
>> This holds on a single line.
>>
>>> + node->cur_frm = buf;
>>> + node->next_frm = buf;
>>> + list_del(&buf->list);
>>> + spin_unlock_irqrestore(&node->dma_queue_lock, flags);
>>> +
>>> + buffer_addr[i] =
>>> + vb2_dma_contig_plane_dma_addr(&buf->vb.vb2_buf, 0);
>>
>> This too.
>>
>>> +
>>> + unicam_start_rx(dev, buffer_addr);
>>> +
>>> + ret = v4l2_subdev_call(&dev->sd, video, s_stream, 1);
>>> + if (ret < 0) {
>>> + dev_err(dev->v4l2_dev.dev, "stream on failed in subdev\n");
>>> + goto err_disable_unicam;
>>> + }
>>
>> Do I understand correctly that all this will run twice, once when
>> VIDIOC_STREAMON is called on the image node, and once when it's called
>> on the metadata node ? That doesn't seem right. The stop_streaming
>> function is affected similarly.
>>
>>> +
>>> + dev->clocks_enabled = true;
>>> + return 0;
>>> +
>>> +err_disable_unicam:
>>> + unicam_disable(dev);
>>> + clk_disable_unprepare(dev->clock);
>>> +err_vpu_clock:
>>> + if (clk_set_min_rate(dev->vpu_clock, 0))
>>> + dev_err(dev->v4l2_dev.dev, "failed to reset the VPU clock\n");
>>> + clk_disable_unprepare(dev->vpu_clock);
>>> +error_pipeline:
>>> + media_pipeline_stop(node->video_dev.entity.pads);
>>> +err_pm_put:
>>> + unicam_runtime_put(dev);
>>> +err_streaming:
>>> + unicam_return_buffers(node, VB2_BUF_STATE_QUEUED);
>>> + node->streaming = false;
>>> +
>>> + return ret;
>>> +}
>>> +
>>> +static void unicam_stop_streaming(struct vb2_queue *vq)
>>> +{
>>> + struct unicam_node *node = vb2_get_drv_priv(vq);
>>> + struct unicam_device *dev = node->dev;
>>> +
>>> + node->streaming = false;
>>> + /*
>>> + * Stop streaming the sensor and disable the peripheral.
>>> + * We cannot continue streaming embedded data with the
>>> + * image pad disabled.
>>> + */
>>> + if (v4l2_subdev_call(&dev->sd, video, s_stream, 0) < 0)
>>> + dev_err(dev->v4l2_dev.dev, "stream off failed in subdev\n");
>>> +
>>> + unicam_disable(dev);
>>> +
>>> + media_pipeline_stop(node->video_dev.entity.pads);
>>> +
>>> + if (dev->clocks_enabled) {
>>
>> .stop_streaming() can only be called after a successful
>> .start_streaming() call, so you can drop the clocks_enabled field.
>>
>>> + if (clk_set_min_rate(dev->vpu_clock, 0))
>>> + dev_err(dev->v4l2_dev.dev, "failed to reset the VPU
>>> clock\n");
>>> + clk_disable_unprepare(dev->vpu_clock);
>>> + clk_disable_unprepare(dev->clock);
>>> + dev->clocks_enabled = false;
>>> + }
>>> + unicam_runtime_put(dev);
>>> +
>>> + if (is_metadata_node(node)) {
>>> + /*
>>> + * Allow the hardware to spin in the dummy buffer.
>>> + * This is only really needed if the embedded data pad is
>>> + * disabled before the image pad.
>>> + */
>>> + unicam_wr_dma_addr(dev, node->dummy_buf_dma_addr,
>>> + DUMMY_BUF_SIZE, METADATA_NODE);
>>> + }
>>> +
>>> + /* Clear all queued buffers for the node */
>>> + unicam_return_buffers(node, VB2_BUF_STATE_ERROR);
>>> +}
>>> +
>>> +static const struct vb2_ops unicam_video_qops = {
>>> + .wait_prepare = vb2_ops_wait_prepare,
>>> + .wait_finish = vb2_ops_wait_finish,
>>> + .queue_setup = unicam_queue_setup,
>>> + .buf_prepare = unicam_buffer_prepare,
>>> + .buf_queue = unicam_buffer_queue,
>>> + .start_streaming = unicam_start_streaming,
>>> + .stop_streaming = unicam_stop_streaming,
>>> +};
>>> +
>>> +/* unicam capture driver file operations */
>>> +static const struct v4l2_file_operations unicam_fops = {
>>> + .owner = THIS_MODULE,
>>> + .open = v4l2_fh_open,
>>> + .release = vb2_fop_release,
>>> + .poll = vb2_fop_poll,
>>> + .unlocked_ioctl = video_ioctl2, /* V4L2 ioctl handler */
>>
>> Drop the comment.
>>
>>> + .mmap = vb2_fop_mmap,
>>> +};
>>> +
>>> +static int
>>> +unicam_async_bound(struct v4l2_async_notifier *notifier,
>>> + struct v4l2_subdev *subdev,
>>> + struct v4l2_async_subdev *asd)
>>> +{
>>> + struct unicam_device *unicam =
>>> to_unicam_device(notifier->v4l2_dev);
>>> +
>>> + if (unicam->sensor) {
>>> + dev_info(unicam->v4l2_dev.dev, "subdev %s (Already set!!)",
>>> + subdev->name);
>>> + return 0;
>>> + }
>>
>> This can't happen as we register a single asd only.
>>
>>> +
>>> + unicam->sensor = subdev;
>>> + dev_dbg(unicam->v4l2_dev.dev, "Using sensor %s for capture\n",
>>> subdev->name);
>>
>> Line wrap.
>>
>>> +
>>> + return 0;
>>> +}
>>> +
>>> +static void unicam_release(struct kref *kref)
>>> +{
>>> + struct unicam_device *unicam =
>>> + container_of(kref, struct unicam_device, kref);
>>> +
>>> + v4l2_ctrl_handler_free(&unicam->ctrl_handler);
>>> + media_device_cleanup(&unicam->mdev);
>>> +
>>> + if (unicam->sensor_state)
>>> + __v4l2_subdev_state_free(unicam->sensor_state);
>>> +
>>> + kfree(unicam);
>>> +}
>>> +
>>> +static void unicam_put(struct unicam_device *unicam)
>>> +{
>>> + kref_put(&unicam->kref, unicam_release);
>>> +}
>>> +
>>> +static void unicam_get(struct unicam_device *unicam)
>>> +{
>>> + kref_get(&unicam->kref);
>>> +}
>>> +
>>> +static void unicam_node_release(struct video_device *vdev)
>>> +{
>>> + struct unicam_node *node = video_get_drvdata(vdev);
>>> +
>>> + unicam_put(node->dev);
>>> +}
>>> +
>>> +static void unicam_mc_set_default_format(struct unicam_node *node)
>>> +{
>>> + dev_dbg(node->dev->v4l2_dev.dev, "Set default format for %s
>>> node\n",
>>
>> node->dev->dev should do, no need to go through v4l2_dev here. But I'd
>> drop the message, I don't think it brings much value.
>>
>>> + node->node_id == IMAGE_NODE ? "image" : "metadata");
>>> +
>>> + if (is_image_node(node)) {
>>> + struct v4l2_pix_format *pix_fmt = &node->v_fmt.fmt.pix;
>>> +
>>> + pix_fmt->width = 640;
>>> + pix_fmt->height = 480;
>>> + pix_fmt->field = V4L2_FIELD_NONE;
>>> + pix_fmt->colorspace = V4L2_COLORSPACE_SRGB;
>>> + pix_fmt->ycbcr_enc = V4L2_YCBCR_ENC_601;
>>> + pix_fmt->quantization = V4L2_QUANTIZATION_LIM_RANGE;
>>> + pix_fmt->xfer_func = V4L2_XFER_FUNC_SRGB;
>>> + pix_fmt->pixelformat = formats[0].fourcc;
>>> + unicam_calc_format_size_bpl(node->dev, &formats[0],
>>> + &node->v_fmt);
>>> + node->v_fmt.type = V4L2_BUF_TYPE_VIDEO_CAPTURE;
>>> +
>>> + node->fmt = &formats[0];
>>> + } else {
>>> + const struct unicam_fmt *fmt;
>>> +
>>> + /* Fix this node format as embedded data. */
>>> + fmt = find_format_by_code(MEDIA_BUS_FMT_METADATA_8);
>>> + node->v_fmt.fmt.meta.dataformat = fmt->fourcc;
>>> + node->fmt = fmt;
>>
>> Move this to the end to group all v_fmt initialization together.
>>
>>> + node->v_fmt.fmt.meta.buffersize = UNICAM_EMBEDDED_SIZE;
>>> + node->embedded_lines = 1;
>>
>> This belongs to the called.
>>
>>> + node->v_fmt.type = V4L2_BUF_TYPE_META_CAPTURE;
>>> + }
>>> +}
>>> +
>>> +static int register_node(struct unicam_device *unicam, struct
>>> unicam_node *node,
>>> + enum v4l2_buf_type type)
>>
>> Replace the type parameter with a enum unicam_node_type, that will look
>> cleaner in
>>
>>> +{
>>> + struct video_device *vdev;
>>> + struct vb2_queue *q;
>>> + int ret;
>>> +
>>> + node->dev = unicam;
>>> + node->node_id = type == V4L2_BUF_TYPE_VIDEO_CAPTURE ?
>>> + IMAGE_NODE :
>>> + METADATA_NODE;
>>> +
>>> + spin_lock_init(&node->dma_queue_lock);
>>> + mutex_init(&node->lock);
>>> +
>>> + vdev = &node->video_dev;
>>> + /*
>>> + * If the sensor subdevice has any controls, associate the node
>>> + * with the ctrl handler to allow access from userland.
>>> + */
>>> + if (!list_empty(&unicam->ctrl_handler.ctrls))
>>> + vdev->ctrl_handler = &unicam->ctrl_handler;
>>
>> Drop this, and move the vdev assignment below, just before it gets used.
>>
>> /* Initialize the videobuf2 queue. */
>>> +
>>> + q = &node->buffer_queue;
>>> + q->type = type;
>>> + q->io_modes = VB2_MMAP | VB2_DMABUF | VB2_READ;
>>
>> Drop VB2_READ, that shouldn't be used.
>>
>>> + q->drv_priv = node;
>>> + q->ops = &unicam_video_qops;
>>> + q->mem_ops = &vb2_dma_contig_memops;
>>> + q->buf_struct_size = sizeof(struct unicam_buffer);
>>> + q->timestamp_flags = V4L2_BUF_FLAG_TIMESTAMP_MONOTONIC;
>>> + q->lock = &node->lock;
>>> + q->min_buffers_needed = 1;
>>> + q->dev = &unicam->pdev->dev;
>>> +
>>> + ret = vb2_queue_init(q);
>>> + if (ret) {
>>> + dev_err(unicam->v4l2_dev.dev, "vb2_queue_init() failed\n");
>>> + return ret;
>>> + }
>>> +
>>> + INIT_LIST_HEAD(&node->dma_queue);
>>
>> Move this to the beginning of the function with the rest of the node
>> initialization.
>>
>> /* Initialize the video device. */
>>
>>> +
>>> + vdev->release = unicam_node_release;
>>> + vdev->fops = &unicam_fops;
>>> + vdev->ioctl_ops = &unicam_mc_ioctl_ops;
>>> + vdev->v4l2_dev = &unicam->v4l2_dev;
>>> + vdev->vfl_dir = VFL_DIR_RX;
>>> + vdev->queue = q;
>>> + vdev->lock = &node->lock;
>>> + vdev->device_caps = (type == V4L2_BUF_TYPE_VIDEO_CAPTURE) ?
>>
>> No need for parentheses.
>>
>>> + V4L2_CAP_VIDEO_CAPTURE : V4L2_CAP_META_CAPTURE;
>>> + vdev->device_caps |= V4L2_CAP_READWRITE | V4L2_CAP_STREAMING |
>>> V4L2_CAP_IO_MC;
>>
>> Drop V4L2_CAP_READWRITE.
>>
>>> + vdev->entity.ops = &unicam_mc_entity_ops;
>>> +
>>> + /* Define the device names */
>>> + snprintf(vdev->name, sizeof(vdev->name), "%s-%s",
>>> UNICAM_MODULE_NAME,
>>> + type == V4L2_BUF_TYPE_VIDEO_CAPTURE ? "image" : "embedded");
>>> +
>>> + video_set_drvdata(vdev, node);
>>> + if (type == V4L2_BUF_TYPE_VIDEO_CAPTURE)
>>> + vdev->entity.flags |= MEDIA_ENT_FL_DEFAULT;
>>> + node->pad.flags = MEDIA_PAD_FL_SINK;
>>> + media_entity_pads_init(&vdev->entity, 1, &node->pad);
>>
>> Error check.
>>
>>> +
>>> + node->dummy_buf_cpu_addr = dma_alloc_coherent(&unicam->pdev->dev,
>>> + DUMMY_BUF_SIZE,
>>> + &node->dummy_buf_dma_addr,
>>> + GFP_KERNEL);
>>> + if (!node->dummy_buf_cpu_addr) {
>>> + dev_err(unicam->v4l2_dev.dev, "Unable to allocate dummy
>>> buffer.\n");
>>> + return -ENOMEM;
>>> + }
>>> +
>>> + ret = video_register_device(vdev, VFL_TYPE_VIDEO, -1);
>>> + if (ret) {
>>> + dev_err(unicam->v4l2_dev.dev, "Unable to register video
>>> device %s\n",
>>> + vdev->name);
>>> + return ret;
>>> + }
>>> +
>>> + /*
>>> + * Acquire a reference to unicam, which will be released when
>>> the video
>>> + * device will be unregistered and userspace will have closed
>>> all open
>>> + * file handles.
>>> + */
>>> + unicam_get(unicam);
>>
>> This is error-prone. References should be acquired when stored, so this
>> should be moved to the beginning of the function, and written as
>>
>> node->dev = unicam_get(unicam);
>>
>> (you'll need to modify the function prototype to return the pointer).
>> You can then drop the comment. You will also to add error handling to
>> this function, as the reference will be released by
>> unicam_node_release(), which is called only after successful
>> registration of the video device. All error paths before registration
>> will thus need to release the reference (possibly with a goto
>> err_unicam_put)..
>>
>>> + node->registered = true;
>>> +
>>> + ret = media_create_pad_link(&unicam->sd.entity,
>>> + node->src_pad_id,
>>
>> unicam->sd is the unicam sd, node->src_pad_id is documented as being the
>> source pad of the sensor. This is wrong.
>>
>>> + &node->video_dev.entity,
>>> + 0,
>>> + MEDIA_LNK_FL_ENABLED |
>>> + MEDIA_LNK_FL_IMMUTABLE);
>>> + if (ret)
>>> + dev_err(unicam->v4l2_dev.dev, "Unable to create pad link for
>>> %s",
>>> + unicam->sensor->name);
>>
>> return ret;
>> }
>>
>>> +
>>> + unicam_mc_set_default_format(node);
>>
>> Call this before registering the video device, otherwise it can race
>> with userspace.
>>
>>> +
>>> + return ret;
>>
>> return 0;
>>
>>> +}
>>> +
>>> +static void unregister_nodes(struct unicam_device *unicam)
>>> +{
>>> + unsigned int i;
>>> +
>>> + for (i = 0; i < ARRAY_SIZE(unicam->node); i++) {
>>> + struct unicam_node *node = &unicam->node[i];
>>> +
>>> + if (node->dummy_buf_cpu_addr) {
>>> + dma_free_coherent(&unicam->pdev->dev, DUMMY_BUF_SIZE,
>>> + node->dummy_buf_cpu_addr,
>>> + node->dummy_buf_dma_addr);
>>> + }
>>
>> No need for curly braces.
>>
>>> +
>>> + if (node->registered) {
>>> + node->registered = false;
>>> + video_unregister_device(&node->video_dev);
>>
>> Swap the two lines.
>>
>>> + }
>>> + }
>>> +}
>>> +
>>> +static int unicam_async_complete(struct v4l2_async_notifier *notifier)
>>> +{
>>> + struct unicam_device *unicam =
>>> to_unicam_device(notifier->v4l2_dev);
>>> + unsigned int i, source_pads = 0;
>>> + int ret;
>>> + static struct lock_class_key key;
>>> +
>>> + unicam->v4l2_dev.notify = unicam_notify;
>>> +
>>> + unicam->sensor_state = __v4l2_subdev_state_alloc(unicam->sensor,
>>> + "unicam:state->lock",
>>> + &key);
>>
>> You never use sensor_state.
>>
>>> + if (!unicam->sensor_state)
>>> + return -ENOMEM;
>>> +
>>> + for (i = 0; i < unicam->sd.entity.num_pads; i++) {
>>> + if (unicam->sd.entity.pads[i].flags & MEDIA_PAD_FL_SOURCE) {
>>> + if (source_pads < MAX_NODES) {
>>> + unicam->node[source_pads].src_pad_id = i;
>>> + dev_dbg(unicam->v4l2_dev.dev, "source pad %u is
>>> index %u\n",
>>> + source_pads, i);
>>> + }
>>> + source_pads++;
>>> + }
>>> + }
>>> + if (!source_pads) {
>>> + dev_err(unicam->v4l2_dev.dev, "No source pads on sensor.\n");
>>> + goto unregister;
>>> + }
>>
>> You're not looking at the sensor here, unicam->sd is the unicam subdev.
>> Furthermore, you're hardcoding the source pad of the sensor to 0 below
>> when creating the link. And in any case, that's not how it's supposed to
>> be done, a source subdev could have multiple source pads, you need to
>> identify the correct one from the device tree.
>>
>> The right way to handle this is to replace the call to
>> media_create_pad_link() with v4l2_create_fwnode_links_to_pad(). This can
>> be done in the .bound() handler.
>>
>>> +
>>> + ret = register_node(unicam, &unicam->node[IMAGE_NODE],
>>> + V4L2_BUF_TYPE_VIDEO_CAPTURE);
>>
>> Replace the type parameter with a enum unicam_node_type, so this will
>> become
>>
>> ret = register_node(unicam, &unicam->node[IMAGE_NODE], IMAGE_NODE);
>>
>> which is less error-prone. You could even drop the second parameters:
>>
>> ret = register_node(unicam, IMAGE_NODE);
>>
>>> + if (ret) {
>>> + dev_err(unicam->v4l2_dev.dev, "Unable to register image
>>> video device.\n");
>>> + goto unregister;
>>> + }
>>> +
>>> + /* \todo: check before :-) */
>>> + unicam->sensor_embedded_data = true;
>>
>> Check what ? In any case I don't think this belongs here.
>>
>>> +
>>> + ret = register_node(unicam, &unicam->node[METADATA_NODE],
>>> + V4L2_BUF_TYPE_META_CAPTURE);
>>> + if (ret) {
>>> + dev_err(unicam->v4l2_dev.dev, "Unable to register metadata
>>> video device.\n");
>>> + goto unregister;
>>> + }
>>> +
>>> + ret = media_create_pad_link(&unicam->sensor->entity,
>>> + 0,
>>> + &unicam->sd.entity,
>>> + UNICAM_SD_PAD_SINK,
>>> + MEDIA_LNK_FL_ENABLED |
>>> + MEDIA_LNK_FL_IMMUTABLE);
>>> + if (ret)
>>> + dev_err(unicam->v4l2_dev.dev, "Unable to create pad link for
>>> %s",
>>> + unicam->sensor->name);
>>> +
>>> + ret = v4l2_device_register_subdev_nodes(&unicam->v4l2_dev);
>>> + if (ret) {
>>> + dev_err(unicam->v4l2_dev.dev, "Unable to register subdev
>>> nodes.\n");
>>> + goto unregister;
>>> + }
>>> +
>>> + /*
>>> + * Release the initial reference, all references are now owned
>>> by the
>>> + * video devices.
>>> + */
>>> + unicam_put(unicam);
>>> + return 0;
>>> +
>>> +unregister:
>>> + unregister_nodes(unicam);
>>> + unicam_put(unicam);
>>> +
>>> + return ret;
>>> +}
>>> +
>>> +static const struct v4l2_async_notifier_operations unicam_async_ops = {
>>> + .bound = unicam_async_bound,
>>> + .complete = unicam_async_complete,
>>> +};
>>> +
>>> +static int of_unicam_connect_subdevs(struct unicam_device *dev)
>>> +{
>>> + struct platform_device *pdev = dev->pdev;
>>> + struct v4l2_fwnode_endpoint ep = { };
>>> + struct device_node *ep_node;
>>> + struct device_node *sensor_node;
>>> + unsigned int lane;
>>> + int ret = -EINVAL;
>>> +
>>> + if (of_property_read_u32(pdev->dev.of_node, "num-data-lanes",
>>> + &dev->max_data_lanes) < 0) {
>>> + dev_err(dev->v4l2_dev.dev, "number of data lanes not set\n");
>>> + return -EINVAL;
>>> + }
>>> +
>>> + /* Get the local endpoint and remote device. */
>>> + ep_node = of_graph_get_next_endpoint(pdev->dev.of_node, NULL);
>>
>> Use of_graph_get_endpoint_by_regs(node, 0, -1).
>>
>>> + if (!ep_node) {
>>> + dev_dbg(dev->v4l2_dev.dev, "can't get next endpoint\n");
>>
>> dev_dbg(dev->v4l2_dev.dev, "No endpoint\n");
>>
>>> + return -EINVAL;
>>> + }
>>> +
>>> + dev_dbg(dev->v4l2_dev.dev, "ep_node is %pOF\n", ep_node);
>>
>> You can drop this.
>>
>>> +
>>> + sensor_node = of_graph_get_remote_port_parent(ep_node);
>>> + if (!sensor_node) {
>>> + dev_dbg(dev->v4l2_dev.dev, "can't get remote parent\n");
>>> + goto cleanup_exit;
>>> + }
>>> +
>>> + dev_dbg(dev->v4l2_dev.dev, "found subdevice %pOF\n", sensor_node);
>>
>> There's no subdev yet, we're dealing with DT here.
>>
>> dev_dbg(dev->v4l2_dev.dev, "found source device %pOF\n",
>> sensor_node);
>>
>> But I would actually drop all this. The sensor_node is only used in
>> debug or error messages, so you could remove it and use ep_node in those
>> messages instead. Or even better, drop the "subdevice %pOF: " prefix in
>> the messages, that will make the code simpler.
>>
>>> +
>>> + /* Parse the local endpoint and validate its configuration. */
>>> + v4l2_fwnode_endpoint_parse(of_fwnode_handle(ep_node), &ep);
>>
>> Use v4l2_fwnode_endpoint_alloc_parse() and check for errors (and update
>> the error cleanup path accordingly).
>>
>>> +
>>> + dev_dbg(dev->v4l2_dev.dev, "parsed local endpoint, bus_type %u\n",
>>> + ep.bus_type);
>>> +
>>> + dev->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(dev->v4l2_dev.dev, "subdevice %pOF: %u data
>>> lanes not supported\n",
>>> + sensor_node,
>>> + 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(dev->v4l2_dev.dev, "subdevice %pOF: data
>>> lanes reordering not supported\n",
>>> + sensor_node);
>>> + goto cleanup_exit;
>>> + }
>>> + }
>>> +
>>> + if (ep.bus.mipi_csi2.num_data_lanes > dev->max_data_lanes) {
>>> + dev_err(dev->v4l2_dev.dev, "subdevice requires %u data
>>> lanes when %u are supported\n",
>>
>> s/subdevice/endpoint/
>>
>>> + ep.bus.mipi_csi2.num_data_lanes,
>>> + dev->max_data_lanes);
>>> + }
>>> +
>>> + dev->max_data_lanes = ep.bus.mipi_csi2.num_data_lanes;
>>> + dev->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(dev->v4l2_dev.dev, "subdevice %pOF: unsupported
>>> lanes configuration\n",
>>> + sensor_node);
>>> + goto cleanup_exit;
>>> + }
>>> +
>>> + dev->max_data_lanes = 1;
>>> + dev->bus_flags = ep.bus.mipi_csi1.strobe;
>>> + break;
>>
>> I wonder if this driver will ever be used with a CCP2 sensor.
>>
>>> +
>>> + default:
>>> + /* Unsupported bus type */
>>> + dev_err(dev->v4l2_dev.dev, "subdevice %pOF: unsupported bus
>>> type %u\n",
>>> + sensor_node, ep.bus_type);
>>> + goto cleanup_exit;
>>> + }
>>> +
>>> + dev_dbg(dev->v4l2_dev.dev, "subdevice %pOF: %s bus, %u data
>>> lanes, flags=0x%08x\n",
>>> + sensor_node,
>>> + dev->bus_type == V4L2_MBUS_CSI2_DPHY ? "CSI-2" : "CCP2",
>>> + dev->max_data_lanes, dev->bus_flags);
>>> +
>>> + /* Initialize and register the async notifier. */
>>> + v4l2_async_nf_init(&dev->notifier);
>>> + dev->notifier.ops = &unicam_async_ops;
>>> +
>>> + dev->asd.match_type = V4L2_ASYNC_MATCH_FWNODE;
>>> + dev->asd.match.fwnode =
>>> fwnode_graph_get_remote_endpoint(of_fwnode_handle(ep_node));
>>> + ret = v4l2_async_nf_add_subdev(&dev->notifier, &dev->asd);
>>
>> The asd must be dynamically allocated as stated by the documentation of
>> __v4l2_async_nf_add_subdev(). You can use
>> v4l2_async_nf_add_fwnode_remote() to simplify this.
>>
>>> + if (ret) {
>>> + dev_err(dev->v4l2_dev.dev, "Error adding subdevice: %d\n",
>>> ret);
>>> + goto cleanup_exit;
>>> + }
>>> +
>>> + ret = v4l2_async_nf_register(&dev->v4l2_dev, &dev->notifier);
>>> + if (ret) {
>>> + dev_err(dev->v4l2_dev.dev, "Error registering async
>>> notifier: %d\n", ret);
>>> + ret = -EINVAL;
>>> + }
>>> +
>>> +cleanup_exit:
>>> + of_node_put(sensor_node);
>>> + of_node_put(ep_node);
>>> +
>>> + return ret;
>>> +}
>>> +
>>> +static int bcm2835_media_dev_init(struct unicam_device *unicam,
>>> + struct platform_device *pdev)
>>> +{
>>> + int ret = 0;
>>> +
>>> + unicam->mdev.dev = &pdev->dev;
>>> + strscpy(unicam->mdev.model, UNICAM_MODULE_NAME,
>>> + sizeof(unicam->mdev.model));
>>> + strscpy(unicam->mdev.serial, "", sizeof(unicam->mdev.serial));
>>> + snprintf(unicam->mdev.bus_info, sizeof(unicam->mdev.bus_info),
>>> + "platform:%s", dev_name(&pdev->dev));
>>> + unicam->mdev.hw_revision = 0;
>>> +
>>> + media_device_init(&unicam->mdev);
>>> +
>>> + unicam->v4l2_dev.mdev = &unicam->mdev;
>>> +
>>> + ret = v4l2_device_register(&pdev->dev, &unicam->v4l2_dev);
>>
>> return v4l2_device_register(&pdev->dev, &unicam->v4l2_dev);
>>
>>> +
>>> + return ret;
>>> +}
>>> +
>>> +/* Internal subdev */
>>> +
>>> +static int _unicam_subdev_set_routing(struct v4l2_subdev *sd,
>>
>> __ is more conventional.
>>
>>> + struct v4l2_subdev_state *state,
>>> + struct v4l2_subdev_krouting *routing)
>>> +{
>>> + static const struct v4l2_mbus_framefmt format = {
>>> + .width = 640,
>>> + .height = 480,
>>> + .code = MEDIA_BUS_FMT_UYVY8_2X8,
>>> + .field = V4L2_FIELD_NONE,
>>> + .colorspace = V4L2_COLORSPACE_SRGB,
>>> + .ycbcr_enc = V4L2_YCBCR_ENC_601,
>>> + .quantization = V4L2_QUANTIZATION_LIM_RANGE,
>>> + .xfer_func = V4L2_XFER_FUNC_SRGB,
>>> + .flags = 0,
>>> + };
>>
>> As this should match the values in unicam_mc_set_default_format(), can
>> you move this structure out of the function, and use it to initialize
>> the pix_fmt fields in unicam_mc_set_default_format() ?
>>
>>> + int ret;
>>> +
>>> + ret = v4l2_subdev_routing_validate_1_to_1(routing);
>>> + if (ret)
>>> + return ret;
>>> +
>>> + v4l2_subdev_lock_state(state);
>>> +
>>> + ret = v4l2_subdev_set_routing_with_fmt(sd, state, routing,
>>> &format);
>>> +
>>> + v4l2_subdev_unlock_state(state);
>>> +
>>> + if (ret)
>>> + return ret;
>>> +
>>> + return 0;
>>> +}
>>> +
>>> +static int unicam_subdev_set_routing(struct v4l2_subdev *sd,
>>> + struct v4l2_subdev_state *state,
>>> + enum v4l2_subdev_format_whence which,
>>> + struct v4l2_subdev_krouting *routing)
>>> +{
>>> + struct unicam_device *unicam = sd_to_unicam_device(sd);
>>> +
>>> + if (which == V4L2_SUBDEV_FORMAT_ACTIVE && unicam->streaming)
>>> + return -EBUSY;
>>> +
>>> + return _unicam_subdev_set_routing(sd, state, routing);
>>> +}
>>> +
>>> +static int unicam_subdev_init_cfg(struct v4l2_subdev *sd,
>>> + struct v4l2_subdev_state *state)
>>> +{
>>> + struct v4l2_subdev_route routes[] = {
>>> + {
>>> + .sink_pad = 0,
>>> + .sink_stream = 0,
>>> + .source_pad = 1,
>>
>> You have macros you can use for the pad numbers.
>>
>>> + .source_stream = 0,
>>> + .flags = V4L2_SUBDEV_ROUTE_FL_ACTIVE,
>>> + },
>>> + };
>>> +
>>> + struct v4l2_subdev_krouting routing = {
>>> + .num_routes = ARRAY_SIZE(routes),
>>> + .routes = routes,
>>> + };
>>> +
>>> + /* Initialize routing to single route to the fist source pad */
>>> + return _unicam_subdev_set_routing(sd, state, &routing);
>>> +}
>>> +
>>> +static int unicam_subdev_enum_mbus_code(struct v4l2_subdev *sd,
>>> + struct v4l2_subdev_state *state,
>>> + struct v4l2_subdev_mbus_code_enum *code)
>>> +{
>>> + int ret = 0;
>>> +
>>> + v4l2_subdev_lock_state(state);
>>> +
>>> + /* No transcoding, source and sink codes must match. */
>>> + if (unicam_sd_pad_is_source(code->pad)) {
>>> + struct v4l2_mbus_framefmt *fmt;
>>> +
>>> + if (code->index > 0) {
>>> + ret = -EINVAL;
>>> + goto out;
>>> + }
>>> +
>>> + fmt = v4l2_subdev_state_get_opposite_stream_format(state,
>>> + code->pad,
>>> + code->stream);
>>> + if (!fmt) {
>>> + ret = -EINVAL;
>>> + goto out;
>>> + }
>>> +
>>> + code->code = fmt->code;
>>> + } else {
>>> + if (code->index >= ARRAY_SIZE(formats)) {
>>> + ret = -EINVAL;
>>> + goto out;
>>> + }
>>> +
>>> + code->code = formats[code->index].code;
>>> + }
>>> +
>>> +out:
>>> + v4l2_subdev_unlock_state(state);
>>> +
>>> + return ret;
>>> +}
>>> +
>>> +static int unicam_subdev_start_streaming(struct unicam_device *unicam)
>>> +{
>>> + const struct v4l2_subdev_krouting *routing;
>>> + struct v4l2_subdev_state *state;
>>> + int ret = 0;
>>
>> No need to initialize ret to 0.
>>
>>> +
>>> + state = v4l2_subdev_lock_active_state(&unicam->sd);
>>> +
>>> + routing = &state->routing;
>>> +
>>> + v4l2_subdev_unlock_state(state);
>>
>> As a piece of conceptual art, maybe, but as useful code, I have some
>> doubts.
>>
>>> +
>>> + unicam->streaming = true;
>>> +
>>> + ret = v4l2_subdev_call(unicam->sensor, video, s_stream, 1);
>>> + if (ret) {
>>> + v4l2_subdev_call(unicam->sensor, video, s_stream, 0);
>>
>> If .s_stream(1) fails, why do you need to call .s_stream(0) ?
>>
>>> + unicam->streaming = false;
>>
>> Drop this, and move the unicam->streaming = true line below.
>>
>>> + return ret;
>>> + }
>>> +
>>> + return 0;
>>> +}
>>> +
>>> +static int unicam_subdev_stop_streaming(struct unicam_device *unicam)
>>> +{
>>> + v4l2_subdev_call(unicam->sensor, video, s_stream, 0);
>>> +
>>> + unicam->streaming = false;
>>> +
>>> + return 0;
>>> +}
>>> +
>>> +static int unicam_subdev_s_stream(struct v4l2_subdev *sd, int enable)
>>> +{
>>> + struct unicam_device *unicam = sd_to_unicam_device(sd);
>>> +
>>> + if (enable)
>>> + return unicam_subdev_start_streaming(unicam);
>>> + else
>>> + return unicam_subdev_stop_streaming(unicam);
>>> + return 0;
>>
>> Drop the last line.
>>
>>> +}
>>> +
>>> +static int unicam_subdev_set_pad_format(struct v4l2_subdev *sd,
>>> + struct v4l2_subdev_state *state,
>>> + struct v4l2_subdev_format *format)
>>> +{
>>> + struct unicam_device *unicam = sd_to_unicam_device(sd);
>>> + struct v4l2_mbus_framefmt *fmt;
>>> + int ret = 0;
>>> +
>>> + if (format->which == V4L2_SUBDEV_FORMAT_ACTIVE &&
>>> unicam->streaming)
>>> + return -EBUSY;
>>> +
>>> + /* No transcoding, source and sink formats must match. */
>>> + if (unicam_sd_pad_is_source(format->pad))
>>> + return v4l2_subdev_get_fmt(sd, state, format);
>>> +
>>> + // TODO: implement fmt validation
>>
>> That's a candidate for v3.
>>
>>> +
>>> + v4l2_subdev_lock_state(state);
>>> +
>>> + fmt = v4l2_state_get_stream_format(state, format->pad,
>>> format->stream);
>>> + if (!fmt) {
>>> + ret = -EINVAL;
>>> + goto out;
>>> + }
>>> +
>>> + *fmt = format->format;
>>> +
>>> + fmt = v4l2_subdev_state_get_opposite_stream_format(state,
>>> format->pad,
>>> + format->stream);
>>> + if (!fmt) {
>>> + ret = -EINVAL;
>>> + goto out;
>>> + }
>>> +
>>> + *fmt = format->format;
>>
>> Let's avoid applying the format on one pad to then fail on the other
>> pad.
>>
>> sink_format = v4l2_state_get_stream_format(state, format->pad,
>> format->stream);
>> source_format = v4l2_subdev_state_get_opposite_stream_format(state,
>> format->pad,
>> format->stream);
>> if (!sink_format || !source_format) {
>> ret = -EINVAL;
>> goto out;
>> }
>>
>> *sink_format = format->format;
>> *source_format = format->format;
>>
>>> +
>>> +out:
>>> + v4l2_subdev_unlock_state(state);
>>> +
>>> + return ret;
>>> +}
>>> +
>>> +static int unicam_subdev_enum_frame_size(struct v4l2_subdev *sd,
>>> + struct v4l2_subdev_state *state,
>>> + struct v4l2_subdev_frame_size_enum *fse)
>>> +{
>>> + const struct unicam_fmt *fmtinfo;
>>> + int ret = 0;
>>> +
>>> + if (fse->index > 0)
>>> + return -EINVAL;
>>> +
>>> + v4l2_subdev_lock_state(state);
>>> +
>>> + /* No transcoding, source and sink formats must match. */
>>> + if (unicam_sd_pad_is_source(fse->pad)) {
>>> + struct v4l2_mbus_framefmt *fmt;
>>> +
>>> + fmt = v4l2_subdev_state_get_opposite_stream_format(state,
>>> + fse->pad,
>>> + fse->stream);
>>> +
>>
>> You can drop this blank line.
>>
>>> + if (!fmt) {
>>> + ret = -EINVAL;
>>> + goto out;
>>> + }
>>> +
>>> + if (fse->code != fmt->code) {
>>> + ret = -EINVAL;
>>> + goto out;
>>> + }
>>> +
>>> + fse->min_width = fmt->width;
>>> + fse->max_width = fmt->width;
>>> + fse->min_height = fmt->height;
>>> + fse->max_height = fmt->height;
>>> + } else {
>>> + fmtinfo = find_format_by_code(fse->code);
>>> + if (!fmtinfo) {
>>> + ret = -EINVAL;
>>> + goto out;
>>> + }
>>> +
>>> + fse->min_width = MIN_WIDTH * 8 / ALIGN(fmtinfo->depth, 8);
>>> + fse->max_width = MAX_WIDTH * 8 / ALIGN(fmtinfo->depth, 8);
>>> + fse->min_height = MIN_HEIGHT;
>>> + fse->max_height = MAX_HEIGHT;
>>> + }
>>> +
>>> +out:
>>> + v4l2_subdev_unlock_state(state);
>>> +
>>> + return ret;
>>> +}
>>> +
>>> +static const struct v4l2_subdev_video_ops unicam_subdev_video_ops = {
>>> + .s_stream = unicam_subdev_s_stream,
>>> +};
>>> +
>>> +static const struct v4l2_subdev_pad_ops unicam_subdev_pad_ops = {
>>> + .init_cfg = unicam_subdev_init_cfg,
>>> + .enum_mbus_code = unicam_subdev_enum_mbus_code,
>>> + .get_fmt = v4l2_subdev_get_fmt,
>>> + .set_fmt = unicam_subdev_set_pad_format,
>>> + .set_routing = unicam_subdev_set_routing,
>>> + .enum_frame_size = unicam_subdev_enum_frame_size,
>>> +};
>>> +
>>> +static const struct v4l2_subdev_ops unicam_subdev_ops = {
>>> + .video = &unicam_subdev_video_ops,
>>> + .pad = &unicam_subdev_pad_ops,
>>> +};
>>> +
>>> +static struct media_entity_operations unicam_subdev_media_ops = {
>>
>> static const
>>
>>> + .link_validate = v4l2_subdev_link_validate,
>>> + .has_route = v4l2_subdev_has_route,
>>> +};
>>> +
>>> +static int unicam_probe(struct platform_device *pdev)
>>> +{
>>> + struct unicam_device *unicam;
>>> + int ret = 0;
>>> + int i;
>>> +
>>> + unicam = kzalloc(sizeof(*unicam), GFP_KERNEL);
>>> + if (!unicam)
>>> + return -ENOMEM;
>>> +
>>> + kref_init(&unicam->kref);
>>> + unicam->pdev = pdev;
>>> +
>>> + unicam->base = devm_platform_ioremap_resource(pdev, 0);
>>> + if (IS_ERR(unicam->base)) {
>>> + dev_err(unicam->v4l2_dev.dev, "Failed to get main io block\n");
>>
>> Let's use unicam->dev to access the struct device, especially given that
>> unicam->v4l2_dev is only initialized below when calling
>> bcm2835_media_dev_init().
