From: Lachlan Michael <Lachlan.Michael@sony.com>
To: Jai Luthra <jai.luthra@ideasonboard.com>,
devicetree@vger.kernel.org, hverkuil+cisco@kernel.org,
laurent.pinchart@ideasonboard.com, linux-media@vger.kernel.org,
mchehab@kernel.org, sakari.ailus@linux.intel.com
Cc: robh@kernel.org, krzk+dt@kernel.org, conor+dt@kernel.org,
kieran.bingham@ideasonboard.com, Ryuichi.Tadano@sony.com,
Kengo.Hayasaka@sony.com, Tim Bird <Tim.Bird2@sony.com>,
Kazumi.A.Sato@sony.com, Yuji.John.Takahashi@sony.com,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH v3 2/2] media: i2c: Add Sony IMX908 image sensor driver
Date: Thu, 1 Oct 2026 17:36:08 +0900 [thread overview]
Message-ID: <9c7e136c-f092-4a6b-ab1d-cbafdea26706@sony.com> (raw)
In-Reply-To: <179000809637.3701737.17026819145311474886@freya>
Dear Jai,
On 9/22/2026 1:28 AM, Jai Luthra wrote:
> Hi Lachlan, Thank you for your patch. And thanks for fixing my comments
> from v2. I have a few more, some I missed the last time around, but some
> are new. Quoting Lachlan Michael (2026-08-28 12: 18: 43) > The Sony
> IMX908 is an 8. 39 megapixel
> Hi Lachlan,
>
> Thank you for your patch.
>
> And thanks for fixing my comments from v2. I have a few more, some I missed
> the last time around, but some are new.
Appreciate the time you have spent to review it.
Since Laurent made some additional comments, for those I will reply to
Laurent's e-mail.
> Quoting Lachlan Michael (2026-08-28 12:18:43)
>> The Sony IMX908 is an 8.39 megapixel (3856x2176) CMOS image sensor
>> with a MIPI CSI-2 output interface, configurable as either 2 or 4
>> data lanes.
>>
>> Add a V4L2 sub-device driver for the sensor. The driver supports
>> RAW10 and RAW12 output formats, exposure and analogue gain controls,
>> horizontal and vertical flipping, horizontal and vertical blanking
>> controls, window cropping and test pattern generation.
>>
>> HDR modes and RAW16 output are not currently supported.
>>
>> Signed-off-by: Lachlan Michael <lachlan.michael@sony.com>
>> ---
>> Changes in v3:
>> - Correct link-frequency indexing mapping to match the sensor
>> DATARATE_SEL values and the frequency ordering
>> - Changed the pixel_rate to 16-pixel-per-clocks based on datasheet
>> mode values
>> - Clean up the constant names, register definitions and comments in
>> response to review feedback.
>> - Remove various (uneeded) checks in code based on feedback
>> - Remove cached vmax/hmax and recalculated where necessary
>> - Drop pixel_rate and test_pattern pointers from struct
>> - Remove read-only controls set_ctrl switch
>> - Remove event handlers
>> - Remove uneeded delays at startup
>> - Simplify runtime_resume/suspend functions
>> - Remove the 4k recording-area in favour of larger active area
>> - Use the same constraints for VMAX for crop and all pixel mode to
>> simplify the driver logic by removing branching
>> - Rework HMAX based on the fact that the whole line is always read
>> even in crop mode. HMAX varies in all modes now, but not with
>> horizontal crop.
>> - Minor edits and bugfixes
>> ---
>> MAINTAINERS | 1 +
>> drivers/media/i2c/Kconfig | 11 +
>> drivers/media/i2c/Makefile | 1 +
>> drivers/media/i2c/imx908.c | 1367 ++++++++++++++++++++++++++++++++++++
>> 4 files changed, 1380 insertions(+)
>> create mode 100644 drivers/media/i2c/imx908.c
>>
>> diff --git a/MAINTAINERS b/MAINTAINERS
>> index bb7adcfd4767..a17c31234ec7 100644
>> --- a/MAINTAINERS
>> +++ b/MAINTAINERS
>> @@ -25542,6 +25542,7 @@ SONY IMX908 SENSOR DRIVER
>> M: Lachlan Michael <lachlan.michael@sony.com>
>> S: Maintained
>> F: Documentation/devicetree/bindings/media/i2c/sony,imx908.yaml
>> +F: drivers/media/i2c/imx908.c
>>
>> SONY MEMORYSTICK SUBSYSTEM
>> M: Maxim Levitsky <maximlevitsky@gmail.com>
>> diff --git a/drivers/media/i2c/Kconfig b/drivers/media/i2c/Kconfig
>> index 5c52007f9cbe..cfa1359d8c7d 100644
>> --- a/drivers/media/i2c/Kconfig
>> +++ b/drivers/media/i2c/Kconfig
>> @@ -310,6 +310,17 @@ config VIDEO_IMX678
>> To compile this driver as a module, choose M here: the
>> module will be called imx678.
>>
>> +config VIDEO_IMX908
>> + tristate "Sony IMX908 sensor support"
>> + depends on GPIOLIB
>> + select V4L2_CCI_I2C
>> + help
>> + This is a Video4Linux2 sensor driver for the Sony
>> + IMX908 CMOS image sensor.
>> +
>> + To compile this driver as a module, choose M here: the
>> + module will be called imx908.
>> +
>> config VIDEO_MAX9271_LIB
>> tristate
>>
>> diff --git a/drivers/media/i2c/Makefile b/drivers/media/i2c/Makefile
>> index d04bd5724552..d91ed1bc6427 100644
>> --- a/drivers/media/i2c/Makefile
>> +++ b/drivers/media/i2c/Makefile
>> @@ -63,6 +63,7 @@ obj-$(CONFIG_VIDEO_IMX412) += imx412.o
>> obj-$(CONFIG_VIDEO_IMX415) += imx415.o
>> obj-$(CONFIG_VIDEO_IMX678) += imx678.o
>> obj-$(CONFIG_VIDEO_IMX471) += imx471.o
>> +obj-$(CONFIG_VIDEO_IMX908) += imx908.o
>> obj-$(CONFIG_VIDEO_IR_I2C) += ir-kbd-i2c.o
>> obj-$(CONFIG_VIDEO_ISL7998X) += isl7998x.o
>> obj-$(CONFIG_VIDEO_KS0127) += ks0127.o
>> diff --git a/drivers/media/i2c/imx908.c b/drivers/media/i2c/imx908.c
>> new file mode 100644
>> index 000000000000..ca85a5cd058a
>> --- /dev/null
>> +++ b/drivers/media/i2c/imx908.c
>> @@ -0,0 +1,1367 @@
>> +// SPDX-License-Identifier: GPL-2.0
>> +/*
>> + * V4L2 driver for Sony IMX908
>> + *
>> + * Diagonal 6.42 mm (Type 1/2.8) CMOS image sensor with 8.39 M pixels.