>>
>> This message can actually br dropped, devm_platform_ioremap_resource()
>> already prints an error.
>>
>>> + ret = PTR_ERR(unicam->base);
>>> + goto err_unicam_put;
>>> + }
>>> +
>>> + unicam->clk_gate_base = devm_platform_ioremap_resource(pdev, 1);
>>> + if (IS_ERR(unicam->clk_gate_base)) {
>>> + dev_err(unicam->v4l2_dev.dev, "Failed to get 2nd io block\n");
>>
>> Same here.
>>
>>> + ret = PTR_ERR(unicam->clk_gate_base);
>>> + goto err_unicam_put;
>>> + }
>>> +
>>> + unicam->clock = devm_clk_get(&pdev->dev, "lp");
>>> + if (IS_ERR(unicam->clock)) {
>>> + dev_err(unicam->v4l2_dev.dev, "Failed to get lp clock\n");
>>> + ret = PTR_ERR(unicam->clock);
>>> + goto err_unicam_put;
>>> + }
>>> +
>>> + unicam->vpu_clock = devm_clk_get(&pdev->dev, "vpu");
>>> + if (IS_ERR(unicam->vpu_clock)) {
>>> + dev_err(unicam->v4l2_dev.dev, "Failed to get vpu clock\n");
>>> + ret = PTR_ERR(unicam->vpu_clock);
>>> + goto err_unicam_put;
>>> + }
>>
>> Could the clock bulk API help here ?
>>
>>> +
>>> + ret = platform_get_irq(pdev, 0);
>>> + if (ret <= 0) {
>>> + dev_err(&pdev->dev, "No IRQ resource\n");
>>> + ret = -EINVAL;
>>> + goto err_unicam_put;
>>> + }
>>> +
>>> + ret = devm_request_irq(&pdev->dev, ret, unicam_isr, 0,
>>> + "unicam_capture0", unicam);
>>> + if (ret) {
>>> + dev_err(&pdev->dev, "Unable to request interrupt\n");
>>> + ret = -EINVAL;
>>> + goto err_unicam_put;
>>> + }
>>> +
>>> + ret = bcm2835_media_dev_init(unicam, pdev);
>>> + if (ret) {
>>> + dev_err(unicam->v4l2_dev.dev,
>>> + "Unable to register v4l2 device.\n");
>>> + goto err_unicam_put;
>>> + }
>>> +
>>> + ret = media_device_register(&unicam->mdev);
>>> + if (ret < 0) {
>>> + dev_err(unicam->v4l2_dev.dev,
>>> + "Unable to register media-controller device.\n");
>>> + goto err_v4l2_unregister;
>>> + }
>>> +
>>> + /* Reserve space for the controls */
>>> + ret = v4l2_ctrl_handler_init(&unicam->ctrl_handler, 16);
>>> + if (ret < 0)
>>> + goto err_media_unregister;
>>
>> The control handler is unused, drop it.
>>
>>> +
>>> + /* set the driver data in platform device */
>>> + platform_set_drvdata(pdev, unicam);
>>> +
>>> + dev_dbg(unicam->v4l2_dev.dev, "Initialize internal subdev");
>>
>> You can drop this.
>>
>>> + v4l2_subdev_init(&unicam->sd, &unicam_subdev_ops);
>>> + v4l2_set_subdevdata(&unicam->sd, unicam);
>>> + unicam->sd.entity.function = MEDIA_ENT_F_VID_IF_BRIDGE;
>>> + unicam->sd.dev = &pdev->dev;
>>> + unicam->sd.owner = THIS_MODULE;
>>> + unicam->sd.flags = V4L2_SUBDEV_FL_HAS_DEVNODE |
>>> V4L2_SUBDEV_FL_MULTIPLEXED;
>>> + snprintf(unicam->sd.name, sizeof(unicam->sd.name),
>>> "unicam-subdev");
>>> +
>>> + unicam->pads[UNICAM_SD_PAD_SINK].flags = MEDIA_PAD_FL_SINK;
>>> +
>>> + for (i = UNICAM_SD_PAD_FIRST_SOURCE; i < UNICAM_SD_NUM_PADS; ++i)
>>> + unicam->pads[i].flags = MEDIA_PAD_FL_SOURCE;
>>
>> unicam->pads[UNICAM_SD_PAD_SOURCE_IMAGE].flags = MEDIA_PAD_FL_SOURCE;
>> unicam->pads[UNICAM_SD_PAD_SOURCE_META].flags = MEDIA_PAD_FL_SOURCE;
>>
>>> + unicam->sd.entity.ops = &unicam_subdev_media_ops;
>>> + ret = media_entity_pads_init(&unicam->sd.entity,
>>> + ARRAY_SIZE(unicam->pads), unicam->pads);
>>> + if (ret) {
>>> + dev_err(unicam->v4l2_dev.dev, "Could not init media
>>> controller for subdev");
>>> + goto err_subdev_unregister;
>>> + }
>>> +
>>> + ret = v4l2_subdev_init_finalize(&unicam->sd);
>>> + if (ret) {
>>> + dev_err(unicam->v4l2_dev.dev, "Could not finalize init for
>>> subdev");
>>> + goto err_entity_cleanup;
>>> + }
>>> +
>>> + ret = v4l2_device_register_subdev(&unicam->v4l2_dev, &unicam->sd);
>>> + if (ret) {
>>> + dev_err(&pdev->dev, "Failed to register internal subdev\n");
>>> + goto err_subdev_unregister;
>>> + }
>>
>> Let's split initialization and registration of the subdev to a separate
>> function, placed above with the subdev operations.
>>
>>> +
>>> + ret = of_unicam_connect_subdevs(unicam);
>>> + if (ret) {
>>> + dev_err(&pdev->dev, "Failed to connect subdevs\n");
>>> + goto err_subdev_unregister;
>>> + }
>>> +
>>> + /* Enable the block power domain */
>>> + pm_runtime_enable(&pdev->dev);
>>> +
>>> + return 0;
>>> +
>>> +err_subdev_unregister:
>>> + v4l2_subdev_cleanup(&unicam->sd);
>>> +err_entity_cleanup:
>>> + media_entity_cleanup(&unicam->sd.entity);
>>> +err_media_unregister:
>>> + media_device_unregister(&unicam->mdev);
>>> +err_v4l2_unregister:
>>> + v4l2_device_unregister(&unicam->v4l2_dev);
>>> +err_unicam_put:
>>> + unicam_put(unicam);
>>> +
>>> + return ret;
>>> +}
>>> +
>>> +static int unicam_remove(struct platform_device *pdev)
>>> +{
>>> + struct unicam_device *unicam = platform_get_drvdata(pdev);
>>> +
>>> + v4l2_async_nf_unregister(&unicam->notifier);
>>> + v4l2_device_unregister(&unicam->v4l2_dev);
>>> + media_device_unregister(&unicam->mdev);
>>> + unregister_nodes(unicam);
>>
>> I'm a bit worried there could be race conditions with userspace here.
>> For instance, calling v4l2_async_nf_unregister() will result in
>> v4l2_device_unregister_subdev() being called on the sensor subdev, which
>> may race with userspace starting capture on a video node. The following
>> order would, I think be safer:
>>
>> unregister_nodes(unicam);
>> v4l2_device_unregister(&unicam->v4l2_dev);
>> media_device_unregister(&unicam->mdev);
>> v4l2_async_nf_unregister(&unicam->notifier);
>>
>> But this will cause a problem, when unregistering device nodes, the last
>> reference to unicam would be dropped. I think you could drop the calls
>> to unicam_put() from unicam_async_complete(), and add a unicam_put()
>> call here. Dave, does that sound good to you ?
>>
>>> +
>>> + pm_runtime_disable(&pdev->dev);
>>> +
>>> + return 0;
>>> +}
>>> +
>>> +static const struct of_device_id unicam_of_match[] = {
>>> + { .compatible = "brcm,bcm2835-unicam", },
>>> + { /* sentinel */ },
>>> +};
>>> +MODULE_DEVICE_TABLE(of, unicam_of_match);
>>> +
>>> +static struct platform_driver unicam_driver = {
>>> + .probe = unicam_probe,
>>> + .remove = unicam_remove,
>>> + .driver = {
>>> + .name = UNICAM_MODULE_NAME,
>>> + .of_match_table = of_match_ptr(unicam_of_match),
>>> + },
>>> +};
>>> +
>>> +module_platform_driver(unicam_driver);
>>> +
>>> +MODULE_AUTHOR("Dave Stevenson <dave.stevenson@raspberrypi.com>");
>>> +MODULE_DESCRIPTION("BCM2835 Unicam driver");
>>> +MODULE_LICENSE("GPL");
>>> +MODULE_VERSION(UNICAM_VERSION);
>>> diff --git a/drivers/media/platform/bcm2835/vc4-regs-unicam.h
>>> b/drivers/media/platform/bcm2835/vc4-regs-unicam.h
>>> new file mode 100644
>>> index 000000000000..ae059a171d0f
>>> --- /dev/null
>>> +++ b/drivers/media/platform/bcm2835/vc4-regs-unicam.h
>>> @@ -0,0 +1,253 @@
>>> +/* SPDX-License-Identifier: GPL-2.0-only */
>>> +
>>> +/*
>>> + * Copyright (C) 2017-2020 Raspberry Pi Trading.
>>> + * Dave Stevenson <dave.stevenson@raspberrypi.com>
>>> + */
>>> +
>>> +#ifndef VC4_REGS_UNICAM_H
>>> +#define VC4_REGS_UNICAM_H
>>> +
>>> +/*
>>> + * The following values are taken from files found within the code drop
>>> + * made by Broadcom for the BCM21553 Graphics Driver, predominantly in
>>> + * brcm_usrlib/dag/vmcsx/vcinclude/hardware_vc4.h.
>>> + * They have been modified to be only the register offset.
>>> + */
>>> +#define UNICAM_CTRL 0x000
>>> +#define UNICAM_STA 0x004
>>> +#define UNICAM_ANA 0x008
>>> +#define UNICAM_PRI 0x00c
>>> +#define UNICAM_CLK 0x010
>>> +#define UNICAM_CLT 0x014
>>> +#define UNICAM_DAT0 0x018
>>> +#define UNICAM_DAT1 0x01c
>>> +#define UNICAM_DAT2 0x020
>>> +#define UNICAM_DAT3 0x024
>>> +#define UNICAM_DLT 0x028
>>> +#define UNICAM_CMP0 0x02c
>>> +#define UNICAM_CMP1 0x030
>>> +#define UNICAM_CAP0 0x034
>>> +#define UNICAM_CAP1 0x038
>>> +#define UNICAM_ICTL 0x100
>>> +#define UNICAM_ISTA 0x104
>>> +#define UNICAM_IDI0 0x108
>>> +#define UNICAM_IPIPE 0x10c
>>> +#define UNICAM_IBSA0 0x110
>>> +#define UNICAM_IBEA0 0x114
>>> +#define UNICAM_IBLS 0x118
>>> +#define UNICAM_IBWP 0x11c
>>> +#define UNICAM_IHWIN 0x120
>>> +#define UNICAM_IHSTA 0x124
>>> +#define UNICAM_IVWIN 0x128
>>> +#define UNICAM_IVSTA 0x12c
>>> +#define UNICAM_ICC 0x130
>>> +#define UNICAM_ICS 0x134
>>> +#define UNICAM_IDC 0x138
>>> +#define UNICAM_IDPO 0x13c
>>> +#define UNICAM_IDCA 0x140
>>> +#define UNICAM_IDCD 0x144
>>> +#define UNICAM_IDS 0x148
>>> +#define UNICAM_DCS 0x200
>>> +#define UNICAM_DBSA0 0x204
>>> +#define UNICAM_DBEA0 0x208
>>> +#define UNICAM_DBWP 0x20c
>>> +#define UNICAM_DBCTL 0x300
>>> +#define UNICAM_IBSA1 0x304
>>> +#define UNICAM_IBEA1 0x308
>>> +#define UNICAM_IDI1 0x30c
>>> +#define UNICAM_DBSA1 0x310
>>> +#define UNICAM_DBEA1 0x314
>>> +#define UNICAM_MISC 0x400
>>
>> Please add one tab before the value, to match the indentation of the
>> rest of the file.
>>
>>> +
>>> +/*
>>> + * The following bitmasks are from the kernel released by Broadcom
>>> + * for Android - https://android.googlesource.com/kernel/bcm/
>>> + * The Rhea, Hawaii, and Java chips all contain the same VideoCore4
>>> + * Unicam block as BCM2835, as defined in eg
>>> + * arch/arm/mach-rhea/include/mach/rdb_A0/brcm_rdb_cam.h and similar.
>>> + * Values reworked to use the kernel BIT and GENMASK macros.
>>> + *
>>> + * Some of the bit mnenomics have been amended to match the datasheet.
>>> + */
>>> +/* UNICAM_CTRL Register */
>>> +#define UNICAM_CPE BIT(0)
>>> +#define UNICAM_MEM BIT(1)
>>> +#define UNICAM_CPR BIT(2)
>>> +#define UNICAM_CPM_MASK GENMASK(3, 3)
>>> +#define UNICAM_CPM_CSI2 0
>>> +#define UNICAM_CPM_CCP2 1
>>> +#define UNICAM_SOE BIT(4)
>>> +#define UNICAM_DCM_MASK GENMASK(5, 5)
>>> +#define UNICAM_DCM_STROBE 0
>>> +#define UNICAM_DCM_DATA 1
>>> +#define UNICAM_SLS BIT(6)
>>> +#define UNICAM_PFT_MASK GENMASK(11, 8)
>>> +#define UNICAM_OET_MASK GENMASK(20, 12)
>>> +
>>> +/* UNICAM_STA Register */
>>> +#define UNICAM_SYN BIT(0)
>>> +#define UNICAM_CS BIT(1)
>>> +#define UNICAM_SBE BIT(2)
>>> +#define UNICAM_PBE BIT(3)
>>> +#define UNICAM_HOE BIT(4)
>>> +#define UNICAM_PLE BIT(5)
>>> +#define UNICAM_SSC BIT(6)
>>> +#define UNICAM_CRCE BIT(7)
>>> +#define UNICAM_OES BIT(8)
>>> +#define UNICAM_IFO BIT(9)
>>> +#define UNICAM_OFO BIT(10)
>>> +#define UNICAM_BFO BIT(11)
>>> +#define UNICAM_DL BIT(12)
>>> +#define UNICAM_PS BIT(13)
>>> +#define UNICAM_IS BIT(14)
>>> +#define UNICAM_PI0 BIT(15)
>>> +#define UNICAM_PI1 BIT(16)
>>> +#define UNICAM_FSI_S BIT(17)
>>> +#define UNICAM_FEI_S BIT(18)
>>> +#define UNICAM_LCI_S BIT(19)
>>> +#define UNICAM_BUF0_RDY BIT(20)
>>> +#define UNICAM_BUF0_NO BIT(21)
>>> +#define UNICAM_BUF1_RDY BIT(22)
>>> +#define UNICAM_BUF1_NO BIT(23)
>>> +#define UNICAM_DI BIT(24)
>>> +
>>> +#define UNICAM_STA_MASK_ALL \
>>> + (UNICAM_DL + \
>>> + UNICAM_SBE + \
>>> + UNICAM_PBE + \
>>> + UNICAM_HOE + \
>>> + UNICAM_PLE + \
>>> + UNICAM_SSC + \
>>> + UNICAM_CRCE + \
>>> + UNICAM_IFO + \
>>> + UNICAM_OFO + \
>>> + UNICAM_PS + \
>>> + UNICAM_PI0 + \
>>> + UNICAM_PI1)
>>
>> I'd use | instead of + to combine bits. Same below.
>>
>>> +
>>> +/* UNICAM_ANA Register */
>>> +#define UNICAM_APD BIT(0)
>>> +#define UNICAM_BPD BIT(1)
>>> +#define UNICAM_AR BIT(2)
>>> +#define UNICAM_DDL BIT(3)
>>> +#define UNICAM_CTATADJ_MASK GENMASK(7, 4)
>>> +#define UNICAM_PTATADJ_MASK GENMASK(11, 8)
>>> +
>>> +/* UNICAM_PRI Register */
>>> +#define UNICAM_PE BIT(0)
>>> +#define UNICAM_PT_MASK GENMASK(2, 1)
>>> +#define UNICAM_NP_MASK GENMASK(7, 4)
>>> +#define UNICAM_PP_MASK GENMASK(11, 8)
>>> +#define UNICAM_BS_MASK GENMASK(15, 12)
>>> +#define UNICAM_BL_MASK GENMASK(17, 16)
>>> +
>>> +/* UNICAM_CLK Register */
>>> +#define UNICAM_CLE BIT(0)
>>> +#define UNICAM_CLPD BIT(1)
>>> +#define UNICAM_CLLPE BIT(2)
>>> +#define UNICAM_CLHSE BIT(3)
>>> +#define UNICAM_CLTRE BIT(4)
>>> +#define UNICAM_CLAC_MASK GENMASK(8, 5)
>>> +#define UNICAM_CLSTE BIT(29)
>>> +
>>> +/* UNICAM_CLT Register */
>>> +#define UNICAM_CLT1_MASK GENMASK(7, 0)
>>> +#define UNICAM_CLT2_MASK GENMASK(15, 8)
>>> +
>>> +/* UNICAM_DATn Registers */
>>> +#define UNICAM_DLE BIT(0)
>>> +#define UNICAM_DLPD BIT(1)
>>> +#define UNICAM_DLLPE BIT(2)
>>> +#define UNICAM_DLHSE BIT(3)
>>> +#define UNICAM_DLTRE BIT(4)
>>> +#define UNICAM_DLSM BIT(5)
>>> +#define UNICAM_DLFO BIT(28)
>>> +#define UNICAM_DLSTE BIT(29)
>>> +
>>> +#define UNICAM_DAT_MASK_ALL (UNICAM_DLSTE + UNICAM_DLFO)
>>
>> This also needs an indentation fix, as well as UNICAM_ISTA_MASK_ALL.
>>
>>> +
>>> +/* UNICAM_DLT Register */
>>> +#define UNICAM_DLT1_MASK GENMASK(7, 0)
>>> +#define UNICAM_DLT2_MASK GENMASK(15, 8)
>>> +#define UNICAM_DLT3_MASK GENMASK(23, 16)
>>> +
>>> +/* UNICAM_ICTL Register */
>>> +#define UNICAM_FSIE BIT(0)
>>> +#define UNICAM_FEIE BIT(1)
>>> +#define UNICAM_IBOB BIT(2)
>>> +#define UNICAM_FCM BIT(3)
>>> +#define UNICAM_TFC BIT(4)
>>> +#define UNICAM_LIP_MASK GENMASK(6, 5)
>>> +#define UNICAM_LCIE_MASK GENMASK(28, 16)
>>> +
>>> +/* UNICAM_IDI0/1 Register */
>>> +#define UNICAM_ID0_MASK GENMASK(7, 0)
>>> +#define UNICAM_ID1_MASK GENMASK(15, 8)
>>> +#define UNICAM_ID2_MASK GENMASK(23, 16)
>>> +#define UNICAM_ID3_MASK GENMASK(31, 24)
>>> +
>>> +/* UNICAM_ISTA Register */
>>> +#define UNICAM_FSI BIT(0)
>>> +#define UNICAM_FEI BIT(1)
>>> +#define UNICAM_LCI BIT(2)
>>> +
>>> +#define UNICAM_ISTA_MASK_ALL (UNICAM_FSI + UNICAM_FEI + UNICAM_LCI)
>>> +
>>> +/* UNICAM_IPIPE Register */
>>> +#define UNICAM_PUM_MASK GENMASK(2, 0)
>>> + /* Unpacking modes */
>>> + #define UNICAM_PUM_NONE 0
>>> + #define UNICAM_PUM_UNPACK6 1
>>> + #define UNICAM_PUM_UNPACK7 2
>>> + #define UNICAM_PUM_UNPACK8 3
>>> + #define UNICAM_PUM_UNPACK10 4
>>> + #define UNICAM_PUM_UNPACK12 5
>>> + #define UNICAM_PUM_UNPACK14 6
>>> + #define UNICAM_PUM_UNPACK16 7
>>> +#define UNICAM_DDM_MASK GENMASK(6, 3)
>>> +#define UNICAM_PPM_MASK GENMASK(9, 7)
>>> + /* Packing modes */
>>> + #define UNICAM_PPM_NONE 0
>>> + #define UNICAM_PPM_PACK8 1
>>> + #define UNICAM_PPM_PACK10 2
>>> + #define UNICAM_PPM_PACK12 3
>>> + #define UNICAM_PPM_PACK14 4
>>> + #define UNICAM_PPM_PACK16 5
>>
>> I'd drop the leading tab.
>>
>>> +#define UNICAM_DEM_MASK GENMASK(11, 10)
>>> +#define UNICAM_DEBL_MASK GENMASK(14, 12)
>>> +#define UNICAM_ICM_MASK GENMASK(16, 15)
>>> +#define UNICAM_IDM_MASK GENMASK(17, 17)
>>> +
>>> +/* UNICAM_ICC Register */
>>> +#define UNICAM_ICFL_MASK GENMASK(4, 0)
>>> +#define UNICAM_ICFH_MASK GENMASK(9, 5)
>>> +#define UNICAM_ICST_MASK GENMASK(12, 10)
>>> +#define UNICAM_ICLT_MASK GENMASK(15, 13)
>>> +#define UNICAM_ICLL_MASK GENMASK(31, 16)
>>> +
>>> +/* UNICAM_DCS Register */
>>> +#define UNICAM_DIE BIT(0)
>>> +#define UNICAM_DIM BIT(1)
>>> +#define UNICAM_DBOB BIT(3)
>>> +#define UNICAM_FDE BIT(4)
>>> +#define UNICAM_LDP BIT(5)
>>> +#define UNICAM_EDL_MASK GENMASK(15, 8)
>>> +
>>> +/* UNICAM_DBCTL Register */
>>> +#define UNICAM_DBEN BIT(0)
>>> +#define UNICAM_BUF0_IE BIT(1)
>>> +#define UNICAM_BUF1_IE BIT(2)
>>> +
>>> +/* UNICAM_CMP[0,1] register */
>>> +#define UNICAM_PCE BIT(31)
>>> +#define UNICAM_GI BIT(9)
>>> +#define UNICAM_CPH BIT(8)
>>> +#define UNICAM_PCVC_MASK GENMASK(7, 6)
>>> +#define UNICAM_PCDT_MASK GENMASK(5, 0)
>>> +
>>> +/* UNICAM_MISC register */
>>> +#define UNICAM_FL0 BIT(6)
>>> +#define UNICAM_FL1 BIT(9)
>>> +
>>> +#endif
>>
^ permalink raw reply
* [PATCH 1/1] dt-bindings: net: fsl,fec: Add nvmem-cells / nvmem-cell-names properties
From: Alexander Stein @ 2022-01-26 14:47 UTC (permalink / raw)
To: Joakim Zhang, David S . Miller, Jakub Kicinski, Rob Herring
Cc: Alexander Stein, netdev, devicetree
These properties are inherited from ethernet-controller.yaml.
This fixes the dt_binding_check warning:
imx8mm-tqma8mqml-mba8mx.dt.yaml: ethernet@30be0000: 'nvmem-cell-names',
'nvmem-cells' do not match any of the regexes: 'pinctrl-[0-9]+'
Signed-off-by: Alexander Stein <alexander.stein@ew.tq-group.com>
---
Documentation/devicetree/bindings/net/fsl,fec.yaml | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/Documentation/devicetree/bindings/net/fsl,fec.yaml b/Documentation/devicetree/bindings/net/fsl,fec.yaml
index daa2f79a294f..73616924fa29 100644
--- a/Documentation/devicetree/bindings/net/fsl,fec.yaml
+++ b/Documentation/devicetree/bindings/net/fsl,fec.yaml
@@ -121,6 +121,10 @@ properties:
mac-address: true
+ nvmem-cells: true
+
+ nvmem-cell-names: true
+
tx-internal-delay-ps:
enum: [0, 2000]
--
2.25.1
^ permalink raw reply related
* Re: [RFC PATCH v2 4/7] media: bcm2835-unicam: Add support for for CCP2/CSI2 camera interface
From: Jean-Michel Hautbois @ 2022-01-26 14:21 UTC (permalink / raw)
To: Laurent Pinchart
Cc: dave.stevenson, devicetree, kernel-list, linux-arm-kernel,
linux-kernel, linux-media, linux-rpi-kernel, lukasz, mchehab,
naush, robh, tomi.valkeinen
In-Reply-To: <YeySvkxEfpzTj+pC@pendragon.ideasonboard.com>
Hi Laurent, thanks for the review !
A question for Dave below :-)
On 23/01/2022 00:26, Laurent Pinchart wrote:
> Hi Jean-Michel,
>
> Thank you for the patch.
>
> Dave, Naush, there are a few questions for you below.
>
> On Fri, Jan 21, 2022 at 09:18:07AM +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@raspberrypi.com>
>> Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
>> Signed-off-by: Jean-Michel Hautbois <jeanmichel.hautbois@ideasonboard.com>
>>
>> ---
>> in v2: Remove the unicam_{info,debug,error} macros and use
>> dev_dbg/dev_err instead.
>> ---
>> MAINTAINERS | 1 +
>> drivers/media/platform/Kconfig | 1 +
>> drivers/media/platform/Makefile | 2 +
>> drivers/media/platform/bcm2835/Kconfig | 21 +
>> drivers/media/platform/bcm2835/Makefile | 3 +
>> .../media/platform/bcm2835/bcm2835-unicam.c | 2678 +++++++++++++++++
>> .../media/platform/bcm2835/vc4-regs-unicam.h | 253 ++
>> 7 files changed, 2959 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.c
>> create mode 100644 drivers/media/platform/bcm2835/vc4-regs-unicam.h
>>
>> diff --git a/MAINTAINERS b/MAINTAINERS
>> index 07f238fd5ff9..b17bb533e007 100644
>> --- a/MAINTAINERS
>> +++ b/MAINTAINERS
>> @@ -3684,6 +3684,7 @@ M: Raspberry Pi Kernel Maintenance <kernel-list@raspberrypi.com>
>> L: linux-media@vger.kernel.org
>> S: Maintained
>> F: Documentation/devicetree/bindings/media/brcm,bcm2835-unicam.yaml
>> +F: drivers/media/platform/bcm2835/
>>
>> BROADCOM BCM47XX MIPS ARCHITECTURE
>> M: Hauke Mehrtens <hauke@hauke-m.de>
>> diff --git a/drivers/media/platform/Kconfig b/drivers/media/platform/Kconfig
>> index 9fbdba0fd1e7..033b0358fbb8 100644
>> --- a/drivers/media/platform/Kconfig
>> +++ b/drivers/media/platform/Kconfig
>> @@ -170,6 +170,7 @@ source "drivers/media/platform/am437x/Kconfig"
>> source "drivers/media/platform/xilinx/Kconfig"
>> source "drivers/media/platform/rcar-vin/Kconfig"
>> source "drivers/media/platform/atmel/Kconfig"
>> +source "drivers/media/platform/bcm2835/Kconfig"
>> source "drivers/media/platform/sunxi/Kconfig"
>>
>> config VIDEO_TI_CAL
>> diff --git a/drivers/media/platform/Makefile b/drivers/media/platform/Makefile
>> index 19bcbced7382..689e64857db1 100644
>> --- a/drivers/media/platform/Makefile
>> +++ b/drivers/media/platform/Makefile
>> @@ -85,6 +85,8 @@ obj-$(CONFIG_VIDEO_QCOM_CAMSS) += qcom/camss/
>>
>> obj-$(CONFIG_VIDEO_QCOM_VENUS) += qcom/venus/
>>
>> +obj-y += bcm2835/
>
> One day it would be nice to clean the media Kconfig and Makefile
> infrastructure to use a consistent style for all drivers. One can always
> dream :-)
>
>> +
>> obj-y += sunxi/
>>
>> obj-$(CONFIG_VIDEO_MESON_GE2D) += meson/ge2d/
>> diff --git a/drivers/media/platform/bcm2835/Kconfig b/drivers/media/platform/bcm2835/Kconfig
>> new file mode 100644
>> index 000000000000..bd1370199650
>> --- /dev/null
>> +++ b/drivers/media/platform/bcm2835/Kconfig
>> @@ -0,0 +1,21 @@
>> +# Broadcom VideoCore4 V4L2 camera support
>> +
>> +config VIDEO_BCM2835_UNICAM
>> + tristate "Broadcom BCM283x/BCM271x Unicam video capture driver"
>> + depends on VIDEO_V4L2
>> + depends on ARCH_BCM2835 || COMPILE_TEST
>
> I would put the arch dependency first, but that may be just me.
>
>> + select VIDEO_V4L2_SUBDEV_API
>> + select MEDIA_CONTROLLER
>> + select VIDEOBUF2_DMA_CONTIG
>> + select V4L2_FWNODE
>
> And sort there alphabetically.
>
>> + help
>> + Say Y here to enable support for the BCM283x/BCM271x CSI-2 receiver.
>> + This is a V4L2 driver that controls the CSI-2 receiver directly,
>> + independently from the VC4 firmware.
>> + This driver is mutually exclusive with the use of bcm2835-camera. The
>> + firmware will disable all access to the peripheral from within the
>> + firmware if it finds a DT node using it, and bcm2835-camera will
>> + therefore fail to probe.
>
> Dave, Naush, what should we do with the bcm2835-camera driver in staging
> ? Do we need to keep it to support older Pi models ?
>
>> +
>> + To compile this driver as a module, choose M here. The module will be
>> + called bcm2835-unicam.
>> diff --git a/drivers/media/platform/bcm2835/Makefile b/drivers/media/platform/bcm2835/Makefile
>> new file mode 100644
>> index 000000000000..a98aba03598a
>> --- /dev/null
>> +++ b/drivers/media/platform/bcm2835/Makefile
>> @@ -0,0 +1,3 @@
>> +# Makefile for BCM2835 Unicam driver
>> +
>> +obj-$(CONFIG_VIDEO_BCM2835_UNICAM) += bcm2835-unicam.o
>> diff --git a/drivers/media/platform/bcm2835/bcm2835-unicam.c b/drivers/media/platform/bcm2835/bcm2835-unicam.c
>> new file mode 100644
>> index 000000000000..7636496be00a
>> --- /dev/null
>> +++ b/drivers/media/platform/bcm2835/bcm2835-unicam.c
>> @@ -0,0 +1,2678 @@
>> +// SPDX-License-Identifier: GPL-2.0-only
>> +/*
>> + * BCM283x / BCM271x Unicam Capture Driver
>> + *
>> + * Copyright (C) 2017-2020 - Raspberry Pi (Trading) Ltd.
>> + *
>> + * Dave Stevenson <dave.stevenson@raspberrypi.com>
>> + *
>> + * Based on TI am437x driver by
>> + * Benoit Parrot <bparrot@ti.com>
>> + * Lad, Prabhakar <prabhakar.csengg@gmail.com>
>> + *
>> + * and TI CAL camera interface driver by
>> + * Benoit Parrot <bparrot@ti.com>
>> + *
>> + *
>> + * There are two camera drivers in the kernel for BCM283x - this one
>> + * and bcm2835-camera (currently in staging).
>> + *
>> + * This driver directly controls the Unicam peripheral - there is no
>> + * involvement with the VideoCore firmware. Unicam receives CSI-2 or
>> + * CCP2 data and writes it into SDRAM.
>> + * The only potential processing options are to repack Bayer data into an
>> + * alternate format, and applying windowing.
>> + * The repacking does not shift the data, so can repack V4L2_PIX_FMT_Sxxxx10P
>> + * to V4L2_PIX_FMT_Sxxxx10, or V4L2_PIX_FMT_Sxxxx12P to V4L2_PIX_FMT_Sxxxx12,
>> + * but not generically up to V4L2_PIX_FMT_Sxxxx16. The driver will add both
>> + * formats where the relevant formats are defined, and will automatically
>> + * configure the repacking as required.
>> + * Support for windowing may be added later.
>> + *
>> + * It should be possible to connect this driver to any sensor with a
>> + * suitable output interface and V4L2 subdevice driver.
>> + *
>> + * bcm2835-camera uses the VideoCore firmware to control the sensor,
>> + * Unicam, ISP, and all tuner control loops. Fully processed frames are
>> + * delivered to the driver by the firmware. It only has sensor drivers
>> + * for Omnivision OV5647, and Sony IMX219 sensors.
>> + *
>> + * The two drivers are mutually exclusive for the same Unicam instance.
>> + * The VideoCore firmware checks the device tree configuration during boot.
>> + * If it finds device tree nodes called csi0 or csi1 it will block the
>> + * firmware from accessing the peripheral, and bcm2835-camera will
>> + * not be able to stream data.
>> + */
>> +
>> +#include <linux/clk.h>
>> +#include <linux/delay.h>
>> +#include <linux/device.h>
>> +#include <linux/dma-mapping.h>
>> +#include <linux/err.h>
>> +#include <linux/init.h>
>> +#include <linux/interrupt.h>
>> +#include <linux/io.h>
>> +#include <linux/module.h>
>> +#include <linux/of_device.h>
>> +#include <linux/of_graph.h>
>> +#include <linux/pinctrl/consumer.h>
>> +#include <linux/platform_device.h>
>> +#include <linux/pm_runtime.h>
>> +#include <linux/slab.h>
>> +#include <linux/uaccess.h>
>> +#include <linux/videodev2.h>
>> +
>> +#include <media/v4l2-common.h>
>> +#include <media/v4l2-ctrls.h>
>> +#include <media/v4l2-dev.h>
>> +#include <media/v4l2-device.h>
>> +#include <media/v4l2-dv-timings.h>
>> +#include <media/v4l2-event.h>
>> +#include <media/v4l2-ioctl.h>
>> +#include <media/v4l2-fwnode.h>
>> +#include <media/videobuf2-dma-contig.h>
>> +
>> +#include <media/v4l2-async.h>
>> +#define v4l2_async_nf_add_subdev __v4l2_async_nf_add_subdev
>
> Don't do that ! It's the kind of pattern that a malicious actor would
> use to obfuscate code and get vulnerabilities merged.
>
>> +
>> +#include "vc4-regs-unicam.h"
>> +
>> +#define UNICAM_MODULE_NAME "unicam"
>> +#define UNICAM_VERSION "0.1.0"
>
> I doubt the version would ever be incremented, so let's just drop it.
>
>> +
>> +static int debug;
>> +module_param(debug, int, 0644);
>> +MODULE_PARM_DESC(debug, "Debug level 0-3");
>
> Not used.
>
>> +
>> +/*
>> + * Unicam must request a minimum of 250Mhz from the VPU clock.
>> + * Otherwise the input FIFOs overrun and cause image corruption.
>> + */
>> +#define MIN_VPU_CLOCK_RATE (250 * 1000 * 1000)
>> +/*
>> + * To protect against a dodgy sensor driver never returning an error from
>> + * enum_mbus_code, set a maximum index value to be used.
>> + */
>> +#define MAX_ENUM_MBUS_CODE 128
>> +
>> +/*
>> + * Stride is a 16 bit register, but also has to be a multiple of 32.
>> + */
>> +#define BPL_ALIGNMENT 32
>> +#define MAX_BYTESPERLINE ((1 << 16) - BPL_ALIGNMENT)
>> +/*
>> + * Max width is therefore determined by the max stride divided by
>> + * the number of bits per pixel. Take 32bpp as a
>> + * worst case.
>> + * No imposed limit on the height, so adopt a square image for want
>> + * of anything better.
>> + */
>> +#define MAX_WIDTH (MAX_BYTESPERLINE / 4)
>> +#define MAX_HEIGHT MAX_WIDTH
>> +/* Define a nominal minimum image size */
>> +#define MIN_WIDTH 16
>> +#define MIN_HEIGHT 16
>> +/* Default size of the embedded buffer */
>> +#define UNICAM_EMBEDDED_SIZE 16384
>> +
>> +/*
>> + * Size of the dummy buffer. Can be any size really, but the DMA
>> + * allocation works in units of page sizes.
>> + */
>> +#define DUMMY_BUF_SIZE (PAGE_SIZE)
>
> No need for parentheses.
>
> Some of those macros have toogeneric names and may result in namespace
> clashes. Please prefix them all with UNICAM_.
>
>> +
>> +#define UNICAM_SD_PAD_SINK 0
>> +#define UNICAM_SD_PAD_FIRST_SOURCE 1
>
> I'd name the source pads explicitly.
>
> #define UNICAM_SD_PAD_SOURCE_IMAGE 1
> #define UNICAM_SD_PAD_SOURCE_METADATA 2
>
> (or META instead of METADATA, up to you, as long as we use consistent
> naming through the driver).
>
>> +#define UNICAM_SD_NUM_SOURCE_PADS 2
>> +#define UNICAM_SD_NUM_PADS (1 + UNICAM_SD_NUM_SOURCE_PADS)
>> +
>> +static inline bool unicam_sd_pad_is_sink(u32 pad)
>> +{
>> + /* Camera RX has 1 sink pad, and N source pads */
>> + return pad == 0;
>> +}
>> +
>> +static inline bool unicam_sd_pad_is_source(u32 pad)
>> +{
>> + /* Camera RX has 1 sink pad, and N source pads */
>> + return pad >= UNICAM_SD_PAD_FIRST_SOURCE &&
>> + pad <= UNICAM_SD_NUM_SOURCE_PADS;
>
> return pad != UNICAM_SD_PAD_SINK;
>
> and drop UNICAM_SD_PAD_FIRST_SOURCE and UNICAM_SD_NUM_SOURCE_PADS.
>
>> +}
>> +
>> +enum node_types {
>
> enum unicam_node_type {
>
>> + IMAGE_NODE,
>> + METADATA_NODE,
>> + MAX_NODES
>> +};
>
> UNICAM_ prefixes please.
>
>> +
>> +#define MASK_CS_DEFAULT BIT(V4L2_COLORSPACE_DEFAULT)
>
> This and several of the macros below are not used.
>
>> +#define MASK_CS_SMPTE170M BIT(V4L2_COLORSPACE_SMPTE170M)
>> +#define MASK_CS_SMPTE240M BIT(V4L2_COLORSPACE_SMPTE240M)
>> +#define MASK_CS_REC709 BIT(V4L2_COLORSPACE_REC709)
>> +#define MASK_CS_BT878 BIT(V4L2_COLORSPACE_BT878)
>> +#define MASK_CS_470_M BIT(V4L2_COLORSPACE_470_SYSTEM_M)
>> +#define MASK_CS_470_BG BIT(V4L2_COLORSPACE_470_SYSTEM_BG)
>> +#define MASK_CS_JPEG BIT(V4L2_COLORSPACE_JPEG)
>> +#define MASK_CS_SRGB BIT(V4L2_COLORSPACE_SRGB)
>> +#define MASK_CS_OPRGB BIT(V4L2_COLORSPACE_OPRGB)
>> +#define MASK_CS_BT2020 BIT(V4L2_COLORSPACE_BT2020)
>> +#define MASK_CS_RAW BIT(V4L2_COLORSPACE_RAW)
>> +#define MASK_CS_DCI_P3 BIT(V4L2_COLORSPACE_DCI_P3)
>> +
>> +#define MAX_COLORSPACE 32
>> +
>> +/*
>> + * struct unicam_fmt - Unicam media bus format information
>> + * @pixelformat: V4L2 pixel format FCC identifier. 0 if n/a.
>> + * @repacked_fourcc: V4L2 pixel format FCC identifier if the data is expanded
>> + * out to 16bpp. 0 if n/a.
>> + * @code: V4L2 media bus format code.
>> + * @depth: Bits per pixel as delivered from the source.
>> + * @csi_dt: CSI data type.
>> + * @valid_colorspaces: Bitmask of valid colorspaces so that the Media Controller
>> + * centric try_fmt can validate the colorspace and pass
>> + * v4l2-compliance.
>
> I'd actually drop colorspace support completely. Unicam doesn't affect
> the colorspace. It will output whatever it receives from the sensor. You
> can accept any colorspace on the sink pad of the subdev, and just
> propagate that to the source pad.
>
> If that causes v4l2-compliance failures, then we have a problem either
> in the sensor drivers, or in v4l2-compliance.
>
>> + * @check_variants: Flag to denote that there are multiple mediabus formats
>> + * still in the list that could match this V4L2 format.
>> + * @mc_skip: Media Controller shouldn't list this format via ENUM_FMT as it is
>> + * a duplicate of an earlier format.
>> + * @metadata_fmt: This format only applies to the metadata pad.
>> + */
>> +struct unicam_fmt {
>> + u32 fourcc;
>> + u32 repacked_fourcc;
>> + u32 code;
>> + u8 depth;
>> + u8 csi_dt;
>> + u32 valid_colorspaces;
>> + u8 check_variants:1;
>> + u8 mc_skip:1;
>> + u8 metadata_fmt:1;
>> +};
>> +
>> +static const struct unicam_fmt formats[] = {
>> + /* YUV Formats */
>> + {
>> + .fourcc = V4L2_PIX_FMT_YUYV,
>> + .code = MEDIA_BUS_FMT_YUYV8_2X8,
>
> The convention for CSI-2 buses is to used the _1X16 media bus codes for
> YUV. You can drop the first four entries, and drop the check_variants
> and mc_skip fields.
>
>> + .depth = 16,
>> + .csi_dt = 0x1e,
>> + .check_variants = 1,
>> + .valid_colorspaces = MASK_CS_SMPTE170M | MASK_CS_REC709 |
>> + MASK_CS_JPEG,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_UYVY,
>> + .code = MEDIA_BUS_FMT_UYVY8_2X8,
>> + .depth = 16,
>> + .csi_dt = 0x1e,
>> + .check_variants = 1,
>> + .valid_colorspaces = MASK_CS_SMPTE170M | MASK_CS_REC709 |
>> + MASK_CS_JPEG,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_YVYU,
>> + .code = MEDIA_BUS_FMT_YVYU8_2X8,
>> + .depth = 16,
>> + .csi_dt = 0x1e,
>> + .check_variants = 1,
>> + .valid_colorspaces = MASK_CS_SMPTE170M | MASK_CS_REC709 |
>> + MASK_CS_JPEG,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_VYUY,
>> + .code = MEDIA_BUS_FMT_VYUY8_2X8,
>> + .depth = 16,
>> + .csi_dt = 0x1e,
>> + .check_variants = 1,
>> + .valid_colorspaces = MASK_CS_SMPTE170M | MASK_CS_REC709 |
>> + MASK_CS_JPEG,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_YUYV,
>> + .code = MEDIA_BUS_FMT_YUYV8_1X16,
>> + .depth = 16,
>> + .csi_dt = 0x1e,
>> + .mc_skip = 1,
>> + .valid_colorspaces = MASK_CS_SMPTE170M | MASK_CS_REC709 |
>> + MASK_CS_JPEG,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_UYVY,
>> + .code = MEDIA_BUS_FMT_UYVY8_1X16,
>> + .depth = 16,
>> + .csi_dt = 0x1e,
>> + .mc_skip = 1,
>> + .valid_colorspaces = MASK_CS_SMPTE170M | MASK_CS_REC709 |
>> + MASK_CS_JPEG,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_YVYU,
>> + .code = MEDIA_BUS_FMT_YVYU8_1X16,
>> + .depth = 16,
>> + .csi_dt = 0x1e,
>> + .mc_skip = 1,
>> + .valid_colorspaces = MASK_CS_SMPTE170M | MASK_CS_REC709 |
>> + MASK_CS_JPEG,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_VYUY,
>> + .code = MEDIA_BUS_FMT_VYUY8_1X16,
>> + .depth = 16,
>> + .csi_dt = 0x1e,
>> + .mc_skip = 1,
>> + .valid_colorspaces = MASK_CS_SMPTE170M | MASK_CS_REC709 |
>> + MASK_CS_JPEG,
>> + }, {
>> + /* RGB Formats */
>> + .fourcc = V4L2_PIX_FMT_RGB565, /* gggbbbbb rrrrrggg */
>> + .code = MEDIA_BUS_FMT_RGB565_2X8_LE,
>> + .depth = 16,
>> + .csi_dt = 0x22,
>> + .valid_colorspaces = MASK_CS_SRGB,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_RGB565X, /* rrrrrggg gggbbbbb */
>> + .code = MEDIA_BUS_FMT_RGB565_2X8_BE,
>> + .depth = 16,
>> + .csi_dt = 0x22,
>> + .valid_colorspaces = MASK_CS_SRGB,
>
> CSI-2 specifies the RGB565 format unambiguously, there are no little
> endian of big endian variants. Let's drop one of the two entries, and
> use MEDIA_BUS_FMT_RGB565_1X15. Dave, Naush, what is the correct pixel
> format for this, is data layed out in big endian or little endian ?
>
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_RGB555, /* gggbbbbb arrrrrgg */
>> + .code = MEDIA_BUS_FMT_RGB555_2X8_PADHI_LE,
>> + .depth = 16,
>> + .csi_dt = 0x21,
>> + .valid_colorspaces = MASK_CS_SRGB,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_RGB555X, /* arrrrrgg gggbbbbb */
>> + .code = MEDIA_BUS_FMT_RGB555_2X8_PADHI_BE,
>> + .depth = 16,
>> + .csi_dt = 0x21,
>> + .valid_colorspaces = MASK_CS_SRGB,
>
> Same here, but we need to define a new MEDIA_BUS_FMT_RGB555_1X16 format.
> The CSI-2 RGB555 data type has the padding bit between the green and
> blue components, so I wonder if Unicam maps that to one of
> V4L2_PIX_FMT_RGB555 or V4L2_PIX_FMT_RGB555X, or if a new pixel format is
> needed too.
>
> Another option would be to drop RGB555 support until someone needs it
> (with a comment in the table to tell the entry is missing).
>
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_RGB24, /* rgb */
>> + .code = MEDIA_BUS_FMT_RGB888_1X24,
>> + .depth = 24,
>> + .csi_dt = 0x24,
>> + .valid_colorspaces = MASK_CS_SRGB,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_BGR24, /* bgr */
>> + .code = MEDIA_BUS_FMT_BGR888_1X24,
>> + .depth = 24,
>> + .csi_dt = 0x24,
>> + .valid_colorspaces = MASK_CS_SRGB,
>
> This is possibly more tricky, as CSI-2 defined RGB888 as being
> transmitted in BGR order, but sensors could possibly also support RGB as
> a non-standard extension (the same may CSI-2 standardizes on UYVY, but
> many sensors can swap components). We can keep both entry I suppose.
>
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_RGB32, /* argb */
>> + .code = MEDIA_BUS_FMT_ARGB8888_1X32,
>> + .depth = 32,
>> + .csi_dt = 0x0,
>> + .valid_colorspaces = MASK_CS_SRGB,
>
> There's no 32-bit RGB in CSI-2, should this be dropped ?
>
>> + }, {
>> + /* Bayer Formats */
>> + .fourcc = V4L2_PIX_FMT_SBGGR8,
>> + .code = MEDIA_BUS_FMT_SBGGR8_1X8,
>> + .depth = 8,
>> + .csi_dt = 0x2a,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_SGBRG8,
>> + .code = MEDIA_BUS_FMT_SGBRG8_1X8,
>> + .depth = 8,
>> + .csi_dt = 0x2a,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_SGRBG8,
>> + .code = MEDIA_BUS_FMT_SGRBG8_1X8,
>> + .depth = 8,
>> + .csi_dt = 0x2a,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_SRGGB8,
>> + .code = MEDIA_BUS_FMT_SRGGB8_1X8,
>> + .depth = 8,
>> + .csi_dt = 0x2a,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_SBGGR10P,
>> + .repacked_fourcc = V4L2_PIX_FMT_SBGGR10,
>> + .code = MEDIA_BUS_FMT_SBGGR10_1X10,
>> + .depth = 10,
>> + .csi_dt = 0x2b,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_SGBRG10P,
>> + .repacked_fourcc = V4L2_PIX_FMT_SGBRG10,
>> + .code = MEDIA_BUS_FMT_SGBRG10_1X10,
>> + .depth = 10,
>> + .csi_dt = 0x2b,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_SGRBG10P,
>> + .repacked_fourcc = V4L2_PIX_FMT_SGRBG10,
>> + .code = MEDIA_BUS_FMT_SGRBG10_1X10,
>> + .depth = 10,
>> + .csi_dt = 0x2b,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_SRGGB10P,
>> + .repacked_fourcc = V4L2_PIX_FMT_SRGGB10,
>> + .code = MEDIA_BUS_FMT_SRGGB10_1X10,
>> + .depth = 10,
>> + .csi_dt = 0x2b,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_SBGGR12P,
>> + .repacked_fourcc = V4L2_PIX_FMT_SBGGR12,
>> + .code = MEDIA_BUS_FMT_SBGGR12_1X12,
>> + .depth = 12,
>> + .csi_dt = 0x2c,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_SGBRG12P,
>> + .repacked_fourcc = V4L2_PIX_FMT_SGBRG12,
>> + .code = MEDIA_BUS_FMT_SGBRG12_1X12,
>> + .depth = 12,
>> + .csi_dt = 0x2c,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_SGRBG12P,
>> + .repacked_fourcc = V4L2_PIX_FMT_SGRBG12,
>> + .code = MEDIA_BUS_FMT_SGRBG12_1X12,
>> + .depth = 12,
>> + .csi_dt = 0x2c,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_SRGGB12P,
>> + .repacked_fourcc = V4L2_PIX_FMT_SRGGB12,
>> + .code = MEDIA_BUS_FMT_SRGGB12_1X12,
>> + .depth = 12,
>> + .csi_dt = 0x2c,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_SBGGR14P,
>> + .repacked_fourcc = V4L2_PIX_FMT_SBGGR14,
>> + .code = MEDIA_BUS_FMT_SBGGR14_1X14,
>> + .depth = 14,
>> + .csi_dt = 0x2d,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_SGBRG14P,
>> + .repacked_fourcc = V4L2_PIX_FMT_SGBRG14,
>> + .code = MEDIA_BUS_FMT_SGBRG14_1X14,
>> + .depth = 14,
>> + .csi_dt = 0x2d,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_SGRBG14P,
>> + .repacked_fourcc = V4L2_PIX_FMT_SGRBG14,
>> + .code = MEDIA_BUS_FMT_SGRBG14_1X14,
>> + .depth = 14,
>> + .csi_dt = 0x2d,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_SRGGB14P,
>> + .repacked_fourcc = V4L2_PIX_FMT_SRGGB14,
>> + .code = MEDIA_BUS_FMT_SRGGB14_1X14,
>> + .depth = 14,
>> + .csi_dt = 0x2d,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + /*
>> + * 16 bit Bayer formats could be supported, but there is no CSI2
>> + * data_type defined for raw 16, and no sensors that produce it at
>> + * present.
>> + */
>
> RAW16 is now defined in CSI-2, but that can be handled later.
>
>> +
>> + /* Greyscale formats */
>> + .fourcc = V4L2_PIX_FMT_GREY,
>> + .code = MEDIA_BUS_FMT_Y8_1X8,
>> + .depth = 8,
>> + .csi_dt = 0x2a,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_Y10P,
>> + .repacked_fourcc = V4L2_PIX_FMT_Y10,
>> + .code = MEDIA_BUS_FMT_Y10_1X10,
>> + .depth = 10,
>> + .csi_dt = 0x2b,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_Y12P,
>> + .repacked_fourcc = V4L2_PIX_FMT_Y12,
>> + .code = MEDIA_BUS_FMT_Y12_1X12,
>> + .depth = 12,
>> + .csi_dt = 0x2c,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + }, {
>> + .fourcc = V4L2_PIX_FMT_Y14P,
>> + .repacked_fourcc = V4L2_PIX_FMT_Y14,
>> + .code = MEDIA_BUS_FMT_Y14_1X14,
>> + .depth = 14,
>> + .csi_dt = 0x2d,
>> + .valid_colorspaces = MASK_CS_RAW,
>> + },
>> + /* Embedded data format */
>> + {
>> + .fourcc = V4L2_META_FMT_8,
>> + .code = MEDIA_BUS_FMT_METADATA_8,
>> + .depth = 8,
>> + .metadata_fmt = 1,
>> + }
>> +};
>
> Let's reorder the code a bit to improve readability:
>
> - Macro definitions
> - Structure definitions (with the container_of wrappers following the
> respective structure)
> - Misc helper functions (from is_metadata_node() to reg_write_field())
> - Format data table and helper functions (from find_format_by_code() to
> unicam_calc_format_size_bpl())
> - Hardware handling:
> - From unicam_wr_dma_addr() to unicam_isr()
> - From unicam_set_packing_config() to unicam_disable()
> - V4L2 subdev (ops & initialization & registration)
> - vb2 queue opq
> - V4L2 video device (ops & initialization & registration)
> - Probe
>
> I recommend adding a comment header in front of each section after the
> structure definitions with a distinctive visual appearance, to delimit
> code blocks very clearly. I usually use the following style:
>
> /* -----------------------------------------------------------------------------
> * V4L2 Subdev Operations
> */
>
>> +
>> +struct unicam_buffer {
>> + struct vb2_v4l2_buffer vb;
>> + struct list_head list;
>> +};
>> +
>> +static inline struct unicam_buffer *to_unicam_buffer(struct vb2_buffer *vb)
>> +{
>> + return container_of(vb, struct unicam_buffer, vb.vb2_buf);
>> +}
>> +
>> +struct unicam_node {
>> + bool registered;
>> + int open;
>
> Not used. Please go through all structure fields and delete the ones
> that are not used.
>
>> + bool streaming;
>> + unsigned int node_id;
>> + /* Source pad id on the sensor for this node */
>> + unsigned int src_pad_id;
>> + /* Pointer pointing to current v4l2_buffer */
>> + struct unicam_buffer *cur_frm;
>> + /* Pointer pointing to next v4l2_buffer */
>> + struct unicam_buffer *next_frm;
>> + /* video capture */
>> + const struct unicam_fmt *fmt;
>> + /* Used to store current pixel format */
>> + struct v4l2_format v_fmt;
>> + /* Used to store current mbus frame format */
>> + struct v4l2_mbus_framefmt m_fmt;
>> + /* Buffer queue used in video-buf */
>> + struct vb2_queue buffer_queue;
>> + /* Queue of filled frames */
>> + struct list_head dma_queue;
>> + /* IRQ lock for DMA queue */
>> + spinlock_t dma_queue_lock;
>> + /* lock used to access this structure */
>> + struct mutex lock;
>> + /* Identifies video device for this channel */
>> + struct video_device video_dev;
>> + /* Pointer to the parent handle */
>> + struct unicam_device *dev;
>> + struct media_pad pad;
>> + unsigned int embedded_lines;
>> + struct media_pipeline pipe;
>> + /*
>> + * Dummy buffer intended to be used by unicam
>> + * if we have no other queued buffers to swap to.
>> + */
>> + void *dummy_buf_cpu_addr;
>> + dma_addr_t dummy_buf_dma_addr;
>> +};
>> +
>> +struct unicam_device {
>> + struct kref kref;
>> +
>> + /* V4l2 specific parameters */
>> + struct v4l2_async_subdev asd;
>> +
>> + /* peripheral base address */
>> + void __iomem *base;
>> + /* clock gating base address */
>> + void __iomem *clk_gate_base;
>> + /* lp clock handle */
>> + struct clk *clock;
>> + /* vpu clock handle */
>> + struct clk *vpu_clock;
>> + /* vpu clock request */
>> + struct clk_request *vpu_req;
>
> Not used (and that may be a problem).
In the original linux-rpi tree, there is this portion of code in
unicam_start_streaming:
dev->vpu_req = clk_request_start(dev->vpu_clock, MIN_VPU_CLOCK_RATE);
if (!dev->vpu_req) {
unicam_err(dev, "failed to set up VPU clock\n");
goto error_pipeline;
}
ret = clk_prepare_enable(dev->vpu_clock);
if (ret) {
unicam_err(dev, "Failed to enable VPU clock: %d\n", ret);
goto error_pipeline;
}
And this is used as this because it depends on the non-merged series
"clk: [PATCH v2 0/3] clk: Implement a clock request API" [1]
[1]: https://lore.kernel.org/all/20210914093515.260031-1-maxime@cerno.tech/
That's why I modified the code and called:
clk_set_min_rate(dev->vpu_clock, UNICAM_MIN_VPU_CLOCK_RATE);
Dave, is it ok or do we need absolutely the clock request API ?
JM
>
>> + /* clock status for error handling */
>> + bool clocks_enabled;
>> + /* V4l2 device */
>> + struct v4l2_device v4l2_dev;
>> + struct media_device mdev;
>> +
>> + /* parent device */
>> + struct platform_device *pdev;
>
> The pdev field is only used to access pdev.dev, so I'd replace it with
> the struct device.
>
>> + /* subdevice async Notifier */
>> + struct v4l2_async_notifier notifier;
>> + unsigned int sequence;
>> +
>> + /* ptr to sub device */
>> + struct v4l2_subdev *sensor;
>> + /* Pad config for the sensor */
>> + struct v4l2_subdev_state *sensor_state;
>> +
>> + /* Internal subdev */
>> + struct v4l2_subdev sd;
>> + struct media_pad pads[UNICAM_SD_NUM_PADS];
>> +
>> + bool streaming;
>
> "streaming" is ambiguous, it's not clear it relates to the subdev. I'd
> group all the fields related to the subdev in a structure:
>
> struct {
> struct v4l2_subdev sd;
> struct media_pad pads[UNICAM_SD_NUM_PADS];
> bool streaming;
> } subdev;
>
> (there may be more, I haven't checked)
>
>> +
>> + enum v4l2_mbus_type bus_type;
>> + /*
>> + * Stores bus.mipi_csi2.flags for CSI2 sensors, or
>> + * bus.mipi_csi1.strobe for CCP2.
>> + */
>> + unsigned int bus_flags;
>> + unsigned int max_data_lanes;
>> + unsigned int active_data_lanes;
>> + bool sensor_embedded_data;
>> +
>> + struct unicam_node node[MAX_NODES];
>> + struct v4l2_ctrl_handler ctrl_handler;
>> +
>> + bool mc_api;
>> +};
>> +
>> +static inline struct unicam_device *
>> +to_unicam_device(struct v4l2_device *v4l2_dev)
>
> I'd rename this to v4l2_device_to_unicam_device(), as there's another
> container_of wrapper.
>
>> +{
>> + return container_of(v4l2_dev, struct unicam_device, v4l2_dev);
>> +}
>> +
>> +static inline struct unicam_device *
>> +sd_to_unicam_device(struct v4l2_subdev *sd)
>> +{
>> + return container_of(sd, struct unicam_device, sd);
>> +}
>> +
>> +static inline bool is_metadata_node(struct unicam_node *node)
>> +{
>> + return node->video_dev.device_caps & V4L2_CAP_META_CAPTURE;
>> +}
>> +
>> +static inline bool is_image_node(struct unicam_node *node)
>> +{
>> + return node->video_dev.device_caps & V4L2_CAP_VIDEO_CAPTURE;
>> +}
>> +
>> +/* Hardware access */
>> +static inline void clk_write(struct unicam_device *dev, u32 val)
>> +{
>> + writel(val | 0x5a000000, dev->clk_gate_base);
>> +}
>
> The name of this function and the functions below are too generic. Add a
> unicam_ prefix.
>
>> +
>> +static inline u32 reg_read(struct unicam_device *dev, u32 offset)
>> +{
>> + return readl(dev->base + offset);
>> +}
>> +
>> +static inline void reg_write(struct unicam_device *dev, u32 offset, u32 val)
>> +{
>> + writel(val, dev->base + offset);
>> +}
>> +
>> +static inline int get_field(u32 value, u32 mask)
>> +{
>> + return (value & mask) >> __ffs(mask);
>> +}
>> +
>> +static inline void set_field(u32 *valp, u32 field, u32 mask)
>> +{
>> + u32 val = *valp;
>> +
>> + val &= ~mask;
>> + val |= (field << __ffs(mask)) & mask;
>> + *valp = val;
>> +}
>> +
>> +static inline u32 reg_read_field(struct unicam_device *dev, u32 offset,
>> + u32 mask)
>
> Not used.
>
>> +{
>> + return get_field(reg_read(dev, offset), mask);
>> +}
>> +
>> +static inline void reg_write_field(struct unicam_device *dev, u32 offset,
>> + u32 field, u32 mask)
>> +{
>> + u32 val = reg_read(dev, offset);
>> +
>> + set_field(&val, field, mask);
>> + reg_write(dev, offset, val);
>> +}
>
> I'm really not a bit fan of the read-modify-update patterns, more often
> than not they can be replaced by direct writes. No need to fix this
> though, it's ok.
>
>> +
>> +/* Power management functions */
>> +static inline int unicam_runtime_get(struct unicam_device *dev)
>> +{
>> + return pm_runtime_get_sync(&dev->pdev->dev);
>
> Use pm_runtime_resume_and_get() instead, as the error handling is broken
> in the driver otherwise.
>
>> +}
>> +
>> +static inline void unicam_runtime_put(struct unicam_device *dev)
>> +{
>> + pm_runtime_put_sync(&dev->pdev->dev);
>> +}
>
> Drop these wrappers and call the functions directly.
>
>> +
>> +/* Format setup functions */
>> +static const struct unicam_fmt *find_format_by_code(u32 code)
>> +{
>> + unsigned int i;
>> +
>> + for (i = 0; i < ARRAY_SIZE(formats); i++) {
>> + if (formats[i].code == code)
>> + return &formats[i];
>> + }
>> +
>> + return NULL;
>> +}
>> +
>> +static const struct unicam_fmt *find_format_by_fourcc(u32 fourcc)
>> +{
>> + unsigned int i;
>> +
>> + for (i = 0; i < ARRAY_SIZE(formats); ++i) {
>> + if (formats[i].fourcc == fourcc)
>> + return &formats[i];
>> + }
>> +
>> + return NULL;
>> +}
>> +
>> +static unsigned int bytes_per_line(u32 width, const struct unicam_fmt *fmt,
>> + u32 v4l2_fourcc)
>> +{
>> + if (v4l2_fourcc == fmt->repacked_fourcc)
>> + /* Repacking always goes to 16bpp */
>> + return ALIGN(width << 1, BPL_ALIGNMENT);
>> + else
>> + return ALIGN((width * fmt->depth) >> 3, BPL_ALIGNMENT);
>> +}
>> +
>> +static int unicam_calc_format_size_bpl(struct unicam_device *dev,
>> + const struct unicam_fmt *fmt,
>> + struct v4l2_format *f)
>> +{
>> + unsigned int min_bytesperline;
>> +
>> + v4l_bound_align_image(&f->fmt.pix.width, MIN_WIDTH, MAX_WIDTH, 2,
>> + &f->fmt.pix.height, MIN_HEIGHT, MAX_HEIGHT, 0,
>> + 0);
>> +
>> + min_bytesperline = bytes_per_line(f->fmt.pix.width, fmt,
>> + f->fmt.pix.pixelformat);
>> +
>> + if (f->fmt.pix.bytesperline > min_bytesperline &&
>> + f->fmt.pix.bytesperline <= MAX_BYTESPERLINE)
>> + f->fmt.pix.bytesperline = ALIGN(f->fmt.pix.bytesperline,
>> + BPL_ALIGNMENT);
>> + else
>> + f->fmt.pix.bytesperline = min_bytesperline;
>> +
>> + f->fmt.pix.sizeimage = f->fmt.pix.height * f->fmt.pix.bytesperline;
>> +
>> + dev_dbg(dev->v4l2_dev.dev, "%s: fourcc: %08X size: %dx%d bpl:%d img_size:%d\n",
>> + __func__,
>> + f->fmt.pix.pixelformat,
>> + f->fmt.pix.width, f->fmt.pix.height,
>> + f->fmt.pix.bytesperline, f->fmt.pix.sizeimage);
>> +
>> + return 0;
>> +}
>> +
>> +static void unicam_wr_dma_addr(struct unicam_device *dev, dma_addr_t dmaaddr,
>> + unsigned int buffer_size, int node_id)
>> +{
>> + dma_addr_t endaddr = dmaaddr + buffer_size;
>> +
>> + if (node_id == IMAGE_NODE) {
>> + reg_write(dev, UNICAM_IBSA0, dmaaddr);
>> + reg_write(dev, UNICAM_IBEA0, endaddr);
>> + } else {
>> + reg_write(dev, UNICAM_DBSA0, dmaaddr);
>> + reg_write(dev, UNICAM_DBEA0, endaddr);
>> + }
>> +}
>> +
>> +static unsigned int unicam_get_lines_done(struct unicam_device *dev)
>> +{
>> + dma_addr_t start_addr, cur_addr;
>> + unsigned int stride = dev->node[IMAGE_NODE].v_fmt.fmt.pix.bytesperline;
>> + struct unicam_buffer *frm = dev->node[IMAGE_NODE].cur_frm;
>> +
>> + if (!frm)
>> + return 0;
>> +
>> + start_addr = vb2_dma_contig_plane_dma_addr(&frm->vb.vb2_buf, 0);
>> + cur_addr = reg_read(dev, UNICAM_IBWP);
>> + return (unsigned int)(cur_addr - start_addr) / stride;
>> +}
>> +
>> +static void unicam_schedule_next_buffer(struct unicam_node *node)
>> +{
>> + struct unicam_device *dev = node->dev;
>> + struct unicam_buffer *buf;
>> + unsigned int size;
>> + dma_addr_t addr;
>> +
>> + buf = list_first_entry(&node->dma_queue, struct unicam_buffer, list);
>> + node->next_frm = buf;
>> + list_del(&buf->list);
>> +
>> + addr = vb2_dma_contig_plane_dma_addr(&buf->vb.vb2_buf, 0);
>> + if (is_image_node(node)) {
>> + size = node->v_fmt.fmt.pix.sizeimage;
>> + unicam_wr_dma_addr(dev, addr, size, IMAGE_NODE);
>> + } else {
>> + size = node->v_fmt.fmt.meta.buffersize;
>> + unicam_wr_dma_addr(dev, addr, size, METADATA_NODE);
>> + }
>> +}
>> +
>> +static void unicam_schedule_dummy_buffer(struct unicam_node *node)
>> +{
>> + struct unicam_device *dev = node->dev;
>> + int node_id = is_image_node(node) ? IMAGE_NODE : METADATA_NODE;
>> +
>> + dev_dbg(dev->v4l2_dev.dev, "Scheduling dummy buffer for node %d\n", node_id);
>> +
>> + unicam_wr_dma_addr(dev, node->dummy_buf_dma_addr, DUMMY_BUF_SIZE,
>> + node_id);
>> + node->next_frm = NULL;
>> +}
>> +
>> +static void unicam_process_buffer_complete(struct unicam_node *node,
>> + unsigned int sequence)
>> +{
>> + node->cur_frm->vb.field = node->m_fmt.field;
>> + node->cur_frm->vb.sequence = sequence;
>> +
>> + vb2_buffer_done(&node->cur_frm->vb.vb2_buf, VB2_BUF_STATE_DONE);
>> +}
>> +
>> +static void unicam_queue_event_sof(struct unicam_device *unicam)
>> +{
>> + struct v4l2_event event = {
>> + .type = V4L2_EVENT_FRAME_SYNC,
>> + .u.frame_sync.frame_sequence = unicam->sequence,
>> + };
>> +
>> + v4l2_event_queue(&unicam->node[IMAGE_NODE].video_dev, &event);
>> +}
>> +
>> +/*
>> + * unicam_isr : ISR handler for unicam capture
>> + * @irq: irq number
>> + * @dev_id: dev_id ptr
>> + *
>> + * It changes status of the captured buffer, takes next buffer from the queue
>> + * and sets its address in unicam registers
>> + */
>> +static irqreturn_t unicam_isr(int irq, void *dev)
>> +{
>> + struct unicam_device *unicam = dev;
>> + unsigned int lines_done = unicam_get_lines_done(dev);
>> + unsigned int sequence = unicam->sequence;
>> + unsigned int i;
>> + u32 ista, sta;
>> + bool fe;
>> + u64 ts;
>> +
>> + sta = reg_read(unicam, UNICAM_STA);
>> + /* Write value back to clear the interrupts */
>> + reg_write(unicam, UNICAM_STA, sta);
>> +
>> + ista = reg_read(unicam, UNICAM_ISTA);
>> + /* Write value back to clear the interrupts */
>> + reg_write(unicam, UNICAM_ISTA, ista);
>> +
>> + dev_dbg(unicam->v4l2_dev.dev, "ISR: ISTA: 0x%X, STA: 0x%X, sequence %d, lines done %d",
>> + ista, sta, sequence, lines_done);
>> +
>> + if (!(sta & (UNICAM_IS | UNICAM_PI0)))
>> + return IRQ_HANDLED;
>> +
>> + /*
>> + * Look for either the Frame End interrupt or the Packet Capture status
>> + * to signal a frame end.
>> + */
>> + fe = (ista & UNICAM_FEI || sta & UNICAM_PI0);
>
> No need for the outer parentheses.
>
>> +
>> + /*
>> + * We must run the frame end handler first. If we have a valid next_frm
>> + * and we get a simultaneout FE + FS interrupt, running the FS handler
>> + * first would null out the next_frm ptr and we would have lost the
>> + * buffer forever.
>> + */
>> + if (fe) {
>> + /*
>> + * Ensure we have swapped buffers already as we can't
>> + * stop the peripheral. If no buffer is available, use a
>> + * dummy buffer to dump out frames until we get a new buffer
>> + * to use.
>> + */
>> + for (i = 0; i < ARRAY_SIZE(unicam->node); i++) {
>> + if (!unicam->node[i].streaming)
>> + continue;
>> +
>> + /*
>> + * If cur_frm == next_frm, it means we have not had
>> + * a chance to swap buffers, likely due to having
>> + * multiple interrupts occurring simultaneously (like FE
>> + * + FS + LS). In this case, we cannot signal the buffer
>> + * as complete, as the HW will reuse that buffer.
>> + */
>> + if (unicam->node[i].cur_frm &&
>> + unicam->node[i].cur_frm != unicam->node[i].next_frm)
>> + unicam_process_buffer_complete(&unicam->node[i],
>> + sequence);
>> + unicam->node[i].cur_frm = unicam->node[i].next_frm;
>> + }
>> + unicam->sequence++;
>> + }
>> +
>> + if (ista & UNICAM_FSI) {
>> + /*
>> + * Timestamp is to be when the first data byte was captured,
>> + * aka frame start.
>> + */
>> + ts = ktime_get_ns();
>> + for (i = 0; i < ARRAY_SIZE(unicam->node); i++) {
>> + if (!unicam->node[i].streaming)
>> + continue;
>> +
>> + if (unicam->node[i].cur_frm)
>> + unicam->node[i].cur_frm->vb.vb2_buf.timestamp =
>> + ts;
>> + else
>> + dev_dbg(unicam->v4l2_dev.dev, "ISR: [%d] Dropping frame, buffer not available at FS\n",
>> + i);
>> + /*
>> + * Set the next frame output to go to a dummy frame
>> + * if we have not managed to obtain another frame
>> + * from the queue.
>> + */
>> + unicam_schedule_dummy_buffer(&unicam->node[i]);
>> + }
>> +
>> + unicam_queue_event_sof(unicam);
>> + }
>> +
>> + /*
>> + * Cannot swap buffer at frame end, there may be a race condition
>> + * where the HW does not actually swap it if the new frame has
>> + * already started.
>> + */
>> + if (ista & (UNICAM_FSI | UNICAM_LCI) && !fe) {
>> + for (i = 0; i < ARRAY_SIZE(unicam->node); i++) {
>> + if (!unicam->node[i].streaming)
>> + continue;
>> +
>> + spin_lock(&unicam->node[i].dma_queue_lock);
>> + if (!list_empty(&unicam->node[i].dma_queue) &&
>> + !unicam->node[i].next_frm)
>> + unicam_schedule_next_buffer(&unicam->node[i]);
>> + spin_unlock(&unicam->node[i].dma_queue_lock);
>> + }
>> + }
>> +
>> + if (reg_read(unicam, UNICAM_ICTL) & UNICAM_FCM) {
>> + /* Switch out of trigger mode if selected */
>> + reg_write_field(unicam, UNICAM_ICTL, 1, UNICAM_TFC);
>> + reg_write_field(unicam, UNICAM_ICTL, 0, UNICAM_FCM);
>> + }
>> + return IRQ_HANDLED;
>> +}
>> +
>> +/* V4L2 Common IOCTLs */
>> +static int unicam_querycap(struct file *file, void *priv,
>> + struct v4l2_capability *cap)
>> +{
>> + struct unicam_node *node = video_drvdata(file);
>> + struct unicam_device *dev = node->dev;
>> +
>> + strscpy(cap->driver, UNICAM_MODULE_NAME, sizeof(cap->driver));
>> + strscpy(cap->card, UNICAM_MODULE_NAME, sizeof(cap->card));
>> +
>> + snprintf(cap->bus_info, sizeof(cap->bus_info),
>> + "platform:%s", dev_name(&dev->pdev->dev));
>> +
>> + cap->capabilities |= V4L2_CAP_VIDEO_CAPTURE | V4L2_CAP_META_CAPTURE;
>> +
>> + return 0;
>> +}
>> +
>> +static int unicam_log_status(struct file *file, void *fh)
>> +{
>> + struct unicam_node *node = video_drvdata(file);
>> + struct unicam_device *dev = node->dev;
>> + u32 reg;
>> +
>> + /* status for sub devices */
>> + v4l2_device_call_all(&dev->v4l2_dev, 0, core, log_status);
>
> I'm tempted to drop this, bit I won't insist.
>
>> +
>> + dev_info(dev->v4l2_dev.dev, "-----Receiver status-----\n");
>> + dev_info(dev->v4l2_dev.dev, "V4L2 width/height: %ux%u\n",
>> + node->v_fmt.fmt.pix.width, node->v_fmt.fmt.pix.height);
>> + dev_info(dev->v4l2_dev.dev, "Mediabus format: %08x\n", node->fmt->code);
>> + dev_info(dev->v4l2_dev.dev, "V4L2 format: %08x\n",
>> + node->v_fmt.fmt.pix.pixelformat);
>> + reg = reg_read(dev, UNICAM_IPIPE);
>> + dev_info(dev->v4l2_dev.dev, "Unpacking/packing: %u / %u\n",
>> + get_field(reg, UNICAM_PUM_MASK),
>> + get_field(reg, UNICAM_PPM_MASK));
>> + dev_info(dev->v4l2_dev.dev, "----Live data----\n");
>> + dev_info(dev->v4l2_dev.dev, "Programmed stride: %4u\n",
>> + reg_read(dev, UNICAM_IBLS));
>> + dev_info(dev->v4l2_dev.dev, "Detected resolution: %ux%u\n",
>> + reg_read(dev, UNICAM_IHSTA),
>> + reg_read(dev, UNICAM_IVSTA));
>> + dev_info(dev->v4l2_dev.dev, "Write pointer: %08x\n",
>> + reg_read(dev, UNICAM_IBWP));
>> +
>> + return 0;
>> +}
>> +
>> +static int unicam_subscribe_event(struct v4l2_fh *fh,
>> + const struct v4l2_event_subscription *sub)
>> +{
>> + switch (sub->type) {
>> + case V4L2_EVENT_FRAME_SYNC:
>> + return v4l2_event_subscribe(fh, sub, 2, NULL);
>> + case V4L2_EVENT_SOURCE_CHANGE:
>> + return v4l2_event_subscribe(fh, sub, 4, NULL);
>
> The drive doesn't generate V4L2_EVENT_SOURCE_CHANGE events, drop it.