>> + *
>> + * Copyright 2026 Sony Semiconductor Solutions Corporation
>> + *
>> + */
>> +
>> +#include <linux/align.h>
>> +#include <linux/array_size.h>
>> +#include <linux/bitops.h>
>> +#include <linux/clk.h>
>> +#include <linux/container_of.h>
>> +#include <linux/delay.h>
>> +#include <linux/err.h>
>> +#include <linux/gpio/consumer.h>
>> +#include <linux/i2c.h>
>> +#include <linux/math.h> /* DIV_ROUND_UP */
>> +#include <linux/math64.h> /* DIV64_U64_ROUND_UP */
>> +#include <linux/minmax.h>
>> +#include <linux/module.h>
>> +#include <linux/pm_runtime.h>
>> +#include <linux/property.h>
>> +#include <linux/regulator/consumer.h>
>> +
>> +#include <media/media-entity.h>
>> +#include <media/v4l2-cci.h>
>> +#include <media/v4l2-ctrls.h>
>> +#include <media/v4l2-fwnode.h>
>> +#include <media/v4l2-mediabus.h>
>> +#include <media/v4l2-rect.h>
>> +#include <media/v4l2-subdev.h>
>> +
>> +/* ---- IMX908 Registers --------- */
>> +#define IMX908_REG_STANDBY CCI_REG8(0x3000)
>> +#define IMX908_STANDBY_EN BIT(0) /* Standby mode */
>> +#define IMX908_STANDBY_CANCEL 0x00 /* Operating mode */
>> +#define IMX908_REG_XMSTA CCI_REG8(0x3002) /* [0] Init 1h */
>> +#define IMX908_XMSTA_START 0x00
>> +#define IMX908_XMSTA_STOP 0x01
>> +#define IMX908_REG_XMASTER CCI_REG8(0x3003) /* [0] Init 0h */
>> +#define IMX908_CONTROLLER_MODE 0x00
>> +#define IMX908_REG_INCK_SEL CCI_REG8(0x3014) /* 0:74.25 1:37.125 2:72 3:27 4:24 */
>> +#define IMX908_REG_DATARATE_SEL CCI_REG8(0x3015)
>> +
>> +/* ---- Crop */
>> +#define IMX908_REG_WINMODE CCI_REG8(0x3018) /* [3:0] */
>> +#define IMX908_WINMODE_ALLPIX 0x00 /* All-pixel readout mode */
>> +#define IMX908_WINMODE_CROP 0x04 /* Window cropping mode */
>> +
>> +/* ---- Flip */
>> +#define IMX908_REG_HREVERSE CCI_REG8(0x3020)
>> +#define IMX908_HREVERSE_NORMAL 0x00
>> +#define IMX908_HREVERSE_INV 0x01
>> +#define IMX908_REG_VREVERSE CCI_REG8(0x3021)
>> +#define IMX908_VREVERSE_NORMAL 0x00
>> +#define IMX908_VREVERSE_INV 0x01
>> +
>> +/* ---- Internal Analog to Digital conversion bit width */
>> +#define IMX908_REG_ADBIT CCI_REG8(0x3022)
>> +#define IMX908_ADBIT_10BIT 0x00 /* 10-bit */
>> +#define IMX908_ADBIT_12BIT 0x01 /* 11-bit + digital dither */
>> +
>> +#define IMX908_REG_MDBIT CCI_REG8(0x3023)
>> +#define IMX908_MDBIT_RAW10 0x00
>> +#define IMX908_MDBIT_RAW12 0x01
>> +
>> +/* ---- VMAX Frame length in lines */
>> +#define IMX908_REG_VMAX CCI_REG24_LE(0x3028) /* [19:0] */
>> +#define IMX908_VMAX_MAX 0xfffff
>> +#define IMX908_VMAX_DEFAULT 2250
>> +/*
>> + * VMAX restrictions, documented for window cropping mode:
>> + * VMAX >= PIX_VWIDTH + 70
>> + * VMAX >= 1206
>> + */
>> +#define IMX908_VMAX_MARGIN 70
>> +#define IMX908_VMAX_MIN 1206
>> +
>> +/* ---- HMAX Line length in clocks */
>> +#define IMX908_REG_HMAX CCI_REG16_LE(0x302c) /* [15:0] */
>> +#define IMX908_HMAX_MAX 0xffff
>> +#define IMX908_HMAX_DEFAULT 1100
>> +
>> +/* ---- Cropping related */
>> +#define IMX908_REG_PIX_HST CCI_REG16_LE(0x303c) /* [12:0] Init 0h */
>> +#define IMX908_REG_PIX_HWIDTH CCI_REG16_LE(0x303e) /* [12:0] Init 0F10h */
>> +#define IMX908_REG_LANEMODE CCI_REG8(0x3040)
>> +#define IMX908_LANEMODE_2LANE 0x01
>> +#define IMX908_LANEMODE_4LANE 0x03
>> +#define IMX908_REG_PIX_VST CCI_REG16_LE(0x3044) /* [12:0] Init 0h */
>> +#define IMX908_REG_PIX_VWIDTH CCI_REG16_LE(0x3046) /* [12:0] Init 0884h */
>> +#define IMX908_CROP_ALIGN_VSTART 4
>> +#define IMX908_CROP_ALIGN_VWIDTH 4
>> +#define IMX908_CROP_ALIGN_HSTART 4
>> +#define IMX908_CROP_ALIGN_HWIDTH 16
>> +#define IMX908_CROP_MIN_HEIGHT 1136
>> +#define IMX908_CROP_MIN_WIDTH 1040
>> +
>> +/* ---- Shutter */
>> +#define IMX908_REG_SHR0 CCI_REG24_LE(0x3050) /* [19:0] */
>> +#define IMX908_MIN_SHR0 7
>> +
>> +/*
>> + * Analogue gain control
>> + * Range is from 0 to 100 (0dB - 30dB) with 0.3dB step size
>> + * Values from 101 to 240 are valid but correspond to additional digital gain
>> + * (0.3dB - 42dB) so don't expose it to userspace
>> + */
>> +#define IMX908_REG_GAIN CCI_REG16_LE(0x3070) /* [10:0] */
>> +#define IMX908_ANA_GAIN_MAX 100
>> +#define IMX908_ANA_GAIN_MIN 0
>> +#define IMX908_ANA_GAIN_STEP 1
>> +#define IMX908_ANA_GAIN_DEFAULT 0
>> +
>> +/* ---- XVS/XHS output mode */
>> +#define IMX908_XXS_DRV CCI_REG8(0x30a6) /* [1:0] XVS_DRV [3:2] XHS_DRV */
>> +
>> +#define IMX908_REG_BLKLEVEL CCI_REG16_LE(0x30dc) /* [15:0] black level */
>> +
>> +/* ---- Test Pattern Generator (TPG) registers */
>> +#define IMX908_REG_TPG_EN CCI_REG8(0x30e0)
>> +#define IMX908_TPG_EN_BIT BIT(0)
>> +#define IMX908_REG_TPG_PATSEL CCI_REG8(0x30e2) /* [4:0] pattern idx */
>> +#define IMX908_REG_TESTCLKEN CCI_REG8(0x4900)
>> +#define IMX908_TESTCLKEN_BIT BIT(3)
>> +
>> +#define IMX908_REG_TYPE_ID CCI_REG16_LE(0x4c0c)
>> +#define IMX908_CHIP_ID 0x038c
>> +
>> +/* ---- IMX908 Common Register List ---- */
>> +static const struct cci_reg_sequence imx908_common_regs[] = {
>> + { IMX908_XXS_DRV, IMX908_CONTROLLER_MODE },
>> + { CCI_REG8(0x039c), 0x03 },
>> + { CCI_REG8(0x3416), 0x20 },
>> + { CCI_REG8(0x3417), 0x00 },
>> + { CCI_REG8(0x3456), 0xf4 },
>> + { CCI_REG8(0x345d), 0x01 },
>> + { CCI_REG8(0x3460), 0x00 },
>> + { CCI_REG8(0x3461), 0x0b },
>> + { CCI_REG8(0x3471), 0x00 },
>> + { CCI_REG8(0x3472), 0x23 },
>> + { CCI_REG8(0x347b), 0x02 },
>> + { CCI_REG8(0x3481), 0x01 },
>> + { CCI_REG8(0x380c), 0x00 },
>> + { CCI_REG8(0x380f), 0x0c },
>> + { CCI_REG8(0x381c), 0x11 },
>> + { CCI_REG8(0x3820), 0x22 },
>> + { CCI_REG8(0x3824), 0x33 },
>> + { CCI_REG8(0x3828), 0x22 },
>> + { CCI_REG8(0x382c), 0x33 },
>> + { CCI_REG8(0x3830), 0x33 },
>> + { CCI_REG8(0x3838), 0x05 },
>> + { CCI_REG8(0x383c), 0x07 },
>> + { CCI_REG8(0x3840), 0x06 },
>> + { CCI_REG8(0x3848), 0x17 },
>> + { CCI_REG8(0x384c), 0x0b },
>> + { CCI_REG8(0x3850), 0x10 },
>> + { CCI_REG8(0x3854), 0x12 },
>> + { CCI_REG8(0x385c), 0x20 },
>> + { CCI_REG8(0x38c4), 0x64 },
>> + { CCI_REG8(0x38c5), 0x64 },
>> + { CCI_REG8(0x38c6), 0x64 },
>> + { CCI_REG8(0x3c4a), 0x15 },
>> + { CCI_REG8(0x3c4c), 0x13 },
>> + { CCI_REG8(0x3c4d), 0x13 },
>> + { CCI_REG8(0x3c4e), 0x13 },
>> + { CCI_REG8(0x3c50), 0x77 },
>> + { CCI_REG8(0x3c51), 0x07 },
>> + { CCI_REG8(0x3db4), 0x00 },
>> + { CCI_REG8(0x4419), 0x06 },
>> + { CCI_REG8(0x441c), 0x00 },
>> + { CCI_REG8(0x4426), 0x00 },
>> + { CCI_REG8(0x4538), 0x20 },
>> + { CCI_REG8(0x4539), 0x19 },
>> + { CCI_REG8(0x453a), 0x19 },
>> + { CCI_REG8(0x453b), 0x19 },
>> + { CCI_REG8(0x453c), 0x19 },
>> + { CCI_REG8(0x453d), 0x19 },
>> + { CCI_REG8(0x453e), 0x19 },
>> + { CCI_REG8(0x453f), 0x19 },
>> + { CCI_REG8(0x4540), 0x19 },
>> + { CCI_REG8(0x4544), 0x11 },
>> + { CCI_REG8(0x4545), 0x11 },
>> + { CCI_REG8(0x4546), 0x11 },
>> + { CCI_REG8(0x463c), 0x20 },
>> + { CCI_REG8(0x465e), 0xcf },
>> + { CCI_REG8(0x4684), 0x20 },
>> + { CCI_REG8(0x46a6), 0xcf },
>> + { CCI_REG8(0x46e2), 0xf3 },
>> +};
>> +
>> +static const u32 imx908_bit_depth_regs[] = {
>> + CCI_REG8(0x3d78),
>> + CCI_REG8(0x3d79),
>> + CCI_REG8(0x3d80),
>> + CCI_REG8(0x3d81),
>> + CCI_REG8(0x3d88),
>> + CCI_REG8(0x3d89),
>> + CCI_REG8(0x3d90),
>> + CCI_REG8(0x3d91),
>> +};
>> +
>> +#define IMX908_EXPOSURE_MIN 1
>> +#define IMX908_EXPOSURE_STEP 1
>> +
>> +/* ---- Internal image-data interface clock */
>> +#define IMX908_VTPXCK_HZ 74250000ULL
>> +
>> +/* Fixed pixel rate: 16 pixels per video timing pixel clock cycle */
>> +#define IMX908_PIX_PER_CLK 16
>> +#define IMX908_PIXEL_RATE (IMX908_VTPXCK_HZ * IMX908_PIX_PER_CLK)
>> +
>> +/* ---- Subdev Pads */
>> +#define IMX908_SOURCE_PAD 0
>> +
>> +#define IMX908_DEFAULT_MBUS_CODE MEDIA_BUS_FMT_SRGGB10_1X10
>> +
>> +/*
>> + * IMX908 total area includes active area height plus
>> + * 4 pixels effective pixel ignored area
>> + * 10 pixels vertical direction effective OB
>> + * 10 pixels OB side ignored area
>> + */
>> +static const struct v4l2_rect imx908_total_area = {
>> + .top = 0,
>> + .left = 0,
>> + .width = 3856,
>> + .height = 2200,
>> +};
>> +
>
> The "Pixel Arrangement" section in the datasheet for IMX678 (Starvis 2,
> same resolution) mentioned a horizontal margin of 8 and 9, and vertical
> margin of 10 + 10 + 4 + 8 and 9, thus the total came out to be 3857x2201
>
> Which also explained why the CFA readout stays R->G->G->B irrespective of
> VFLIP and HFLIP, the extra pixel allowed shifting the readout start.