>
>> + }
>> +
>> + return v4l2_ctrl_subscribe_event(fh, sub);
>> +}
>> +
>> +static void unicam_notify(struct v4l2_subdev *sd,
>> + unsigned int notification, void *arg)
>> +{
>> + struct unicam_device *dev = to_unicam_device(sd->v4l2_dev);
>
> Use sd_to_unicam_device().
>
>> +
>> + switch (notification) {
>> + case V4L2_DEVICE_NOTIFY_EVENT:
>> + v4l2_event_queue(&dev->node[IMAGE_NODE].video_dev, arg);
>> + break;
>> + default:
>> + break;
>> + }
>> +}
>
> Not needed, drop it.
>
>> +
>> +/* V4L2 Media Controller Centric IOCTLs */
>> +
>> +static int unicam_mc_enum_fmt_vid_cap(struct file *file, void *priv,
>> + struct v4l2_fmtdesc *f)
>
> Drop "mc" in all function names as that's the only option now.
>
>> +{
>> + int i, j;
>
> Never negative, can be unsigned int.
>
> j is not a loop counter but the format index, I'd rename it to index to
> make it clearer.
>
>> +
>> + for (i = 0, j = 0; i < ARRAY_SIZE(formats); i++) {
>> + if (f->mbus_code && formats[i].code != f->mbus_code)
>> + continue;
>> + if (formats[i].mc_skip || formats[i].metadata_fmt)
>> + continue;
>> +
>> + if (formats[i].fourcc) {
>
> All entries have a fourcc so you can drop this check.
>
>> + if (j == f->index) {
>> + f->pixelformat = formats[i].fourcc;
>> + f->type = V4L2_BUF_TYPE_VIDEO_CAPTURE;
>
> Not needed, the type is an input parameter. Same below.
>
>> + return 0;
>> + }
>> + j++;
>> + }
>> + if (formats[i].repacked_fourcc) {
>> + if (j == f->index) {
>> + f->pixelformat = formats[i].repacked_fourcc;
>> + f->type = V4L2_BUF_TYPE_VIDEO_CAPTURE;
>> + return 0;
>> + }
>> + j++;
>> + }
>> + }
>> +
>> + return -EINVAL;
>> +}
>> +
>> +static int unicam_mc_g_fmt_vid_cap(struct file *file, void *priv,
>> + struct v4l2_format *f)
>> +{
>> + struct unicam_node *node = video_drvdata(file);
>> +
>> + if (!is_image_node(node))
>> + return -EINVAL;
>
> Can this happen ?
>
>> +
>> + *f = node->v_fmt;
>> +
>> + return 0;
>> +}
>> +
>> +static void unicam_mc_try_fmt(struct unicam_node *node, struct v4l2_format *f,
>> + const struct unicam_fmt **ret_fmt)
>> +{
>> + struct v4l2_pix_format *v4l2_format = &f->fmt.pix;
>> + struct unicam_device *dev = node->dev;
>> + const struct unicam_fmt *fmt;
>> + int is_rgb;
>
> bool.
>
>> +
>> + /*
>> + * Default to the first format if the requested pixel format code isn't
>> + * supported.
>> + */
>> + fmt = find_format_by_fourcc(v4l2_format->pixelformat);
>> + if (!fmt) {
>> + fmt = &formats[0];
>> + v4l2_format->pixelformat = fmt->fourcc;
>> + }
>> +
>> + unicam_calc_format_size_bpl(dev, fmt, f);
>> +
>> + if (v4l2_format->field == V4L2_FIELD_ANY)
>> + v4l2_format->field = V4L2_FIELD_NONE;
>> +
>> + if (v4l2_format->colorspace >= MAX_COLORSPACE ||
>> + !(fmt->valid_colorspaces & (1 << v4l2_format->colorspace))) {
>> + v4l2_format->colorspace = __ffs(fmt->valid_colorspaces);
>> +
>> + v4l2_format->xfer_func =
>> + V4L2_MAP_XFER_FUNC_DEFAULT(v4l2_format->colorspace);
>> + v4l2_format->ycbcr_enc =
>> + V4L2_MAP_YCBCR_ENC_DEFAULT(v4l2_format->colorspace);
>> + is_rgb = v4l2_format->colorspace == V4L2_COLORSPACE_SRGB;
>> + v4l2_format->quantization =
>> + V4L2_MAP_QUANTIZATION_DEFAULT(is_rgb,
>> + v4l2_format->colorspace,
>> + v4l2_format->ycbcr_enc);
>> + }
>
> Drop this, we can accept any colorspace produced by the sensor.
>
>> +
>> + if (ret_fmt)
>> + *ret_fmt = fmt;
>
> How about returning the format from the function instead ?
>
>> +
>> + dev_dbg(dev->v4l2_dev.dev, "%s: %08x %ux%u (bytesperline %u sizeimage %u)\n",
>> + __func__, v4l2_format->pixelformat,
>> + v4l2_format->width, v4l2_format->height,
>> + v4l2_format->bytesperline, v4l2_format->sizeimage);
>> +}
>> +
>> +static int unicam_mc_try_fmt_vid_cap(struct file *file, void *priv,
>> + struct v4l2_format *f)
>> +{
>> + struct unicam_node *node = video_drvdata(file);
>> +
>> + unicam_mc_try_fmt(node, f, NULL);
>> + return 0;
>> +}
>> +
>> +static int unicam_mc_s_fmt_vid_cap(struct file *file, void *priv,
>> + struct v4l2_format *f)
>> +{
>> + struct unicam_node *node = video_drvdata(file);
>> + struct unicam_device *dev = node->dev;
>> + const struct unicam_fmt *fmt;
>> +
>> + if (vb2_is_busy(&node->buffer_queue)) {
>> + dev_dbg(dev->v4l2_dev.dev, "%s device busy\n", __func__);
>> + return -EBUSY;
>> + }
>> +
>> + unicam_mc_try_fmt(node, f, &fmt);
>> +
>> + node->v_fmt = *f;
>> + node->fmt = fmt;
>> +
>> + return 0;
>> +}
>> +
>> +static int unicam_mc_enum_framesizes(struct file *file, void *fh,
>> + struct v4l2_frmsizeenum *fsize)
>> +{
>> + struct unicam_node *node = video_drvdata(file);
>> + struct unicam_device *dev = node->dev;
>> +
>> + if (fsize->index > 0)
>> + return -EINVAL;
>> +
>> + if (!find_format_by_fourcc(fsize->pixel_format)) {
>> + dev_dbg(dev->v4l2_dev.dev, "Invalid pixel format 0x%08x\n",
>> + fsize->pixel_format);
>
> I'd drop this message.
>
>> + return -EINVAL;
>> + }
>> +
>> + fsize->type = V4L2_FRMSIZE_TYPE_STEPWISE;
>> + fsize->stepwise.min_width = MIN_WIDTH;
>> + fsize->stepwise.max_width = MAX_WIDTH;
>> + fsize->stepwise.step_width = 1;
>> + fsize->stepwise.min_height = MIN_HEIGHT;
>> + fsize->stepwise.max_height = MAX_HEIGHT;
>> + fsize->stepwise.step_height = 1;
>
> Is this valid for the metadata node too ?
>
>> +
>> + return 0;
>> +}
>> +
>> +static int unicam_mc_enum_fmt_meta_cap(struct file *file, void *priv,
>> + struct v4l2_fmtdesc *f)
>> +{
>> + struct unicam_node *node = video_drvdata(file);
>> + int i, j;
>> +
>> + if (!is_metadata_node(node))
>> + return -EINVAL;
>
> Same comments as for video format enumeration.
>
>> +
>> + for (i = 0, j = 0; i < ARRAY_SIZE(formats); i++) {
>> + if (f->mbus_code && formats[i].code != f->mbus_code)
>> + continue;
>> + if (!formats[i].metadata_fmt)
>> + continue;
>> +
>> + if (formats[i].fourcc) {
>> + if (j == f->index) {
>> + f->pixelformat = formats[i].fourcc;
>> + f->type = V4L2_BUF_TYPE_META_CAPTURE;
>> + return 0;
>> + }
>> + j++;
>> + }
>> + }
>> +
>> + return -EINVAL;
>> +}
>> +
>> +static int unicam_mc_g_fmt_meta_cap(struct file *file, void *priv,
>> + struct v4l2_format *f)
>> +{
>> + struct unicam_node *node = video_drvdata(file);
>> +
>> + if (!is_metadata_node(node))
>> + return -EINVAL;
>
> Can this happen ? Same below.
>
>> +
>> + *f = node->v_fmt;
>> +
>> + return 0;
>> +}
>> +
>> +static int unicam_mc_try_fmt_meta_cap(struct file *file, void *priv,
>> + struct v4l2_format *f)
>> +{
>> + struct unicam_node *node = video_drvdata(file);
>> +
>> + if (!is_metadata_node(node))
>> + return -EINVAL;
>> +
>> + f->fmt.meta.dataformat = V4L2_META_FMT_8;
>> +
>> + return 0;
>> +}
>> +
>> +static int unicam_mc_s_fmt_meta_cap(struct file *file, void *priv,
>> + struct v4l2_format *f)
>> +{
>> + struct unicam_node *node = video_drvdata(file);
>> +
>> + if (!is_metadata_node(node))
>> + return -EINVAL;
>> +
>> + unicam_mc_try_fmt_meta_cap(file, priv, f);
>> +
>> + node->v_fmt = *f;
>> +
>> + return 0;
>> +}
>> +
>> +static const struct v4l2_ioctl_ops unicam_mc_ioctl_ops = {
>> + .vidioc_querycap = unicam_querycap,
>> + .vidioc_enum_fmt_vid_cap = unicam_mc_enum_fmt_vid_cap,
>> + .vidioc_g_fmt_vid_cap = unicam_mc_g_fmt_vid_cap,
>> + .vidioc_try_fmt_vid_cap = unicam_mc_try_fmt_vid_cap,
>> + .vidioc_s_fmt_vid_cap = unicam_mc_s_fmt_vid_cap,
>> +
>> + .vidioc_enum_fmt_meta_cap = unicam_mc_enum_fmt_meta_cap,
>> + .vidioc_g_fmt_meta_cap = unicam_mc_g_fmt_meta_cap,
>> + .vidioc_try_fmt_meta_cap = unicam_mc_try_fmt_meta_cap,
>> + .vidioc_s_fmt_meta_cap = unicam_mc_s_fmt_meta_cap,
>> +
>> + .vidioc_enum_framesizes = unicam_mc_enum_framesizes,
>> + .vidioc_reqbufs = vb2_ioctl_reqbufs,
>> + .vidioc_create_bufs = vb2_ioctl_create_bufs,
>> + .vidioc_prepare_buf = vb2_ioctl_prepare_buf,
>> + .vidioc_querybuf = vb2_ioctl_querybuf,
>> + .vidioc_qbuf = vb2_ioctl_qbuf,
>> + .vidioc_dqbuf = vb2_ioctl_dqbuf,
>> + .vidioc_expbuf = vb2_ioctl_expbuf,
>> + .vidioc_streamon = vb2_ioctl_streamon,
>> + .vidioc_streamoff = vb2_ioctl_streamoff,
>> +
>> + .vidioc_log_status = unicam_log_status,
>> + .vidioc_subscribe_event = unicam_subscribe_event,
>> + .vidioc_unsubscribe_event = v4l2_event_unsubscribe,
>> +};
>> +
>> +static const struct media_entity_operations unicam_mc_entity_ops = {
>> + .link_validate = v4l2_subdev_link_validate,
>> + .has_route = v4l2_subdev_has_route,
>> +};
>
> I don't think this is needed for the video nodes.
>
>> +
>> +/* videobuf2 Operations */
>> +
>> +static int unicam_queue_setup(struct vb2_queue *vq,
>> + unsigned int *nbuffers,
>> + unsigned int *nplanes,
>> + unsigned int sizes[],
>> + struct device *alloc_devs[])
>> +{
>> + struct unicam_node *node = vb2_get_drv_priv(vq);
>> + struct unicam_device *dev = node->dev;
>> + unsigned int size = is_image_node(node) ?
>> + node->v_fmt.fmt.pix.sizeimage :
>> + node->v_fmt.fmt.meta.buffersize;
>> +
>> + if (vq->num_buffers + *nbuffers < 3)
>> + *nbuffers = 3 - vq->num_buffers;
>> +
>> + if (*nplanes) {
>> + if (sizes[0] < size) {
>> + dev_err(dev->v4l2_dev.dev, "sizes[0] %i < size %u\n", sizes[0],
>> + size);
>
> This can be a debug message, let's not give a way to unprivileged
> userspace to flood the kernel log.
>
>> + return -EINVAL;
>> + }
>> + size = sizes[0];
>> + }
>> +
>> + *nplanes = 1;
>> + sizes[0] = size;
>> +
>> + return 0;
>> +}
>> +
>> +static int unicam_buffer_prepare(struct vb2_buffer *vb)
>> +{
>> + struct unicam_node *node = vb2_get_drv_priv(vb->vb2_queue);
>> + struct unicam_device *dev = node->dev;
>> + struct unicam_buffer *buf = to_unicam_buffer(vb);
>> + unsigned long size;
>> +
>> + if (WARN_ON(!node->fmt))
>> + return -EINVAL;
>> +
>> + size = is_image_node(node) ? node->v_fmt.fmt.pix.sizeimage :
>> + node->v_fmt.fmt.meta.buffersize;
>> + if (vb2_plane_size(vb, 0) < size) {
>> + dev_err(dev->v4l2_dev.dev, "data will not fit into plane (%lu < %lu)\n",
>> + vb2_plane_size(vb, 0), size);
>
> Debug message here too.
>
>> + return -EINVAL;
>> + }
>> +
>> + vb2_set_plane_payload(&buf->vb.vb2_buf, 0, size);
>> + return 0;
>> +}
>> +
>> +static void unicam_buffer_queue(struct vb2_buffer *vb)
>> +{
>> + struct unicam_node *node = vb2_get_drv_priv(vb->vb2_queue);
>> + struct unicam_buffer *buf = to_unicam_buffer(vb);
>> + unsigned long flags;
>> +
>> + spin_lock_irqsave(&node->dma_queue_lock, flags);
>
> Small optimization, as this is guaranteed to be called with interrupts
> enabled, you can use the spin_lock_irq() variant.
>
>> + list_add_tail(&buf->list, &node->dma_queue);
>> + spin_unlock_irqrestore(&node->dma_queue_lock, flags);
>> +}
>> +
>> +static void unicam_set_packing_config(struct unicam_device *dev)
>> +{
>> + u32 pack, unpack;
>> + u32 val;
>> +
>> + if (dev->node[IMAGE_NODE].v_fmt.fmt.pix.pixelformat ==
>> + dev->node[IMAGE_NODE].fmt->fourcc) {
>> + unpack = UNICAM_PUM_NONE;
>> + pack = UNICAM_PPM_NONE;
>> + } else {
>> + switch (dev->node[IMAGE_NODE].fmt->depth) {
>> + case 8:
>> + unpack = UNICAM_PUM_UNPACK8;
>> + break;
>> + case 10:
>> + unpack = UNICAM_PUM_UNPACK10;
>> + break;
>> + case 12:
>> + unpack = UNICAM_PUM_UNPACK12;
>> + break;
>> + case 14:
>> + unpack = UNICAM_PUM_UNPACK14;
>> + break;
>> + case 16:
>> + unpack = UNICAM_PUM_UNPACK16;
>> + break;
>> + default:
>> + unpack = UNICAM_PUM_NONE;
>> + break;
>> + }
>> +
>> + /* Repacking is always to 16bpp */
>> + pack = UNICAM_PPM_PACK16;
>> + }
>> +
>> + val = 0;
>> + set_field(&val, unpack, UNICAM_PUM_MASK);
>> + set_field(&val, pack, UNICAM_PPM_MASK);
>> + reg_write(dev, UNICAM_IPIPE, val);
>> +}
>> +
>> +static void unicam_cfg_image_id(struct unicam_device *dev)
>> +{
>> + if (dev->bus_type == V4L2_MBUS_CSI2_DPHY) {
>> + /* CSI2 mode, hardcode VC 0 for now. */
>> + reg_write(dev, UNICAM_IDI0,
>> + (0 << 6) | dev->node[IMAGE_NODE].fmt->csi_dt);
>> + } else {
>> + /* CCP2 mode */
>> + reg_write(dev, UNICAM_IDI0,
>> + 0x80 | dev->node[IMAGE_NODE].fmt->csi_dt);
>> + }
>> +}
>> +
>> +static void unicam_enable_ed(struct unicam_device *dev)
>> +{
>> + u32 val = reg_read(dev, UNICAM_DCS);
>> +
>> + set_field(&val, 2, UNICAM_EDL_MASK);
>> + /* Do not wrap at the end of the embedded data buffer */
>> + set_field(&val, 0, UNICAM_DBOB);
>> +
>> + reg_write(dev, UNICAM_DCS, val);
>> +}
>> +
>> +static void unicam_start_rx(struct unicam_device *dev, dma_addr_t *addr)
>> +{
>> + int line_int_freq = dev->node[IMAGE_NODE].v_fmt.fmt.pix.height >> 2;
>> + unsigned int size, i;
>> + u32 val;
>> +
>> + if (line_int_freq < 128)
>> + line_int_freq = 128;
>> +
>> + /* Enable lane clocks */
>> + val = 1;
>> + for (i = 0; i < dev->active_data_lanes; i++)
>> + val = val << 2 | 1;
>> + clk_write(dev, val);
>> +
>> + /* Basic init */
>> + reg_write(dev, UNICAM_CTRL, UNICAM_MEM);
>> +
>> + /* Enable analogue control, and leave in reset. */
>> + val = UNICAM_AR;
>> + set_field(&val, 7, UNICAM_CTATADJ_MASK);
>> + set_field(&val, 7, UNICAM_PTATADJ_MASK);
>> + reg_write(dev, UNICAM_ANA, val);
>> + usleep_range(1000, 2000);
>> +
>> + /* Come out of reset */
>> + reg_write_field(dev, UNICAM_ANA, 0, UNICAM_AR);
>> +
>> + /* Peripheral reset */
>> + reg_write_field(dev, UNICAM_CTRL, 1, UNICAM_CPR);
>> + reg_write_field(dev, UNICAM_CTRL, 0, UNICAM_CPR);
>> +
>> + reg_write_field(dev, UNICAM_CTRL, 0, UNICAM_CPE);
>> +
>> + /* Enable Rx control. */
>> + val = reg_read(dev, UNICAM_CTRL);
>> + if (dev->bus_type == V4L2_MBUS_CSI2_DPHY) {
>> + set_field(&val, UNICAM_CPM_CSI2, UNICAM_CPM_MASK);
>> + set_field(&val, UNICAM_DCM_STROBE, UNICAM_DCM_MASK);
>> + } else {
>> + set_field(&val, UNICAM_CPM_CCP2, UNICAM_CPM_MASK);
>> + set_field(&val, dev->bus_flags, UNICAM_DCM_MASK);
>> + }
>> + /* Packet framer timeout */
>> + set_field(&val, 0xf, UNICAM_PFT_MASK);
>> + set_field(&val, 128, UNICAM_OET_MASK);
>> + reg_write(dev, UNICAM_CTRL, val);
>> +
>> + reg_write(dev, UNICAM_IHWIN, 0);
>> + reg_write(dev, UNICAM_IVWIN, 0);
>> +
>> + /* AXI bus access QoS setup */
>> + val = reg_read(dev, UNICAM_PRI);
>> + set_field(&val, 0, UNICAM_BL_MASK);
>> + set_field(&val, 0, UNICAM_BS_MASK);
>> + set_field(&val, 0xe, UNICAM_PP_MASK);
>> + set_field(&val, 8, UNICAM_NP_MASK);
>> + set_field(&val, 2, UNICAM_PT_MASK);
>> + set_field(&val, 1, UNICAM_PE);
>> + reg_write(dev, UNICAM_PRI, val);
>> +
>> + reg_write_field(dev, UNICAM_ANA, 0, UNICAM_DDL);
>> +
>> + /* Always start in trigger frame capture mode (UNICAM_FCM set) */
>> + val = UNICAM_FSIE | UNICAM_FEIE | UNICAM_FCM | UNICAM_IBOB;
>> + set_field(&val, line_int_freq, UNICAM_LCIE_MASK);
>> + reg_write(dev, UNICAM_ICTL, val);
>> + reg_write(dev, UNICAM_STA, UNICAM_STA_MASK_ALL);
>> + reg_write(dev, UNICAM_ISTA, UNICAM_ISTA_MASK_ALL);
>> +
>> + /* tclk_term_en */
>> + reg_write_field(dev, UNICAM_CLT, 2, UNICAM_CLT1_MASK);
>> + /* tclk_settle */
>> + reg_write_field(dev, UNICAM_CLT, 6, UNICAM_CLT2_MASK);
>> + /* td_term_en */
>> + reg_write_field(dev, UNICAM_DLT, 2, UNICAM_DLT1_MASK);
>> + /* ths_settle */
>> + reg_write_field(dev, UNICAM_DLT, 6, UNICAM_DLT2_MASK);
>> + /* trx_enable */
>> + reg_write_field(dev, UNICAM_DLT, 0, UNICAM_DLT3_MASK);
>> +
>> + reg_write_field(dev, UNICAM_CTRL, 0, UNICAM_SOE);
>> +
>> + /* Packet compare setup - required to avoid missing frame ends */
>> + val = 0;
>> + set_field(&val, 1, UNICAM_PCE);
>> + set_field(&val, 1, UNICAM_GI);
>> + set_field(&val, 1, UNICAM_CPH);
>> + set_field(&val, 0, UNICAM_PCVC_MASK);
>> + set_field(&val, 1, UNICAM_PCDT_MASK);
>> + reg_write(dev, UNICAM_CMP0, val);
>> +
>> + /* Enable clock lane and set up terminations */
>> + val = 0;
>> + if (dev->bus_type == V4L2_MBUS_CSI2_DPHY) {
>> + /* CSI2 */
>> + set_field(&val, 1, UNICAM_CLE);
>> + set_field(&val, 1, UNICAM_CLLPE);
>> + if (dev->bus_flags & V4L2_MBUS_CSI2_CONTINUOUS_CLOCK) {
>> + set_field(&val, 1, UNICAM_CLTRE);
>> + set_field(&val, 1, UNICAM_CLHSE);
>> + }
>> + } else {
>> + /* CCP2 */
>> + set_field(&val, 1, UNICAM_CLE);
>> + set_field(&val, 1, UNICAM_CLHSE);
>> + set_field(&val, 1, UNICAM_CLTRE);
>> + }
>> + reg_write(dev, UNICAM_CLK, val);
>> +
>> + /*
>> + * Enable required data lanes with appropriate terminations.
>> + * The same value needs to be written to UNICAM_DATn registers for
>> + * the active lanes, and 0 for inactive ones.
>> + */
>> + val = 0;
>> + if (dev->bus_type == V4L2_MBUS_CSI2_DPHY) {
>> + /* CSI2 */
>> + set_field(&val, 1, UNICAM_DLE);
>> + set_field(&val, 1, UNICAM_DLLPE);
>> + if (dev->bus_flags & V4L2_MBUS_CSI2_CONTINUOUS_CLOCK) {
>> + set_field(&val, 1, UNICAM_DLTRE);
>> + set_field(&val, 1, UNICAM_DLHSE);
>> + }
>> + } else {
>> + /* CCP2 */
>> + set_field(&val, 1, UNICAM_DLE);
>> + set_field(&val, 1, UNICAM_DLHSE);
>> + set_field(&val, 1, UNICAM_DLTRE);
>> + }
>> + reg_write(dev, UNICAM_DAT0, val);
>> +
>> + if (dev->active_data_lanes == 1)
>> + val = 0;
>> + reg_write(dev, UNICAM_DAT1, val);
>> +
>> + if (dev->max_data_lanes > 2) {
>> + /*
>> + * Registers UNICAM_DAT2 and UNICAM_DAT3 only valid if the
>> + * instance supports more than 2 data lanes.
>> + */
>> + if (dev->active_data_lanes == 2)
>> + val = 0;
>> + reg_write(dev, UNICAM_DAT2, val);
>> +
>> + if (dev->active_data_lanes == 3)
>> + val = 0;
>> + reg_write(dev, UNICAM_DAT3, val);
>> + }
>> +
>> + reg_write(dev, UNICAM_IBLS,
>> + dev->node[IMAGE_NODE].v_fmt.fmt.pix.bytesperline);
>> + size = dev->node[IMAGE_NODE].v_fmt.fmt.pix.sizeimage;
>> + unicam_wr_dma_addr(dev, addr[IMAGE_NODE], size, IMAGE_NODE);
>> + unicam_set_packing_config(dev);
>> + unicam_cfg_image_id(dev);
>> +
>> + val = reg_read(dev, UNICAM_MISC);
>> + set_field(&val, 1, UNICAM_FL0);
>> + set_field(&val, 1, UNICAM_FL1);
>> + reg_write(dev, UNICAM_MISC, val);
>> +
>> + if (dev->node[METADATA_NODE].streaming && dev->sensor_embedded_data) {
>> + dev_dbg(dev->v4l2_dev.dev, "enable metadata dma\n");
>> + size = dev->node[METADATA_NODE].v_fmt.fmt.meta.buffersize;
>> + unicam_enable_ed(dev);
>> + unicam_wr_dma_addr(dev, addr[METADATA_NODE], size, METADATA_NODE);
>> + }
>> +
>> + /* Enable peripheral */
>> + reg_write_field(dev, UNICAM_CTRL, 1, UNICAM_CPE);
>> +
>> + /* Load image pointers */
>> + reg_write_field(dev, UNICAM_ICTL, 1, UNICAM_LIP_MASK);
>> +
>> + /* Load embedded data buffer pointers if needed */
>> + if (dev->node[METADATA_NODE].streaming && dev->sensor_embedded_data)
>> + reg_write_field(dev, UNICAM_DCS, 1, UNICAM_LDP);
>> +
>> + /*
>> + * Enable trigger only for the first frame to
>> + * sync correctly to the FS from the source.
>> + */
>> + reg_write_field(dev, UNICAM_ICTL, 1, UNICAM_TFC);
>> +}
>> +
>> +static void unicam_disable(struct unicam_device *dev)
>> +{
>> + /* Analogue lane control disable */
>> + reg_write_field(dev, UNICAM_ANA, 1, UNICAM_DDL);
>> +
>> + /* Stop the output engine */
>> + reg_write_field(dev, UNICAM_CTRL, 1, UNICAM_SOE);
>> +
>> + /* Disable the data lanes. */
>> + reg_write(dev, UNICAM_DAT0, 0);
>> + reg_write(dev, UNICAM_DAT1, 0);
>> +
>> + if (dev->max_data_lanes > 2) {
>> + reg_write(dev, UNICAM_DAT2, 0);
>> + reg_write(dev, UNICAM_DAT3, 0);
>> + }
>> +
>> + /* Peripheral reset */
>> + reg_write_field(dev, UNICAM_CTRL, 1, UNICAM_CPR);
>> + usleep_range(50, 100);
>> + reg_write_field(dev, UNICAM_CTRL, 0, UNICAM_CPR);
>> +
>> + /* Disable peripheral */
>> + reg_write_field(dev, UNICAM_CTRL, 0, UNICAM_CPE);
>> +
>> + /* Clear ED setup */
>> + reg_write(dev, UNICAM_DCS, 0);
>> +
>> + /* Disable all lane clocks */
>> + clk_write(dev, 0);
>> +}
>> +
>> +static void unicam_return_buffers(struct unicam_node *node,
>> + enum vb2_buffer_state state)
>> +{
>> + struct unicam_buffer *buf, *tmp;
>> + unsigned long flags;
>> +
>> + spin_lock_irqsave(&node->dma_queue_lock, flags);
>
> Same here, spin_lock_irq() would do. But isn't this function meant to be
> called with the hardware stopped ? Shouldn't it then be safe to drop the
> spinlock completely ?
>
>> + list_for_each_entry_safe(buf, tmp, &node->dma_queue, list) {
>> + list_del(&buf->list);
>> + vb2_buffer_done(&buf->vb.vb2_buf, state);
>> + }
>> +
>> + if (node->cur_frm)
>> + vb2_buffer_done(&node->cur_frm->vb.vb2_buf,
>> + state);
>> + if (node->next_frm && node->cur_frm != node->next_frm)
>> + vb2_buffer_done(&node->next_frm->vb.vb2_buf,
>> + state);
>> +
>> + node->cur_frm = NULL;
>> + node->next_frm = NULL;
>> + spin_unlock_irqrestore(&node->dma_queue_lock, flags);
>> +}
>> +
>> +static int unicam_video_check_format(struct unicam_node *node)
>> +{
>> + const struct v4l2_mbus_framefmt *format;
>> + struct media_pad *remote_pad;
>> + struct v4l2_subdev_state *state;
>> + int ret = 0;
>> +
>> + remote_pad = media_entity_remote_pad(&node->pad);
>> + if (!remote_pad)
>> + return -ENODEV;
>
> This could be done in unicam_async_complete() and cached in the
> unicam_device structure, the same way we cache the sensor subdev
> pointer.
>
>> +
>> + state = v4l2_subdev_lock_active_state(node->dev->sensor);
>
> That's not correct, the video node is connected to the subdev, not to
> the sensor.
>
>> +
>> + format = v4l2_state_get_stream_format(state,
>> + remote_pad->index, 0);
>
> This holds on a single line.
>
>> + if (!format) {
>> + ret = -EINVAL;
>> + goto out;
>> + }
>> +
>> + if (node->fmt->code != format->code ||
>> + node->v_fmt.fmt.pix.height != format->height ||
>> + node->v_fmt.fmt.pix.width != format->width ||
>> + node->v_fmt.fmt.pix.field != format->field) {
>
> A debug message that prints the configuration on both sides could be
> useful for debugging here.
>
>> + ret = -EPIPE;
>> + goto out;
>> + }
>> +
>> +out:
>> + v4l2_subdev_unlock_state(state);
>> +
>> + return ret;
>> +}
>> +
>> +static int unicam_start_streaming(struct vb2_queue *vq, unsigned int count)
>> +{
>> + struct unicam_node *node = vb2_get_drv_priv(vq);
>> + struct unicam_device *dev = node->dev;
>> + dma_addr_t buffer_addr[MAX_NODES] = { 0 };
>> + struct unicam_buffer *buf;
>> + unsigned long flags;
>> + unsigned int i;
>> + int ret;
>> +
>> + node->streaming = true;
>
> Locking seems to be completely missing from the driver. I'll comment
> more on that on v3, it will be easier after all the other cleanups.
>
>> +
>> + dev->sequence = 0;
>> + ret = unicam_runtime_get(dev);
>> + if (ret < 0) {
>> + dev_dbg(dev->v4l2_dev.dev, "unicam_runtime_get failed\n");
>
> This is really not supposed to happen, dev_err() is appropriate.
>
>> + goto err_streaming;
>> + }
>> +
>> + ret = media_pipeline_start(node->video_dev.entity.pads, &node->pipe);
>> + if (ret < 0) {
>> + dev_err(dev->v4l2_dev.dev, "Failed to start media pipeline: %d\n", ret);
>
> dev_dbg() here too.
>
>> + goto err_pm_put;
>> + }
>> +
>> + dev->active_data_lanes = dev->max_data_lanes;
>
> You need to get the number of data lanes from the sensor subdev and
> validate it to ensure it doesn't exceed the maximum.
>
>> +
>> + unicam_video_check_format(node);
>
> Where's the return value check ?
>
>> +
>> + dev_dbg(dev->v4l2_dev.dev, "Running with %u data lanes\n",
>> + dev->active_data_lanes);
>> +
>> + ret = clk_set_min_rate(dev->vpu_clock, MIN_VPU_CLOCK_RATE);
>> + if (ret) {
>> + dev_err(dev->v4l2_dev.dev, "failed to set up VPU clock\n");
>> + goto error_pipeline;
>> + }
>> +
>> + ret = clk_prepare_enable(dev->vpu_clock);
>> + if (ret) {
>> + dev_err(dev->v4l2_dev.dev, "Failed to enable VPU clock: %d\n", ret);
>> + goto error_pipeline;
>> + }
>> +
>> + ret = clk_set_rate(dev->clock, 100 * 1000 * 1000);
>> + if (ret) {
>> + dev_err(dev->v4l2_dev.dev, "failed to set up CSI clock\n");
>> + goto err_vpu_clock;
>> + }
>> +
>> + ret = clk_prepare_enable(dev->clock);
>> + if (ret) {
>> + dev_err(dev->v4l2_dev.dev, "Failed to enable CSI clock: %d\n", ret);
>> + goto err_vpu_clock;
>> + }
>> +
>> + dev_dbg(dev->v4l2_dev.dev, "node %s\n", node->video_dev.name);
>> +
>> + spin_lock_irqsave(&node->dma_queue_lock, flags);
>> + buf = list_first_entry(&node->dma_queue,
>> + struct unicam_buffer, list);
>
> This holds on a single line.
>
>> + node->cur_frm = buf;
>> + node->next_frm = buf;
>> + list_del(&buf->list);
>> + spin_unlock_irqrestore(&node->dma_queue_lock, flags);
>> +
>> + buffer_addr[i] =
>> + vb2_dma_contig_plane_dma_addr(&buf->vb.vb2_buf, 0);
>
> This too.
>
>> +
>> + unicam_start_rx(dev, buffer_addr);
>> +
>> + ret = v4l2_subdev_call(&dev->sd, video, s_stream, 1);
>> + if (ret < 0) {
>> + dev_err(dev->v4l2_dev.dev, "stream on failed in subdev\n");
>> + goto err_disable_unicam;
>> + }
>
> Do I understand correctly that all this will run twice, once when
> VIDIOC_STREAMON is called on the image node, and once when it's called
> on the metadata node ? That doesn't seem right. The stop_streaming
> function is affected similarly.
>
>> +
>> + dev->clocks_enabled = true;
>> + return 0;
>> +
>> +err_disable_unicam:
>> + unicam_disable(dev);
>> + clk_disable_unprepare(dev->clock);
>> +err_vpu_clock:
>> + if (clk_set_min_rate(dev->vpu_clock, 0))
>> + dev_err(dev->v4l2_dev.dev, "failed to reset the VPU clock\n");
>> + clk_disable_unprepare(dev->vpu_clock);
>> +error_pipeline:
>> + media_pipeline_stop(node->video_dev.entity.pads);
>> +err_pm_put:
>> + unicam_runtime_put(dev);
>> +err_streaming:
>> + unicam_return_buffers(node, VB2_BUF_STATE_QUEUED);
>> + node->streaming = false;
>> +
>> + return ret;
>> +}
>> +
>> +static void unicam_stop_streaming(struct vb2_queue *vq)
>> +{
>> + struct unicam_node *node = vb2_get_drv_priv(vq);
>> + struct unicam_device *dev = node->dev;
>> +
>> + node->streaming = false;
>> + /*
>> + * Stop streaming the sensor and disable the peripheral.
>> + * We cannot continue streaming embedded data with the
>> + * image pad disabled.
>> + */
>> + if (v4l2_subdev_call(&dev->sd, video, s_stream, 0) < 0)
>> + dev_err(dev->v4l2_dev.dev, "stream off failed in subdev\n");
>> +
>> + unicam_disable(dev);
>> +
>> + media_pipeline_stop(node->video_dev.entity.pads);
>> +
>> + if (dev->clocks_enabled) {
>
> .stop_streaming() can only be called after a successful
> .start_streaming() call, so you can drop the clocks_enabled field.
>
>> + if (clk_set_min_rate(dev->vpu_clock, 0))
>> + dev_err(dev->v4l2_dev.dev, "failed to reset the VPU clock\n");
>> + clk_disable_unprepare(dev->vpu_clock);
>> + clk_disable_unprepare(dev->clock);
>> + dev->clocks_enabled = false;
>> + }
>> + unicam_runtime_put(dev);
>> +
>> + if (is_metadata_node(node)) {
>> + /*
>> + * Allow the hardware to spin in the dummy buffer.
>> + * This is only really needed if the embedded data pad is
>> + * disabled before the image pad.
>> + */
>> + unicam_wr_dma_addr(dev, node->dummy_buf_dma_addr,
>> + DUMMY_BUF_SIZE, METADATA_NODE);
>> + }
>> +
>> + /* Clear all queued buffers for the node */
>> + unicam_return_buffers(node, VB2_BUF_STATE_ERROR);
>> +}
>> +
>> +static const struct vb2_ops unicam_video_qops = {
>> + .wait_prepare = vb2_ops_wait_prepare,
>> + .wait_finish = vb2_ops_wait_finish,
>> + .queue_setup = unicam_queue_setup,
>> + .buf_prepare = unicam_buffer_prepare,
>> + .buf_queue = unicam_buffer_queue,
>> + .start_streaming = unicam_start_streaming,
>> + .stop_streaming = unicam_stop_streaming,
>> +};
>> +
>> +/* unicam capture driver file operations */
>> +static const struct v4l2_file_operations unicam_fops = {
>> + .owner = THIS_MODULE,
>> + .open = v4l2_fh_open,
>> + .release = vb2_fop_release,
>> + .poll = vb2_fop_poll,
>> + .unlocked_ioctl = video_ioctl2, /* V4L2 ioctl handler */
>
> Drop the comment.
>
>> + .mmap = vb2_fop_mmap,
>> +};
>> +
>> +static int
>> +unicam_async_bound(struct v4l2_async_notifier *notifier,
>> + struct v4l2_subdev *subdev,
>> + struct v4l2_async_subdev *asd)
>> +{
>> + struct unicam_device *unicam = to_unicam_device(notifier->v4l2_dev);
>> +
>> + if (unicam->sensor) {
>> + dev_info(unicam->v4l2_dev.dev, "subdev %s (Already set!!)",
>> + subdev->name);
>> + return 0;
>> + }
>
> This can't happen as we register a single asd only.
>
>> +
>> + unicam->sensor = subdev;
>> + dev_dbg(unicam->v4l2_dev.dev, "Using sensor %s for capture\n", subdev->name);
>
> Line wrap.