I re-checked the IMX908 datasheet under the Pixel Arrangement and I
confirmed the current width and height are correct.
In both the vertical and horizontal directions, the final "Effective
margin for error processing" is written as 9 (pixels) in the diagram,
however in the comments it says "The last effective line and column are
not read out."
If we represent the not reading out as a -1 the numbers work out.
- the horizontal width is: 0+8+3840+9(-1)+0 = 3856
- the vertical height is: 10+10+4+8+2160+9(-1) = 2200
>> +static const struct v4l2_rect imx908_active_area = {
>> + .top = 0,
>> + .left = 0,
>> + .width = 3856,
>> + .height = 2176,
>
> Shouldn't this have .top = 24 (10 + 10 + 4)?
>
> At least on IMX678, while the diagram shows the OB area at bottom, that's
> where the scan starts from (N12 pin) when HREVERSE/VREVERSE are 0.
>
> It would be good to confirm if IMX908 is indeed different in its behaviour
> from Starvis2 sensors.
You are right. I will change this to .top = 24,
>> +};
>> +
>> +static const char *const imx908_supply_names[] = {
>> + "avdd", /* Analog (3.3V) supply */
>> + "dvdd", /* Digital Core (1.1V) supply */
>> + "ovdd", /* IF (1.8V) supply */
>> +};
>> +
>> +static const u32 imx908_mbus_codes[] = {
>> + MEDIA_BUS_FMT_SRGGB10_1X10, /* RAW10 */
>> + MEDIA_BUS_FMT_SRGGB12_1X12, /* RAW12 */
>> +};
>> +
>> +/*
>> + * Link frequency indices, ordered so that the index is also the
>> + * DATARATE_SEL register value. Keep in descending frequency order.
>> + */
>> +enum {
>> + IMX908_LINK_FREQ_1188MHZ = 0x00,
>> + IMX908_LINK_FREQ_1039MHZ = 0x01,
>> + IMX908_LINK_FREQ_891MHZ = 0x02,
>> + IMX908_LINK_FREQ_720MHZ = 0x03,
>> + IMX908_LINK_FREQ_594MHZ = 0x04,
>> + IMX908_LINK_FREQ_445MHZ = 0x05,
>> + IMX908_LINK_FREQ_360MHZ = 0x06,
>> + IMX908_LINK_FREQ_297MHZ = 0x07,
>> +};
>> +
>> +static const s64 imx908_link_freqs[] = {
>> + [IMX908_LINK_FREQ_1188MHZ] = 1188000000LL,
>> + [IMX908_LINK_FREQ_1039MHZ] = 1039500000LL,
>> + [IMX908_LINK_FREQ_891MHZ] = 891000000LL,
>> + [IMX908_LINK_FREQ_720MHZ] = 720000000LL,
>> + [IMX908_LINK_FREQ_594MHZ] = 594000000LL,
>> + [IMX908_LINK_FREQ_445MHZ] = 445500000LL,
>> + [IMX908_LINK_FREQ_360MHZ] = 360000000LL,
>> + [IMX908_LINK_FREQ_297MHZ] = 297000000LL,
>> +};
>> +
>> +/* Allowed clock values in Hz */
>> +static const u32 imx908_inck_table[] = {
>> + 74250000,
>> + 37125000,
>> + 72000000,
>> + 27000000,
>> + 24000000,
>> +};
>> +
>> +struct imx908 {
>> + struct v4l2_subdev sd;
>> + struct media_pad pad;
>> + struct device *dev;
>> +
>> + struct regmap *cci;
>> +
>> + struct clk *xclk;
>> + struct gpio_desc *reset_gpio;
>> + struct regulator_bulk_data supplies[ARRAY_SIZE(imx908_supply_names)];
>> +
>> + u8 inck_sel;
>> +
>> + u8 num_lanes;
>> + unsigned long link_freq_bitmap;
>> + unsigned int link_freq_idx;
>> +
>> + struct {
>> + struct v4l2_ctrl_handler handler;
>> +
>> + struct v4l2_ctrl *exposure;
>> + struct v4l2_ctrl *vblank;
>> + struct v4l2_ctrl *hblank;
>> + } ctrls;
>> +};
>> +
>> +static inline struct imx908 *to_imx908(struct v4l2_subdev *_sd)
>> +{
>> + return container_of(_sd, struct imx908, sd);
>> +}
>> +
>> +/* ----------------------- Basic Helper Functions --------------------------- */
>> +static bool imx908_mbus_code_supported(struct imx908 *imx, u32 code)
>> +{
>> + for (unsigned int i = 0; i < ARRAY_SIZE(imx908_mbus_codes); i++) {
>> + if (imx908_mbus_codes[i] == code)
>> + return true;
>> + }
>> +
>> + return false;
>> +}
>> +
>> +static u16 imx908_calc_hmax(u32 width, u32 hblank)
>> +{
>> + u32 hmax = DIV_ROUND_UP(width + hblank, IMX908_PIX_PER_CLK);
>> +
>> + return min_t(u32, hmax, IMX908_HMAX_MAX);
>> +}
>> +
>> +/* Horizontal cropping is digital, so the full width is always read out */
>> +static u16 imx908_calc_min_hmax(struct imx908 *imx, u8 bpp)
>> +{
>> + u64 link_hz = imx908_link_freqs[imx->link_freq_idx];
>> + u64 num = (u64)imx908_active_area.width * bpp * IMX908_VTPXCK_HZ;
>> + u64 den = (u64)imx->num_lanes * link_hz * 2; /* DDR */
>> +
>> + /*
>> + * den can exceed 32 bits (e.g. 4 lanes * 720 MHz * 2 = 5.76 GHz), so
>> + * DIV_ROUND_UP_ULL / do_div would truncate the divisor to u32. Use a
>> + * full 64/64 division.
>> + */
>> + return DIV64_U64_ROUND_UP(num, den);
>> +}
>> +
>> +static u32 imx908_calc_min_vblank(const struct v4l2_rect *crop)
>> +{
>> + u32 min_vmax = max_t(u32, crop->height + IMX908_VMAX_MARGIN,
>> + IMX908_VMAX_MIN);
>> +
>> + return min_vmax - crop->height;
>> +}
>> +
>> +static int imx908_set_exposure_lines(struct imx908 *imx, u32 exposure_lines)
>> +{
>> + const struct v4l2_mbus_framefmt *format;
>> + struct v4l2_subdev_state *state;
>> + u32 vmax, max_lines;
>> +
>> + state = v4l2_subdev_get_locked_active_state(&imx->sd);
>> + format = v4l2_subdev_state_get_format(state, IMX908_SOURCE_PAD);
>> +
>> + vmax = format->height + imx->ctrls.vblank->val;
>> + max_lines = vmax - IMX908_MIN_SHR0;
>> +
>> + /* Guards SHR0 underflow when VBLANK+EXPOSURE are set together */
>> + exposure_lines = clamp_t(u32, exposure_lines, 1, max_lines);
>> +
>> + u32 shr0 = vmax - exposure_lines;
>> +
>> + return cci_write(imx->cci, IMX908_REG_SHR0, shr0, NULL);
>> +}
>> +
>> +static u8 imx908_bits_per_pixel(u32 code)
>> +{
>> + switch (code) {
>> + case MEDIA_BUS_FMT_SRGGB12_1X12:
>> + return 12;
>> + case MEDIA_BUS_FMT_SRGGB10_1X10:
>> + default:
>> + return 10;
>> + }
>> +}
>> +
>> +static u32 imx908_hmax_to_hblank(u16 hmax, u32 width)
>> +{
>> + return hmax * IMX908_PIX_PER_CLK - width;
>> +}
>> +
>> +static int imx908_update_hblank_limits(struct imx908 *imx, u32 old_width,
>> + u32 width, u8 bpp)
>> +{
>> + u16 min_hmax = imx908_calc_min_hmax(imx, bpp);
>> + u32 max_hblank = imx908_hmax_to_hblank(IMX908_HMAX_MAX, width);
>> + u32 min_hblank = imx908_hmax_to_hblank(min_hmax, width);
>> + /* Re-derive the line period at the old width to preserve line time */
>> + u16 hmax = imx908_calc_hmax(old_width, imx->ctrls.hblank->val);
>> + u32 hblank = imx908_hmax_to_hblank(hmax, width);
>> + int ret;
>> +
>> + hblank = clamp_t(u32, hblank, min_hblank, max_hblank);
>> +
>> + ret = __v4l2_ctrl_modify_range(imx->ctrls.hblank, min_hblank,
>> + max_hblank, IMX908_PIX_PER_CLK, hblank);
>> + if (ret)
>> + return ret;
>> +
>> + if (imx->ctrls.hblank->val != hblank)
>> + __v4l2_ctrl_s_ctrl(imx->ctrls.hblank, hblank);
>> +
>> + return 0;
>> +}
>> +
>> +static int imx908_update_vblank_limits(struct imx908 *imx,
>> + const struct v4l2_rect *crop,
>> + u32 old_height)
>> +{
>> + u32 min_vblank = imx908_calc_min_vblank(crop);
>> + u32 max_vblank = IMX908_VMAX_MAX - crop->height;
>> + /* Preserve frame length across height changes when possible. */
>> + u32 vmax = old_height + imx->ctrls.vblank->val;
>> + u32 vblank = clamp_t(u32, vmax - crop->height,
>> + min_vblank, max_vblank);
>> + int ret;
>> +
>> + ret = __v4l2_ctrl_modify_range(imx->ctrls.vblank, min_vblank,
>> + max_vblank, 1, vblank);
>> + if (ret)
>> + return ret;
>> +
>> + if (imx->ctrls.vblank->val != vblank)
>> + __v4l2_ctrl_s_ctrl(imx->ctrls.vblank, vblank);
>> +
>> + return 0;
>> +}
>> +
>> +/* ------------------ Pattern Generator (TPG) ----------------------- */
>> +static const char * const imx908_tpg_menu[] = {
>> + "Disabled",
>> + "All 000h",
>> + "All FFFh",
>> + "All 555h",
>> + "All AAAh",
>> + "555h / AAAh Toggle",
>> + "AAAh / 555h Toggle",
>> + "000h / 555h Toggle",
>> + "555h / 000h Toggle",
>> + "000h / FFFh Toggle",
>> + "FFFh / 000h Toggle",
>> + "Horiz Color Bars",
>> + "Vert Color Bars",
>> +};
>> +
>> +static int imx908_update_test_pattern(struct imx908 *imx, u32 index)
>> +{
>> + int ret = 0;
>> +
>> + if (index == 0) {
>> + /* Picture (not TPG) output setting */
>> + cci_update_bits(imx->cci, IMX908_REG_TPG_EN, IMX908_TPG_EN_BIT,
>> + 0x00, &ret);
>> +
>> + /* Disable TESTCLKEN (write-only register) */
>> + cci_write(imx->cci, IMX908_REG_TESTCLKEN, 0x00, &ret);
>> +
>> + return ret;
>> + }
>> +
>> + cci_write(imx->cci, IMX908_REG_TESTCLKEN, IMX908_TESTCLKEN_BIT, &ret);
>> + cci_write(imx->cci, IMX908_REG_TPG_PATSEL, index - 1, &ret);
>> + cci_update_bits(imx->cci, IMX908_REG_TPG_EN, IMX908_TPG_EN_BIT,
>> + IMX908_TPG_EN_BIT, &ret);
>> +
>> + return ret;
>> +}
>> +
>> +static int imx908_update_framing_limits(struct imx908 *imx,
>> + struct v4l2_subdev_state *state,
>> + u32 old_width,
>> + u32 old_height)
>> +{
>> + const struct v4l2_mbus_framefmt *fmt;
>> + const struct v4l2_rect *crop;
>> + int ret;
>> +
>> + fmt = v4l2_subdev_state_get_format(state, IMX908_SOURCE_PAD);
>> + crop = v4l2_subdev_state_get_crop(state, IMX908_SOURCE_PAD);
>> + u8 bpp = imx908_bits_per_pixel(fmt->code);
>> +
>> + ret = imx908_update_hblank_limits(imx, old_width, fmt->width, bpp);
>> + if (ret)
>> + return ret;
>> +
>> + ret = imx908_update_vblank_limits(imx, crop, old_height);
>> + if (ret)
>> + return ret;
>> +
>> + u32 vmax = fmt->height + clamp_t(u32, imx->ctrls.vblank->val,
>> + imx908_calc_min_vblank(crop),
>> + IMX908_VMAX_MAX - fmt->height);
>> +
>> + return __v4l2_ctrl_modify_range(imx->ctrls.exposure,
>> + IMX908_EXPOSURE_MIN,
>> + vmax - IMX908_MIN_SHR0,
>> + IMX908_EXPOSURE_STEP,
>> + imx->ctrls.exposure->default_value);
>> +}
>> +
>> +/* ----------------------- HW Programming ---------------------- */
>> +static int imx908_set_mode(struct imx908 *imx,
>> + const struct v4l2_mbus_framefmt *fmt)
>> +{
>> + u8 adbit, mdbit, value;
>> + u16 blklevel;
>> + int ret = 0;
>> +
>> + switch (fmt->code) {
>> + case MEDIA_BUS_FMT_SRGGB10_1X10:
>> + adbit = IMX908_ADBIT_10BIT;
>> + mdbit = IMX908_MDBIT_RAW10;
>> + blklevel = 50; /* datasheet recommended value for 10-bit */
>> + value = 0x0c; /* register map value for 10-bit */
>> + break;
>> + case MEDIA_BUS_FMT_SRGGB12_1X12:
>> + default:
>> + adbit = IMX908_ADBIT_12BIT;
>> + mdbit = IMX908_MDBIT_RAW12;
>> + blklevel = 200; /* datasheet recommended value for 12-bit */
>> + value = 0x05; /* register map value for 12-bit */
>> + break;
>> + }
>> +
>> + cci_write(imx->cci, IMX908_REG_XMASTER, IMX908_CONTROLLER_MODE, &ret);
>> + cci_write(imx->cci, IMX908_REG_ADBIT, adbit, &ret);
>> + cci_write(imx->cci, IMX908_REG_MDBIT, mdbit, &ret);
>> +
>> + for (unsigned int i = 0; i < ARRAY_SIZE(imx908_bit_depth_regs); ++i)
>> + cci_write(imx->cci, imx908_bit_depth_regs[i], value, &ret);
>> +
>> + cci_write(imx->cci, IMX908_REG_INCK_SEL, imx->inck_sel, &ret);
>> + cci_write(imx->cci, IMX908_REG_DATARATE_SEL, imx->link_freq_idx, &ret);
>> + cci_write(imx->cci, IMX908_REG_LANEMODE, imx->num_lanes == 2 ?
>> + IMX908_LANEMODE_2LANE : IMX908_LANEMODE_4LANE, &ret);
>> + cci_write(imx->cci, IMX908_REG_BLKLEVEL, blklevel, &ret);
>
> Has the BLKLEVEL register handling changed from Starvis2 -> Starvis3?
>
> For IMX678 the default value recommended by datasheet were the same, 50 and
> 200, but with a note that increasing register by 1 in 12-bit mode meant an
> increase of 4 in the output. Thus the register always reads 50 (or 0x32) by
> default.
In the IMX908 register map (excel file), the clearly defined values are
50 for 10bit and 200 for 12bit. I have followed this. I also ran a
simple experiment to check the values for a completely dark (black) RAW
capture and the result was 3200 for both cases as expected. So I think
the current code is correct.