>
>> +
>> + return 0;
>> +}
>> +
>> +static void unicam_release(struct kref *kref)
>> +{
>> + struct unicam_device *unicam =
>> + container_of(kref, struct unicam_device, kref);
>> +
>> + v4l2_ctrl_handler_free(&unicam->ctrl_handler);
>> + media_device_cleanup(&unicam->mdev);
>> +
>> + if (unicam->sensor_state)
>> + __v4l2_subdev_state_free(unicam->sensor_state);
>> +
>> + kfree(unicam);
>> +}
>> +
>> +static void unicam_put(struct unicam_device *unicam)
>> +{
>> + kref_put(&unicam->kref, unicam_release);
>> +}
>> +
>> +static void unicam_get(struct unicam_device *unicam)
>> +{
>> + kref_get(&unicam->kref);
>> +}
>> +
>> +static void unicam_node_release(struct video_device *vdev)
>> +{
>> + struct unicam_node *node = video_get_drvdata(vdev);
>> +
>> + unicam_put(node->dev);
>> +}
>> +
>> +static void unicam_mc_set_default_format(struct unicam_node *node)
>> +{
>> + dev_dbg(node->dev->v4l2_dev.dev, "Set default format for %s node\n",
>
> node->dev->dev should do, no need to go through v4l2_dev here. But I'd
> drop the message, I don't think it brings much value.
>
>> + node->node_id == IMAGE_NODE ? "image" : "metadata");
>> +
>> + if (is_image_node(node)) {
>> + struct v4l2_pix_format *pix_fmt = &node->v_fmt.fmt.pix;
>> +
>> + pix_fmt->width = 640;
>> + pix_fmt->height = 480;
>> + pix_fmt->field = V4L2_FIELD_NONE;
>> + pix_fmt->colorspace = V4L2_COLORSPACE_SRGB;
>> + pix_fmt->ycbcr_enc = V4L2_YCBCR_ENC_601;
>> + pix_fmt->quantization = V4L2_QUANTIZATION_LIM_RANGE;
>> + pix_fmt->xfer_func = V4L2_XFER_FUNC_SRGB;
>> + pix_fmt->pixelformat = formats[0].fourcc;
>> + unicam_calc_format_size_bpl(node->dev, &formats[0],
>> + &node->v_fmt);
>> + node->v_fmt.type = V4L2_BUF_TYPE_VIDEO_CAPTURE;
>> +
>> + node->fmt = &formats[0];
>> + } else {
>> + const struct unicam_fmt *fmt;
>> +
>> + /* Fix this node format as embedded data. */
>> + fmt = find_format_by_code(MEDIA_BUS_FMT_METADATA_8);
>> + node->v_fmt.fmt.meta.dataformat = fmt->fourcc;
>> + node->fmt = fmt;
>
> Move this to the end to group all v_fmt initialization together.
>
>> + node->v_fmt.fmt.meta.buffersize = UNICAM_EMBEDDED_SIZE;
>> + node->embedded_lines = 1;
>
> This belongs to the called.
>
>> + node->v_fmt.type = V4L2_BUF_TYPE_META_CAPTURE;
>> + }
>> +}
>> +
>> +static int register_node(struct unicam_device *unicam, struct unicam_node *node,
>> + enum v4l2_buf_type type)
>
> Replace the type parameter with a enum unicam_node_type, that will look
> cleaner in
>
>> +{
>> + struct video_device *vdev;
>> + struct vb2_queue *q;
>> + int ret;
>> +
>> + node->dev = unicam;
>> + node->node_id = type == V4L2_BUF_TYPE_VIDEO_CAPTURE ?
>> + IMAGE_NODE :
>> + METADATA_NODE;
>> +
>> + spin_lock_init(&node->dma_queue_lock);
>> + mutex_init(&node->lock);
>> +
>> + vdev = &node->video_dev;
>> + /*
>> + * If the sensor subdevice has any controls, associate the node
>> + * with the ctrl handler to allow access from userland.
>> + */
>> + if (!list_empty(&unicam->ctrl_handler.ctrls))
>> + vdev->ctrl_handler = &unicam->ctrl_handler;
>
> Drop this, and move the vdev assignment below, just before it gets used.
>
> /* Initialize the videobuf2 queue. */
>> +
>> + q = &node->buffer_queue;
>> + q->type = type;
>> + q->io_modes = VB2_MMAP | VB2_DMABUF | VB2_READ;
>
> Drop VB2_READ, that shouldn't be used.
>
>> + q->drv_priv = node;
>> + q->ops = &unicam_video_qops;
>> + q->mem_ops = &vb2_dma_contig_memops;
>> + q->buf_struct_size = sizeof(struct unicam_buffer);
>> + q->timestamp_flags = V4L2_BUF_FLAG_TIMESTAMP_MONOTONIC;
>> + q->lock = &node->lock;
>> + q->min_buffers_needed = 1;
>> + q->dev = &unicam->pdev->dev;
>> +
>> + ret = vb2_queue_init(q);
>> + if (ret) {
>> + dev_err(unicam->v4l2_dev.dev, "vb2_queue_init() failed\n");
>> + return ret;
>> + }
>> +
>> + INIT_LIST_HEAD(&node->dma_queue);
>
> Move this to the beginning of the function with the rest of the node
> initialization.
>
> /* Initialize the video device. */
>
>> +
>> + vdev->release = unicam_node_release;
>> + vdev->fops = &unicam_fops;
>> + vdev->ioctl_ops = &unicam_mc_ioctl_ops;
>> + vdev->v4l2_dev = &unicam->v4l2_dev;
>> + vdev->vfl_dir = VFL_DIR_RX;
>> + vdev->queue = q;
>> + vdev->lock = &node->lock;
>> + vdev->device_caps = (type == V4L2_BUF_TYPE_VIDEO_CAPTURE) ?
>
> No need for parentheses.
>
>> + V4L2_CAP_VIDEO_CAPTURE : V4L2_CAP_META_CAPTURE;
>> + vdev->device_caps |= V4L2_CAP_READWRITE | V4L2_CAP_STREAMING | V4L2_CAP_IO_MC;
>
> Drop V4L2_CAP_READWRITE.
>
>> + vdev->entity.ops = &unicam_mc_entity_ops;
>> +
>> + /* Define the device names */
>> + snprintf(vdev->name, sizeof(vdev->name), "%s-%s", UNICAM_MODULE_NAME,
>> + type == V4L2_BUF_TYPE_VIDEO_CAPTURE ? "image" : "embedded");
>> +
>> + video_set_drvdata(vdev, node);
>> + if (type == V4L2_BUF_TYPE_VIDEO_CAPTURE)
>> + vdev->entity.flags |= MEDIA_ENT_FL_DEFAULT;
>> + node->pad.flags = MEDIA_PAD_FL_SINK;
>> + media_entity_pads_init(&vdev->entity, 1, &node->pad);
>
> Error check.
>
>> +
>> + node->dummy_buf_cpu_addr = dma_alloc_coherent(&unicam->pdev->dev,
>> + DUMMY_BUF_SIZE,
>> + &node->dummy_buf_dma_addr,
>> + GFP_KERNEL);
>> + if (!node->dummy_buf_cpu_addr) {
>> + dev_err(unicam->v4l2_dev.dev, "Unable to allocate dummy buffer.\n");
>> + return -ENOMEM;
>> + }
>> +
>> + ret = video_register_device(vdev, VFL_TYPE_VIDEO, -1);
>> + if (ret) {
>> + dev_err(unicam->v4l2_dev.dev, "Unable to register video device %s\n",
>> + vdev->name);
>> + return ret;
>> + }
>> +
>> + /*
>> + * Acquire a reference to unicam, which will be released when the video
>> + * device will be unregistered and userspace will have closed all open
>> + * file handles.
>> + */
>> + unicam_get(unicam);
>
> This is error-prone. References should be acquired when stored, so this
> should be moved to the beginning of the function, and written as
>
> node->dev = unicam_get(unicam);
>
> (you'll need to modify the function prototype to return the pointer).
> You can then drop the comment. You will also to add error handling to
> this function, as the reference will be released by
> unicam_node_release(), which is called only after successful
> registration of the video device. All error paths before registration
> will thus need to release the reference (possibly with a goto
> err_unicam_put)..
>
>> + node->registered = true;
>> +
>> + ret = media_create_pad_link(&unicam->sd.entity,
>> + node->src_pad_id,
>
> unicam->sd is the unicam sd, node->src_pad_id is documented as being the
> source pad of the sensor. This is wrong.
>
>> + &node->video_dev.entity,
>> + 0,
>> + MEDIA_LNK_FL_ENABLED |
>> + MEDIA_LNK_FL_IMMUTABLE);
>> + if (ret)
>> + dev_err(unicam->v4l2_dev.dev, "Unable to create pad link for %s",
>> + unicam->sensor->name);
>
> return ret;
> }
>
>> +
>> + unicam_mc_set_default_format(node);
>
> Call this before registering the video device, otherwise it can race
> with userspace.
>
>> +
>> + return ret;
>
> return 0;
>
>> +}
>> +
>> +static void unregister_nodes(struct unicam_device *unicam)
>> +{
>> + unsigned int i;
>> +
>> + for (i = 0; i < ARRAY_SIZE(unicam->node); i++) {
>> + struct unicam_node *node = &unicam->node[i];
>> +
>> + if (node->dummy_buf_cpu_addr) {
>> + dma_free_coherent(&unicam->pdev->dev, DUMMY_BUF_SIZE,
>> + node->dummy_buf_cpu_addr,
>> + node->dummy_buf_dma_addr);
>> + }
>
> No need for curly braces.
>
>> +
>> + if (node->registered) {
>> + node->registered = false;
>> + video_unregister_device(&node->video_dev);
>
> Swap the two lines.
>
>> + }
>> + }
>> +}
>> +
>> +static int unicam_async_complete(struct v4l2_async_notifier *notifier)
>> +{
>> + struct unicam_device *unicam = to_unicam_device(notifier->v4l2_dev);
>> + unsigned int i, source_pads = 0;
>> + int ret;
>> + static struct lock_class_key key;
>> +
>> + unicam->v4l2_dev.notify = unicam_notify;
>> +
>> + unicam->sensor_state = __v4l2_subdev_state_alloc(unicam->sensor,
>> + "unicam:state->lock",
>> + &key);
>
> You never use sensor_state.
>
>> + if (!unicam->sensor_state)
>> + return -ENOMEM;
>> +
>> + for (i = 0; i < unicam->sd.entity.num_pads; i++) {
>> + if (unicam->sd.entity.pads[i].flags & MEDIA_PAD_FL_SOURCE) {
>> + if (source_pads < MAX_NODES) {
>> + unicam->node[source_pads].src_pad_id = i;
>> + dev_dbg(unicam->v4l2_dev.dev, "source pad %u is index %u\n",
>> + source_pads, i);
>> + }
>> + source_pads++;
>> + }
>> + }
>> + if (!source_pads) {
>> + dev_err(unicam->v4l2_dev.dev, "No source pads on sensor.\n");
>> + goto unregister;
>> + }
>
> You're not looking at the sensor here, unicam->sd is the unicam subdev.
> Furthermore, you're hardcoding the source pad of the sensor to 0 below
> when creating the link. And in any case, that's not how it's supposed to
> be done, a source subdev could have multiple source pads, you need to
> identify the correct one from the device tree.
>
> The right way to handle this is to replace the call to
> media_create_pad_link() with v4l2_create_fwnode_links_to_pad(). This can
> be done in the .bound() handler.
>
>> +
>> + ret = register_node(unicam, &unicam->node[IMAGE_NODE],
>> + V4L2_BUF_TYPE_VIDEO_CAPTURE);
>
> Replace the type parameter with a enum unicam_node_type, so this will
> become
>
> ret = register_node(unicam, &unicam->node[IMAGE_NODE], IMAGE_NODE);
>
> which is less error-prone. You could even drop the second parameters:
>
> ret = register_node(unicam, IMAGE_NODE);
>
>> + if (ret) {
>> + dev_err(unicam->v4l2_dev.dev, "Unable to register image video device.\n");
>> + goto unregister;
>> + }
>> +
>> + /* \todo: check before :-) */
>> + unicam->sensor_embedded_data = true;
>
> Check what ? In any case I don't think this belongs here.
>
>> +
>> + ret = register_node(unicam, &unicam->node[METADATA_NODE],
>> + V4L2_BUF_TYPE_META_CAPTURE);
>> + if (ret) {
>> + dev_err(unicam->v4l2_dev.dev, "Unable to register metadata video device.\n");
>> + goto unregister;
>> + }
>> +
>> + ret = media_create_pad_link(&unicam->sensor->entity,
>> + 0,
>> + &unicam->sd.entity,
>> + UNICAM_SD_PAD_SINK,
>> + MEDIA_LNK_FL_ENABLED |
>> + MEDIA_LNK_FL_IMMUTABLE);
>> + if (ret)
>> + dev_err(unicam->v4l2_dev.dev, "Unable to create pad link for %s",
>> + unicam->sensor->name);
>> +
>> + ret = v4l2_device_register_subdev_nodes(&unicam->v4l2_dev);
>> + if (ret) {
>> + dev_err(unicam->v4l2_dev.dev, "Unable to register subdev nodes.\n");
>> + goto unregister;
>> + }
>> +
>> + /*
>> + * Release the initial reference, all references are now owned by the
>> + * video devices.
>> + */
>> + unicam_put(unicam);
>> + return 0;
>> +
>> +unregister:
>> + unregister_nodes(unicam);
>> + unicam_put(unicam);
>> +
>> + return ret;
>> +}
>> +
>> +static const struct v4l2_async_notifier_operations unicam_async_ops = {
>> + .bound = unicam_async_bound,
>> + .complete = unicam_async_complete,
>> +};
>> +
>> +static int of_unicam_connect_subdevs(struct unicam_device *dev)
>> +{
>> + struct platform_device *pdev = dev->pdev;
>> + struct v4l2_fwnode_endpoint ep = { };
>> + struct device_node *ep_node;
>> + struct device_node *sensor_node;
>> + unsigned int lane;
>> + int ret = -EINVAL;
>> +
>> + if (of_property_read_u32(pdev->dev.of_node, "num-data-lanes",
>> + &dev->max_data_lanes) < 0) {
>> + dev_err(dev->v4l2_dev.dev, "number of data lanes not set\n");
>> + return -EINVAL;
>> + }
>> +
>> + /* Get the local endpoint and remote device. */
>> + ep_node = of_graph_get_next_endpoint(pdev->dev.of_node, NULL);
>
> Use of_graph_get_endpoint_by_regs(node, 0, -1).
>
>> + if (!ep_node) {
>> + dev_dbg(dev->v4l2_dev.dev, "can't get next endpoint\n");
>
> dev_dbg(dev->v4l2_dev.dev, "No endpoint\n");
>
>> + return -EINVAL;
>> + }
>> +
>> + dev_dbg(dev->v4l2_dev.dev, "ep_node is %pOF\n", ep_node);
>
> You can drop this.
>
>> +
>> + sensor_node = of_graph_get_remote_port_parent(ep_node);
>> + if (!sensor_node) {
>> + dev_dbg(dev->v4l2_dev.dev, "can't get remote parent\n");
>> + goto cleanup_exit;
>> + }
>> +
>> + dev_dbg(dev->v4l2_dev.dev, "found subdevice %pOF\n", sensor_node);
>
> There's no subdev yet, we're dealing with DT here.
>
> dev_dbg(dev->v4l2_dev.dev, "found source device %pOF\n", sensor_node);
>
> But I would actually drop all this. The sensor_node is only used in
> debug or error messages, so you could remove it and use ep_node in those
> messages instead. Or even better, drop the "subdevice %pOF: " prefix in
> the messages, that will make the code simpler.
>
>> +
>> + /* Parse the local endpoint and validate its configuration. */
>> + v4l2_fwnode_endpoint_parse(of_fwnode_handle(ep_node), &ep);
>
> Use v4l2_fwnode_endpoint_alloc_parse() and check for errors (and update
> the error cleanup path accordingly).
>
>> +
>> + dev_dbg(dev->v4l2_dev.dev, "parsed local endpoint, bus_type %u\n",
>> + ep.bus_type);
>> +
>> + dev->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(dev->v4l2_dev.dev, "subdevice %pOF: %u data lanes not supported\n",
>> + sensor_node,
>> + 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(dev->v4l2_dev.dev, "subdevice %pOF: data lanes reordering not supported\n",
>> + sensor_node);
>> + goto cleanup_exit;
>> + }
>> + }
>> +
>> + if (ep.bus.mipi_csi2.num_data_lanes > dev->max_data_lanes) {
>> + dev_err(dev->v4l2_dev.dev, "subdevice requires %u data lanes when %u are supported\n",
>
> s/subdevice/endpoint/
>
>> + ep.bus.mipi_csi2.num_data_lanes,
>> + dev->max_data_lanes);
>> + }
>> +
>> + dev->max_data_lanes = ep.bus.mipi_csi2.num_data_lanes;
>> + dev->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(dev->v4l2_dev.dev, "subdevice %pOF: unsupported lanes configuration\n",
>> + sensor_node);
>> + goto cleanup_exit;
>> + }
>> +
>> + dev->max_data_lanes = 1;
>> + dev->bus_flags = ep.bus.mipi_csi1.strobe;
>> + break;
>
> I wonder if this driver will ever be used with a CCP2 sensor.
>
>> +
>> + default:
>> + /* Unsupported bus type */
>> + dev_err(dev->v4l2_dev.dev, "subdevice %pOF: unsupported bus type %u\n",
>> + sensor_node, ep.bus_type);
>> + goto cleanup_exit;
>> + }
>> +
>> + dev_dbg(dev->v4l2_dev.dev, "subdevice %pOF: %s bus, %u data lanes, flags=0x%08x\n",
>> + sensor_node,
>> + dev->bus_type == V4L2_MBUS_CSI2_DPHY ? "CSI-2" : "CCP2",
>> + dev->max_data_lanes, dev->bus_flags);
>> +
>> + /* Initialize and register the async notifier. */
>> + v4l2_async_nf_init(&dev->notifier);
>> + dev->notifier.ops = &unicam_async_ops;
>> +
>> + dev->asd.match_type = V4L2_ASYNC_MATCH_FWNODE;
>> + dev->asd.match.fwnode = fwnode_graph_get_remote_endpoint(of_fwnode_handle(ep_node));
>> + ret = v4l2_async_nf_add_subdev(&dev->notifier, &dev->asd);
>
> The asd must be dynamically allocated as stated by the documentation of
> __v4l2_async_nf_add_subdev(). You can use
> v4l2_async_nf_add_fwnode_remote() to simplify this.
>
>> + if (ret) {
>> + dev_err(dev->v4l2_dev.dev, "Error adding subdevice: %d\n", ret);
>> + goto cleanup_exit;
>> + }
>> +
>> + ret = v4l2_async_nf_register(&dev->v4l2_dev, &dev->notifier);
>> + if (ret) {
>> + dev_err(dev->v4l2_dev.dev, "Error registering async notifier: %d\n", ret);
>> + ret = -EINVAL;
>> + }
>> +
>> +cleanup_exit:
>> + of_node_put(sensor_node);
>> + of_node_put(ep_node);
>> +
>> + return ret;
>> +}
>> +
>> +static int bcm2835_media_dev_init(struct unicam_device *unicam,
>> + struct platform_device *pdev)
>> +{
>> + int ret = 0;
>> +
>> + unicam->mdev.dev = &pdev->dev;
>> + strscpy(unicam->mdev.model, UNICAM_MODULE_NAME,
>> + sizeof(unicam->mdev.model));
>> + strscpy(unicam->mdev.serial, "", sizeof(unicam->mdev.serial));
>> + snprintf(unicam->mdev.bus_info, sizeof(unicam->mdev.bus_info),
>> + "platform:%s", dev_name(&pdev->dev));
>> + unicam->mdev.hw_revision = 0;
>> +
>> + media_device_init(&unicam->mdev);
>> +
>> + unicam->v4l2_dev.mdev = &unicam->mdev;
>> +
>> + ret = v4l2_device_register(&pdev->dev, &unicam->v4l2_dev);
>
> return v4l2_device_register(&pdev->dev, &unicam->v4l2_dev);
>
>> +
>> + return ret;
>> +}
>> +
>> +/* Internal subdev */
>> +
>> +static int _unicam_subdev_set_routing(struct v4l2_subdev *sd,
>
> __ is more conventional.
>
>> + struct v4l2_subdev_state *state,
>> + struct v4l2_subdev_krouting *routing)
>> +{
>> + static const struct v4l2_mbus_framefmt format = {
>> + .width = 640,
>> + .height = 480,
>> + .code = MEDIA_BUS_FMT_UYVY8_2X8,
>> + .field = V4L2_FIELD_NONE,
>> + .colorspace = V4L2_COLORSPACE_SRGB,
>> + .ycbcr_enc = V4L2_YCBCR_ENC_601,
>> + .quantization = V4L2_QUANTIZATION_LIM_RANGE,
>> + .xfer_func = V4L2_XFER_FUNC_SRGB,
>> + .flags = 0,
>> + };
>
> As this should match the values in unicam_mc_set_default_format(), can
> you move this structure out of the function, and use it to initialize
> the pix_fmt fields in unicam_mc_set_default_format() ?
>
>> + int ret;
>> +
>> + ret = v4l2_subdev_routing_validate_1_to_1(routing);
>> + if (ret)
>> + return ret;
>> +
>> + v4l2_subdev_lock_state(state);
>> +
>> + ret = v4l2_subdev_set_routing_with_fmt(sd, state, routing, &format);
>> +
>> + v4l2_subdev_unlock_state(state);
>> +
>> + if (ret)
>> + return ret;
>> +
>> + return 0;
>> +}
>> +
>> +static int unicam_subdev_set_routing(struct v4l2_subdev *sd,
>> + struct v4l2_subdev_state *state,
>> + enum v4l2_subdev_format_whence which,
>> + struct v4l2_subdev_krouting *routing)
>> +{
>> + struct unicam_device *unicam = sd_to_unicam_device(sd);
>> +
>> + if (which == V4L2_SUBDEV_FORMAT_ACTIVE && unicam->streaming)
>> + return -EBUSY;
>> +
>> + return _unicam_subdev_set_routing(sd, state, routing);
>> +}
>> +
>> +static int unicam_subdev_init_cfg(struct v4l2_subdev *sd,
>> + struct v4l2_subdev_state *state)
>> +{
>> + struct v4l2_subdev_route routes[] = {
>> + {
>> + .sink_pad = 0,
>> + .sink_stream = 0,
>> + .source_pad = 1,
>
> You have macros you can use for the pad numbers.
>
>> + .source_stream = 0,
>> + .flags = V4L2_SUBDEV_ROUTE_FL_ACTIVE,
>> + },
>> + };
>> +
>> + struct v4l2_subdev_krouting routing = {
>> + .num_routes = ARRAY_SIZE(routes),
>> + .routes = routes,
>> + };
>> +
>> + /* Initialize routing to single route to the fist source pad */
>> + return _unicam_subdev_set_routing(sd, state, &routing);
>> +}
>> +
>> +static int unicam_subdev_enum_mbus_code(struct v4l2_subdev *sd,
>> + struct v4l2_subdev_state *state,
>> + struct v4l2_subdev_mbus_code_enum *code)
>> +{
>> + int ret = 0;
>> +
>> + v4l2_subdev_lock_state(state);
>> +
>> + /* No transcoding, source and sink codes must match. */
>> + if (unicam_sd_pad_is_source(code->pad)) {
>> + struct v4l2_mbus_framefmt *fmt;
>> +
>> + if (code->index > 0) {
>> + ret = -EINVAL;
>> + goto out;
>> + }
>> +
>> + fmt = v4l2_subdev_state_get_opposite_stream_format(state,
>> + code->pad,
>> + code->stream);
>> + if (!fmt) {
>> + ret = -EINVAL;
>> + goto out;
>> + }
>> +
>> + code->code = fmt->code;
>> + } else {
>> + if (code->index >= ARRAY_SIZE(formats)) {
>> + ret = -EINVAL;
>> + goto out;
>> + }
>> +
>> + code->code = formats[code->index].code;
>> + }
>> +
>> +out:
>> + v4l2_subdev_unlock_state(state);
>> +
>> + return ret;
>> +}
>> +
>> +static int unicam_subdev_start_streaming(struct unicam_device *unicam)
>> +{
>> + const struct v4l2_subdev_krouting *routing;
>> + struct v4l2_subdev_state *state;
>> + int ret = 0;
>
> No need to initialize ret to 0.
>
>> +
>> + state = v4l2_subdev_lock_active_state(&unicam->sd);
>> +
>> + routing = &state->routing;
>> +
>> + v4l2_subdev_unlock_state(state);
>
> As a piece of conceptual art, maybe, but as useful code, I have some
> doubts.
>
>> +
>> + unicam->streaming = true;
>> +
>> + ret = v4l2_subdev_call(unicam->sensor, video, s_stream, 1);
>> + if (ret) {
>> + v4l2_subdev_call(unicam->sensor, video, s_stream, 0);
>
> If .s_stream(1) fails, why do you need to call .s_stream(0) ?
>
>> + unicam->streaming = false;
>
> Drop this, and move the unicam->streaming = true line below.
>
>> + return ret;
>> + }
>> +
>> + return 0;
>> +}
>> +
>> +static int unicam_subdev_stop_streaming(struct unicam_device *unicam)
>> +{
>> + v4l2_subdev_call(unicam->sensor, video, s_stream, 0);
>> +
>> + unicam->streaming = false;
>> +
>> + return 0;
>> +}
>> +
>> +static int unicam_subdev_s_stream(struct v4l2_subdev *sd, int enable)
>> +{
>> + struct unicam_device *unicam = sd_to_unicam_device(sd);
>> +
>> + if (enable)
>> + return unicam_subdev_start_streaming(unicam);
>> + else
>> + return unicam_subdev_stop_streaming(unicam);
>> + return 0;
>
> Drop the last line.
>
>> +}
>> +
>> +static int unicam_subdev_set_pad_format(struct v4l2_subdev *sd,
>> + struct v4l2_subdev_state *state,
>> + struct v4l2_subdev_format *format)
>> +{
>> + struct unicam_device *unicam = sd_to_unicam_device(sd);
>> + struct v4l2_mbus_framefmt *fmt;
>> + int ret = 0;
>> +
>> + if (format->which == V4L2_SUBDEV_FORMAT_ACTIVE && unicam->streaming)
>> + return -EBUSY;
>> +
>> + /* No transcoding, source and sink formats must match. */
>> + if (unicam_sd_pad_is_source(format->pad))
>> + return v4l2_subdev_get_fmt(sd, state, format);
>> +
>> + // TODO: implement fmt validation
>
> That's a candidate for v3.
>
>> +
>> + v4l2_subdev_lock_state(state);
>> +
>> + fmt = v4l2_state_get_stream_format(state, format->pad, format->stream);
>> + if (!fmt) {
>> + ret = -EINVAL;
>> + goto out;
>> + }
>> +
>> + *fmt = format->format;
>> +
>> + fmt = v4l2_subdev_state_get_opposite_stream_format(state, format->pad,
>> + format->stream);
>> + if (!fmt) {
>> + ret = -EINVAL;
>> + goto out;
>> + }
>> +
>> + *fmt = format->format;
>
> Let's avoid applying the format on one pad to then fail on the other
> pad.
>
> sink_format = v4l2_state_get_stream_format(state, format->pad,
> format->stream);
> source_format = v4l2_subdev_state_get_opposite_stream_format(state,
> format->pad,
> format->stream);
> if (!sink_format || !source_format) {
> ret = -EINVAL;
> goto out;
> }
>
> *sink_format = format->format;
> *source_format = format->format;
>
>> +
>> +out:
>> + v4l2_subdev_unlock_state(state);
>> +
>> + return ret;
>> +}
>> +
>> +static int unicam_subdev_enum_frame_size(struct v4l2_subdev *sd,
>> + struct v4l2_subdev_state *state,
>> + struct v4l2_subdev_frame_size_enum *fse)
>> +{
>> + const struct unicam_fmt *fmtinfo;
>> + int ret = 0;
>> +
>> + if (fse->index > 0)
>> + return -EINVAL;
>> +
>> + v4l2_subdev_lock_state(state);
>> +
>> + /* No transcoding, source and sink formats must match. */
>> + if (unicam_sd_pad_is_source(fse->pad)) {
>> + struct v4l2_mbus_framefmt *fmt;
>> +
>> + fmt = v4l2_subdev_state_get_opposite_stream_format(state,
>> + fse->pad,
>> + fse->stream);
>> +
>
> You can drop this blank line.
>
>> + if (!fmt) {
>> + ret = -EINVAL;
>> + goto out;
>> + }
>> +
>> + if (fse->code != fmt->code) {
>> + ret = -EINVAL;
>> + goto out;
>> + }
>> +
>> + fse->min_width = fmt->width;
>> + fse->max_width = fmt->width;
>> + fse->min_height = fmt->height;
>> + fse->max_height = fmt->height;
>> + } else {
>> + fmtinfo = find_format_by_code(fse->code);
>> + if (!fmtinfo) {
>> + ret = -EINVAL;
>> + goto out;
>> + }
>> +
>> + fse->min_width = MIN_WIDTH * 8 / ALIGN(fmtinfo->depth, 8);
>> + fse->max_width = MAX_WIDTH * 8 / ALIGN(fmtinfo->depth, 8);
>> + fse->min_height = MIN_HEIGHT;
>> + fse->max_height = MAX_HEIGHT;
>> + }
>> +
>> +out:
>> + v4l2_subdev_unlock_state(state);
>> +
>> + return ret;
>> +}
>> +
>> +static const struct v4l2_subdev_video_ops unicam_subdev_video_ops = {
>> + .s_stream = unicam_subdev_s_stream,
>> +};
>> +
>> +static const struct v4l2_subdev_pad_ops unicam_subdev_pad_ops = {
>> + .init_cfg = unicam_subdev_init_cfg,
>> + .enum_mbus_code = unicam_subdev_enum_mbus_code,
>> + .get_fmt = v4l2_subdev_get_fmt,
>> + .set_fmt = unicam_subdev_set_pad_format,
>> + .set_routing = unicam_subdev_set_routing,
>> + .enum_frame_size = unicam_subdev_enum_frame_size,
>> +};
>> +
>> +static const struct v4l2_subdev_ops unicam_subdev_ops = {
>> + .video = &unicam_subdev_video_ops,
>> + .pad = &unicam_subdev_pad_ops,
>> +};
>> +
>> +static struct media_entity_operations unicam_subdev_media_ops = {
>
> static const
>
>> + .link_validate = v4l2_subdev_link_validate,
>> + .has_route = v4l2_subdev_has_route,
>> +};
>> +
>> +static int unicam_probe(struct platform_device *pdev)
>> +{
>> + struct unicam_device *unicam;
>> + int ret = 0;
>> + int i;
>> +
>> + unicam = kzalloc(sizeof(*unicam), GFP_KERNEL);
>> + if (!unicam)
>> + return -ENOMEM;
>> +
>> + kref_init(&unicam->kref);
>> + unicam->pdev = pdev;
>> +
>> + unicam->base = devm_platform_ioremap_resource(pdev, 0);
>> + if (IS_ERR(unicam->base)) {
>> + dev_err(unicam->v4l2_dev.dev, "Failed to get main io block\n");
>
> Let's use unicam->dev to access the struct device, especially given that
> unicam->v4l2_dev is only initialized below when calling
> bcm2835_media_dev_init().
>
> This message can actually br dropped, devm_platform_ioremap_resource()
> already prints an error.
>
>> + ret = PTR_ERR(unicam->base);
>> + goto err_unicam_put;
>> + }
>> +
>> + unicam->clk_gate_base = devm_platform_ioremap_resource(pdev, 1);
>> + if (IS_ERR(unicam->clk_gate_base)) {
>> + dev_err(unicam->v4l2_dev.dev, "Failed to get 2nd io block\n");
>
> Same here.
>
>> + ret = PTR_ERR(unicam->clk_gate_base);
>> + goto err_unicam_put;
>> + }
>> +
>> + unicam->clock = devm_clk_get(&pdev->dev, "lp");
>> + if (IS_ERR(unicam->clock)) {
>> + dev_err(unicam->v4l2_dev.dev, "Failed to get lp clock\n");
>> + ret = PTR_ERR(unicam->clock);
>> + goto err_unicam_put;
>> + }
>> +
>> + unicam->vpu_clock = devm_clk_get(&pdev->dev, "vpu");
>> + if (IS_ERR(unicam->vpu_clock)) {
>> + dev_err(unicam->v4l2_dev.dev, "Failed to get vpu clock\n");
>> + ret = PTR_ERR(unicam->vpu_clock);
>> + goto err_unicam_put;
>> + }
>
> Could the clock bulk API help here ?
>
>> +
>> + ret = platform_get_irq(pdev, 0);
>> + if (ret <= 0) {
>> + dev_err(&pdev->dev, "No IRQ resource\n");
>> + ret = -EINVAL;
>> + goto err_unicam_put;
>> + }
>> +
>> + ret = devm_request_irq(&pdev->dev, ret, unicam_isr, 0,
>> + "unicam_capture0", unicam);
>> + if (ret) {
>> + dev_err(&pdev->dev, "Unable to request interrupt\n");
>> + ret = -EINVAL;
>> + goto err_unicam_put;
>> + }
>> +
>> + ret = bcm2835_media_dev_init(unicam, pdev);
>> + if (ret) {
>> + dev_err(unicam->v4l2_dev.dev,
>> + "Unable to register v4l2 device.\n");
>> + goto err_unicam_put;
>> + }
>> +
>> + ret = media_device_register(&unicam->mdev);
>> + if (ret < 0) {
>> + dev_err(unicam->v4l2_dev.dev,
>> + "Unable to register media-controller device.\n");
>> + goto err_v4l2_unregister;
>> + }
>> +
>> + /* Reserve space for the controls */
>> + ret = v4l2_ctrl_handler_init(&unicam->ctrl_handler, 16);
>> + if (ret < 0)
>> + goto err_media_unregister;
>
> The control handler is unused, drop it.
>
>> +
>> + /* set the driver data in platform device */
>> + platform_set_drvdata(pdev, unicam);
>> +
>> + dev_dbg(unicam->v4l2_dev.dev, "Initialize internal subdev");
>
> You can drop this.
>
>> + v4l2_subdev_init(&unicam->sd, &unicam_subdev_ops);
>> + v4l2_set_subdevdata(&unicam->sd, unicam);
>> + unicam->sd.entity.function = MEDIA_ENT_F_VID_IF_BRIDGE;
>> + unicam->sd.dev = &pdev->dev;
>> + unicam->sd.owner = THIS_MODULE;
>> + unicam->sd.flags = V4L2_SUBDEV_FL_HAS_DEVNODE | V4L2_SUBDEV_FL_MULTIPLEXED;
>> + snprintf(unicam->sd.name, sizeof(unicam->sd.name), "unicam-subdev");
>> +
>> + unicam->pads[UNICAM_SD_PAD_SINK].flags = MEDIA_PAD_FL_SINK;
>> +
>> + for (i = UNICAM_SD_PAD_FIRST_SOURCE; i < UNICAM_SD_NUM_PADS; ++i)
>> + unicam->pads[i].flags = MEDIA_PAD_FL_SOURCE;
>
> unicam->pads[UNICAM_SD_PAD_SOURCE_IMAGE].flags = MEDIA_PAD_FL_SOURCE;
> unicam->pads[UNICAM_SD_PAD_SOURCE_META].flags = MEDIA_PAD_FL_SOURCE;
>
>> + unicam->sd.entity.ops = &unicam_subdev_media_ops;
>> + ret = media_entity_pads_init(&unicam->sd.entity,
>> + ARRAY_SIZE(unicam->pads), unicam->pads);
>> + if (ret) {
>> + dev_err(unicam->v4l2_dev.dev, "Could not init media controller for subdev");
>> + goto err_subdev_unregister;
>> + }
>> +
>> + ret = v4l2_subdev_init_finalize(&unicam->sd);
>> + if (ret) {
>> + dev_err(unicam->v4l2_dev.dev, "Could not finalize init for subdev");
>> + goto err_entity_cleanup;
>> + }
>> +
>> + ret = v4l2_device_register_subdev(&unicam->v4l2_dev, &unicam->sd);
>> + if (ret) {
>> + dev_err(&pdev->dev, "Failed to register internal subdev\n");
>> + goto err_subdev_unregister;
>> + }
>
> Let's split initialization and registration of the subdev to a separate
> function, placed above with the subdev operations.
>
>> +
>> + ret = of_unicam_connect_subdevs(unicam);
>> + if (ret) {
>> + dev_err(&pdev->dev, "Failed to connect subdevs\n");
>> + goto err_subdev_unregister;
>> + }
>> +
>> + /* Enable the block power domain */
>> + pm_runtime_enable(&pdev->dev);
>> +
>> + return 0;
>> +
>> +err_subdev_unregister:
>> + v4l2_subdev_cleanup(&unicam->sd);
>> +err_entity_cleanup:
>> + media_entity_cleanup(&unicam->sd.entity);
>> +err_media_unregister:
>> + media_device_unregister(&unicam->mdev);
>> +err_v4l2_unregister:
>> + v4l2_device_unregister(&unicam->v4l2_dev);
>> +err_unicam_put:
>> + unicam_put(unicam);
>> +
>> + return ret;
>> +}
>> +
>> +static int unicam_remove(struct platform_device *pdev)
>> +{
>> + struct unicam_device *unicam = platform_get_drvdata(pdev);
>> +
>> + v4l2_async_nf_unregister(&unicam->notifier);
>> + v4l2_device_unregister(&unicam->v4l2_dev);
>> + media_device_unregister(&unicam->mdev);
>> + unregister_nodes(unicam);
>
> I'm a bit worried there could be race conditions with userspace here.
> For instance, calling v4l2_async_nf_unregister() will result in
> v4l2_device_unregister_subdev() being called on the sensor subdev, which
> may race with userspace starting capture on a video node. The following
> order would, I think be safer:
>
> unregister_nodes(unicam);
> v4l2_device_unregister(&unicam->v4l2_dev);
> media_device_unregister(&unicam->mdev);
> v4l2_async_nf_unregister(&unicam->notifier);
>
> But this will cause a problem, when unregistering device nodes, the last
> reference to unicam would be dropped. I think you could drop the calls
> to unicam_put() from unicam_async_complete(), and add a unicam_put()
> call here. Dave, does that sound good to you ?