I do note older versions of the Software Reference Manual (e.g. v0.2)
had the same value (032h) for both 10bit and 12bit, but this has been
corrected in v1.0 to the following:
Recommended setting
10-bit output: 032h (50d in LSB units)
12-bit output: 0C8h (200d in LSB units)
16-bit output: C80h (3200d in LSB units)
>> +
>> + if (ret)
>> + dev_err(imx->dev, " register write failed: %d\n", ret);
>> +
>> + return ret;
>> +}
>> +
>> +static int imx908_program_window(struct imx908 *imx,
>> + const struct v4l2_rect *crop)
>> +{
>> + bool all_pixel_mode = v4l2_rect_equal(crop, &imx908_active_area);
>> + int ret = 0;
>> +
>> + cci_write(imx->cci, IMX908_REG_WINMODE,
>> + all_pixel_mode ? IMX908_WINMODE_ALLPIX : IMX908_WINMODE_CROP,
>> + &ret);
>> +
>> + if (!all_pixel_mode) {
>> + cci_write(imx->cci, IMX908_REG_PIX_HST, crop->left, &ret);
>> + cci_write(imx->cci, IMX908_REG_PIX_HWIDTH, crop->width, &ret);
>> + cci_write(imx->cci, IMX908_REG_PIX_VST, crop->top, &ret);
>> + cci_write(imx->cci, IMX908_REG_PIX_VWIDTH, crop->height, &ret);
>> + }
>> +
>> + return ret;
>> +}
>> +
>> +/* ----------------------- Start/Stop streaming ---------------------- */
>> +static int imx908_start_streaming(struct imx908 *imx,
>> + struct v4l2_subdev_state *sd_state)
>> +{
>> + const struct v4l2_mbus_framefmt *format;
>> + const struct v4l2_rect *crop;
>> + int ret = 0;
>> +
>> + crop = v4l2_subdev_state_get_crop(sd_state, IMX908_SOURCE_PAD);
>> + format = v4l2_subdev_state_get_format(sd_state, IMX908_SOURCE_PAD);
>> +
>> + cci_multi_reg_write(imx->cci, imx908_common_regs,
>> + ARRAY_SIZE(imx908_common_regs), &ret);
>> + if (ret)
>> + return ret;
>> +
>> + ret = imx908_set_mode(imx, format);
>> + if (ret)
>> + return ret;
>> +
>> + ret = imx908_program_window(imx, crop);
>> + if (ret)
>> + return ret;
>> +
>> + ret = __v4l2_ctrl_handler_setup(imx->sd.ctrl_handler);
>> + if (ret)
>> + return ret;
>> +
>> + cci_write(imx->cci, IMX908_REG_STANDBY, IMX908_STANDBY_CANCEL, &ret);
>> + fsleep(24 * USEC_PER_MSEC); /* >=24 ms stabilization after standby cancel */
>> +
>> + cci_write(imx->cci, IMX908_REG_XMSTA, IMX908_XMSTA_START, &ret);
>> +
>> + return ret;
>> +}
>> +
>> +static int imx908_stop_streaming(struct imx908 *imx)
>> +{
>> + int ret = 0;
>> +
>> + cci_write(imx->cci, IMX908_REG_STANDBY, IMX908_STANDBY_EN, &ret);
>> + cci_write(imx->cci, IMX908_REG_XMSTA, IMX908_XMSTA_STOP, &ret);
>> +
>> + return ret;
>> +}
>> +
>> +/* --------------------------- V4L2 controls ------------------------------ */
>> +static int imx908_set_ctrl(struct v4l2_ctrl *ctrl)
>> +{
>> + struct imx908 *imx = container_of(ctrl->handler, struct imx908,
>> + ctrls.handler);
>> + const struct v4l2_mbus_framefmt *format;
>> + struct v4l2_subdev_state *state;
>> + int ret = 0;
>> +
>> + state = v4l2_subdev_get_locked_active_state(&imx->sd);
>> + format = v4l2_subdev_state_get_format(state, IMX908_SOURCE_PAD);
>> +
>> + /* Update exposure control limits even if the sensor is not streaming */
>> + if (ctrl->id == V4L2_CID_VBLANK) {
>> + const struct v4l2_rect *crop;
>> +
>> + crop = v4l2_subdev_state_get_crop(state, IMX908_SOURCE_PAD);
>> +
>> + u32 min_vblank = imx908_calc_min_vblank(crop);
>> + u32 max_vblank = IMX908_VMAX_MAX - format->height;
>> + u32 vblank = clamp_t(u32, ctrl->val, min_vblank, max_vblank);
>
> The above 3 lines are redundant for simple VBLANK control updates. V4L2
> control framework would ensure new values are within the defined min/max
> range before the drivers .s_ctrl is triggered.
Ok.
>> +
>> + u32 vmax = format->height + vblank;
>> + u32 max_exposure = vmax - IMX908_MIN_SHR0;
>> + u32 current_exposure = clamp_t(u32, imx->ctrls.exposure->cur.val,
>> + IMX908_EXPOSURE_MIN, max_exposure);
>
> Same here. No reason the exposure->cur.val would be out of bounds when user
> requests an update to VBLANK value. So this should be:
>
> u32 current_exposure = imx->ctrls.exposure->cur.val
See my reply to Laurent's comment.
>> +
>> + ret = __v4l2_ctrl_modify_range(imx->ctrls.exposure,
>> + IMX908_EXPOSURE_MIN,
>> + max_exposure,
>> + IMX908_EXPOSURE_STEP,
>> + current_exposure);
>> + if (ret)
>> + return ret;
>> + }
>> +
>> + /* Hardware writes only when powered; cached ctrls applied on resume */
>> + if (!pm_runtime_get_if_active(imx->dev))
>> + return 0;
>> +
>> + switch (ctrl->id) {
>> + case V4L2_CID_EXPOSURE:
>> + ret = imx908_set_exposure_lines(imx, ctrl->val);
>> + break;
>> +
>> + case V4L2_CID_ANALOGUE_GAIN:
>> + cci_write(imx->cci, IMX908_REG_GAIN, ctrl->val, &ret);
>> + break;
>> +
>> + case V4L2_CID_VBLANK: {
>> + u32 vmax = format->height + ctrl->val;
>> +
>> + ret = cci_write(imx->cci, IMX908_REG_VMAX, vmax, NULL);
>> + /* SHR0 derived from VMAX, re-apply exposure after changes */
>> + if (!ret)
>> + ret = imx908_set_exposure_lines(imx,
>> + imx->ctrls.exposure->val);
>> + break;
>> + }
>> +
>> + case V4L2_CID_HBLANK: {
>> + u16 hmax = imx908_calc_hmax(format->width, ctrl->val);
>> +
>> + cci_write(imx->cci, IMX908_REG_HMAX, hmax, &ret);
>> + break;
>> + }
>> +
>> + case V4L2_CID_TEST_PATTERN:
>> + ret = imx908_update_test_pattern(imx, ctrl->val);
>> + break;
>> +
>> + case V4L2_CID_HFLIP:
>> + cci_write(imx->cci, IMX908_REG_HREVERSE,
>> + ctrl->val ? IMX908_HREVERSE_INV
>> + : IMX908_HREVERSE_NORMAL, &ret);
>> + break;
>> +
>> + case V4L2_CID_VFLIP:
>> + cci_write(imx->cci, IMX908_REG_VREVERSE,
>> + ctrl->val ? IMX908_VREVERSE_INV
>> + : IMX908_VREVERSE_NORMAL, &ret);
>> + break;
>> +
>
> Any reason the register values for HREVERSE/VREVERSE are inverted now w.r.t
> the userspace HFLIP/VFLIP settings?
>
> Ah, and is this why imx908_active_area.top was 0 and not 24 ?
>
> What v2 did (writing ctrl->val directly to the registers) seemed okay to
> me. It's usually a good idea to mention functional changes in the
> changelog. If there is a valid reason for the swap, maybe also add a code
> comment here explaining why.
See reply to Laurent's additional comment.
>> + default:
>> + dev_warn(imx->dev,
>> + "ctrl(id:0x%x,val:0x%x) is not handled\n",
>> + ctrl->id, ctrl->val);
>> + break;
>> + }
>> +
>> + pm_runtime_put(imx->dev);
>> +
>> + return ret;
>> +}
>> +
>> +static const struct v4l2_ctrl_ops imx908_ctrl_ops = {
>> + .s_ctrl = imx908_set_ctrl,
>> +};
>> +
>> +/* ------------------------ Pad/video ops ------------------ */
>> +static int imx908_enum_mbus_code(struct v4l2_subdev *sd,
>> + struct v4l2_subdev_state *sd_state,
>> + struct v4l2_subdev_mbus_code_enum *code)
>> +{
>> + if (code->index >= ARRAY_SIZE(imx908_mbus_codes))
>> + return -EINVAL;
>> + code->code = imx908_mbus_codes[code->index];
>> + return 0;
>> +}
>> +
>> +static int imx908_enum_frame_size(struct v4l2_subdev *sd,
>> + struct v4l2_subdev_state *sd_state,
>> + struct v4l2_subdev_frame_size_enum *fse)
>> +{
>> + struct imx908 *imx = to_imx908(sd);
>> + const struct v4l2_rect *crop;
>> +
>> + if (fse->index > 0)
>> + return -EINVAL;
>> +
>> + if (!imx908_mbus_code_supported(imx, fse->code))
>> + return -EINVAL;
>> +
>> + crop = v4l2_subdev_state_get_crop(sd_state, IMX908_SOURCE_PAD);
>> +
>> + fse->min_width = crop->width;
>> + fse->max_width = crop->width;
>> + fse->min_height = crop->height;
>> + fse->max_height = crop->height;
>> +
>> + return 0;
>> +}
>> +
>> +static int imx908_set_pad_format(struct v4l2_subdev *sd,
>> + struct v4l2_subdev_state *sd_state,
>> + struct v4l2_subdev_format *fmt)
>> +{
>> + struct imx908 *imx = to_imx908(sd);
>> + struct v4l2_mbus_framefmt *format;
>> +
>> + if (fmt->which == V4L2_SUBDEV_FORMAT_ACTIVE &&
>> + v4l2_subdev_is_streaming(sd))
>> + return -EBUSY;
>> +
>> + format = v4l2_subdev_state_get_format(sd_state, fmt->pad);
>> +
>> + if (imx908_mbus_code_supported(imx, fmt->format.code))
>> + format->code = fmt->format.code;
>> + else
>> + format->code = IMX908_DEFAULT_MBUS_CODE;
>> +
>> + /* Fully define metadata */
>> + format->field = V4L2_FIELD_NONE;
>> + format->colorspace = V4L2_COLORSPACE_RAW;
>> + format->ycbcr_enc = V4L2_YCBCR_ENC_DEFAULT;
>> + format->quantization = V4L2_QUANTIZATION_FULL_RANGE;
>> + format->xfer_func = V4L2_XFER_FUNC_NONE;
>> +
>> + /* Limits track the real controls only for the ACTIVE state, not TRY */
>> + if (fmt->which == V4L2_SUBDEV_FORMAT_ACTIVE) {
>> + /* set_pad_format() never changes width, so old == new here */
>> + u32 old_width = format->width;
>> + u32 old_height = format->height;
>> + int ret = imx908_update_framing_limits(imx, sd_state,
>> + old_width, old_height);
>> +
>> + if (ret)
>> + return ret;
>> + }
>
> I see program_window() only looks at crop size, so changing format
> width/height to something else using S_FMT will not change anything on the
> CSI-2 bus.