>
>> +
>> + pm_runtime_disable(&pdev->dev);
>> +
>> + return 0;
>> +}
>> +
>> +static const struct of_device_id unicam_of_match[] = {
>> + { .compatible = "brcm,bcm2835-unicam", },
>> + { /* sentinel */ },
>> +};
>> +MODULE_DEVICE_TABLE(of, unicam_of_match);
>> +
>> +static struct platform_driver unicam_driver = {
>> + .probe = unicam_probe,
>> + .remove = unicam_remove,
>> + .driver = {
>> + .name = UNICAM_MODULE_NAME,
>> + .of_match_table = of_match_ptr(unicam_of_match),
>> + },
>> +};
>> +
>> +module_platform_driver(unicam_driver);
>> +
>> +MODULE_AUTHOR("Dave Stevenson <dave.stevenson@raspberrypi.com>");
>> +MODULE_DESCRIPTION("BCM2835 Unicam driver");
>> +MODULE_LICENSE("GPL");
>> +MODULE_VERSION(UNICAM_VERSION);
>> diff --git a/drivers/media/platform/bcm2835/vc4-regs-unicam.h b/drivers/media/platform/bcm2835/vc4-regs-unicam.h
>> new file mode 100644
>> index 000000000000..ae059a171d0f
>> --- /dev/null
>> +++ b/drivers/media/platform/bcm2835/vc4-regs-unicam.h
>> @@ -0,0 +1,253 @@
>> +/* SPDX-License-Identifier: GPL-2.0-only */
>> +
>> +/*
>> + * Copyright (C) 2017-2020 Raspberry Pi Trading.
>> + * Dave Stevenson <dave.stevenson@raspberrypi.com>
>> + */
>> +
>> +#ifndef VC4_REGS_UNICAM_H
>> +#define VC4_REGS_UNICAM_H
>> +
>> +/*
>> + * The following values are taken from files found within the code drop
>> + * made by Broadcom for the BCM21553 Graphics Driver, predominantly in
>> + * brcm_usrlib/dag/vmcsx/vcinclude/hardware_vc4.h.
>> + * They have been modified to be only the register offset.
>> + */
>> +#define UNICAM_CTRL 0x000
>> +#define UNICAM_STA 0x004
>> +#define UNICAM_ANA 0x008
>> +#define UNICAM_PRI 0x00c
>> +#define UNICAM_CLK 0x010
>> +#define UNICAM_CLT 0x014
>> +#define UNICAM_DAT0 0x018
>> +#define UNICAM_DAT1 0x01c
>> +#define UNICAM_DAT2 0x020
>> +#define UNICAM_DAT3 0x024
>> +#define UNICAM_DLT 0x028
>> +#define UNICAM_CMP0 0x02c
>> +#define UNICAM_CMP1 0x030
>> +#define UNICAM_CAP0 0x034
>> +#define UNICAM_CAP1 0x038
>> +#define UNICAM_ICTL 0x100
>> +#define UNICAM_ISTA 0x104
>> +#define UNICAM_IDI0 0x108
>> +#define UNICAM_IPIPE 0x10c
>> +#define UNICAM_IBSA0 0x110
>> +#define UNICAM_IBEA0 0x114
>> +#define UNICAM_IBLS 0x118
>> +#define UNICAM_IBWP 0x11c
>> +#define UNICAM_IHWIN 0x120
>> +#define UNICAM_IHSTA 0x124
>> +#define UNICAM_IVWIN 0x128
>> +#define UNICAM_IVSTA 0x12c
>> +#define UNICAM_ICC 0x130
>> +#define UNICAM_ICS 0x134
>> +#define UNICAM_IDC 0x138
>> +#define UNICAM_IDPO 0x13c
>> +#define UNICAM_IDCA 0x140
>> +#define UNICAM_IDCD 0x144
>> +#define UNICAM_IDS 0x148
>> +#define UNICAM_DCS 0x200
>> +#define UNICAM_DBSA0 0x204
>> +#define UNICAM_DBEA0 0x208
>> +#define UNICAM_DBWP 0x20c
>> +#define UNICAM_DBCTL 0x300
>> +#define UNICAM_IBSA1 0x304
>> +#define UNICAM_IBEA1 0x308
>> +#define UNICAM_IDI1 0x30c
>> +#define UNICAM_DBSA1 0x310
>> +#define UNICAM_DBEA1 0x314
>> +#define UNICAM_MISC 0x400
>
> Please add one tab before the value, to match the indentation of the
> rest of the file.
>
>> +
>> +/*
>> + * The following bitmasks are from the kernel released by Broadcom
>> + * for Android - https://android.googlesource.com/kernel/bcm/
>> + * The Rhea, Hawaii, and Java chips all contain the same VideoCore4
>> + * Unicam block as BCM2835, as defined in eg
>> + * arch/arm/mach-rhea/include/mach/rdb_A0/brcm_rdb_cam.h and similar.
>> + * Values reworked to use the kernel BIT and GENMASK macros.
>> + *
>> + * Some of the bit mnenomics have been amended to match the datasheet.
>> + */
>> +/* UNICAM_CTRL Register */
>> +#define UNICAM_CPE BIT(0)
>> +#define UNICAM_MEM BIT(1)
>> +#define UNICAM_CPR BIT(2)
>> +#define UNICAM_CPM_MASK GENMASK(3, 3)
>> +#define UNICAM_CPM_CSI2 0
>> +#define UNICAM_CPM_CCP2 1
>> +#define UNICAM_SOE BIT(4)
>> +#define UNICAM_DCM_MASK GENMASK(5, 5)
>> +#define UNICAM_DCM_STROBE 0
>> +#define UNICAM_DCM_DATA 1
>> +#define UNICAM_SLS BIT(6)
>> +#define UNICAM_PFT_MASK GENMASK(11, 8)
>> +#define UNICAM_OET_MASK GENMASK(20, 12)
>> +
>> +/* UNICAM_STA Register */
>> +#define UNICAM_SYN BIT(0)
>> +#define UNICAM_CS BIT(1)
>> +#define UNICAM_SBE BIT(2)
>> +#define UNICAM_PBE BIT(3)
>> +#define UNICAM_HOE BIT(4)
>> +#define UNICAM_PLE BIT(5)
>> +#define UNICAM_SSC BIT(6)
>> +#define UNICAM_CRCE BIT(7)
>> +#define UNICAM_OES BIT(8)
>> +#define UNICAM_IFO BIT(9)
>> +#define UNICAM_OFO BIT(10)
>> +#define UNICAM_BFO BIT(11)
>> +#define UNICAM_DL BIT(12)
>> +#define UNICAM_PS BIT(13)
>> +#define UNICAM_IS BIT(14)
>> +#define UNICAM_PI0 BIT(15)
>> +#define UNICAM_PI1 BIT(16)
>> +#define UNICAM_FSI_S BIT(17)
>> +#define UNICAM_FEI_S BIT(18)
>> +#define UNICAM_LCI_S BIT(19)
>> +#define UNICAM_BUF0_RDY BIT(20)
>> +#define UNICAM_BUF0_NO BIT(21)
>> +#define UNICAM_BUF1_RDY BIT(22)
>> +#define UNICAM_BUF1_NO BIT(23)
>> +#define UNICAM_DI BIT(24)
>> +
>> +#define UNICAM_STA_MASK_ALL \
>> + (UNICAM_DL + \
>> + UNICAM_SBE + \
>> + UNICAM_PBE + \
>> + UNICAM_HOE + \
>> + UNICAM_PLE + \
>> + UNICAM_SSC + \
>> + UNICAM_CRCE + \
>> + UNICAM_IFO + \
>> + UNICAM_OFO + \
>> + UNICAM_PS + \
>> + UNICAM_PI0 + \
>> + UNICAM_PI1)
>
> I'd use | instead of + to combine bits. Same below.
>
>> +
>> +/* UNICAM_ANA Register */
>> +#define UNICAM_APD BIT(0)
>> +#define UNICAM_BPD BIT(1)
>> +#define UNICAM_AR BIT(2)
>> +#define UNICAM_DDL BIT(3)
>> +#define UNICAM_CTATADJ_MASK GENMASK(7, 4)
>> +#define UNICAM_PTATADJ_MASK GENMASK(11, 8)
>> +
>> +/* UNICAM_PRI Register */
>> +#define UNICAM_PE BIT(0)
>> +#define UNICAM_PT_MASK GENMASK(2, 1)
>> +#define UNICAM_NP_MASK GENMASK(7, 4)
>> +#define UNICAM_PP_MASK GENMASK(11, 8)
>> +#define UNICAM_BS_MASK GENMASK(15, 12)
>> +#define UNICAM_BL_MASK GENMASK(17, 16)
>> +
>> +/* UNICAM_CLK Register */
>> +#define UNICAM_CLE BIT(0)
>> +#define UNICAM_CLPD BIT(1)
>> +#define UNICAM_CLLPE BIT(2)
>> +#define UNICAM_CLHSE BIT(3)
>> +#define UNICAM_CLTRE BIT(4)
>> +#define UNICAM_CLAC_MASK GENMASK(8, 5)
>> +#define UNICAM_CLSTE BIT(29)
>> +
>> +/* UNICAM_CLT Register */
>> +#define UNICAM_CLT1_MASK GENMASK(7, 0)
>> +#define UNICAM_CLT2_MASK GENMASK(15, 8)
>> +
>> +/* UNICAM_DATn Registers */
>> +#define UNICAM_DLE BIT(0)
>> +#define UNICAM_DLPD BIT(1)
>> +#define UNICAM_DLLPE BIT(2)
>> +#define UNICAM_DLHSE BIT(3)
>> +#define UNICAM_DLTRE BIT(4)
>> +#define UNICAM_DLSM BIT(5)
>> +#define UNICAM_DLFO BIT(28)
>> +#define UNICAM_DLSTE BIT(29)
>> +
>> +#define UNICAM_DAT_MASK_ALL (UNICAM_DLSTE + UNICAM_DLFO)
>
> This also needs an indentation fix, as well as UNICAM_ISTA_MASK_ALL.
>
>> +
>> +/* UNICAM_DLT Register */
>> +#define UNICAM_DLT1_MASK GENMASK(7, 0)
>> +#define UNICAM_DLT2_MASK GENMASK(15, 8)
>> +#define UNICAM_DLT3_MASK GENMASK(23, 16)
>> +
>> +/* UNICAM_ICTL Register */
>> +#define UNICAM_FSIE BIT(0)
>> +#define UNICAM_FEIE BIT(1)
>> +#define UNICAM_IBOB BIT(2)
>> +#define UNICAM_FCM BIT(3)
>> +#define UNICAM_TFC BIT(4)
>> +#define UNICAM_LIP_MASK GENMASK(6, 5)
>> +#define UNICAM_LCIE_MASK GENMASK(28, 16)
>> +
>> +/* UNICAM_IDI0/1 Register */
>> +#define UNICAM_ID0_MASK GENMASK(7, 0)
>> +#define UNICAM_ID1_MASK GENMASK(15, 8)
>> +#define UNICAM_ID2_MASK GENMASK(23, 16)
>> +#define UNICAM_ID3_MASK GENMASK(31, 24)
>> +
>> +/* UNICAM_ISTA Register */
>> +#define UNICAM_FSI BIT(0)
>> +#define UNICAM_FEI BIT(1)
>> +#define UNICAM_LCI BIT(2)
>> +
>> +#define UNICAM_ISTA_MASK_ALL (UNICAM_FSI + UNICAM_FEI + UNICAM_LCI)
>> +
>> +/* UNICAM_IPIPE Register */
>> +#define UNICAM_PUM_MASK GENMASK(2, 0)
>> + /* Unpacking modes */
>> + #define UNICAM_PUM_NONE 0
>> + #define UNICAM_PUM_UNPACK6 1
>> + #define UNICAM_PUM_UNPACK7 2
>> + #define UNICAM_PUM_UNPACK8 3
>> + #define UNICAM_PUM_UNPACK10 4
>> + #define UNICAM_PUM_UNPACK12 5
>> + #define UNICAM_PUM_UNPACK14 6
>> + #define UNICAM_PUM_UNPACK16 7
>> +#define UNICAM_DDM_MASK GENMASK(6, 3)
>> +#define UNICAM_PPM_MASK GENMASK(9, 7)
>> + /* Packing modes */
>> + #define UNICAM_PPM_NONE 0
>> + #define UNICAM_PPM_PACK8 1
>> + #define UNICAM_PPM_PACK10 2
>> + #define UNICAM_PPM_PACK12 3
>> + #define UNICAM_PPM_PACK14 4
>> + #define UNICAM_PPM_PACK16 5
>
> I'd drop the leading tab.
>
>> +#define UNICAM_DEM_MASK GENMASK(11, 10)
>> +#define UNICAM_DEBL_MASK GENMASK(14, 12)
>> +#define UNICAM_ICM_MASK GENMASK(16, 15)
>> +#define UNICAM_IDM_MASK GENMASK(17, 17)
>> +
>> +/* UNICAM_ICC Register */
>> +#define UNICAM_ICFL_MASK GENMASK(4, 0)
>> +#define UNICAM_ICFH_MASK GENMASK(9, 5)
>> +#define UNICAM_ICST_MASK GENMASK(12, 10)
>> +#define UNICAM_ICLT_MASK GENMASK(15, 13)
>> +#define UNICAM_ICLL_MASK GENMASK(31, 16)
>> +
>> +/* UNICAM_DCS Register */
>> +#define UNICAM_DIE BIT(0)
>> +#define UNICAM_DIM BIT(1)
>> +#define UNICAM_DBOB BIT(3)
>> +#define UNICAM_FDE BIT(4)
>> +#define UNICAM_LDP BIT(5)
>> +#define UNICAM_EDL_MASK GENMASK(15, 8)
>> +
>> +/* UNICAM_DBCTL Register */
>> +#define UNICAM_DBEN BIT(0)
>> +#define UNICAM_BUF0_IE BIT(1)
>> +#define UNICAM_BUF1_IE BIT(2)
>> +
>> +/* UNICAM_CMP[0,1] register */
>> +#define UNICAM_PCE BIT(31)
>> +#define UNICAM_GI BIT(9)
>> +#define UNICAM_CPH BIT(8)
>> +#define UNICAM_PCVC_MASK GENMASK(7, 6)
>> +#define UNICAM_PCDT_MASK GENMASK(5, 0)
>> +
>> +/* UNICAM_MISC register */
>> +#define UNICAM_FL0 BIT(6)
>> +#define UNICAM_FL1 BIT(9)
>> +
>> +#endif
>
^ permalink raw reply
* [PATCH v4 4/4] arm64: dts: imx8mp: Add memory for USB3 glue layer to usb3 nodes
From: Alexander Stein @ 2022-01-26 14:13 UTC (permalink / raw)
To: Greg Kroah-Hartman, Rob Herring, Shawn Guo, Sascha Hauer,
Fabio Estevam
Cc: Alexander Stein, NXP Linux Team, linux-usb, devicetree,
linux-arm-kernel, Li Jun
In-Reply-To: <20220126141340.234125-1-alexander.stein@ew.tq-group.com>
The USB3 glue layer has 2 areas in the register set, see RM Rev.1
section 11.2.5.2.1 GLUE_usb3 memory map:
* USB3 control/status
* PHY control/status
Provide the memory area to the usb3 nodes for accessing the features
in the USB3 control area.
Signed-off-by: Alexander Stein <alexander.stein@ew.tq-group.com>
---
arch/arm64/boot/dts/freescale/imx8mp.dtsi | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/arch/arm64/boot/dts/freescale/imx8mp.dtsi b/arch/arm64/boot/dts/freescale/imx8mp.dtsi
index e61ac5f136ad..d4aadb434d36 100644
--- a/arch/arm64/boot/dts/freescale/imx8mp.dtsi
+++ b/arch/arm64/boot/dts/freescale/imx8mp.dtsi
@@ -1030,7 +1030,8 @@ usb3_phy0: usb-phy@381f0040 {
usb3_0: usb@32f10100 {
compatible = "fsl,imx8mp-dwc3";
- reg = <0x32f10100 0x8>;
+ reg = <0x32f10100 0x8>,
+ <0x381f0000 0x20>;
clocks = <&clk IMX8MP_CLK_HSIO_ROOT>,
<&clk IMX8MP_CLK_USB_ROOT>;
clock-names = "hsio", "suspend";
@@ -1070,7 +1071,8 @@ usb3_phy1: usb-phy@382f0040 {
usb3_1: usb@32f10108 {
compatible = "fsl,imx8mp-dwc3";
- reg = <0x32f10108 0x8>;
+ reg = <0x32f10108 0x8>,
+ <0x382f0000 0x20>;
clocks = <&clk IMX8MP_CLK_HSIO_ROOT>,
<&clk IMX8MP_CLK_USB_ROOT>;
clock-names = "hsio", "suspend";
--
2.25.1
^ permalink raw reply related
* [PATCH v4 3/4] usb: dwc3: imx8mp: Add support for setting SOC specific flags
From: Alexander Stein @ 2022-01-26 14:13 UTC (permalink / raw)
To: Greg Kroah-Hartman, Rob Herring, Shawn Guo, Sascha Hauer,
Fabio Estevam
Cc: Alexander Stein, NXP Linux Team, linux-usb, devicetree,
linux-arm-kernel, Li Jun
In-Reply-To: <20220126141340.234125-1-alexander.stein@ew.tq-group.com>
The i.MX8MP glue layer has support for the following flags:
* over-current polarity
* PWR pad polarity
* controlling PPC flag in HCCPARAMS register
* permanent port attach for usb2 & usb3 port
Allow setting these flags by supporting specific flags in the glue node.
In order to get this to work an additional IORESOURCE_MEM and clock is
necessary. For backward compatibility this is purely optional.
Signed-off-by: Alexander Stein <alexander.stein@ew.tq-group.com>
---
drivers/usb/dwc3/dwc3-imx8mp.c | 62 ++++++++++++++++++++++++++++++++++
1 file changed, 62 insertions(+)
diff --git a/drivers/usb/dwc3/dwc3-imx8mp.c b/drivers/usb/dwc3/dwc3-imx8mp.c
index 1c8fe657b3a9..5a96b66bf237 100644
--- a/drivers/usb/dwc3/dwc3-imx8mp.c
+++ b/drivers/usb/dwc3/dwc3-imx8mp.c
@@ -36,10 +36,22 @@
#define USB_WAKEUP_EN_MASK GENMASK(5, 0)
+/* USB glue registers */
+#define USB_CTRL0 0x00
+#define USB_CTRL1 0x04
+
+#define USB_CTRL0_PORTPWR_EN BIT(12) /* 1 - PPC enabled (default) */
+#define USB_CTRL0_USB3_FIXED BIT(22) /* 1 - USB3 permanent attached */
+#define USB_CTRL0_USB2_FIXED BIT(23) /* 1 - USB2 permanent attached */
+
+#define USB_CTRL1_OC_POLARITY BIT(16) /* 0 - HIGH / 1 - LOW */
+#define USB_CTRL1_PWR_POLARITY BIT(17) /* 0 - HIGH / 1 - LOW */
+
struct dwc3_imx8mp {
struct device *dev;
struct platform_device *dwc3;
void __iomem *hsio_blk_base;
+ void __iomem *glue_base;
struct clk *hsio_clk;
struct clk *suspend_clk;
int irq;
@@ -47,6 +59,42 @@ struct dwc3_imx8mp {
bool wakeup_pending;
};
+static void imx8mp_configure_glue(struct dwc3_imx8mp *dwc3_imx)
+{
+ struct device *dev = dwc3_imx->dev;
+ u32 value;
+
+ if (!dwc3_imx->glue_base)
+ return;
+
+ value = readl(dwc3_imx->glue_base + USB_CTRL0);
+
+ if (device_property_read_bool(dev, "fsl,permanently-attached"))
+ value |= (USB_CTRL0_USB2_FIXED | USB_CTRL0_USB3_FIXED);
+ else
+ value &= ~(USB_CTRL0_USB2_FIXED | USB_CTRL0_USB3_FIXED);
+
+ if (device_property_read_bool(dev, "fsl,disable-port-power-control"))
+ value &= ~(USB_CTRL0_PORTPWR_EN);
+ else
+ value |= USB_CTRL0_PORTPWR_EN;
+
+ writel(value, dwc3_imx->glue_base + USB_CTRL0);
+
+ value = readl(dwc3_imx->glue_base + USB_CTRL1);
+ if (device_property_read_bool(dev, "fsl,over-current-active-low"))
+ value |= USB_CTRL1_OC_POLARITY;
+ else
+ value &= ~USB_CTRL1_OC_POLARITY;
+
+ if (device_property_read_bool(dev, "fsl,power-active-low"))
+ value |= USB_CTRL1_PWR_POLARITY;
+ else
+ value &= ~USB_CTRL1_PWR_POLARITY;
+
+ writel(value, dwc3_imx->glue_base + USB_CTRL1);
+}
+
static void dwc3_imx8mp_wakeup_enable(struct dwc3_imx8mp *dwc3_imx)
{
struct dwc3 *dwc3 = platform_get_drvdata(dwc3_imx->dwc3);
@@ -100,6 +148,7 @@ static int dwc3_imx8mp_probe(struct platform_device *pdev)
struct device *dev = &pdev->dev;
struct device_node *dwc3_np, *node = dev->of_node;
struct dwc3_imx8mp *dwc3_imx;
+ struct resource *res;
int err, irq;
if (!node) {
@@ -119,6 +168,15 @@ static int dwc3_imx8mp_probe(struct platform_device *pdev)
if (IS_ERR(dwc3_imx->hsio_blk_base))
return PTR_ERR(dwc3_imx->hsio_blk_base);
+ res = platform_get_resource(pdev, IORESOURCE_MEM, 1);
+ if (!res) {
+ dev_warn(dev, "Base address for glue layer missing. Continuing without, some features are missing though.");
+ } else {
+ dwc3_imx->glue_base = devm_ioremap_resource(dev, res);
+ if (IS_ERR(dwc3_imx->glue_base))
+ return PTR_ERR(dwc3_imx->glue_base);
+ }
+
dwc3_imx->hsio_clk = devm_clk_get(dev, "hsio");
if (IS_ERR(dwc3_imx->hsio_clk)) {
err = PTR_ERR(dwc3_imx->hsio_clk);
@@ -152,6 +210,8 @@ static int dwc3_imx8mp_probe(struct platform_device *pdev)
}
dwc3_imx->irq = irq;
+ imx8mp_configure_glue(dwc3_imx);
+
pm_runtime_set_active(dev);
pm_runtime_enable(dev);
err = pm_runtime_get_sync(dev);
@@ -252,6 +312,8 @@ static int __maybe_unused dwc3_imx8mp_resume(struct dwc3_imx8mp *dwc3_imx,
dwc3_imx8mp_wakeup_disable(dwc3_imx);
dwc3_imx->pm_suspended = false;
+ imx8mp_configure_glue(dwc3_imx);
+
if (dwc3_imx->wakeup_pending) {
dwc3_imx->wakeup_pending = false;
if (dwc->current_dr_role == DWC3_GCTL_PRTCAP_DEVICE) {
--
2.25.1
^ permalink raw reply related
* [PATCH v4 2/4] dt-bindings: usb: dwc3-imx8mp: Add imx8mp specific flags
From: Alexander Stein @ 2022-01-26 14:13 UTC (permalink / raw)
To: Greg Kroah-Hartman, Rob Herring, Shawn Guo, Sascha Hauer,
Fabio Estevam
Cc: Alexander Stein, NXP Linux Team, linux-usb, devicetree,
linux-arm-kernel, Li Jun
In-Reply-To: <20220126141340.234125-1-alexander.stein@ew.tq-group.com>
This adds bindings for features in the USB glue block. They allow
setting polarity of PWR and OC as well as disabling port power control.
Also permanently attached can be annotated as well.
Additional IO address and clock are needed.
Signed-off-by: Alexander Stein <alexander.stein@ew.tq-group.com>
---
.../bindings/usb/fsl,imx8mp-dwc3.yaml | 31 ++++++++++++++++---
1 file changed, 27 insertions(+), 4 deletions(-)
diff --git a/Documentation/devicetree/bindings/usb/fsl,imx8mp-dwc3.yaml b/Documentation/devicetree/bindings/usb/fsl,imx8mp-dwc3.yaml
index 974032b1fda0..01ab0f922ae8 100644
--- a/Documentation/devicetree/bindings/usb/fsl,imx8mp-dwc3.yaml
+++ b/Documentation/devicetree/bindings/usb/fsl,imx8mp-dwc3.yaml
@@ -15,9 +15,9 @@ properties:
const: fsl,imx8mp-dwc3
reg:
- maxItems: 1
- description: Address and length of the register set for the wrapper of
- dwc3 core on the SOC.
+ items:
+ - description: Address and length of the register set for HSIO Block Control
+ - description: Address and length of the register set for the wrapper of dwc3 core on the SOC.
"#address-cells":
enum: [ 1, 2 ]
@@ -49,6 +49,28 @@ properties:
- const: hsio
- const: suspend
+ fsl,permanently-attached:
+ type: boolean
+ description:
+ Indicates if the device atached to a downstream port is
+ permanently attached.
+
+ fsl,disable-port-power-control:
+ type: boolean
+ description:
+ Indicates whether the host controller implementation includes port
+ power control. Defines Bit 3 in capability register (HCCPARAMS).
+
+ fsl,over-current-active-low:
+ type: boolean
+ description:
+ Over current signal polarity is active low.
+
+ fsl,power-active-low:
+ type: boolean
+ description:
+ Power pad (PWR) polarity is active low.
+
# Required child node:
patternProperties:
@@ -74,7 +96,8 @@ examples:
#include <dt-bindings/interrupt-controller/arm-gic.h>
usb3_0: usb@32f10100 {
compatible = "fsl,imx8mp-dwc3";
- reg = <0x32f10100 0x8>;
+ reg = <0x32f10100 0x8>,
+ <0x381f0000 0x20>;
clocks = <&clk IMX8MP_CLK_HSIO_ROOT>,
<&clk IMX8MP_CLK_USB_ROOT>;
clock-names = "hsio", "suspend";
--
2.25.1
^ permalink raw reply related
* [PATCH v4 1/4] usb: dwc3: imx8mp: rename iomem base pointer
From: Alexander Stein @ 2022-01-26 14:13 UTC (permalink / raw)
To: Greg Kroah-Hartman, Rob Herring, Shawn Guo, Sascha Hauer,
Fabio Estevam
Cc: Alexander Stein, NXP Linux Team, linux-usb, devicetree,
linux-arm-kernel, Li Jun
In-Reply-To: <20220126141340.234125-1-alexander.stein@ew.tq-group.com>
Until now the iomem used is not USB glue as the name suggests, but
HSIO BLK_CTL. Rename the struct member accordingly. This is a preparing
patch for when USB glue is actually used.
Signed-off-by: Alexander Stein <alexander.stein@ew.tq-group.com>
---
drivers/usb/dwc3/dwc3-imx8mp.c | 16 ++++++++--------
1 file changed, 8 insertions(+), 8 deletions(-)
diff --git a/drivers/usb/dwc3/dwc3-imx8mp.c b/drivers/usb/dwc3/dwc3-imx8mp.c
index d328d20abfbc..1c8fe657b3a9 100644
--- a/drivers/usb/dwc3/dwc3-imx8mp.c
+++ b/drivers/usb/dwc3/dwc3-imx8mp.c
@@ -39,7 +39,7 @@
struct dwc3_imx8mp {
struct device *dev;
struct platform_device *dwc3;
- void __iomem *glue_base;
+ void __iomem *hsio_blk_base;
struct clk *hsio_clk;
struct clk *suspend_clk;
int irq;
@@ -55,7 +55,7 @@ static void dwc3_imx8mp_wakeup_enable(struct dwc3_imx8mp *dwc3_imx)
if (!dwc3)
return;
- val = readl(dwc3_imx->glue_base + USB_WAKEUP_CTRL);
+ val = readl(dwc3_imx->hsio_blk_base + USB_WAKEUP_CTRL);
if ((dwc3->current_dr_role == DWC3_GCTL_PRTCAP_HOST) && dwc3->xhci)
val |= USB_WAKEUP_EN | USB_WAKEUP_SS_CONN |
@@ -64,16 +64,16 @@ static void dwc3_imx8mp_wakeup_enable(struct dwc3_imx8mp *dwc3_imx)
val |= USB_WAKEUP_EN | USB_WAKEUP_VBUS_EN |
USB_WAKEUP_VBUS_SRC_SESS_VAL;
- writel(val, dwc3_imx->glue_base + USB_WAKEUP_CTRL);
+ writel(val, dwc3_imx->hsio_blk_base + USB_WAKEUP_CTRL);
}
static void dwc3_imx8mp_wakeup_disable(struct dwc3_imx8mp *dwc3_imx)
{
u32 val;
- val = readl(dwc3_imx->glue_base + USB_WAKEUP_CTRL);
+ val = readl(dwc3_imx->hsio_blk_base + USB_WAKEUP_CTRL);
val &= ~(USB_WAKEUP_EN | USB_WAKEUP_EN_MASK);
- writel(val, dwc3_imx->glue_base + USB_WAKEUP_CTRL);
+ writel(val, dwc3_imx->hsio_blk_base + USB_WAKEUP_CTRL);
}
static irqreturn_t dwc3_imx8mp_interrupt(int irq, void *_dwc3_imx)
@@ -115,9 +115,9 @@ static int dwc3_imx8mp_probe(struct platform_device *pdev)
dwc3_imx->dev = dev;
- dwc3_imx->glue_base = devm_platform_ioremap_resource(pdev, 0);
- if (IS_ERR(dwc3_imx->glue_base))
- return PTR_ERR(dwc3_imx->glue_base);
+ dwc3_imx->hsio_blk_base = devm_platform_ioremap_resource(pdev, 0);
+ if (IS_ERR(dwc3_imx->hsio_blk_base))
+ return PTR_ERR(dwc3_imx->hsio_blk_base);
dwc3_imx->hsio_clk = devm_clk_get(dev, "hsio");
if (IS_ERR(dwc3_imx->hsio_clk)) {
--
2.25.1
^ permalink raw reply related
* [PATCH v4 0/4] i.MX8MP: more USB3 glue layer feature support
From: Alexander Stein @ 2022-01-26 14:13 UTC (permalink / raw)
To: Greg Kroah-Hartman, Rob Herring, Shawn Guo, Sascha Hauer,
Fabio Estevam
Cc: Alexander Stein, NXP Linux Team, linux-usb, devicetree,
linux-arm-kernel, Li Jun
Hi all,
Thanks for the feedback on v3 [1].
This patchset aims to support flags for e.g. over-current active low or port
permanantly attached which are provided in the USB3 glue layer.
In v4 the patchset depends on Lucas' i.MX8MP power-domains patchset [2].
Because of that the 'phy' clock usage can be dropped, as now power domains
take care of the clocks. Due to powering up/down the glue layer settings
will have to be reapplied during resume.
Changes in v4:
* Depends on Lucas' i.MX8MP power-domains patchset
* removed 'phy' clock usage from dwc3-imx8mp.c
* Reapply glue layer settings upon resume
Changes in v3:
* Rename existing member for clarity
* Moved feature implementation from phy-fsl-imx8mq-usb.c to dwc3-imx8mp.c
[1] https://patchwork.kernel.org/project/linux-arm-kernel/cover/20220118131626.926394-1-alexander.stein@ew.tq-group.com/
[2] https://patchwork.kernel.org/project/linux-arm-kernel/cover/20220119134027.2931945-1-l.stach@pengutronix.de/
Alexander Stein (4):
usb: dwc3: imx8mp: rename iomem base pointer
dt-bindings: usb: dwc3-imx8mp: Add imx8mp specific flags
usb: dwc3: imx8mp: Add support for setting SOC specific flags
arm64: dts: imx8mp: Add memory for USB3 glue layer to usb3 nodes
.../bindings/usb/fsl,imx8mp-dwc3.yaml | 31 +++++++-
arch/arm64/boot/dts/freescale/imx8mp.dtsi | 6 +-
drivers/usb/dwc3/dwc3-imx8mp.c | 76 +++++++++++++++++--
3 files changed, 100 insertions(+), 13 deletions(-)
--
2.25.1
^ permalink raw reply
* Re: (EXT) [PATCH 7/9] arm64: dts: imx8mp: add HSIO power-domains
From: Alexander Stein @ 2022-01-26 14:02 UTC (permalink / raw)
To: Lucas Stach
Cc: Shawn Guo, Rob Herring, linux-arm-kernel, Pengutronix Kernel Team,
NXP Linux Team, Fabio Estevam, devicetree, linux-arm-kernel,
patchwork-lst
In-Reply-To: <20220119134027.2931945-8-l.stach@pengutronix.de>
Hi Lucas,
Am Mittwoch, 19. Januar 2022, 14:40:25 CET schrieb Lucas Stach:
> This adds the GPC and HSIO blk-ctrl nodes providing power control for
> the high-speed (USB and PCIe) IOs.
>
> Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
$ make dtbs_check DT_SCHEMA_FILES=Documentation/devicetree/bindings/usb/
fsl,imx8mp-dwc3.yaml gives the warnings:
> usb@32f10100: 'power-domains' does not match any of the regexes:
'^usb@[0-9a-f]+$', 'pinctrl-[0-9]+'
> From schema: linux/Documentation/devicetree/bindings/usb/fsl,imx8mp-
dwc3.yaml
> usb@32f10108: 'power-domains' does not match any of the regexes:
'^usb@[0-9a-f]+$', 'pinctrl-[0-9]+'
> From schema: linux/Documentation/devicetree/bindings/usb/fsl,imx8mp-
dwc3.yaml
Alexander
> ---
> arch/arm64/boot/dts/freescale/imx8mp.dtsi | 63 ++++++++++++++++++++---
> 1 file changed, 57 insertions(+), 6 deletions(-)
>
> diff --git a/arch/arm64/boot/dts/freescale/imx8mp.dtsi
> b/arch/arm64/boot/dts/freescale/imx8mp.dtsi index
> 04d259de5667..b76af96b9b5c 100644
> --- a/arch/arm64/boot/dts/freescale/imx8mp.dtsi
> +++ b/arch/arm64/boot/dts/freescale/imx8mp.dtsi
> @@ -4,6 +4,7 @@
> */
>
> #include <dt-bindings/clock/imx8mp-clock.h>
> +#include <dt-bindings/power/imx8mp-power.h>
> #include <dt-bindings/gpio/gpio.h>
> #include <dt-bindings/input/input.h>
> #include <dt-bindings/interrupt-controller/arm-gic.h>
> @@ -443,6 +444,44 @@ src: reset-controller@30390000 {
> interrupts = <GIC_SPI 89
IRQ_TYPE_LEVEL_HIGH>;
> #reset-cells = <1>;
> };
> +
> + gpc: gpc@303a0000 {
> + compatible = "fsl,imx8mp-gpc";
> + reg = <0x303a0000 0x10000>;
> + interrupt-parent = <&gic>;
> + interrupt-controller;
> + #interrupt-cells = <3>;
> +
> + pgc {
> + #address-cells = <1>;
> + #size-cells = <0>;
> +
> + pgc_pcie_phy: power-
domain@1 {
> + #power-
domain-cells = <0>;
> + reg =
<IMX8MP_POWER_DOMAIN_PCIE_PHY>;
> + };
> +
> + pgc_usb1_phy: power-
domain@2 {
> + #power-
domain-cells = <0>;
> + reg =
<IMX8MP_POWER_DOMAIN_USB1_PHY>;
> + };
> +
> + pgc_usb2_phy: power-
domain@3 {
> + #power-
domain-cells = <0>;
> + reg =
<IMX8MP_POWER_DOMAIN_USB2_PHY>;
> + };
> +
> + pgc_hsiomix: power-
domains@17 {
> + #power-
domain-cells = <0>;
> + reg =
<IMX8MP_POWER_DOMAIN_HSIOMIX>;
> + clocks =
<&clk IMX8MP_CLK_HSIO_AXI>,
> +
<&clk IMX8MP_CLK_HSIO_ROOT>;
> + assigned-
clocks = <&clk IMX8MP_CLK_HSIO_AXI>;
> + assigned-
clock-parents = <&clk IMX8MP_SYS_PLL2_500M>;
> + assigned-
clock-rates = <500000000>;
> + };
> + };
> + };
> };
>
> aips2: bus@30400000 {
> @@ -875,6 +914,20 @@ ddr-pmu@3d800000 {
> interrupts = <GIC_SPI 98
IRQ_TYPE_LEVEL_HIGH>;
> };
>
> + hsio_blk_ctrl: blk-ctrl@32f10000 {
> + compatible = "fsl,imx8mp-hsio-blk-ctrl",
"syscon";
> + reg = <0x32f10000 0x24>;
> + clocks = <&clk IMX8MP_CLK_USB_ROOT>,
> + <&clk IMX8MP_CLK_PCIE_ROOT>;
> + clock-names = "usb", "pcie";
> + power-domains = <&pgc_hsiomix>,
<&pgc_hsiomix>,
> + <&pgc_usb1_phy>,
<&pgc_usb2_phy>,
> + <&pgc_hsiomix>,
<&pgc_pcie_phy>;
> + power-domain-names = "bus", "usb", "usb-
phy1",
> + "usb-phy2",
"pcie", "pcie-phy";
> + #power-domain-cells = <1>;
> + };
> +
> usb3_phy0: usb-phy@381f0040 {
> compatible = "fsl,imx8mp-usb-phy";
> reg = <0x381f0040 0x40>;
> @@ -882,6 +935,7 @@ usb3_phy0: usb-phy@381f0040 {
> clock-names = "phy";
> assigned-clocks = <&clk
IMX8MP_CLK_USB_PHY_REF>;
> assigned-clock-parents = <&clk
IMX8MP_CLK_24M>;
> + power-domains = <&hsio_blk_ctrl
IMX8MP_HSIOBLK_PD_USB_PHY1>;
> #phy-cells = <0>;
> status = "disabled";
> };
> @@ -893,6 +947,7 @@ usb3_0: usb@32f10100 {
> <&clk IMX8MP_CLK_USB_ROOT>;
> clock-names = "hsio", "suspend";
> interrupts = <GIC_SPI 148
IRQ_TYPE_LEVEL_HIGH>;
> + power-domains = <&hsio_blk_ctrl
IMX8MP_HSIOBLK_PD_USB>;
> #address-cells = <1>;
> #size-cells = <1>;
> dma-ranges = <0x40000000 0x40000000
0xc0000000>;
> @@ -906,9 +961,6 @@ usb_dwc3_0: usb@38100000 {
> <&clk
IMX8MP_CLK_USB_CORE_REF>,
> <&clk
IMX8MP_CLK_USB_ROOT>;
> clock-names = "bus_early", "ref",
"suspend";
> - assigned-clocks = <&clk
IMX8MP_CLK_HSIO_AXI>;
> - assigned-clock-parents = <&clk
IMX8MP_SYS_PLL2_500M>;
> - assigned-clock-rates =
<500000000>;
> interrupts = <GIC_SPI 40
IRQ_TYPE_LEVEL_HIGH>;
> phys = <&usb3_phy0>, <&usb3_phy0>;
> phy-names = "usb2-phy", "usb3-
phy";
> @@ -924,6 +976,7 @@ usb3_phy1: usb-phy@382f0040 {
> clock-names = "phy";
> assigned-clocks = <&clk
IMX8MP_CLK_USB_PHY_REF>;
> assigned-clock-parents = <&clk
IMX8MP_CLK_24M>;
> + power-domains = <&hsio_blk_ctrl
IMX8MP_HSIOBLK_PD_USB_PHY2>;
> #phy-cells = <0>;
> };
>
> @@ -934,6 +987,7 @@ usb3_1: usb@32f10108 {
> <&clk IMX8MP_CLK_USB_ROOT>;
> clock-names = "hsio", "suspend";
> interrupts = <GIC_SPI 149
IRQ_TYPE_LEVEL_HIGH>;
> + power-domains = <&hsio_blk_ctrl
IMX8MP_HSIOBLK_PD_USB>;
> #address-cells = <1>;
> #size-cells = <1>;
> dma-ranges = <0x40000000 0x40000000
0xc0000000>;
> @@ -947,9 +1001,6 @@ usb_dwc3_1: usb@38200000 {
> <&clk
IMX8MP_CLK_USB_CORE_REF>,
> <&clk
IMX8MP_CLK_USB_ROOT>;
> clock-names = "bus_early", "ref",
"suspend";
> - assigned-clocks = <&clk
IMX8MP_CLK_HSIO_AXI>;
> - assigned-clock-parents = <&clk
IMX8MP_SYS_PLL2_500M>;
> - assigned-clock-rates =
<500000000>;
> interrupts = <GIC_SPI 41
IRQ_TYPE_LEVEL_HIGH>;
> phys = <&usb3_phy1>, <&usb3_phy1>;
> phy-names = "usb2-phy", "usb3-
phy";
^ permalink raw reply
* [PATCH v3 11/12] misc: fastrpc: Add dma handle implementation
From: Srinivas Kandagatla @ 2022-01-26 13:53 UTC (permalink / raw)
To: robh+dt, gregkh
Cc: devicetree, ekangupt, bkumar, linux-kernel, srini,
bjorn.andersson, linux-arm-msm, Vamsi Krishna Gattupalli,
Srinivas Kandagatla
In-Reply-To: <20220126135304.16340-1-srinivas.kandagatla@linaro.org>
From: Vamsi Krishna Gattupalli <quic_vgattupa@quicinc.com>
The remote arguments carry both remote buffers and dma handles. Add proper
dma handle instructions to make it compatible with DSP implementation.