>
> For freely-configurable sensor drivers like these, I think the usual
> pattern is to allow S_SELECTION to configure the window size, and S_FMT
> always snaps format->width = crop->width (same for height).
See reply to Laurent's additional comment.
>> +
>> + fmt->format = *format;
>> +
>> + return 0;
>> +}
>> +
>> +static int imx908_set_selection(struct v4l2_subdev *sd,
>> + struct v4l2_subdev_state *sd_state,
>> + struct v4l2_subdev_selection *sel)
>> +{
>> + struct imx908 *imx = to_imx908(sd);
>> + struct v4l2_mbus_framefmt *format;
>> + struct v4l2_rect *r = &sel->r;
>> + struct v4l2_rect *crop;
>> +
>> + if (sel->target != V4L2_SEL_TGT_CROP)
>> + return -EINVAL;
>> +
>> + if (sel->which == V4L2_SUBDEV_FORMAT_ACTIVE &&
>> + v4l2_subdev_is_streaming(sd))
>> + return -EBUSY;
>> +
>> + crop = v4l2_subdev_state_get_crop(sd_state, sel->pad);
>> + format = v4l2_subdev_state_get_format(sd_state, sel->pad);
>> +
>> + /*
>> + * Round the crop rectangle size down to the hardware alignment
>> + * constraints, then clamp it to the supported size range.
>> + */
>> + r->width = clamp_t(s32, ALIGN_DOWN(r->width, IMX908_CROP_ALIGN_HWIDTH),
>> + IMX908_CROP_MIN_WIDTH, imx908_active_area.width);
>> + r->height = clamp_t(s32, ALIGN_DOWN(r->height, IMX908_CROP_ALIGN_VWIDTH),
>> + IMX908_CROP_MIN_HEIGHT, imx908_active_area.height);
>> +
>> + /*
>> + * Round and clamp the crop position similarly, with the maximum value
>> + * chosen so that the crop rectangle remains fully inside the active
>> + * pixel array.
>> + */
>> + r->left = clamp_t(s32, ALIGN_DOWN(r->left, IMX908_CROP_ALIGN_HSTART), 0,
>> + imx908_active_area.width - r->width);
>> + r->top = clamp_t(s32, ALIGN_DOWN(r->top, IMX908_CROP_ALIGN_VSTART), 0,
>> + imx908_active_area.height - r->height);
>> +
>> + u32 old_width = crop->width;
>> + u32 old_height = crop->height;
>> +
>> + *crop = *r;
>> +
>> + /* IMX908 has no binning, so the output size matches the crop 1:1 */
>> + format->width = crop->width;
>> + format->height = crop->height;
>> +
>> + if (sel->which == V4L2_SUBDEV_FORMAT_ACTIVE) {
>> + int ret = imx908_update_framing_limits(imx, sd_state,
>> + old_width, old_height);
>> +
>> + if (ret)
>> + return ret;
>> + }
>> +
>> + return 0;
>> +}
>> +
>> +static int imx908_get_selection(struct v4l2_subdev *sd,
>> + struct v4l2_subdev_state *sd_state,
>> + struct v4l2_subdev_selection *sel)
>> +{
>> + switch (sel->target) {
>> + case V4L2_SEL_TGT_CROP:
>> + sel->r = *v4l2_subdev_state_get_crop(sd_state,
>> + IMX908_SOURCE_PAD);
>> + return 0;
>> +
>> + case V4L2_SEL_TGT_NATIVE_SIZE:
>> + sel->r = imx908_total_area;
>> + return 0;
>> +
>> + case V4L2_SEL_TGT_CROP_DEFAULT:
>> + sel->r = imx908_active_area;
>> + return 0;
>> +
>> + case V4L2_SEL_TGT_CROP_BOUNDS:
>> + sel->r = imx908_active_area;
>> + return 0;
>
> nitpick: can be simplified as
>
> case V4L2_SEL_TGT_CROP_DEFAULT:
> case V4L2_SEL_TGT_CROP_BOUNDS:
> sel->r = imx908_active_area;
> return 0;
Ok.
>> + default:
>> + return -EINVAL;
>> + }
>> +}
>> +
>> +static int imx908_enable_streams(struct v4l2_subdev *sd,
>> + struct v4l2_subdev_state *sd_state,
>> + u32 pad,
>> + u64 streams_mask)
>> +{
>> + struct imx908 *imx = to_imx908(sd);
>> + int ret;
>> +
>> + ret = pm_runtime_resume_and_get(imx->dev);
>> + if (ret)
>> + return ret;
>> +
>> + ret = imx908_start_streaming(imx, sd_state);
>> + if (ret) {
>> + pm_runtime_put_autosuspend(imx->dev);
>> + return ret;
>> + }
>> +
>> + return 0;
>> +}
>> +
>> +static int imx908_disable_streams(struct v4l2_subdev *sd,
>> + struct v4l2_subdev_state *sd_state,
>> + u32 pad,
>> + u64 streams_mask)
>> +{
>> + struct imx908 *imx = to_imx908(sd);
>> + int ret;
>> +
>> + ret = imx908_stop_streaming(imx);
>> +
>> + pm_runtime_put_autosuspend(imx->dev);
>> +
>> + return ret;
>> +}
>> +
>> +static int imx908_init_state(struct v4l2_subdev *sd,
>> + struct v4l2_subdev_state *sd_state)
>> +{
>> + struct v4l2_subdev_selection sel = {
>> + .which = V4L2_SUBDEV_FORMAT_TRY,
>> + .pad = IMX908_SOURCE_PAD,
>> + .target = V4L2_SEL_TGT_CROP,
>> + .r = imx908_active_area,
>> + };
>> + struct v4l2_subdev_format fmt = {
>> + .which = V4L2_SUBDEV_FORMAT_TRY,
>> + .pad = IMX908_SOURCE_PAD,
>> + .format = {
>> + .code = IMX908_DEFAULT_MBUS_CODE,
>> + .width = imx908_active_area.width,
>> + .height = imx908_active_area.height,
>> + },
>> + };
>> +
>> + imx908_set_selection(sd, sd_state, &sel);
>> + imx908_set_pad_format(sd, sd_state, &fmt);
>> +
>> + return 0;
>> +}
>> +
>> +static const struct v4l2_subdev_video_ops imx908_video_ops = {
>> + .s_stream = v4l2_subdev_s_stream_helper,
>> +};
>> +
>> +static const struct v4l2_subdev_pad_ops imx908_pad_ops = {
>> + .enum_mbus_code = imx908_enum_mbus_code,
>> + .enum_frame_size = imx908_enum_frame_size,
>> + .get_fmt = v4l2_subdev_get_fmt,
>> + .set_fmt = imx908_set_pad_format,
>> + .get_selection = imx908_get_selection,
>> + .set_selection = imx908_set_selection,
>> + .enable_streams = imx908_enable_streams,
>> + .disable_streams = imx908_disable_streams,
>> +};
>> +
>> +static const struct v4l2_subdev_internal_ops imx908_internal_ops = {
>> + .init_state = imx908_init_state,
>> +};
>> +
>> +static const struct v4l2_subdev_ops imx908_subdev_ops = {
>> + .video = &imx908_video_ops,
>> + .pad = &imx908_pad_ops,
>> +};
>> +
>> +/* ----------------------- Power management ---------------------- */
>> +static int imx908_power_on(struct device *dev)
>> +{
>> + struct v4l2_subdev *sd = dev_get_drvdata(dev);
>> + struct imx908 *imx = to_imx908(sd);
>> + int ret;
>> +
>> + ret = regulator_bulk_enable(ARRAY_SIZE(imx908_supply_names),
>> + imx->supplies);
>> + if (ret) {
>> + dev_err(imx->dev, "failed to enable regulators\n");
>> + return ret;
>> + }
>> + /* XCLR already asserted by GPIOD_OUT_HIGH */
>> + udelay(1); /* >= 500ns T_low */
>> + gpiod_set_value_cansleep(imx->reset_gpio, 0); /* Sensor start */
>> +
>> + ret = clk_prepare_enable(imx->xclk);
>> + if (ret) {
>> + dev_err(imx->dev, "failed to enable xclk: %d\n", ret);
>> + goto err_reset;
>> + }
>> +
>> + fsleep(20); /* T_1 >= 20us before first SDA/SCL */
>> +
>> + return 0;
>> +
>> +err_reset:
>> + gpiod_set_value_cansleep(imx->reset_gpio, 1); /* assert reset */
>> + regulator_bulk_disable(ARRAY_SIZE(imx908_supply_names), imx->supplies);
>> + return ret;
>> +}
>> +
>> +static void imx908_power_off(struct device *dev)
>> +{
>> + struct v4l2_subdev *sd = dev_get_drvdata(dev);
>> + struct imx908 *imx = to_imx908(sd);
>> +
>> + clk_disable_unprepare(imx->xclk);
>> + gpiod_set_value_cansleep(imx->reset_gpio, 1);
>> + regulator_bulk_disable(ARRAY_SIZE(imx908_supply_names), imx->supplies);
>> +}
>> +
>> +static int imx908_runtime_resume(struct device *dev)
>> +{
>> + return imx908_power_on(dev);
>> +}
>> +
>> +static int imx908_runtime_suspend(struct device *dev)
>> +{
>> + imx908_power_off(dev);
>> + return 0;
>> +}
>> +
>
> You could use imx908_power_off and imx908_power_on directly as the
> RUNTIME_PM_OPS below.
See comment after Laurent's additional comment.
>> +static DEFINE_RUNTIME_DEV_PM_OPS(imx908_pm_ops,
>> + imx908_runtime_suspend,
>> + imx908_runtime_resume, NULL);
>> +
>> +/* ------------------------------- Probe/remove --------------------------- */
>> +
>> +static int imx908_get_inck_sel(struct imx908 *imx, u32 xclk_freq)
>> +{
>> + for (unsigned int i = 0; i < ARRAY_SIZE(imx908_inck_table); i++) {
>> + if (imx908_inck_table[i] == xclk_freq) {
>> + imx->inck_sel = i;
>> + return 0;
>> + }
>> + }
>> +
>> + return dev_err_probe(imx->dev, -EINVAL, "unsupported xclk %u Hz\n",
>> + xclk_freq);
>> +}
>> +
>> +static int imx908_get_regulators(struct imx908 *imx)
>> +{
>> + for (unsigned int i = 0; i < ARRAY_SIZE(imx908_supply_names); i++)
>> + imx->supplies[i].supply = imx908_supply_names[i];
>> +
>> + return devm_regulator_bulk_get(imx->dev,
>> + ARRAY_SIZE(imx908_supply_names),
>> + imx->supplies);
>> +}
>> +
>> +static int imx908_parse_fwnode(struct imx908 *imx)
>> +{
>> + struct v4l2_fwnode_endpoint bus_cfg = {
>> + .bus_type = V4L2_MBUS_CSI2_DPHY
>> + };
>> + struct fwnode_handle *ep;
>> + int ret = 0;
>> +
>> + ep = fwnode_graph_get_next_endpoint(dev_fwnode(imx->dev), NULL);
>> +
>> + /* Only data-lanes and link-frequencies are used from the endpoint */
>> + ret = v4l2_fwnode_endpoint_alloc_parse(ep, &bus_cfg);
>> + fwnode_handle_put(ep);
>> + if (ret)
>> + return ret;
>> +
>> + imx->num_lanes = bus_cfg.bus.mipi_csi2.num_data_lanes;
>> +
>> + if (imx->num_lanes != 2 && imx->num_lanes != 4) {
>> + dev_err(imx->dev,
>> + "only 2 or 4 CSI-2 data lanes are supported (got %u)\n",
>> + imx->num_lanes);
>> + ret = -EINVAL;
>> + goto out_free;
>> + }
>> +
>> + ret = v4l2_link_freq_to_bitmap(imx->dev,
>> + bus_cfg.link_frequencies,
>> + bus_cfg.nr_of_link_frequencies,
>> + imx908_link_freqs,
>> + ARRAY_SIZE(imx908_link_freqs),
>> + &imx->link_freq_bitmap);
>> + if (ret)
>> + goto out_free;
>
> The above two lines could be dropped.
See reply to Laurent's additional comment.
>> +
>> + imx->link_freq_idx = __ffs(imx->link_freq_bitmap);
>> + dev_dbg(imx->dev, "using %u lanes at link freq %llu Hz\n",
>> + imx->num_lanes, imx908_link_freqs[imx->link_freq_idx]);
>> +
>> +out_free:
>> + v4l2_fwnode_endpoint_free(&bus_cfg);
>> + return ret;
>> +}
>> +
>> +static int imx908_init_controls(struct imx908 *imx)
>> +{
>> + struct v4l2_ctrl_handler *hdl = &imx->ctrls.handler;
>> + struct v4l2_fwnode_device_properties props;
>> + struct v4l2_ctrl *ctrl;
>> + int ret;
>> +
>> + /* Read rotation and orientation properties from the firmware node */
>> + ret = v4l2_fwnode_device_parse(imx->dev, &props);
>> + if (ret)
>> + return ret;
>> +
>> + ret = v4l2_ctrl_handler_init(hdl, 11);
>> + if (ret)
>> + return ret;
>> +
>> + /* PIXEL_RATE is always read-only, so it needs no s_ctrl() handling */
>> + v4l2_ctrl_new_std(hdl, NULL, V4L2_CID_PIXEL_RATE,
>> + IMX908_PIXEL_RATE, IMX908_PIXEL_RATE, 1,
>> + IMX908_PIXEL_RATE);
>> +
>> + /* LINK_FREQ is informational only and has no s_ctrl() handling */
>> + ctrl = v4l2_ctrl_new_int_menu(hdl, NULL, V4L2_CID_LINK_FREQ,
>> + ARRAY_SIZE(imx908_link_freqs) - 1,
>> + imx->link_freq_idx, imx908_link_freqs);
>> + if (ctrl)
>> + ctrl->flags |= V4L2_CTRL_FLAG_READ_ONLY;
>> +
>> + u32 min_vblank = imx908_calc_min_vblank(&imx908_active_area);
>> + u32 max_vblank = IMX908_VMAX_MAX - imx908_active_area.height;
>> + /* Default to the datasheet 30 fps operating point */
>> + u32 def_vblank = IMX908_VMAX_DEFAULT - imx908_active_area.height;
>> +
>> + imx->ctrls.vblank = v4l2_ctrl_new_std(hdl, &imx908_ctrl_ops,
>> + V4L2_CID_VBLANK, min_vblank,
>> + max_vblank, 1, def_vblank);
>> +
>> + u8 bpp = imx908_bits_per_pixel(IMX908_DEFAULT_MBUS_CODE);
>> + u16 min_hmax = imx908_calc_min_hmax(imx, bpp);
>> + u32 min_hblank = imx908_hmax_to_hblank(min_hmax,
>> + imx908_active_area.width);
>> + u32 max_hblank = imx908_hmax_to_hblank(IMX908_HMAX_MAX,
>> + imx908_active_area.width);
>> + u32 hblank = imx908_hmax_to_hblank(IMX908_HMAX_DEFAULT,
>> + imx908_active_area.width);
>> +
>> + /* Default HMAX can be infeasible at low link freqs; clamp into range */
>> + hblank = clamp_t(u32, hblank, min_hblank, max_hblank);
>> +
>> + imx->ctrls.hblank = v4l2_ctrl_new_std(hdl, &imx908_ctrl_ops,
>> + V4L2_CID_HBLANK, min_hblank,
>> + max_hblank, IMX908_PIX_PER_CLK,
>> + hblank);
>> + u32 max_exp = IMX908_VMAX_DEFAULT - IMX908_MIN_SHR0;
>> +
>> + imx->ctrls.exposure = v4l2_ctrl_new_std(hdl, &imx908_ctrl_ops,
>> + V4L2_CID_EXPOSURE,
>> + IMX908_EXPOSURE_MIN,
>> + max_exp,
>> + IMX908_EXPOSURE_STEP,
>> + max_exp / 2);
>> +
>> + v4l2_ctrl_new_std(hdl, &imx908_ctrl_ops, V4L2_CID_ANALOGUE_GAIN,
>> + IMX908_ANA_GAIN_MIN, IMX908_ANA_GAIN_MAX,
>> + IMX908_ANA_GAIN_STEP, IMX908_ANA_GAIN_DEFAULT);
>> +
>> + /* Set test pattern. Menu (13 entries: Disabled + 12 patterns) */
>> + v4l2_ctrl_new_std_menu_items(hdl, &imx908_ctrl_ops,
>> + V4L2_CID_TEST_PATTERN,
>> + ARRAY_SIZE(imx908_tpg_menu) - 1, 0, 0,
>> + imx908_tpg_menu);
>> +
>> + v4l2_ctrl_new_std(hdl, &imx908_ctrl_ops, V4L2_CID_HFLIP, 0, 1, 1, 0);
>> + v4l2_ctrl_new_std(hdl, &imx908_ctrl_ops, V4L2_CID_VFLIP, 0, 1, 1, 0);
>> +
>> + v4l2_ctrl_new_fwnode_properties(hdl, &imx908_ctrl_ops, &props);
>> + if (hdl->error) {
>> + ret = hdl->error;
>> + goto err_free;
>
> Instead of the goto you could directly return here:
>
> v4l2_ctrl_handler_free(hdl);
> return ret;
>
Agreed.
>> + }
>> +
>> + imx->sd.ctrl_handler = hdl;
>> +
>> + return 0;
>> +
>> +err_free:
>> + v4l2_ctrl_handler_free(hdl);
>> + return ret;
>> +}
>> +
>> +static int imx908_identify_model(struct imx908 *imx)
>> +{
>> + u16 chip_id;
>> + u64 val;
>> + int ret;
>> +
>> + /*
>> + * The TYPE_ID registers are not accessible after power-up while
>> + * the device remains in standby. Exit standby and wait for the
>> + * required stabilization period before reading the chip ID.
>> + */
>> + ret = cci_write(imx->cci, IMX908_REG_STANDBY, IMX908_STANDBY_CANCEL, NULL);
>> + if (ret)
>> + return ret;
>> +
>> + fsleep(24 * USEC_PER_MSEC); /* >=24 ms stabilization after standby cancel */
>> +
>> + ret = cci_read(imx->cci, IMX908_REG_TYPE_ID, &val, NULL);
>> + if (ret)
>> + return ret;
>> +
>> + chip_id = val;
>> + dev_dbg(imx->dev, "IMX908 chip ID: 0x%04x\n", chip_id);
>> +
>> + if (chip_id != IMX908_CHIP_ID) {
>> + dev_err(imx->dev, "unexpected chip ID 0x%04x (expected 0x%04x)\n",
>> + chip_id, IMX908_CHIP_ID);
>> + return -ENXIO;
>> + }
>> +
>> + /* Set to standby mode to save power */
>> + ret = cci_write(imx->cci, IMX908_REG_STANDBY, IMX908_STANDBY_EN, NULL);
>> + if (ret) {
>> + dev_err(imx->dev, "failed to enter standby state: %d\n", ret);
>> + return ret;
>> + }
>> +
>> + return 0;
>> +}
>> +
>> +static int imx908_probe(struct i2c_client *client)
>> +{
>> + struct imx908 *imx;
>> + int ret;
>> +
>> + imx = devm_kzalloc(&client->dev, sizeof(*imx), GFP_KERNEL);
>> + if (!imx)
>> + return -ENOMEM;
>> + imx->dev = &client->dev;
>> +
>> + v4l2_i2c_subdev_init(&imx->sd, client, &imx908_subdev_ops);
>> + imx->sd.internal_ops = &imx908_internal_ops;
>> +
>> + imx->cci = devm_cci_regmap_init_i2c(client, 16);
>> + if (IS_ERR(imx->cci))
>> + return dev_err_probe(imx->dev, PTR_ERR(imx->cci),
>> + "CCI regmap init failed\n");
>> +
>> + imx->xclk = devm_v4l2_sensor_clk_get(imx->dev, NULL);
>> + if (IS_ERR(imx->xclk))
>> + return dev_err_probe(imx->dev, PTR_ERR(imx->xclk),
>> + "failed to get clock\n");
>> +
>> + ret = imx908_get_inck_sel(imx, clk_get_rate(imx->xclk));
>> + if (ret)
>> + return ret;
>> +
>> + imx->reset_gpio = devm_gpiod_get_optional(imx->dev, "reset",
>> + GPIOD_OUT_HIGH);
>> +
>> + if (IS_ERR(imx->reset_gpio))
>> + return dev_err_probe(imx->dev, PTR_ERR(imx->reset_gpio),
>> + "failed to get reset gpio\n");
>> +
>> + ret = imx908_get_regulators(imx);
>> + if (ret)
>> + return dev_err_probe(imx->dev, ret, "failed to get regulators\n");
>> +
>> + ret = imx908_parse_fwnode(imx);
>> + if (ret)
>> + return dev_err_probe(imx->dev, ret, "device tree parse failed\n");
>> +
>> + ret = imx908_power_on(imx->dev);
>> + if (ret)
>> + return dev_err_probe(imx->dev, ret, "power-on failed\n");
>> +
>> + ret = imx908_identify_model(imx);
>> + if (ret) {
>> + dev_err(imx->dev, "failed to identify model: %d\n", ret);
>> + goto err_power_off;
>> + }
>> +
>> + /* Device is powered; keep it resumed across registration, release at end */
>> + pm_runtime_set_active(imx->dev);
>> + pm_runtime_get_noresume(imx->dev);
>> + ret = devm_pm_runtime_enable(imx->dev);
>> + if (ret)
>> + goto err_pm_put;
>> +
>> + pm_runtime_set_autosuspend_delay(imx->dev, 1000);
>> + pm_runtime_use_autosuspend(imx->dev);
>> +
>> + ret = imx908_init_controls(imx);
>> + if (ret)
>> + goto err_pm_put;
>> +
>> + imx->sd.flags |= V4L2_SUBDEV_FL_HAS_DEVNODE;
>> + imx->sd.entity.function = MEDIA_ENT_F_CAM_SENSOR;
>> + imx->pad.flags = MEDIA_PAD_FL_SOURCE;
>> + ret = media_entity_pads_init(&imx->sd.entity, 1, &imx->pad);
>> + if (ret)
>> + goto err_hdl;
>> +
>> + /* Share the ctrl handler lock so s_ctrl can access the locked state */
>> + imx->sd.state_lock = imx->ctrls.handler.lock;
>> +
>> + ret = v4l2_subdev_init_finalize(&imx->sd);
>> + if (ret)
>> + goto err_entity;
>> +
>> + ret = v4l2_async_register_subdev_sensor(&imx->sd);
>> + if (ret)
>> + goto err_subdev;
>> +
>> + /* Balance get_noresume; allow the sensor to suspend after probe */
>> + pm_runtime_put_autosuspend(imx->dev);
>> +
>> + return 0;
>> +
>> +err_subdev:
>> + v4l2_subdev_cleanup(&imx->sd);
>> +
>> +err_entity:
>> + media_entity_cleanup(&imx->sd.entity);
>> +
>> +err_hdl:
>> + v4l2_ctrl_handler_free(imx->sd.ctrl_handler);
>> +
>> +err_pm_put:
>> + pm_runtime_put_noidle(imx->dev);
>> +
>> +err_power_off:
>> + imx908_power_off(imx->dev);
>> +
>> + return ret;
>> +}
>> +
>> +static void imx908_remove(struct i2c_client *client)
>> +{
>> + struct v4l2_subdev *sd = i2c_get_clientdata(client);
>> + struct imx908 *imx = to_imx908(sd);
>> +
>> + v4l2_async_unregister_subdev(sd);
>> + v4l2_subdev_cleanup(sd);
>> + media_entity_cleanup(&sd->entity);
>> + v4l2_ctrl_handler_free(sd->ctrl_handler);
>> +
>> + pm_runtime_disable(imx->dev);
>
> This should be dropped given devm_pm_runtime_enable() was used in probe().
Good catch, thank you. Dropped pm_runtime_disable() and simplified
imx908_remove() by delegating the final suspend to
pm_runtime_suspend(imx->dev), leaving runtime PM cleanup to devres.
>> + if (!pm_runtime_status_suspended(imx->dev))
>> + imx908_power_off(imx->dev);
>> + pm_runtime_set_suspended(imx->dev);
>> +}
>> +
>> +static const struct of_device_id imx908_of_ids[] = {
>> + { .compatible = "sony,imx908" },
>> + { /* sentinel */ }
>> +};
>> +MODULE_DEVICE_TABLE(of, imx908_of_ids);
>> +
>> +static struct i2c_driver imx908_i2c_driver = {
>> + .driver = {
>> + .name = "imx908",
>> + .pm = pm_ptr(&imx908_pm_ops),
>> + .of_match_table = imx908_of_ids,
>> + },
>> + .probe = imx908_probe,
>> + .remove = imx908_remove,
>> +};
>> +module_i2c_driver(imx908_i2c_driver);
>> +
>> +MODULE_AUTHOR("Lachlan Michael <Lachlan.Michael@sony.com>");
>> +MODULE_DESCRIPTION("Sony IMX908 Sensor Driver");
>> +MODULE_LICENSE("GPL");
>> --
>> 2.47.3
>>
>
> Thanks,
> Jai
>
next prev parent reply other threads:[~2026-10-01 8:36 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-08-28 6:48 [PATCH v3 0/2] Add bindings and driver for Sony IMX908 Lachlan Michael
2026-08-28 6:48 ` [PATCH v3 1/2] media: dt-bindings: imx908: Add Sony IMX908 sensor Lachlan Michael
2026-08-28 16:32 ` Conor Dooley
2026-09-29 0:28 ` Laurent Pinchart
2026-08-28 6:48 ` [PATCH v3 2/2] media: i2c: Add Sony IMX908 image sensor driver Lachlan Michael
2026-09-21 16:28 ` Jai Luthra
2026-09-29 1:26 ` Laurent Pinchart
2026-10-01 8:40 ` Lachlan Michael
2026-10-01 13:42 ` Laurent Pinchart
2026-10-01 8:36 ` Lachlan Michael [this message]
2026-10-01 12:24 ` Sakari Ailus
2026-10-01 13:31 ` Laurent Pinchart
2026-10-01 13:41 ` Laurent Pinchart
2026-10-01 13:45 ` Sakari Ailus
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=9c7e136c-f092-4a6b-ab1d-cbafdea26706@sony.com \
--to=lachlan.michael@sony.com \
--cc=Kazumi.A.Sato@sony.com \
--cc=Kengo.Hayasaka@sony.com \
--cc=Ryuichi.Tadano@sony.com \
--cc=Tim.Bird2@sony.com \
--cc=Yuji.John.Takahashi@sony.com \
--cc=conor+dt@kernel.org \
--cc=devicetree@vger.kernel.org \
--cc=hverkuil+cisco@kernel.org \
--cc=jai.luthra@ideasonboard.com \
--cc=kieran.bingham@ideasonboard.com \
--cc=krzk+dt@kernel.org \
--cc=laurent.pinchart@ideasonboard.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-media@vger.kernel.org \
--cc=mchehab@kernel.org \
--cc=robh@kernel.org \
--cc=sakari.ailus@linux.intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®