Signed-off-by: Vamsi Krishna Gattupalli <quic_vgattupa@quicinc.com>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
---
drivers/misc/fastrpc.c | 52 +++++++++++++++++++++++++++---------------
1 file changed, 33 insertions(+), 19 deletions(-)
diff --git a/drivers/misc/fastrpc.c b/drivers/misc/fastrpc.c
index 6052a9cb9e2c..56ec7170b698 100644
--- a/drivers/misc/fastrpc.c
+++ b/drivers/misc/fastrpc.c
@@ -100,9 +100,20 @@ struct fastrpc_invoke_buf {
u32 pgidx; /* index to start of contiguous region */
};
-struct fastrpc_remote_arg {
- u64 pv;
- u64 len;
+struct fastrpc_remote_dmahandle {
+ s32 fd; /* dma handle fd */
+ u32 offset; /* dma handle offset */
+ u32 len; /* dma handle length */
+};
+
+struct fastrpc_remote_buf {
+ u64 pv; /* buffer pointer */
+ u64 len; /* length of buffer */
+};
+
+union fastrpc_remote_arg {
+ struct fastrpc_remote_buf buf;
+ struct fastrpc_remote_dmahandle dma;
};
struct fastrpc_mmap_rsp_msg {
@@ -216,7 +227,7 @@ struct fastrpc_invoke_ctx {
struct work_struct put_work;
struct fastrpc_msg msg;
struct fastrpc_user *fl;
- struct fastrpc_remote_arg *rpra;
+ union fastrpc_remote_arg *rpra;
struct fastrpc_map **maps;
struct fastrpc_buf *buf;
struct fastrpc_invoke_args *args;
@@ -766,7 +777,7 @@ static int fastrpc_map_create(struct fastrpc_user *fl, int fd,
* >>>>>> START of METADATA <<<<<<<<<
* +---------------------------------+
* | Arguments |
- * | type:(struct fastrpc_remote_arg)|
+ * | type:(union fastrpc_remote_arg)|
* | (0 - N) |
* +---------------------------------+
* | Invoke Buffer list |
@@ -791,7 +802,7 @@ static int fastrpc_get_meta_size(struct fastrpc_invoke_ctx *ctx)
{
int size = 0;
- size = (sizeof(struct fastrpc_remote_arg) +
+ size = (sizeof(struct fastrpc_remote_buf) +
sizeof(struct fastrpc_invoke_buf) +
sizeof(struct fastrpc_phy_page)) * ctx->nscalars +
sizeof(u64) * FASTRPC_MAX_FDLIST +
@@ -856,7 +867,7 @@ static struct fastrpc_phy_page *fastrpc_phy_page_start(struct fastrpc_invoke_buf
static int fastrpc_get_args(u32 kernel, struct fastrpc_invoke_ctx *ctx)
{
struct device *dev = ctx->fl->sctx->dev;
- struct fastrpc_remote_arg *rpra;
+ union fastrpc_remote_arg *rpra;
struct fastrpc_invoke_buf *list;
struct fastrpc_phy_page *pages;
int inbufs, i, oix, err = 0;
@@ -892,8 +903,8 @@ static int fastrpc_get_args(u32 kernel, struct fastrpc_invoke_ctx *ctx)
i = ctx->olaps[oix].raix;
len = ctx->args[i].length;
- rpra[i].pv = 0;
- rpra[i].len = len;
+ rpra[i].buf.pv = 0;
+ rpra[i].buf.len = len;
list[i].num = len ? 1 : 0;
list[i].pgidx = i;
@@ -903,7 +914,7 @@ static int fastrpc_get_args(u32 kernel, struct fastrpc_invoke_ctx *ctx)
if (ctx->maps[i]) {
struct vm_area_struct *vma = NULL;
- rpra[i].pv = (u64) ctx->args[i].ptr;
+ rpra[i].buf.pv = (u64) ctx->args[i].ptr;
pages[i].addr = ctx->maps[i]->phys;
mmap_read_lock(current->mm);
@@ -930,7 +941,7 @@ static int fastrpc_get_args(u32 kernel, struct fastrpc_invoke_ctx *ctx)
if (rlen < mlen)
goto bail;
- rpra[i].pv = args - ctx->olaps[oix].offset;
+ rpra[i].buf.pv = args - ctx->olaps[oix].offset;
pages[i].addr = ctx->buf->phys -
ctx->olaps[oix].offset +
(pkt_size - rlen);
@@ -944,7 +955,7 @@ static int fastrpc_get_args(u32 kernel, struct fastrpc_invoke_ctx *ctx)
}
if (i < inbufs && !ctx->maps[i]) {
- void *dst = (void *)(uintptr_t)rpra[i].pv;
+ void *dst = (void *)(uintptr_t)rpra[i].buf.pv;
void *src = (void *)(uintptr_t)ctx->args[i].ptr;
if (!kernel) {
@@ -960,12 +971,15 @@ static int fastrpc_get_args(u32 kernel, struct fastrpc_invoke_ctx *ctx)
}
for (i = ctx->nbufs; i < ctx->nscalars; ++i) {
- rpra[i].pv = (u64) ctx->args[i].ptr;
- rpra[i].len = ctx->args[i].length;
list[i].num = ctx->args[i].length ? 1 : 0;
list[i].pgidx = i;
- pages[i].addr = ctx->maps[i]->phys;
- pages[i].size = ctx->maps[i]->size;
+ if (ctx->maps[i]) {
+ pages[i].addr = ctx->maps[i]->phys;
+ pages[i].size = ctx->maps[i]->size;
+ }
+ rpra[i].dma.fd = ctx->args[i].fd;
+ rpra[i].dma.len = ctx->args[i].length;
+ rpra[i].dma.offset = (u64) ctx->args[i].ptr;
}
bail:
@@ -978,7 +992,7 @@ static int fastrpc_get_args(u32 kernel, struct fastrpc_invoke_ctx *ctx)
static int fastrpc_put_args(struct fastrpc_invoke_ctx *ctx,
u32 kernel)
{
- struct fastrpc_remote_arg *rpra = ctx->rpra;
+ union fastrpc_remote_arg *rpra = ctx->rpra;
struct fastrpc_user *fl = ctx->fl;
struct fastrpc_map *mmap = NULL;
struct fastrpc_invoke_buf *list;
@@ -995,9 +1009,9 @@ static int fastrpc_put_args(struct fastrpc_invoke_ctx *ctx,
for (i = inbufs; i < ctx->nbufs; ++i) {
if (!ctx->maps[i]) {
- void *src = (void *)(uintptr_t)rpra[i].pv;
+ void *src = (void *)(uintptr_t)rpra[i].buf.pv;
void *dst = (void *)(uintptr_t)ctx->args[i].ptr;
- u64 len = rpra[i].len;
+ u64 len = rpra[i].buf.len;
if (!kernel) {
if (copy_to_user((void __user *)dst, src, len))
--
2.21.0
^ permalink raw reply related
* [PATCH v3 12/12] arm64: dts: qcom: add non-secure domain property to fastrpc nodes
From: Srinivas Kandagatla @ 2022-01-26 13:53 UTC (permalink / raw)
To: robh+dt, gregkh
Cc: devicetree, ekangupt, bkumar, linux-kernel, srini,
bjorn.andersson, linux-arm-msm, Jeya R, Srinivas Kandagatla
In-Reply-To: <20220126135304.16340-1-srinivas.kandagatla@linaro.org>
From: Jeya R <jeyr@codeaurora.org>
FastRPC DSP domain would be set as secure if non-secure dsp property is not
added to the fastrpc DT node. Add this property to DT files of msm8916,
sdm845, sm8150, sm8250 and sm8350 so that nothing is broken after secure
domain patchset.
This patch is purely for backward compatibility reasons.
Signed-off-by: Jeya R <jeyr@codeaurora.org>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
---
arch/arm64/boot/dts/qcom/msm8916.dtsi | 1 +
arch/arm64/boot/dts/qcom/sdm845.dtsi | 2 ++
arch/arm64/boot/dts/qcom/sm8150.dtsi | 3 +++
arch/arm64/boot/dts/qcom/sm8250.dtsi | 3 +++
arch/arm64/boot/dts/qcom/sm8350.dtsi | 3 +++
5 files changed, 12 insertions(+)
diff --git a/arch/arm64/boot/dts/qcom/msm8916.dtsi b/arch/arm64/boot/dts/qcom/msm8916.dtsi
index 41897eb3736a..a1543012c4fa 100644
--- a/arch/arm64/boot/dts/qcom/msm8916.dtsi
+++ b/arch/arm64/boot/dts/qcom/msm8916.dtsi
@@ -1370,6 +1370,7 @@
compatible = "qcom,fastrpc";
qcom,smd-channels = "fastrpcsmd-apps-dsp";
label = "adsp";
+ qcom,non-secure-domain;
#address-cells = <1>;
#size-cells = <0>;
diff --git a/arch/arm64/boot/dts/qcom/sdm845.dtsi b/arch/arm64/boot/dts/qcom/sdm845.dtsi
index cfdeaa81f1bb..c9d613063966 100644
--- a/arch/arm64/boot/dts/qcom/sdm845.dtsi
+++ b/arch/arm64/boot/dts/qcom/sdm845.dtsi
@@ -838,6 +838,7 @@
compatible = "qcom,fastrpc";
qcom,glink-channels = "fastrpcglink-apps-dsp";
label = "adsp";
+ qcom,non-secure-domain;
#address-cells = <1>;
#size-cells = <0>;
@@ -888,6 +889,7 @@
compatible = "qcom,fastrpc";
qcom,glink-channels = "fastrpcglink-apps-dsp";
label = "cdsp";
+ qcom,non-secure-domain;
#address-cells = <1>;
#size-cells = <0>;
diff --git a/arch/arm64/boot/dts/qcom/sm8150.dtsi b/arch/arm64/boot/dts/qcom/sm8150.dtsi
index 9255982adb69..637c6a6d4054 100644
--- a/arch/arm64/boot/dts/qcom/sm8150.dtsi
+++ b/arch/arm64/boot/dts/qcom/sm8150.dtsi
@@ -1755,6 +1755,7 @@
compatible = "qcom,fastrpc";
qcom,glink-channels = "fastrpcglink-apps-dsp";
label = "sdsp";
+ qcom,non-secure-domain;
#address-cells = <1>;
#size-cells = <0>;
@@ -2997,6 +2998,7 @@
compatible = "qcom,fastrpc";
qcom,glink-channels = "fastrpcglink-apps-dsp";
label = "cdsp";
+ qcom,non-secure-domain;
#address-cells = <1>;
#size-cells = <0>;
@@ -3442,6 +3444,7 @@
compatible = "qcom,fastrpc";
qcom,glink-channels = "fastrpcglink-apps-dsp";
label = "adsp";
+ qcom,non-secure-domain;
#address-cells = <1>;
#size-cells = <0>;
diff --git a/arch/arm64/boot/dts/qcom/sm8250.dtsi b/arch/arm64/boot/dts/qcom/sm8250.dtsi
index c97ea638f6aa..3be4e630c2fe 100644
--- a/arch/arm64/boot/dts/qcom/sm8250.dtsi
+++ b/arch/arm64/boot/dts/qcom/sm8250.dtsi
@@ -2594,6 +2594,7 @@
compatible = "qcom,fastrpc";
qcom,glink-channels = "fastrpcglink-apps-dsp";
label = "sdsp";
+ qcom,non-secure-domain;
#address-cells = <1>;
#size-cells = <0>;
@@ -2659,6 +2660,7 @@
compatible = "qcom,fastrpc";
qcom,glink-channels = "fastrpcglink-apps-dsp";
label = "cdsp";
+ qcom,non-secure-domain;
#address-cells = <1>;
#size-cells = <0>;
@@ -4429,6 +4431,7 @@
compatible = "qcom,fastrpc";
qcom,glink-channels = "fastrpcglink-apps-dsp";
label = "adsp";
+ qcom,non-secure-domain;
#address-cells = <1>;
#size-cells = <0>;
diff --git a/arch/arm64/boot/dts/qcom/sm8350.dtsi b/arch/arm64/boot/dts/qcom/sm8350.dtsi
index 53b39e718fb6..a9a11c747a3a 100644
--- a/arch/arm64/boot/dts/qcom/sm8350.dtsi
+++ b/arch/arm64/boot/dts/qcom/sm8350.dtsi
@@ -1996,6 +1996,7 @@
compatible = "qcom,fastrpc";
qcom,glink-channels = "fastrpcglink-apps-dsp";
label = "sdsp";
+ qcom,non-secure-domain;
#address-cells = <1>;
#size-cells = <0>;
@@ -2065,6 +2066,7 @@
compatible = "qcom,fastrpc";
qcom,glink-channels = "fastrpcglink-apps-dsp";
label = "cdsp";
+ qcom,non-secure-domain;
#address-cells = <1>;
#size-cells = <0>;
@@ -2367,6 +2369,7 @@
compatible = "qcom,fastrpc";
qcom,glink-channels = "fastrpcglink-apps-dsp";
label = "adsp";
+ qcom,non-secure-domain;
#address-cells = <1>;
#size-cells = <0>;
--
2.21.0
^ permalink raw reply related
* [PATCH v3 10/12] misc: fastrpc: Add fdlist implementation
From: Srinivas Kandagatla @ 2022-01-26 13:53 UTC (permalink / raw)
To: robh+dt, gregkh
Cc: devicetree, ekangupt, bkumar, linux-kernel, srini,
bjorn.andersson, linux-arm-msm, Vamsi Krishna Gattupalli,
Srinivas Kandagatla
In-Reply-To: <20220126135304.16340-1-srinivas.kandagatla@linaro.org>
From: Vamsi Krishna Gattupalli <quic_vgattupa@quicinc.com>
Add fdlist implementation to support dma handles. fdlist is populated by
DSP if any map is no longer used and it is freed during put_args.
Signed-off-by: Vamsi Krishna Gattupalli <quic_vgattupa@quicinc.com>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
---
drivers/misc/fastrpc.c | 36 ++++++++++++++++++++++++++++++++----
1 file changed, 32 insertions(+), 4 deletions(-)
diff --git a/drivers/misc/fastrpc.c b/drivers/misc/fastrpc.c
index 1c1815bed2c2..6052a9cb9e2c 100644
--- a/drivers/misc/fastrpc.c
+++ b/drivers/misc/fastrpc.c
@@ -319,7 +319,8 @@ static void fastrpc_map_get(struct fastrpc_map *map)
kref_get(&map->refcount);
}
-static int fastrpc_map_find(struct fastrpc_user *fl, int fd,
+
+static int fastrpc_map_lookup(struct fastrpc_user *fl, int fd,
struct fastrpc_map **ppmap)
{
struct fastrpc_map *map = NULL;
@@ -327,7 +328,6 @@ static int fastrpc_map_find(struct fastrpc_user *fl, int fd,
mutex_lock(&fl->mutex);
list_for_each_entry(map, &fl->maps, node) {
if (map->fd == fd) {
- fastrpc_map_get(map);
*ppmap = map;
mutex_unlock(&fl->mutex);
return 0;
@@ -338,6 +338,17 @@ static int fastrpc_map_find(struct fastrpc_user *fl, int fd,
return -ENOENT;
}
+static int fastrpc_map_find(struct fastrpc_user *fl, int fd,
+ struct fastrpc_map **ppmap)
+{
+ int ret = fastrpc_map_lookup(fl, fd, ppmap);
+
+ if (!ret)
+ fastrpc_map_get(*ppmap);
+
+ return ret;
+}
+
static void fastrpc_buf_free(struct fastrpc_buf *buf)
{
dma_free_coherent(buf->dev, buf->size, buf->virt,
@@ -410,7 +421,7 @@ static void fastrpc_context_free(struct kref *ref)
ctx = container_of(ref, struct fastrpc_invoke_ctx, refcount);
cctx = ctx->cctx;
- for (i = 0; i < ctx->nscalars; i++)
+ for (i = 0; i < ctx->nbufs; i++)
fastrpc_map_put(ctx->maps[i]);
if (ctx->buf)
@@ -968,9 +979,19 @@ static int fastrpc_put_args(struct fastrpc_invoke_ctx *ctx,
u32 kernel)
{
struct fastrpc_remote_arg *rpra = ctx->rpra;
- int i, inbufs;
+ struct fastrpc_user *fl = ctx->fl;
+ struct fastrpc_map *mmap = NULL;
+ struct fastrpc_invoke_buf *list;
+ struct fastrpc_phy_page *pages;
+ u64 *fdlist;
+ int i, inbufs, outbufs, handles;
inbufs = REMOTE_SCALARS_INBUFS(ctx->sc);
+ outbufs = REMOTE_SCALARS_OUTBUFS(ctx->sc);
+ handles = REMOTE_SCALARS_INHANDLES(ctx->sc) + REMOTE_SCALARS_OUTHANDLES(ctx->sc);
+ list = fastrpc_invoke_buf_start(rpra, ctx->nscalars);
+ pages = fastrpc_phy_page_start(list, ctx->nscalars);
+ fdlist = (uint64_t *)(pages + inbufs + outbufs + handles);
for (i = inbufs; i < ctx->nbufs; ++i) {
if (!ctx->maps[i]) {
@@ -987,6 +1008,13 @@ static int fastrpc_put_args(struct fastrpc_invoke_ctx *ctx,
}
}
+ for (i = 0; i < FASTRPC_MAX_FDLIST; i++) {
+ if (!fdlist[i])
+ break;
+ if (!fastrpc_map_lookup(fl, (int)fdlist[i], &mmap))
+ fastrpc_map_put(mmap);
+ }
+
return 0;
}
--
2.21.0
^ permalink raw reply related
* [PATCH v3 08/12] misc: fastrpc: Add support to secure memory map
From: Srinivas Kandagatla @ 2022-01-26 13:53 UTC (permalink / raw)
To: robh+dt, gregkh
Cc: devicetree, ekangupt, bkumar, linux-kernel, srini,
bjorn.andersson, linux-arm-msm, Vamsi Krishna Gattupalli,
Srinivas Kandagatla
In-Reply-To: <20220126135304.16340-1-srinivas.kandagatla@linaro.org>
From: Vamsi Krishna Gattupalli <quic_vgattupa@quicinc.com>
This patch adds support to secure memory allocations for DSP.
It repurposes the reserved field in struct fastrpc_invoke_args
to add attributes to invoke request, for example to setup a secure memory
map for dsp. Secure memory is assigned to DSP Virtual Machine IDs using
Qualcomm SCM calls.
Signed-off-by: Vamsi Krishna Gattupalli <quic_vgattupa@quicinc.com>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
---
drivers/misc/fastrpc.c | 65 ++++++++++++++++++++++++++++++++-----
include/uapi/misc/fastrpc.h | 5 ++-
2 files changed, 60 insertions(+), 10 deletions(-)
diff --git a/drivers/misc/fastrpc.c b/drivers/misc/fastrpc.c
index 8e780e2d5d9d..0090085bfbb9 100644
--- a/drivers/misc/fastrpc.c
+++ b/drivers/misc/fastrpc.c
@@ -17,6 +17,7 @@
#include <linux/rpmsg.h>
#include <linux/scatterlist.h>
#include <linux/slab.h>
+#include <linux/qcom_scm.h>
#include <uapi/misc/fastrpc.h>
#define ADSP_DOMAIN_ID (0)
@@ -25,6 +26,7 @@
#define CDSP_DOMAIN_ID (3)
#define FASTRPC_DEV_MAX 4 /* adsp, mdsp, slpi, cdsp*/
#define FASTRPC_MAX_SESSIONS 13 /*12 compute, 1 cpz*/
+#define FASTRPC_MAX_VMIDS 16
#define FASTRPC_ALIGN 128
#define FASTRPC_MAX_FDLIST 16
#define FASTRPC_MAX_CRCLIST 64
@@ -194,6 +196,7 @@ struct fastrpc_map {
void *va;
u64 len;
u64 raddr;
+ u32 attr;
struct kref refcount;
};
@@ -231,6 +234,9 @@ struct fastrpc_session_ctx {
struct fastrpc_channel_ctx {
int domain_id;
int sesscount;
+ int vmcount;
+ u32 perms;
+ struct qcom_scm_vmperm vmperms[FASTRPC_MAX_VMIDS];
struct rpmsg_device *rpdev;
struct fastrpc_session_ctx session[FASTRPC_MAX_SESSIONS];
spinlock_t lock;
@@ -278,6 +284,20 @@ static void fastrpc_free_map(struct kref *ref)
map = container_of(ref, struct fastrpc_map, refcount);
if (map->table) {
+ if (map->attr & FASTRPC_ATTR_SECUREMAP) {
+ struct qcom_scm_vmperm perm;
+ int err = 0;
+
+ perm.vmid = QCOM_SCM_VMID_HLOS;
+ perm.perm = QCOM_SCM_PERM_RWX;
+ err = qcom_scm_assign_mem(map->phys, map->size,
+ &(map->fl->cctx->vmperms[0].vmid), &perm, 1);
+ if (err) {
+ dev_err(map->fl->sctx->dev, "Failed to assign memory phys 0x%llx size 0x%llx err %d",
+ map->phys, map->size, err);
+ return;
+ }
+ }
dma_buf_unmap_attachment(map->attach, map->table,
DMA_BIDIRECTIONAL);
dma_buf_detach(map->buf, map->attach);
@@ -654,7 +674,7 @@ static const struct dma_buf_ops fastrpc_dma_buf_ops = {
};
static int fastrpc_map_create(struct fastrpc_user *fl, int fd,
- u64 len, struct fastrpc_map **ppmap)
+ u64 len, u32 attr, struct fastrpc_map **ppmap)
{
struct fastrpc_session_ctx *sess = fl->sctx;
struct fastrpc_map *map = NULL;
@@ -696,6 +716,22 @@ static int fastrpc_map_create(struct fastrpc_user *fl, int fd,
map->len = len;
kref_init(&map->refcount);
+ if (attr & FASTRPC_ATTR_SECUREMAP) {
+ /*
+ * If subsystem VMIDs are defined in DTSI, then do
+ * hyp_assign from HLOS to those VM(s)
+ */
+ unsigned int perms = BIT(QCOM_SCM_VMID_HLOS);
+
+ map->attr = attr;
+ err = qcom_scm_assign_mem(map->phys, (u64)map->size, &perms,
+ fl->cctx->vmperms, fl->cctx->vmcount);
+ if (err) {
+ dev_err(sess->dev, "Failed to assign memory with phys 0x%llx size 0x%llx err %d",
+ map->phys, map->size, err);
+ goto map_err;
+ }
+ }
spin_lock(&fl->lock);
list_add_tail(&map->node, &fl->maps);
spin_unlock(&fl->lock);
@@ -780,16 +816,13 @@ static int fastrpc_create_maps(struct fastrpc_invoke_ctx *ctx)
int i, err;
for (i = 0; i < ctx->nscalars; ++i) {
- /* Make sure reserved field is set to 0 */
- if (ctx->args[i].reserved)
- return -EINVAL;
if (ctx->args[i].fd == 0 || ctx->args[i].fd == -1 ||
ctx->args[i].length == 0)
continue;
err = fastrpc_map_create(ctx->fl, ctx->args[i].fd,
- ctx->args[i].length, &ctx->maps[i]);
+ ctx->args[i].length, ctx->args[i].attr, &ctx->maps[i]);
if (err) {
dev_err(dev, "Error Creating map %d\n", err);
return -EINVAL;
@@ -1123,7 +1156,7 @@ static int fastrpc_init_create_process(struct fastrpc_user *fl,
fl->pd = USER_PD;
if (init.filelen && init.filefd) {
- err = fastrpc_map_create(fl, init.filefd, init.filelen, &map);
+ err = fastrpc_map_create(fl, init.filefd, init.filelen, 0, &map);
if (err)
goto err;
}
@@ -1232,7 +1265,6 @@ static int fastrpc_release_current_dsp_process(struct fastrpc_user *fl)
args[0].ptr = (u64)(uintptr_t) &tgid;
args[0].length = sizeof(tgid);
args[0].fd = -1;
- args[0].reserved = 0;
sc = FASTRPC_SCALARS(FASTRPC_RMID_INIT_RELEASE, 1, 0);
return fastrpc_internal_invoke(fl, true, FASTRPC_INIT_HANDLE,
@@ -1373,7 +1405,6 @@ static int fastrpc_init_attach(struct fastrpc_user *fl, int pd)
args[0].ptr = (u64)(uintptr_t) &tgid;
args[0].length = sizeof(tgid);
args[0].fd = -1;
- args[0].reserved = 0;
sc = FASTRPC_SCALARS(FASTRPC_RMID_INIT_ATTACH, 1, 0);
fl->pd = pd;
@@ -1943,9 +1974,10 @@ static int fastrpc_rpmsg_probe(struct rpmsg_device *rpdev)
{
struct device *rdev = &rpdev->dev;
struct fastrpc_channel_ctx *data;
- int i, err, domain_id = -1;
+ int i, err, domain_id = -1, vmcount;
const char *domain;
bool secure_dsp;
+ unsigned int vmids[FASTRPC_MAX_VMIDS];
err = of_property_read_string(rdev->of_node, "label", &domain);
if (err) {
@@ -1965,10 +1997,25 @@ static int fastrpc_rpmsg_probe(struct rpmsg_device *rpdev)
return -EINVAL;
}
+ vmcount = of_property_read_variable_u32_array(rdev->of_node,
+ "qcom,vmids", &vmids[0], 0, FASTRPC_MAX_VMIDS);
+ if (vmcount < 0)
+ vmcount = 0;
+ else if (!qcom_scm_is_available())
+ return -EPROBE_DEFER;
+
data = kzalloc(sizeof(*data), GFP_KERNEL);
if (!data)
return -ENOMEM;
+ if (vmcount) {
+ data->vmcount = vmcount;
+ data->perms = BIT(QCOM_SCM_VMID_HLOS);
+ for (i = 0; i < data->vmcount; i++) {
+ data->vmperms[i].vmid = vmids[i];
+ data->vmperms[i].perm = QCOM_SCM_PERM_RWX;
+ }
+ }
secure_dsp = !(of_property_read_bool(rdev->of_node, "qcom,non-secure-domain"));
data->secure = secure_dsp;
diff --git a/include/uapi/misc/fastrpc.h b/include/uapi/misc/fastrpc.h
index f39edac20305..5e29f2cfa42d 100644
--- a/include/uapi/misc/fastrpc.h
+++ b/include/uapi/misc/fastrpc.h
@@ -63,11 +63,14 @@ enum fastrpc_proc_attr {
FASTRPC_MODE_PRIVILEGED = (1 << 6),
};
+/* Fastrpc attribute for memory protection of buffers */
+#define FASTRPC_ATTR_SECUREMAP (1)
+
struct fastrpc_invoke_args {
__u64 ptr;
__u64 length;
__s32 fd;
- __u32 reserved;
+ __u32 attr;
};
struct fastrpc_invoke {
--
2.21.0
^ permalink raw reply related
* [PATCH v3 09/12] misc: fastrpc: Add helper function to get list and page
From: Srinivas Kandagatla @ 2022-01-26 13:53 UTC (permalink / raw)
To: robh+dt, gregkh
Cc: devicetree, ekangupt, bkumar, linux-kernel, srini,
bjorn.andersson, linux-arm-msm, Vamsi Krishna Gattupalli,
Srinivas Kandagatla
In-Reply-To: <20220126135304.16340-1-srinivas.kandagatla@linaro.org>
From: Vamsi Krishna Gattupalli <quic_vgattupa@quicinc.com>
Add helper functions to get invoke buffer and page start pointers.
Signed-off-by: Vamsi Krishna Gattupalli <quic_vgattupa@quicinc.com>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
---
drivers/misc/fastrpc.c | 15 ++++++++++++---
1 file changed, 12 insertions(+), 3 deletions(-)
diff --git a/drivers/misc/fastrpc.c b/drivers/misc/fastrpc.c
index 0090085bfbb9..1c1815bed2c2 100644
--- a/drivers/misc/fastrpc.c
+++ b/drivers/misc/fastrpc.c
@@ -832,6 +832,16 @@ static int fastrpc_create_maps(struct fastrpc_invoke_ctx *ctx)
return 0;
}
+static struct fastrpc_invoke_buf *fastrpc_invoke_buf_start(union fastrpc_remote_arg *pra, int len)
+{
+ return (struct fastrpc_invoke_buf *)(&pra[len]);
+}
+
+static struct fastrpc_phy_page *fastrpc_phy_page_start(struct fastrpc_invoke_buf *buf, int len)
+{
+ return (struct fastrpc_phy_page *)(&buf[len]);
+}
+
static int fastrpc_get_args(u32 kernel, struct fastrpc_invoke_ctx *ctx)
{
struct device *dev = ctx->fl->sctx->dev;
@@ -859,9 +869,8 @@ static int fastrpc_get_args(u32 kernel, struct fastrpc_invoke_ctx *ctx)
return err;
rpra = ctx->buf->virt;
- list = ctx->buf->virt + ctx->nscalars * sizeof(*rpra);
- pages = ctx->buf->virt + ctx->nscalars * (sizeof(*list) +
- sizeof(*rpra));
+ list = fastrpc_invoke_buf_start(rpra, ctx->nscalars);
+ pages = fastrpc_phy_page_start(list, ctx->nscalars);
args = (uintptr_t)ctx->buf->virt + metalen;
rlen = pkt_size - metalen;
ctx->rpra = rpra;
--
2.21.0
^ permalink raw reply related
* [PATCH v3 07/12] dt-bindings: misc: add fastrpc domain vmid property
From: Srinivas Kandagatla @ 2022-01-26 13:52 UTC (permalink / raw)
To: robh+dt, gregkh
Cc: devicetree, ekangupt, bkumar, linux-kernel, srini,
bjorn.andersson, linux-arm-msm, Vamsi Krishna Gattupalli,
Srinivas Kandagatla
In-Reply-To: <20220126135304.16340-1-srinivas.kandagatla@linaro.org>
From: Vamsi Krishna Gattupalli <quic_vgattupa@quicinc.com>
Add fastrpc domain virtual machine IDs property. This property is used to
setup memory protection for remote processor.
Signed-off-by: Vamsi Krishna Gattupalli <quic_vgattupa@quicinc.com>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
---
Documentation/devicetree/bindings/misc/qcom,fastrpc.txt | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/Documentation/devicetree/bindings/misc/qcom,fastrpc.txt b/Documentation/devicetree/bindings/misc/qcom,fastrpc.txt
index f9a01e2b4c96..5ec124b138a6 100644
--- a/Documentation/devicetree/bindings/misc/qcom,fastrpc.txt
+++ b/Documentation/devicetree/bindings/misc/qcom,fastrpc.txt
@@ -22,6 +22,11 @@ other tasks.
Value type: <boolean>
Definition: Property to specify that dsp domain is non-secure.
+- qcom,vmids:
+ Usage: optional
+ Value type: <u32 array>
+ Definition: Virtual machine IDs for remote processor.
+
- #address-cells
Usage: required
Value type: <u32>
--
2.21.0
^ permalink raw reply related
* [PATCH v3 06/12] misc: fastrpc: check before loading process to the DSP
From: Srinivas Kandagatla @ 2022-01-26 13:52 UTC (permalink / raw)
To: robh+dt, gregkh
Cc: devicetree, ekangupt, bkumar, linux-kernel, srini,
bjorn.andersson, linux-arm-msm, Jeya R, Srinivas Kandagatla
In-Reply-To: <20220126135304.16340-1-srinivas.kandagatla@linaro.org>
From: Jeya R <jeyr@codeaurora.org>
Reject session if DSP domain is secure, device node is non-secure and signed
PD is requested. Secure device node can access DSP without any restriction.
Unsigned PD offload is only allowed for the DSP domain that can support
unsigned offloading.
Signed-off-by: Jeya R <jeyr@codeaurora.org>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
---
drivers/misc/fastrpc.c | 33 +++++++++++++++++++++++++++++++++
include/uapi/misc/fastrpc.h | 17 +++++++++++++++++
2 files changed, 50 insertions(+)
diff --git a/drivers/misc/fastrpc.c b/drivers/misc/fastrpc.c
index 9eadbcf451ef..8e780e2d5d9d 100644
--- a/drivers/misc/fastrpc.c
+++ b/drivers/misc/fastrpc.c
@@ -243,6 +243,7 @@ struct fastrpc_channel_ctx {
struct fastrpc_device *secure_fdevice;
struct fastrpc_device *fdevice;
bool secure;
+ bool unsigned_support;
};
struct fastrpc_device {
@@ -263,6 +264,7 @@ struct fastrpc_user {
int tgid;
int pd;
+ bool is_secure_dev;
/* Lock for lists */
spinlock_t lock;
/* lock for allocations */
@@ -1051,6 +1053,24 @@ static int fastrpc_internal_invoke(struct fastrpc_user *fl, u32 kernel,
return err;
}
+static bool is_session_rejected(struct fastrpc_user *fl, bool unsigned_pd_request)
+{
+ /* Check if the device node is non-secure and channel is secure*/
+ if (!fl->is_secure_dev && fl->cctx->secure) {
+ /*
+ * Allow untrusted applications to offload only to Unsigned PD when
+ * channel is configured as secure and block untrusted apps on channel
+ * that does not support unsigned PD offload
+ */
+ if (!fl->cctx->unsigned_support || !unsigned_pd_request) {
+ dev_err(&fl->cctx->rpdev->dev, "Error: Untrusted application trying to offload to signed PD");
+ return true;
+ }
+ }
+
+ return false;
+}
+
static int fastrpc_init_create_process(struct fastrpc_user *fl,
char __user *argp)
{
@@ -1070,6 +1090,7 @@ static int fastrpc_init_create_process(struct fastrpc_user *fl,
u32 siglen;
} inbuf;
u32 sc;
+ bool unsigned_module = false;
args = kcalloc(FASTRPC_CREATE_PROCESS_NARGS, sizeof(*args), GFP_KERNEL);
if (!args)
@@ -1080,6 +1101,14 @@ static int fastrpc_init_create_process(struct fastrpc_user *fl,
goto err;
}
+ if (init.attrs & FASTRPC_MODE_UNSIGNED_MODULE)
+ unsigned_module = true;
+
+ if (is_session_rejected(fl, unsigned_module)) {
+ err = -ECONNREFUSED;
+ goto err;
+ }
+
if (init.filelen > INIT_FILELEN_MAX) {
err = -EINVAL;
goto err;
@@ -1279,6 +1308,7 @@ static int fastrpc_device_open(struct inode *inode, struct file *filp)
INIT_LIST_HEAD(&fl->user);
fl->tgid = current->tgid;
fl->cctx = cctx;
+ fl->is_secure_dev = fdevice->secure;
fl->sctx = fastrpc_session_alloc(cctx);
if (!fl->sctx) {
@@ -1947,11 +1977,14 @@ static int fastrpc_rpmsg_probe(struct rpmsg_device *rpdev)
case ADSP_DOMAIN_ID:
case MDSP_DOMAIN_ID:
case SDSP_DOMAIN_ID:
+ /* Unsigned PD offloading is only supported on CDSP*/
+ data->unsigned_support = false;
err = fastrpc_device_register(rdev, data, secure_dsp, domains[domain_id]);
if (err)
goto fdev_error;
break;
case CDSP_DOMAIN_ID:
+ data->unsigned_support = true;
/* Create both device nodes so that we can allow both Signed and Unsigned PD */
err = fastrpc_device_register(rdev, data, true, domains[domain_id]);
if (err)
diff --git a/include/uapi/misc/fastrpc.h b/include/uapi/misc/fastrpc.h
index 7cc9d342078a..f39edac20305 100644
--- a/include/uapi/misc/fastrpc.h
+++ b/include/uapi/misc/fastrpc.h
@@ -46,6 +46,23 @@ enum fastrpc_map_flags {
FASTRPC_MAP_MAX,
};
+enum fastrpc_proc_attr {
+ /* Macro for Debug attr */
+ FASTRPC_MODE_DEBUG = (1 << 0),
+ /* Macro for Ptrace */
+ FASTRPC_MODE_PTRACE = (1 << 1),
+ /* Macro for CRC Check */
+ FASTRPC_MODE_CRC = (1 << 2),
+ /* Macro for Unsigned PD */
+ FASTRPC_MODE_UNSIGNED_MODULE = (1 << 3),
+ /* Macro for Adaptive QoS */
+ FASTRPC_MODE_ADAPTIVE_QOS = (1 << 4),
+ /* Macro for System Process */
+ FASTRPC_MODE_SYSTEM_PROCESS = (1 << 5),
+ /* Macro for Prvileged Process */
+ FASTRPC_MODE_PRIVILEGED = (1 << 6),
+};
+
struct fastrpc_invoke_args {
__u64 ptr;
__u64 length;
--
2.21.0
^ permalink raw reply related
* [PATCH v3 05/12] misc: fastrpc: add secure domain support
From: Srinivas Kandagatla @ 2022-01-26 13:52 UTC (permalink / raw)
To: robh+dt, gregkh
Cc: devicetree, ekangupt, bkumar, linux-kernel, srini,
bjorn.andersson, linux-arm-msm, Srinivas Kandagatla
In-Reply-To: <20220126135304.16340-1-srinivas.kandagatla@linaro.org>
ADSP/MDSP/SDSP are by default secured, which means it can only be loaded
with a Signed process.
Where as CDSP can be either be secured/unsecured. non-secured Compute DSP
would allow users to load unsigned process and run hexagon instructions,
but blocking access to secured hardware within the DSP. Where as signed
process with secure CDSP would be allowed to access all the dsp resources.
This patch adds basic code to create device nodes as per device tree property.
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
---
drivers/misc/fastrpc.c | 57 ++++++++++++++++++++++++++++++++++--------
1 file changed, 47 insertions(+), 10 deletions(-)
diff --git a/drivers/misc/fastrpc.c b/drivers/misc/fastrpc.c
index d5fafbe3c709..9eadbcf451ef 100644
--- a/drivers/misc/fastrpc.c
+++ b/drivers/misc/fastrpc.c
@@ -240,12 +240,15 @@ struct fastrpc_channel_ctx {
/* Flag if dsp attributes are cached */
bool valid_attributes;
u32 dsp_attributes[FASTRPC_MAX_DSP_ATTRIBUTES];
+ struct fastrpc_device *secure_fdevice;
struct fastrpc_device *fdevice;
+ bool secure;
};
struct fastrpc_device {
struct fastrpc_channel_ctx *cctx;
struct miscdevice miscdev;
+ bool secure;
};
struct fastrpc_user {
@@ -1683,7 +1686,7 @@ static int fastrpc_req_mem_map(struct fastrpc_user *fl, char __user *argp)
return -EFAULT;
/* create SMMU mapping */
- err = fastrpc_map_create(fl, req.fd, req.length, &map);
+ err = fastrpc_map_create(fl, req.fd, req.length, 0, &map);
if (err) {
dev_err(dev, "failed to map buffer, fd = %d\n", req.fd);
return err;
@@ -1878,7 +1881,7 @@ static struct platform_driver fastrpc_cb_driver = {
};
static int fastrpc_device_register(struct device *dev, struct fastrpc_channel_ctx *cctx,
- const char *domain)
+ bool is_secured, const char *domain)
{
struct fastrpc_device *fdev;
int err;
@@ -1887,15 +1890,21 @@ static int fastrpc_device_register(struct device *dev, struct fastrpc_channel_ct
if (!fdev)
return -ENOMEM;
+ fdev->secure = is_secured;
fdev->cctx = cctx;
fdev->miscdev.minor = MISC_DYNAMIC_MINOR;
fdev->miscdev.fops = &fastrpc_fops;
- fdev->miscdev.name = devm_kasprintf(dev, GFP_KERNEL, "fastrpc-%s", domain);
+ fdev->miscdev.name = devm_kasprintf(dev, GFP_KERNEL, "fastrpc-%s%s",
+ domain, is_secured ? "-secure" : "");
err = misc_register(&fdev->miscdev);
- if (err)
+ if (err) {
kfree(fdev);
- else
- cctx->fdevice = fdev;
+ } else {
+ if (is_secured)
+ cctx->secure_fdevice = fdev;
+ else
+ cctx->fdevice = fdev;
+ }
return err;
}
@@ -1906,6 +1915,7 @@ static int fastrpc_rpmsg_probe(struct rpmsg_device *rpdev)
struct fastrpc_channel_ctx *data;
int i, err, domain_id = -1;
const char *domain;
+ bool secure_dsp;
err = of_property_read_string(rdev->of_node, "label", &domain);
if (err) {
@@ -1929,10 +1939,31 @@ static int fastrpc_rpmsg_probe(struct rpmsg_device *rpdev)
if (!data)
return -ENOMEM;
- err = fastrpc_device_register(rdev, data, domains[domain_id]);
- if (err) {
- kfree(data);
- return err;
+
+ secure_dsp = !(of_property_read_bool(rdev->of_node, "qcom,non-secure-domain"));
+ data->secure = secure_dsp;
+
+ switch (domain_id) {
+ case ADSP_DOMAIN_ID:
+ case MDSP_DOMAIN_ID:
+ case SDSP_DOMAIN_ID:
+ err = fastrpc_device_register(rdev, data, secure_dsp, domains[domain_id]);
+ if (err)
+ goto fdev_error;
+ break;
+ case CDSP_DOMAIN_ID:
+ /* Create both device nodes so that we can allow both Signed and Unsigned PD */
+ err = fastrpc_device_register(rdev, data, true, domains[domain_id]);
+ if (err)
+ goto fdev_error;
+
+ err = fastrpc_device_register(rdev, data, false, domains[domain_id]);
+ if (err)
+ goto fdev_error;
+ break;
+ default:
+ err = -EINVAL;
+ goto fdev_error;
}
kref_init(&data->refcount);
@@ -1946,6 +1977,9 @@ static int fastrpc_rpmsg_probe(struct rpmsg_device *rpdev)
data->rpdev = rpdev;
return of_platform_populate(rdev->of_node, NULL, NULL, rdev);
+fdev_error:
+ kfree(data);
+ return err;
}
static void fastrpc_notify_users(struct fastrpc_user *user)
@@ -1972,6 +2006,9 @@ static void fastrpc_rpmsg_remove(struct rpmsg_device *rpdev)
if (cctx->fdevice)
misc_deregister(&cctx->fdevice->miscdev);
+ if (cctx->secure_fdevice)
+ misc_deregister(&cctx->secure_fdevice->miscdev);
+
of_platform_depopulate(&rpdev->dev);
cctx->rpdev = NULL;
--
2.21.0
^ permalink raw reply related
* [PATCH v3 04/12] dt-bindings: misc: add property to support non-secure DSP
From: Srinivas Kandagatla @ 2022-01-26 13:52 UTC (permalink / raw)
To: robh+dt, gregkh
Cc: devicetree, ekangupt, bkumar, linux-kernel, srini,
bjorn.andersson, linux-arm-msm, Jeya R, Srinivas Kandagatla
In-Reply-To: <20220126135304.16340-1-srinivas.kandagatla@linaro.org>
From: Jeya R <jeyr@codeaurora.org>
Add property to set DSP domain as non-secure.
ADSP/MDSP/SDSP are by default secured, where as CDSP can be either be
secured/unsecured.
non-secured Compute DSP would allow users to load unsigned process
and run hexagon instructions, but limiting access to secured hardware
within the DSP.
Based on this flag device nodes for secured and unsecured are created.
Signed-off-by: Jeya R <jeyr@codeaurora.org>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
---
Documentation/devicetree/bindings/misc/qcom,fastrpc.txt | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/Documentation/devicetree/bindings/misc/qcom,fastrpc.txt b/Documentation/devicetree/bindings/misc/qcom,fastrpc.txt
index 2a1827ab50d2..f9a01e2b4c96 100644
--- a/Documentation/devicetree/bindings/misc/qcom,fastrpc.txt
+++ b/Documentation/devicetree/bindings/misc/qcom,fastrpc.txt
@@ -17,6 +17,11 @@ other tasks.
Definition: should specify the dsp domain name this fastrpc
corresponds to. must be one of this: "adsp", "mdsp", "sdsp", "cdsp"
+- qcom,non-secure-domain:
+ Usage: required
+ Value type: <boolean>
+ Definition: Property to specify that dsp domain is non-secure.
+
- #address-cells
Usage: required
Value type: <u32>
--
2.21.0
^ permalink raw reply related
* [PATCH v3 03/12] misc: fastrpc: Add support to get DSP capabilities
From: Srinivas Kandagatla @ 2022-01-26 13:52 UTC (permalink / raw)
To: robh+dt, gregkh
Cc: devicetree, ekangupt, bkumar, linux-kernel, srini,
bjorn.andersson, linux-arm-msm, Jeya R, Srinivas Kandagatla
In-Reply-To: <20220126135304.16340-1-srinivas.kandagatla@linaro.org>
From: Jeya R <jeyr@codeaurora.org>
Add support to get DSP capabilities. The capability information is cached
on driver.
Signed-off-by: Jeya R <jeyr@codeaurora.org>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
---
drivers/misc/fastrpc.c | 105 ++++++++++++++++++++++++++++++++++++
include/uapi/misc/fastrpc.h | 8 +++
2 files changed, 113 insertions(+)
diff --git a/drivers/misc/fastrpc.c b/drivers/misc/fastrpc.c
index a840b8dabf0e..d5fafbe3c709 100644
--- a/drivers/misc/fastrpc.c
+++ b/drivers/misc/fastrpc.c
@@ -31,10 +31,14 @@
#define FASTRPC_PHYS(p) ((p) & 0xffffffff)
#define FASTRPC_CTX_MAX (256)
#define FASTRPC_INIT_HANDLE 1
+#define FASTRPC_DSP_UTILITIES_HANDLE 2
#define FASTRPC_CTXID_MASK (0xFF0)
#define INIT_FILELEN_MAX (2 * 1024 * 1024)
#define FASTRPC_DEVICE_NAME "fastrpc"
#define ADSP_MMAP_ADD_PAGES 0x1000
+#define DSP_UNSUPPORTED_API (0x80000414)
+/* MAX NUMBER of DSP ATTRIBUTES SUPPORTED */
+#define FASTRPC_MAX_DSP_ATTRIBUTES (256)
/* Retrives number of input buffers from the scalars parameter */
#define REMOTE_SCALARS_INBUFS(sc) (((sc) >> 16) & 0x0ff)
@@ -233,6 +237,9 @@ struct fastrpc_channel_ctx {
struct idr ctx_idr;
struct list_head users;
struct kref refcount;
+ /* Flag if dsp attributes are cached */
+ bool valid_attributes;
+ u32 dsp_attributes[FASTRPC_MAX_DSP_ATTRIBUTES];
struct fastrpc_device *fdevice;
};
@@ -1371,6 +1378,101 @@ static int fastrpc_invoke(struct fastrpc_user *fl, char __user *argp)
return err;
}
+static int fastrpc_get_info_from_dsp(struct fastrpc_user *fl, uint32_t *dsp_attr_buf,
+ uint32_t dsp_attr_buf_len)
+{
+ struct fastrpc_invoke_args args[2] = { 0 };
+
+ /* Capability filled in userspace */
+ dsp_attr_buf[0] = 0;
+
+ args[0].ptr = (u64)(uintptr_t)&dsp_attr_buf_len;
+ args[0].length = sizeof(dsp_attr_buf_len);
+ args[0].fd = -1;
+ args[1].ptr = (u64)(uintptr_t)&dsp_attr_buf[1];
+ args[1].length = dsp_attr_buf_len * sizeof(uint32_t);
+ args[1].fd = -1;
+ fl->pd = 1;
+
+ return fastrpc_internal_invoke(fl, true, FASTRPC_DSP_UTILITIES_HANDLE,
+ FASTRPC_SCALARS(0, 1, 1), args);
+}
+
+static int fastrpc_get_info_from_kernel(struct fastrpc_ioctl_capability *cap,
+ struct fastrpc_user *fl)
+{
+ struct fastrpc_channel_ctx *cctx = fl->cctx;
+ uint32_t attribute_id = cap->attribute_id;
+ uint32_t dsp_attributes[FASTRPC_MAX_DSP_ATTRIBUTES];
+ unsigned long flags;
+ uint32_t domain = cap->domain;
+ int err;
+
+ spin_lock_irqsave(&cctx->lock, flags);
+ /* check if we already have queried dsp for attributes */
+ if (cctx->valid_attributes) {
+ spin_unlock_irqrestore(&cctx->lock, flags);
+ goto done;
+ }
+ spin_unlock_irqrestore(&cctx->lock, flags);
+
+ err = fastrpc_get_info_from_dsp(fl, &dsp_attributes[0], FASTRPC_MAX_DSP_ATTRIBUTES);
+ if (err == DSP_UNSUPPORTED_API) {
+ dev_info(&cctx->rpdev->dev,
+ "Warning: DSP capabilities not supported on domain: %d\n", domain);
+ return -EOPNOTSUPP;
+ } else if (err) {
+ dev_err(&cctx->rpdev->dev, "Error: dsp information is incorrect err: %d\n", err);
+ return err;
+ }
+
+ spin_lock_irqsave(&cctx->lock, flags);
+ memcpy(cctx->dsp_attributes, dsp_attributes, sizeof(u32) * FASTRPC_MAX_DSP_ATTRIBUTES);
+ cctx->valid_attributes = true;
+ spin_unlock_irqrestore(&cctx->lock, flags);
+done:
+ cap->capability = cctx->dsp_attributes[attribute_id];
+
+ return 0;
+}
+
+static int fastrpc_get_dsp_info(struct fastrpc_user *fl, char __user *argp)
+{
+ struct fastrpc_ioctl_capability cap = {0};
+ int err = 0;
+
+ if (copy_from_user(&cap, argp, sizeof(cap)))
+ return -EFAULT;
+
+ cap.capability = 0;
+ if (cap.domain >= FASTRPC_DEV_MAX) {
+ dev_err(&fl->cctx->rpdev->dev, "Error: Invalid domain id:%d, err:%d\n",
+ cap.domain, err);
+ return -ECHRNG;
+ }
+
+ /* Fastrpc Capablities does not support modem domain */
+ if (cap.domain == MDSP_DOMAIN_ID) {
+ dev_err(&fl->cctx->rpdev->dev, "Error: modem not supported %d\n", err);
+ return -ECHRNG;
+ }
+
+ if (cap.attribute_id >= FASTRPC_MAX_DSP_ATTRIBUTES) {
+ dev_err(&fl->cctx->rpdev->dev, "Error: invalid attribute: %d, err: %d\n",
+ cap.attribute_id, err);
+ return -EOVERFLOW;
+ }
+
+ err = fastrpc_get_info_from_kernel(&cap, fl);
+ if (err)
+ return err;
+
+ if (copy_to_user(argp, &cap.capability, sizeof(cap.capability)))
+ return -EFAULT;
+
+ return 0;
+}
+
static int fastrpc_req_munmap_impl(struct fastrpc_user *fl,
struct fastrpc_req_munmap *req)
{
@@ -1677,6 +1779,9 @@ static long fastrpc_device_ioctl(struct file *file, unsigned int cmd,
case FASTRPC_IOCTL_MEM_UNMAP:
err = fastrpc_req_mem_unmap(fl, argp);
break;
+ case FASTRPC_IOCTL_GET_DSP_INFO:
+ err = fastrpc_get_dsp_info(fl, argp);
+ break;
default:
err = -ENOTTY;
break;
diff --git a/include/uapi/misc/fastrpc.h b/include/uapi/misc/fastrpc.h
index d248eeb20e67..7cc9d342078a 100644
--- a/include/uapi/misc/fastrpc.h
+++ b/include/uapi/misc/fastrpc.h
@@ -15,6 +15,7 @@
#define FASTRPC_IOCTL_INIT_ATTACH_SNS _IO('R', 8)
#define FASTRPC_IOCTL_MEM_MAP _IOWR('R', 10, struct fastrpc_mem_map)
#define FASTRPC_IOCTL_MEM_UNMAP _IOWR('R', 11, struct fastrpc_mem_unmap)
+#define FASTRPC_IOCTL_GET_DSP_INFO _IOWR('R', 13, struct fastrpc_ioctl_capability)
/**
* enum fastrpc_map_flags - control flags for mapping memory on DSP user process
@@ -105,4 +106,11 @@ struct fastrpc_mem_unmap {
__s32 reserved[5];
};
+struct fastrpc_ioctl_capability {
+ __u32 domain;
+ __u32 attribute_id;
+ __u32 capability; /* dsp capability */
+ __u32 reserved[4];
+};
+
#endif /* __QCOM_FASTRPC_H__ */
--
2.21.0
^ permalink raw reply related
* [PATCH v3 02/12] misc: fastrpc: add support for FASTRPC_IOCTL_MEM_MAP/UNMAP
From: Srinivas Kandagatla @ 2022-01-26 13:52 UTC (permalink / raw)
To: robh+dt, gregkh
Cc: devicetree, ekangupt, bkumar, linux-kernel, srini,
bjorn.andersson, linux-arm-msm, Jeya R, Srinivas Kandagatla
In-Reply-To: <20220126135304.16340-1-srinivas.kandagatla@linaro.org>
From: Jeya R <jeyr@codeaurora.org>
Add support for IOCTL requests to map and unmap on DSP based on map
flags.
Signed-off-by: Jeya R <jeyr@codeaurora.org>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
---
drivers/misc/fastrpc.c | 155 ++++++++++++++++++++++++++++++++++++
include/uapi/misc/fastrpc.h | 51 ++++++++++++
2 files changed, 206 insertions(+)
diff --git a/drivers/misc/fastrpc.c b/drivers/misc/fastrpc.c
index 00cf0dbb6084..a840b8dabf0e 100644
--- a/drivers/misc/fastrpc.c
+++ b/drivers/misc/fastrpc.c
@@ -72,6 +72,8 @@
#define FASTRPC_RMID_INIT_CREATE 6
#define FASTRPC_RMID_INIT_CREATE_ATTR 7
#define FASTRPC_RMID_INIT_CREATE_STATIC 8
+#define FASTRPC_RMID_INIT_MEM_MAP 10
+#define FASTRPC_RMID_INIT_MEM_UNMAP 11
/* Protection Domain(PD) ids */
#define AUDIO_PD (0) /* also GUEST_OS PD? */
@@ -108,12 +110,29 @@ struct fastrpc_mmap_req_msg {
s32 num;
};
+struct fastrpc_mem_map_req_msg {
+ s32 pgid;
+ s32 fd;
+ s32 offset;
+ u32 flags;
+ u64 vaddrin;
+ s32 num;
+ s32 data_len;
+};
+
struct fastrpc_munmap_req_msg {
s32 pgid;
u64 vaddr;
u64 size;
};
+struct fastrpc_mem_unmap_req_msg {
+ s32 pgid;
+ s32 fd;
+ u64 vaddrin;
+ u64 len;
+};
+
struct fastrpc_msg {
int pid; /* process group id */
int tid; /* thread id */
@@ -170,6 +189,7 @@ struct fastrpc_map {
u64 size;
void *va;
u64 len;
+ u64 raddr;
struct kref refcount;
};
@@ -1493,6 +1513,135 @@ static int fastrpc_req_mmap(struct fastrpc_user *fl, char __user *argp)
return err;
}
+static int fastrpc_req_mem_unmap_impl(struct fastrpc_user *fl, struct fastrpc_mem_unmap *req)
+{
+ struct fastrpc_invoke_args args[1] = { [0] = { 0 } };
+ struct fastrpc_map *map = NULL, *m;
+ struct fastrpc_mem_unmap_req_msg req_msg = { 0 };
+ int err = 0;
+ u32 sc;
+ struct device *dev = fl->sctx->dev;
+
+ spin_lock(&fl->lock);
+ list_for_each_entry_safe(map, m, &fl->maps, node) {
+ if ((req->fd < 0 || map->fd == req->fd) && (map->raddr == req->vaddr))
+ break;
+ map = NULL;
+ }
+
+ spin_unlock(&fl->lock);
+
+ if (!map) {
+ dev_err(dev, "map not in list\n");
+ return -EINVAL;
+ }
+
+ req_msg.pgid = fl->tgid;
+ req_msg.len = map->len;
+ req_msg.vaddrin = map->raddr;
+ req_msg.fd = map->fd;
+
+ args[0].ptr = (u64) &req_msg;
+ args[0].length = sizeof(req_msg);
+
+ sc = FASTRPC_SCALARS(FASTRPC_RMID_INIT_MEM_UNMAP, 1, 0);
+ err = fastrpc_internal_invoke(fl, true, FASTRPC_INIT_HANDLE, sc,
+ &args[0]);
+ fastrpc_map_put(map);
+ if (err)
+ dev_err(dev, "unmmap\tpt fd = %d, 0x%09llx error\n", map->fd, map->raddr);
+
+ return err;
+}
+
+static int fastrpc_req_mem_unmap(struct fastrpc_user *fl, char __user *argp)
+{
+ struct fastrpc_mem_unmap req;
+
+ if (copy_from_user(&req, argp, sizeof(req)))
+ return -EFAULT;
+
+ return fastrpc_req_mem_unmap_impl(fl, &req);
+}
+
+static int fastrpc_req_mem_map(struct fastrpc_user *fl, char __user *argp)
+{
+ struct fastrpc_invoke_args args[4] = { [0 ... 3] = { 0 } };
+ struct fastrpc_mem_map_req_msg req_msg = { 0 };
+ struct fastrpc_mmap_rsp_msg rsp_msg = { 0 };
+ struct fastrpc_mem_unmap req_unmap = { 0 };
+ struct fastrpc_phy_page pages = { 0 };
+ struct fastrpc_mem_map req;
+ struct device *dev = fl->sctx->dev;
+ struct fastrpc_map *map = NULL;
+ int err;
+ u32 sc;
+
+ if (copy_from_user(&req, argp, sizeof(req)))
+ return -EFAULT;
+
+ /* create SMMU mapping */
+ err = fastrpc_map_create(fl, req.fd, req.length, &map);
+ if (err) {
+ dev_err(dev, "failed to map buffer, fd = %d\n", req.fd);
+ return err;
+ }
+
+ req_msg.pgid = fl->tgid;
+ req_msg.fd = req.fd;
+ req_msg.offset = req.offset;
+ req_msg.vaddrin = req.vaddrin;
+ map->va = (void *) req.vaddrin;
+ req_msg.flags = req.flags;
+ req_msg.num = sizeof(pages);
+ req_msg.data_len = 0;
+
+ args[0].ptr = (u64) &req_msg;
+ args[0].length = sizeof(req_msg);
+
+ pages.addr = map->phys;
+ pages.size = map->size;
+
+ args[1].ptr = (u64) &pages;
+ args[1].length = sizeof(pages);
+
+ args[2].ptr = (u64) &pages;
+ args[2].length = 0;
+
+ args[3].ptr = (u64) &rsp_msg;
+ args[3].length = sizeof(rsp_msg);
+
+ sc = FASTRPC_SCALARS(FASTRPC_RMID_INIT_MEM_MAP, 3, 1);
+ err = fastrpc_internal_invoke(fl, true, FASTRPC_INIT_HANDLE, sc, &args[0]);
+ if (err) {
+ dev_err(dev, "mem mmap error, fd %d, vaddr %llx, size %lld\n",
+ req.fd, req.vaddrin, map->size);
+ goto err_invoke;
+ }
+
+ /* update the buffer to be able to deallocate the memory on the DSP */
+ map->raddr = rsp_msg.vaddr;
+
+ /* let the client know the address to use */
+ req.vaddrout = rsp_msg.vaddr;
+
+ if (copy_to_user((void __user *)argp, &req, sizeof(req))) {
+ /* unmap the memory and release the buffer */
+ req_unmap.vaddr = (uintptr_t) rsp_msg.vaddr;
+ req_unmap.length = map->size;
+ fastrpc_req_mem_unmap_impl(fl, &req_unmap);
+ return -EFAULT;
+ }
+
+ return 0;
+
+err_invoke:
+ if (map)
+ fastrpc_map_put(map);
+
+ return err;
+}
+
static long fastrpc_device_ioctl(struct file *file, unsigned int cmd,
unsigned long arg)
{
@@ -1522,6 +1671,12 @@ static long fastrpc_device_ioctl(struct file *file, unsigned int cmd,
case FASTRPC_IOCTL_MUNMAP:
err = fastrpc_req_munmap(fl, argp);
break;
+ case FASTRPC_IOCTL_MEM_MAP:
+ err = fastrpc_req_mem_map(fl, argp);
+ break;
+ case FASTRPC_IOCTL_MEM_UNMAP:
+ err = fastrpc_req_mem_unmap(fl, argp);
+ break;
default:
err = -ENOTTY;
break;
diff --git a/include/uapi/misc/fastrpc.h b/include/uapi/misc/fastrpc.h
index 0a89f95463f6..d248eeb20e67 100644
--- a/include/uapi/misc/fastrpc.h
+++ b/include/uapi/misc/fastrpc.h
@@ -13,6 +13,37 @@
#define FASTRPC_IOCTL_MMAP _IOWR('R', 6, struct fastrpc_req_mmap)
#define FASTRPC_IOCTL_MUNMAP _IOWR('R', 7, struct fastrpc_req_munmap)
#define FASTRPC_IOCTL_INIT_ATTACH_SNS _IO('R', 8)
+#define FASTRPC_IOCTL_MEM_MAP _IOWR('R', 10, struct fastrpc_mem_map)
+#define FASTRPC_IOCTL_MEM_UNMAP _IOWR('R', 11, struct fastrpc_mem_unmap)
+
+/**
+ * enum fastrpc_map_flags - control flags for mapping memory on DSP user process
+ * @FASTRPC_MAP_STATIC: Map memory pages with RW- permission and CACHE WRITEBACK.
+ * The driver is responsible for cache maintenance when passed
+ * the buffer to FastRPC calls. Same virtual address will be
+ * assigned for subsequent FastRPC calls.
+ * @FASTRPC_MAP_RESERVED: Reserved
+ * @FASTRPC_MAP_FD: Map memory pages with RW- permission and CACHE WRITEBACK.
+ * Mapping tagged with a file descriptor. User is responsible for
+ * CPU and DSP cache maintenance for the buffer. Get virtual address
+ * of buffer on DSP using HAP_mmap_get() and HAP_mmap_put() APIs.
+ * @FASTRPC_MAP_FD_DELAYED: Mapping delayed until user call HAP_mmap() and HAP_munmap()
+ * functions on DSP. It is useful to map a buffer with cache modes
+ * other than default modes. User is responsible for CPU and DSP
+ * cache maintenance for the buffer.
+ * @FASTRPC_MAP_FD_NOMAP: This flag is used to skip CPU mapping,
+ * otherwise behaves similar to FASTRPC_MAP_FD_DELAYED flag.
+ * @FASTRPC_MAP_MAX: max count for flags
+ *
+ */
+enum fastrpc_map_flags {
+ FASTRPC_MAP_STATIC = 0,
+ FASTRPC_MAP_RESERVED,
+ FASTRPC_MAP_FD = 2,
+ FASTRPC_MAP_FD_DELAYED,
+ FASTRPC_MAP_FD_NOMAP = 16,
+ FASTRPC_MAP_MAX,
+};
struct fastrpc_invoke_args {
__u64 ptr;
@@ -49,9 +80,29 @@ struct fastrpc_req_mmap {
__u64 vaddrout; /* dsp virtual address */
};
+struct fastrpc_mem_map {
+ __s32 version;
+ __s32 fd; /* fd */
+ __s32 offset; /* buffer offset */
+ __u32 flags; /* flags defined in enum fastrpc_map_flags */
+ __u64 vaddrin; /* buffer virtual address */
+ __u64 length; /* buffer length */
+ __u64 vaddrout; /* [out] remote virtual address */
+ __s32 attrs; /* buffer attributes used for SMMU mapping */
+ __s32 reserved[4];
+};
+
struct fastrpc_req_munmap {
__u64 vaddrout; /* address to unmap */
__u64 size; /* size */
};
+struct fastrpc_mem_unmap {
+ __s32 vesion;
+ __s32 fd; /* fd */
+ __u64 vaddr; /* remote process (dsp) virtual address */
+ __u64 length; /* buffer size */
+ __s32 reserved[5];
+};
+
#endif /* __QCOM_FASTRPC_H__ */
--
2.21.0
^ permalink raw reply related
* [PATCH v3 01/12] misc: fastrpc: separate fastrpc device from channel context
From: Srinivas Kandagatla @ 2022-01-26 13:52 UTC (permalink / raw)
To: robh+dt, gregkh
Cc: devicetree, ekangupt, bkumar, linux-kernel, srini,
bjorn.andersson, linux-arm-msm, Srinivas Kandagatla
In-Reply-To: <20220126135304.16340-1-srinivas.kandagatla@linaro.org>
Currently fastrpc misc device instance is within channel context struct
with a kref. So we have 2 structs with refcount, both of them managing the
same channel context structure.
Separate fastrpc device from channel context and by adding a dedicated
fastrpc_device structure, this should clean the structures a bit and also help
when adding secure device node support.
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
---
drivers/misc/fastrpc.c | 48 ++++++++++++++++++++++++++++++++++--------
1 file changed, 39 insertions(+), 9 deletions(-)
diff --git a/drivers/misc/fastrpc.c b/drivers/misc/fastrpc.c
index 4ccbf43e6bfa..00cf0dbb6084 100644
--- a/drivers/misc/fastrpc.c
+++ b/drivers/misc/fastrpc.c
@@ -78,7 +78,7 @@
#define USER_PD (1)
#define SENSORS_PD (2)
-#define miscdev_to_cctx(d) container_of(d, struct fastrpc_channel_ctx, miscdev)
+#define miscdev_to_fdevice(d) container_of(d, struct fastrpc_device, miscdev)
static const char *domains[FASTRPC_DEV_MAX] = { "adsp", "mdsp",
"sdsp", "cdsp"};
@@ -212,8 +212,13 @@ struct fastrpc_channel_ctx {
spinlock_t lock;
struct idr ctx_idr;
struct list_head users;
- struct miscdevice miscdev;
struct kref refcount;
+ struct fastrpc_device *fdevice;
+};
+
+struct fastrpc_device {
+ struct fastrpc_channel_ctx *cctx;
+ struct miscdevice miscdev;
};
struct fastrpc_user {
@@ -1220,10 +1225,14 @@ static int fastrpc_device_release(struct inode *inode, struct file *file)
static int fastrpc_device_open(struct inode *inode, struct file *filp)
{
- struct fastrpc_channel_ctx *cctx = miscdev_to_cctx(filp->private_data);
+ struct fastrpc_channel_ctx *cctx = NULL;
+ struct fastrpc_device *fdevice = NULL;
struct fastrpc_user *fl = NULL;
unsigned long flags;
+ fdevice = miscdev_to_fdevice(filp->private_data);
+ cctx = fdevice->cctx;
+
fl = kzalloc(sizeof(*fl), GFP_KERNEL);
if (!fl)
return -ENOMEM;
@@ -1608,6 +1617,29 @@ static struct platform_driver fastrpc_cb_driver = {
},
};
+static int fastrpc_device_register(struct device *dev, struct fastrpc_channel_ctx *cctx,
+ const char *domain)
+{
+ struct fastrpc_device *fdev;
+ int err;
+
+ fdev = devm_kzalloc(dev, sizeof(*fdev), GFP_KERNEL);
+ if (!fdev)
+ return -ENOMEM;
+
+ fdev->cctx = cctx;
+ fdev->miscdev.minor = MISC_DYNAMIC_MINOR;
+ fdev->miscdev.fops = &fastrpc_fops;
+ fdev->miscdev.name = devm_kasprintf(dev, GFP_KERNEL, "fastrpc-%s", domain);
+ err = misc_register(&fdev->miscdev);
+ if (err)
+ kfree(fdev);
+ else
+ cctx->fdevice = fdev;
+
+ return err;
+}
+
static int fastrpc_rpmsg_probe(struct rpmsg_device *rpdev)
{
struct device *rdev = &rpdev->dev;
@@ -1637,11 +1669,7 @@ static int fastrpc_rpmsg_probe(struct rpmsg_device *rpdev)
if (!data)
return -ENOMEM;
- data->miscdev.minor = MISC_DYNAMIC_MINOR;
- data->miscdev.name = devm_kasprintf(rdev, GFP_KERNEL, "fastrpc-%s",
- domains[domain_id]);
- data->miscdev.fops = &fastrpc_fops;
- err = misc_register(&data->miscdev);
+ err = fastrpc_device_register(rdev, data, domains[domain_id]);
if (err) {
kfree(data);
return err;
@@ -1681,7 +1709,9 @@ static void fastrpc_rpmsg_remove(struct rpmsg_device *rpdev)
fastrpc_notify_users(user);
spin_unlock_irqrestore(&cctx->lock, flags);
- misc_deregister(&cctx->miscdev);
+ if (cctx->fdevice)
+ misc_deregister(&cctx->fdevice->miscdev);
+
of_platform_depopulate(&rpdev->dev);
cctx->rpdev = NULL;
--
2.21.0
^ permalink raw reply related
* [PATCH v3 00/12] misc: fastrpc: Add missing DSP FastRPC features
From: Srinivas Kandagatla @ 2022-01-26 13:52 UTC (permalink / raw)
To: robh+dt, gregkh
Cc: devicetree, ekangupt, bkumar, linux-kernel, srini,
bjorn.andersson, linux-arm-msm, Srinivas Kandagatla
This patchset adds below DSP FastRPC features that have been missing in
upstream fastrpc driver and also cleans up channel context structure with kref.
- Add ablity to reflect if the DSP domain is secure/unsecure by creating
seperate device nodes for secured domain, this would used by SE policy
to restrict applications loading process on the DSP.
- Add new IOCTL to get DSP capabilites
- Add IOCTL to support mapping memory on the DSP.
- Add support for allocating secure memory for DSP
- Handle fdlist in put args
- Handle dma fds in invoke request.
Tested this series on DragonBoard 845c with TensoFlowLite.
Changes since v2:
- Add support for Secure Memory allocations.
- added handling fdlist and dmalist in and after invoke.
- removed unnecessary debug log
- removed dependency on yaml bindings and added new bindings to .txt
Jeya R (5):
misc: fastrpc: add support for FASTRPC_IOCTL_MEM_MAP/UNMAP
misc: fastrpc: Add support to get DSP capabilities
dt-bindings: misc: add property to support non-secure DSP
misc: fastrpc: check before loading process to the DSP
arm64: dts: qcom: add non-secure domain property to fastrpc nodes
Srinivas Kandagatla (2):
misc: fastrpc: separate fastrpc device from channel context
misc: fastrpc: add secure domain support
Vamsi Krishna Gattupalli (5):
dt-bindings: misc: add fastrpc domain vmid property
misc: fastrpc: Add support to secure memory map
misc: fastrpc: Add helper function to get list and page
misc: fastrpc: Add fdlist implementation
misc: fastrpc: Add dma handle implementation
.../devicetree/bindings/misc/qcom,fastrpc.txt | 10 +
arch/arm64/boot/dts/qcom/msm8916.dtsi | 1 +
arch/arm64/boot/dts/qcom/sdm845.dtsi | 2 +
arch/arm64/boot/dts/qcom/sm8150.dtsi | 3 +
arch/arm64/boot/dts/qcom/sm8250.dtsi | 3 +
arch/arm64/boot/dts/qcom/sm8350.dtsi | 3 +
drivers/misc/fastrpc.c | 552 ++++++++++++++++--
include/uapi/misc/fastrpc.h | 81 ++-
8 files changed, 607 insertions(+), 48 deletions(-)
--
2.21.0
^ permalink raw reply
* Re: (EXT) [PATCH 9/9] arm64: dts: imx8mp: add GPU nodes
From: Alexander Stein @ 2022-01-26 13:51 UTC (permalink / raw)
To: Lucas Stach
Cc: Shawn Guo, Rob Herring, linux-arm-kernel, Pengutronix Kernel Team,
NXP Linux Team, Fabio Estevam, devicetree, linux-arm-kernel,
patchwork-lst
In-Reply-To: <20220119134027.2931945-10-l.stach@pengutronix.de>
Am Mittwoch, 19. Januar 2022, 14:40:27 CET schrieb Lucas Stach:
> Add the DT nodes for both the 3D and 2D GPU cores found on the i.MX8MP.
>
> etnaviv-gpu 38000000.gpu: model: GC7000, revision: 6204
> etnaviv-gpu 38008000.gpu: model: GC520, revision: 5341
> [drm] Initialized etnaviv 1.3.0 20151214 for etnaviv on minor 0
Unfortunately it does not work when CONFIG_DRM_ETNAVIV=m
etnaviv-gpu 38000000.gpu: model: GC0, revision: 0
etnaviv-gpu 38000000.gpu: Unknown GPU model
etnaviv-gpu 38008000.gpu: model: GC0, revision: 0
etnaviv-gpu 38008000.gpu: Unknown GPU model
When I use CONFIG_DRM_ETNAVIV=y, I get the same log message as you. It's not
related to this patch, but I have no clue if the cause is in blk-ctrl or pgc.
I think (don't know for sure yet) my random errors on USB side are gone when
USB drivers (PHY as well) are built-in. But I might be wrong here.
Best regards,
Alexander
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox