Hi Prabhakar, Thank you for the patch. On Mon, Aug 01, 2022 at 10:47:17PM +0100, Lad Prabhakar wrote: > Add MIPI CSI-2 receiver driver for Renesas RZ/G2L. The MIPI > CSI-2 is part of the CRU module found on RZ/G2L family. > > Based on a patch in the BSP by Hien Huynh > <hien.huynh.px@xxxxxxxxxxx> > > Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@xxxxxxxxxxxxxx> > --- > RFC v2 -> v1 > * Fixed initialization sequence of DPHY and link > * Exported DPHY and link initialization functions so that the > CRU core driver can initialize the CRU and CSI2 as per the HW manual. > > RFC v1 -> RFC v2 > * new patch (split up as new driver compared to v1) > --- > drivers/media/platform/renesas/Kconfig | 1 + > drivers/media/platform/renesas/Makefile | 1 + > .../media/platform/renesas/rzg2l-cru/Kconfig | 17 + > .../media/platform/renesas/rzg2l-cru/Makefile | 3 + > .../platform/renesas/rzg2l-cru/rzg2l-csi2.c | 702 ++++++++++++++++++ > .../platform/renesas/rzg2l-cru/rzg2l-csi2.h | 49 ++ > 6 files changed, 773 insertions(+) > create mode 100644 drivers/media/platform/renesas/rzg2l-cru/Kconfig > create mode 100644 drivers/media/platform/renesas/rzg2l-cru/Makefile > create mode 100644 drivers/media/platform/renesas/rzg2l-cru/rzg2l-csi2.c > create mode 100644 drivers/media/platform/renesas/rzg2l-cru/rzg2l-csi2.h > > diff --git a/drivers/media/platform/renesas/Kconfig b/drivers/media/platform/renesas/Kconfig > index 9fd90672ea2d..ed788e991f74 100644 > --- a/drivers/media/platform/renesas/Kconfig > +++ b/drivers/media/platform/renesas/Kconfig > @@ -41,6 +41,7 @@ config VIDEO_SH_VOU > Support for the Video Output Unit (VOU) on SuperH SoCs. > > source "drivers/media/platform/renesas/rcar-vin/Kconfig" > +source "drivers/media/platform/renesas/rzg2l-cru/Kconfig" > > # Mem2mem drivers > > diff --git a/drivers/media/platform/renesas/Makefile b/drivers/media/platform/renesas/Makefile > index 3ec226ef5fd2..55854e868887 100644 > --- a/drivers/media/platform/renesas/Makefile > +++ b/drivers/media/platform/renesas/Makefile > @@ -4,6 +4,7 @@ > # > > obj-y += rcar-vin/ > +obj-y += rzg2l-cru/ > obj-y += vsp1/ > > obj-$(CONFIG_VIDEO_RCAR_DRIF) += rcar_drif.o > diff --git a/drivers/media/platform/renesas/rzg2l-cru/Kconfig b/drivers/media/platform/renesas/rzg2l-cru/Kconfig > new file mode 100644 > index 000000000000..150bf16f167c > --- /dev/null > +++ b/drivers/media/platform/renesas/rzg2l-cru/Kconfig > @@ -0,0 +1,17 @@ > +# SPDX-License-Identifier: GPL-2.0 > + > +config VIDEO_RZG2L_CSI2 > + tristate "RZ/G2L MIPI CSI-2 Receiver" > + depends on V4L_PLATFORM_DRIVERS > + depends on VIDEO_DEV && OF > + depends on ARCH_RENESAS || COMPILE_TEST > + select MEDIA_CONTROLLER > + select VIDEO_V4L2_SUBDEV_API > + select RESET_CONTROLLER > + select V4L2_FWNODE Please order these alphabetically. > + help > + Support for Renesas RZ/G2L (and alike SoC's) MIPI CSI-2 > + Receiver driver. > + > + To compile this driver as a module, choose M here: the > + module will be called rzg2l-csi2. > diff --git a/drivers/media/platform/renesas/rzg2l-cru/Makefile b/drivers/media/platform/renesas/rzg2l-cru/Makefile > new file mode 100644 > index 000000000000..91ea97a944e6 > --- /dev/null > +++ b/drivers/media/platform/renesas/rzg2l-cru/Makefile > @@ -0,0 +1,3 @@ > +# SPDX-License-Identifier: GPL-2.0 > + > +obj-$(CONFIG_VIDEO_RZG2L_CSI2) += rzg2l-csi2.o > diff --git a/drivers/media/platform/renesas/rzg2l-cru/rzg2l-csi2.c b/drivers/media/platform/renesas/rzg2l-cru/rzg2l-csi2.c > new file mode 100644 > index 000000000000..252e33ddb6e8 > --- /dev/null > +++ b/drivers/media/platform/renesas/rzg2l-cru/rzg2l-csi2.c > @@ -0,0 +1,702 @@ > +// SPDX-License-Identifier: GPL-2.0 > +/* > + * Driver for Renesas RZ/G2L MIPI CSI-2 Receiver > + * > + * Copyright (C) 2022 Renesas Electronics Corp. > + */ > + > +#include <linux/clk.h> > +#include <linux/delay.h> > +#include <linux/interrupt.h> > +#include <linux/io.h> > +#include <linux/module.h> > +#include <linux/of.h> > +#include <linux/of_device.h> > +#include <linux/of_graph.h> > +#include <linux/platform_device.h> > +#include <linux/pm_runtime.h> > +#include <linux/reset.h> > +#include <linux/sys_soc.h> > +#include <linux/units.h> > + > +#include <media/v4l2-ctrls.h> > +#include <media/v4l2-device.h> > +#include <media/v4l2-fwnode.h> > +#include <media/v4l2-mc.h> > +#include <media/v4l2-subdev.h> > + > +#include "rzg2l-csi2.h" > + > +/* LINK registers */ > +/* Module Configuration Register */ > +#define CSI2nMCG 0x0 > +#define CSI2nMCG_SDLN GENMASK(11, 8) > + > +/* Module Control Register 0 */ > +#define CSI2nMCT0 0x10 > +#define CSI2nMCT0_VDLN(x) ((x) << 0) > + > +/* Module Control Register 2 */ > +#define CSI2nMCT2 0x18 > +#define CSI2nMCT2_FRRSKW(x) ((x) << 16) > +#define CSI2nMCT2_FRRCLK(x) ((x) << 0) > + > +/* Module Control Register 3 */ > +#define CSI2nMCT3 0x1c > +#define CSI2nMCT3_RXEN BIT(0) > + > +/* Reset Control Register */ > +#define CSI2nRTCT 0x28 > +#define CSI2nRTCT_VSRST BIT(0) > + > +/* Reset Status Register */ > +#define CSI2nRTST 0x2c > +#define CSI2nRTST_VSRSTS BIT(0) > + > +/* Receive Data Type Enable Low Register */ > +#define CSI2nDTEL 0x60 > + > +/* Receive Data Type Enable High Register */ > +#define CSI2nDTEH 0x64 > + > +/* Power Management Status Register */ > +#define CSI2nPMST 0x200 > + > +/* Power Management Status Clear Register */ > +#define CSI2nPMSC 0x204 > + > +/* DPHY registers */ > +/* D-PHY Control Register 0 */ > +#define CSIDPHYCTRL0 0x400 > +#define CSIDPHYCTRL0_EN_LDO1200 BIT(1) > +#define CSIDPHYCTRL0_EN_BGR BIT(0) > + > +/* D-PHY Timing Register 0 */ > +#define CSIDPHYTIM0 0x404 > +#define CSIDPHYTIM0_TCLK_MISS(x) ((x) << 24) > +#define CSIDPHYTIM0_T_INIT(x) ((x) << 0) > + > +/* D-PHY Timing Register 1 */ > +#define CSIDPHYTIM1 0x408 > +#define CSIDPHYTIM1_THS_PREPARE(x) ((x) << 24) > +#define CSIDPHYTIM1_TCLK_PREPARE(x) ((x) << 16) > +#define CSIDPHYTIM1_THS_SETTLE(x) ((x) << 8) > +#define CSIDPHYTIM1_TCLK_SETTLE(x) ((x) << 0) > + > +/* D-PHY Skew Adjustment Function */ > +#define CSIDPHYSKW0 0x460 > +#define CSIDPHYSKW0_UTIL_DL0_SKW_ADJ(x) ((x) & 0x3) > +#define CSIDPHYSKW0_UTIL_DL1_SKW_ADJ(x) (((x) & 0x3) << 4) > +#define CSIDPHYSKW0_UTIL_DL2_SKW_ADJ(x) (((x) & 0x3) << 8) > +#define CSIDPHYSKW0_UTIL_DL3_SKW_ADJ(x) (((x) & 0x3) << 12) > +#define CSIDPHYSKW0_DEFAULT_SKW CSIDPHYSKW0_UTIL_DL0_SKW_ADJ(1) | \ > + CSIDPHYSKW0_UTIL_DL1_SKW_ADJ(1) | \ > + CSIDPHYSKW0_UTIL_DL2_SKW_ADJ(1) | \ > + CSIDPHYSKW0_UTIL_DL3_SKW_ADJ(1) > + > +#define VSRSTS_RETRIES 20 > + > +struct timings { Such a generic name risks creating conflicts later. Could you name the structure rzg2l_csi2_timings ? Same for the dphy_timings enum. > + u32 t_init; > + u32 tclk_miss; > + u32 tclk_settle; > + u32 ths_settle; > + u32 tclk_prepare; > + u32 ths_prepare; > +}; > + > +enum dphy_timings { > + TRANSMISSION_RATE_80_MBPS = 0, > + TRANSMISSION_RATE_125_MBPS, > + TRANSMISSION_RATE_250_MBPS, > + TRANSMISSION_RATE_360_MBPS, > + TRANSMISSION_RATE_360_PLUS_MBPS, > +}; > + > +static const struct timings global_timings[] = { And there, rzg2l_csi2_global_timings. > + [TRANSMISSION_RATE_80_MBPS] = { > + .t_init = 79801, > + .tclk_miss = 4, > + .tclk_settle = 23, > + .ths_settle = 31, > + .tclk_prepare = 10, > + .ths_prepare = 19, > + }, > + [TRANSMISSION_RATE_125_MBPS] = { > + .t_init = 79801, > + .tclk_miss = 4, > + .tclk_settle = 23, > + .ths_settle = 28, > + .tclk_prepare = 10, > + .ths_prepare = 19, > + }, > + [TRANSMISSION_RATE_250_MBPS] = { > + .t_init = 79801, > + .tclk_miss = 4, > + .tclk_settle = 23, > + .ths_settle = 22, > + .tclk_prepare = 10, > + .ths_prepare = 16, > + }, > + [TRANSMISSION_RATE_360_MBPS] = { > + .t_init = 79801, > + .tclk_miss = 4, > + .tclk_settle = 18, > + .ths_settle = 19, > + .tclk_prepare = 10, > + .ths_prepare = 10, > + }, > + [TRANSMISSION_RATE_360_PLUS_MBPS] = { > + .t_init = 79801, > + .tclk_miss = 4, > + .tclk_settle = 18, > + .ths_settle = 18, > + .tclk_prepare = 10, > + .ths_prepare = 10, > + }, > +}; > + > +struct rzg2l_csi2_format { > + u32 code; > + unsigned int datatype; > + unsigned int bpp; > +}; > + > +static const struct rzg2l_csi2_format rzg2l_csi2_formats[] = { > + { .code = MEDIA_BUS_FMT_UYVY8_1X16, .datatype = 0x1e, .bpp = 16 }, > +}; > + > +static const struct rzg2l_csi2_format *rzg2l_csi2_code_to_fmt(unsigned int code) > +{ > + unsigned int i; > + > + for (i = 0; i < ARRAY_SIZE(rzg2l_csi2_formats); i++) > + if (rzg2l_csi2_formats[i].code == code) > + return &rzg2l_csi2_formats[i]; > + > + return NULL; > +} > + > +static inline struct rzg2l_csi2 *notifier_to_csi2(struct v4l2_async_notifier *n) > +{ > + return container_of(n, struct rzg2l_csi2, notifier); > +} > + > +static u32 rzg2l_csi2_read(struct rzg2l_csi2 *csi2, unsigned int reg) > +{ > + return ioread32(csi2->base + reg); > +} > + > +static void rzg2l_csi2_write(struct rzg2l_csi2 *csi2, unsigned int reg, > + u32 data) > +{ > + iowrite32(data, csi2->base + reg); > +} > + > +static void rzg2l_csi2_set(struct rzg2l_csi2 *csi2, unsigned int reg, u32 set) > +{ > + rzg2l_csi2_write(csi2, reg, rzg2l_csi2_read(csi2, reg) | set); > +} > + > +static void rzg2l_csi2_clr(struct rzg2l_csi2 *csi2, unsigned int reg, u32 clr) > +{ > + rzg2l_csi2_write(csi2, reg, rzg2l_csi2_read(csi2, reg) & ~clr); > +} > + > +static int rzg2l_csi2_calc_mbps(struct rzg2l_csi2 *csi2) > +{ > + const struct rzg2l_csi2_format *format; > + struct v4l2_subdev *source; > + struct v4l2_ctrl *ctrl; > + u64 mbps; > + > + if (!csi2->remote) > + return -ENODEV; Can this happen, given that the remote field is set in the bound() handler ? > + > + source = csi2->remote; > + > + /* Read the pixel rate control from remote. */ > + ctrl = v4l2_ctrl_find(source->ctrl_handler, V4L2_CID_PIXEL_RATE); > + if (!ctrl) { > + dev_err(csi2->dev, "no pixel rate control in subdev %s\n", > + source->name); > + return -EINVAL; > + } > + > + format = rzg2l_csi2_code_to_fmt(csi2->mf.code); > + > + /* > + * Calculate hsfreq in Mbps > + * hsfreq = (pixel_rate * bits_per_sample) / number_of_lanes > + */ > + mbps = v4l2_ctrl_g_ctrl_int64(ctrl) * format->bpp; > + do_div(mbps, csi2->lanes * 1000000); > + > + return mbps; > +} > + > +int rzg2l_csi2_cmn_rstb_deassert(struct rzg2l_csi2 *csi2) > +{ > + return reset_control_deassert(csi2->rstc); > +} > +EXPORT_SYMBOL_GPL(rzg2l_csi2_cmn_rstb_deassert); I don't think you need to export this, or the rzg2l_csi2_mipi_link_setting() function. They're called by the CRU driver just before or after calling the .s_stream() operation, so you can move the calls there. > + > +/* ----------------------------------------------------------------------------- > + * DPHY setting > + */ > + > +int rzg2l_csi2_dphy_setting(struct rzg2l_csi2 *csi2, bool on) > +{ > + int lanes, mbps; > + > + dev_dbg(csi2->dev, "Input size (%ux%u%c)\n", > + csi2->mf.width, csi2->mf.height, > + csi2->mf.field == V4L2_FIELD_NONE ? 'p' : 'i'); > + > + /* Checking the maximum lanes support for CSI-2 module */ > + lanes = (rzg2l_csi2_read(csi2, CSI2nMCG) & CSI2nMCG_SDLN) >> 8; > + if (lanes < csi2->lanes) { > + dev_err(csi2->dev, > + "Failed to support %d data lanes\n", csi2->lanes); > + return -EINVAL; > + } You should read and validate the number of lanes at probe time. > + > + mbps = rzg2l_csi2_calc_mbps(csi2); > + if (mbps < 0) > + return mbps; > + > + csi2->hsfreq = mbps; This is only needed in the (on) branch. > + > + if (on) { The two branches share no code. Split them in two separate functions. > + struct timings dphy_timing; Make this a const pointer. > + u32 dphytim0, dphytim1; > + > + /* Set DPHY timing parameters */ > + if (csi2->hsfreq <= 80) > + dphy_timing = global_timings[TRANSMISSION_RATE_80_MBPS]; > + else if (csi2->hsfreq <= 125) > + dphy_timing = global_timings[TRANSMISSION_RATE_125_MBPS]; > + else if (csi2->hsfreq <= 250) > + dphy_timing = global_timings[TRANSMISSION_RATE_250_MBPS]; > + else if (csi2->hsfreq <= 360) > + dphy_timing = global_timings[TRANSMISSION_RATE_360_MBPS]; > + else if (csi2->hsfreq <= 1500) > + dphy_timing = global_timings[TRANSMISSION_RATE_360_PLUS_MBPS]; > + else > + return -EINVAL; > + > + /* Set D-PHY timing parameters */ > + dphytim0 = CSIDPHYTIM0_TCLK_MISS(dphy_timing.tclk_miss) | > + CSIDPHYTIM0_T_INIT(dphy_timing.t_init); > + dphytim1 = CSIDPHYTIM1_THS_PREPARE(dphy_timing.ths_prepare) | > + CSIDPHYTIM1_TCLK_PREPARE(dphy_timing.tclk_prepare) | > + CSIDPHYTIM1_THS_SETTLE(dphy_timing.ths_settle) | > + CSIDPHYTIM1_TCLK_SETTLE(dphy_timing.tclk_settle); > + rzg2l_csi2_write(csi2, CSIDPHYTIM0, dphytim0); > + rzg2l_csi2_write(csi2, CSIDPHYTIM1, dphytim1); > + > + /* Enable D-PHY power control 0 */ > + rzg2l_csi2_write(csi2, CSIDPHYSKW0, CSIDPHYSKW0_DEFAULT_SKW); > + > + /* Set the EN_BGR bit */ > + rzg2l_csi2_set(csi2, CSIDPHYCTRL0, CSIDPHYCTRL0_EN_BGR); > + > + /* Delay 20us to be stable */ > + usleep_range(20, 40); > + > + /* Enable D-PHY power control 1 */ > + rzg2l_csi2_set(csi2, CSIDPHYCTRL0, CSIDPHYCTRL0_EN_LDO1200); > + > + /* Delay 10us to be stable */ > + usleep_range(10, 20); > + > + /* Start supplying the internal clock for the D-PHY block */ > + pm_runtime_get_sync(csi2->dev); Missing error handling (and use pm_runtime_resume_and_get() to make that easier). But it doesn't seem quite right, you're not supposed to resume the device after writing register, but before. > + } else { > + /* Reset the CRU (D-PHY) */ > + reset_control_assert(csi2->rstc); > + > + /* Stop the D-PHY clock */ > + pm_runtime_put(csi2->dev); Same here, that should to after the register writes. > + > + /* Cancel the EN_LDO1200 register setting */ > + rzg2l_csi2_clr(csi2, CSIDPHYCTRL0, CSIDPHYCTRL0_EN_LDO1200); > + > + /* Cancel the EN_BGR register setting */ > + rzg2l_csi2_clr(csi2, CSIDPHYCTRL0, CSIDPHYCTRL0_EN_BGR); > + } > + > + return 0; > +} > +EXPORT_SYMBOL_GPL(rzg2l_csi2_dphy_setting); Would it be a problem to also move this call to .s_stream() to avoid direct calls from the CRU driver ? > + > +void rzg2l_csi2_mipi_link_setting(struct rzg2l_csi2 *csi2, bool on) > +{ > + if (on) { Here too the two branches share no code, split them in two separate functions. > + unsigned long vclk_rate = csi2->vclk_rate / HZ_PER_MHZ; > + u32 frrskw, frrclk, frrskw_coeff, frrclk_coeff; > + > + /* Select data lanes */ > + rzg2l_csi2_write(csi2, CSI2nMCT0, CSI2nMCT0_VDLN(csi2->lanes)); > + > + frrskw_coeff = 3 * vclk_rate * 8; > + frrclk_coeff = frrskw_coeff / 2; > + frrskw = DIV_ROUND_UP(frrskw_coeff, csi2->hsfreq); > + frrclk = DIV_ROUND_UP(frrclk_coeff, csi2->hsfreq); > + rzg2l_csi2_write(csi2, CSI2nMCT2, > + CSI2nMCT2_FRRSKW(frrskw) | CSI2nMCT2_FRRCLK(frrclk)); > + > + /* > + * Select data type. > + * FS, FE, LS, LE, Generic Short Packet Codes 1 to 8, > + * Generic Long Packet Data Types 1 to 4 YUV422 8-bit, > + * RGB565, RGB888, RAW8 to RAW20, User-defined 8-bit > + * data types 1 to 8 > + */ > + rzg2l_csi2_write(csi2, CSI2nDTEL, 0xf778ff0f); > + rzg2l_csi2_write(csi2, CSI2nDTEH, 0x00ffff1f); > + > + /* Enable LINK reception */ > + rzg2l_csi2_set(csi2, CSI2nMCT3, CSI2nMCT3_RXEN); Would it be an issue to use rzg2l_csi2_write() here ? > + } else { > + unsigned int timeout = VSRSTS_RETRIES; > + > + /* Stop LINK reception */ > + rzg2l_csi2_clr(csi2, CSI2nMCT3, CSI2nMCT3_RXEN); Same here, rzg2l_csi2_write(csi2, CSI2nMCT3, 0); Same for the other usage of rzg2l_csi2_set() and rzg2l_csi2_clr() above, to avoid having to read back the value first. > + > + /* Request a software reset of the LINK Video Pixel Interface */ > + rzg2l_csi2_write(csi2, CSI2nRTCT, CSI2nRTCT_VSRST); > + > + /* Make sure CSI2nRTST.VSRSTS bit is cleared */ > + while (timeout--) { > + if (!(rzg2l_csi2_read(csi2, CSI2nRTST) & CSI2nRTST_VSRSTS)) > + break; > + usleep_range(100, 200); > + }; > + } > +} > +EXPORT_SYMBOL_GPL(rzg2l_csi2_mipi_link_setting); > + > +static int rzg2l_csi2_s_stream(struct v4l2_subdev *sd, int enable) > +{ > + struct rzg2l_csi2 *csi2 = sd_to_csi2(sd); > + int ret = 0; > + > + mutex_lock(&csi2->lock); > + > + if (!csi2->remote) { > + ret = -ENODEV; > + goto out; > + } I don't think this can happen either. > + > + if (enable && !csi2->stream_count) { > + ret = v4l2_subdev_call(csi2->remote, video, s_stream, 1); > + if (ret) > + goto out; > + } else if (!enable && csi2->stream_count == 1) { > + v4l2_subdev_call(csi2->remote, video, s_stream, 0); > + } > + > + csi2->stream_count += enable ? 1 : -1; Drop stream_count, .s_stream() is not reference-counted. > +out: > + mutex_unlock(&csi2->lock); > + > + return ret; > +} > + > +static int rzg2l_csi2_set_format(struct v4l2_subdev *sd, > + struct v4l2_subdev_state *state, > + struct v4l2_subdev_format *format) > +{ > + struct rzg2l_csi2 *csi2 = sd_to_csi2(sd); > + struct v4l2_mbus_framefmt *framefmt; > + > + mutex_lock(&csi2->lock); > + > + if (!rzg2l_csi2_code_to_fmt(format->format.code)) > + format->format.code = rzg2l_csi2_formats[0].code; > + > + if (format->which == V4L2_SUBDEV_FORMAT_ACTIVE) { > + csi2->mf = format->format; Please use the subdev active state storage to avoid manually storing the format. See https://lore.kernel.org/linux-media/20220823005822.13805-3-laurent.pinchart@xxxxxxxxxxxxxxxx/ and in particular - How the .get_fmt() and .set_fmt() handlers use v4l2_subdev_get_pad_format() unconditionally to get the format, regardless of whether it's the active or try format. - The v4l2_subdev_lock_and_get_active_state() call in .s_stream() to get the active state, and access the format from it. - You're already calling v4l2_subdev_init_finalize() in probe, good :-) You'll need to implement .init_cfg(), and you'll be able to drop the lock field in rzg2l_csi2. > + } else { > + framefmt = v4l2_subdev_get_try_format(sd, state, 0); > + *framefmt = format->format; > + } > + > + mutex_unlock(&csi2->lock); > + > + return 0; > +} > + > +static int rzg2l_csi2_get_pad_format(struct v4l2_subdev *sd, > + struct v4l2_subdev_state *sd_state, > + struct v4l2_subdev_format *format) > +{ > + struct rzg2l_csi2 *csi2 = sd_to_csi2(sd); > + > + mutex_lock(&csi2->lock); > + if (format->which == V4L2_SUBDEV_FORMAT_ACTIVE) > + format->format = csi2->mf; > + else > + format->format = *v4l2_subdev_get_try_format(sd, sd_state, 0); > + mutex_unlock(&csi2->lock); > + > + return 0; > +} > + > +static const struct v4l2_subdev_video_ops rzg2l_csi2_video_ops = { > + .s_stream = rzg2l_csi2_s_stream, > +}; > + > +static const struct v4l2_subdev_pad_ops rzg2l_csi2_pad_ops = { > + .set_fmt = rzg2l_csi2_set_format, > + .get_fmt = rzg2l_csi2_get_pad_format, Missing enum functions for format and sizes, and missing .init_cfg(). > +}; > + > +static const struct v4l2_subdev_ops rzg2l_csi2_subdev_ops = { > + .video = &rzg2l_csi2_video_ops, > + .pad = &rzg2l_csi2_pad_ops, > +}; > + > +/* ----------------------------------------------------------------------------- > + * Async handling and registration of subdevices and links. > + */ > + > +static int rzg2l_csi2_notify_bound(struct v4l2_async_notifier *notifier, > + struct v4l2_subdev *subdev, > + struct v4l2_async_subdev *asd) > +{ > + struct rzg2l_csi2 *csi2 = notifier_to_csi2(notifier); > + int pad; > + > + pad = media_entity_get_fwnode_pad(&subdev->entity, asd->match.fwnode, > + MEDIA_PAD_FL_SOURCE); > + if (pad < 0) { > + dev_err(csi2->dev, "Failed to find pad for %s\n", subdev->name); > + return pad; > + } I think you can skip this and just hardcode pad to RZG2L_CSI2_SINK in the media_create_pad_link(), there's no other input. > + > + csi2->remote = subdev; > + > + dev_dbg(csi2->dev, "Bound %s pad: %d\n", subdev->name, pad); > + > + return media_create_pad_link(&subdev->entity, pad, > + &csi2->subdev.entity, 0, > + MEDIA_LNK_FL_ENABLED | > + MEDIA_LNK_FL_IMMUTABLE); > +} > + > +static void rzg2l_csi2_notify_unbind(struct v4l2_async_notifier *notifier, > + struct v4l2_subdev *subdev, > + struct v4l2_async_subdev *asd) > +{ > + struct rzg2l_csi2 *csi2 = notifier_to_csi2(notifier); > + > + csi2->remote = NULL; > + > + dev_dbg(csi2->dev, "Unbind %s\n", subdev->name); > +} > + > +static const struct v4l2_async_notifier_operations rzg2l_csi2_notify_ops = { > + .bound = rzg2l_csi2_notify_bound, > + .unbind = rzg2l_csi2_notify_unbind, > +}; > + > +static int rzg2l_csi2_parse_v4l2(struct rzg2l_csi2 *csi2, > + struct v4l2_fwnode_endpoint *vep) > +{ > + /* Only port 0 endpoint 0 is valid. */ > + if (vep->base.port || vep->base.id) > + return -ENOTCONN; > + > + if (vep->bus_type != V4L2_MBUS_CSI2_DPHY) { > + dev_err(csi2->dev, "Unsupported bus: %u\n", vep->bus_type); > + return -EINVAL; > + } I'm not sure you need this check as the DT bindings should make it clear that only CSI-2 sources are valid. Maybe require the bus-type property in DT ? Up to you. > + > + csi2->lanes = vep->bus.mipi_csi2.num_data_lanes; > + if (csi2->lanes != 1 && csi2->lanes != 2 && csi2->lanes != 4) { > + dev_err(csi2->dev, "Unsupported number of data-lanes: %u\n", > + csi2->lanes); > + return -EINVAL; > + } > + > + return 0; > +} > + > +static int rzg2l_csi2_parse_dt(struct rzg2l_csi2 *csi2) > +{ > + struct v4l2_fwnode_endpoint v4l2_ep = { > + .bus_type = V4L2_MBUS_CSI2_DPHY > + }; > + struct v4l2_async_subdev *asd; > + struct fwnode_handle *fwnode; > + struct fwnode_handle *ep; > + int ret; > + > + ep = fwnode_graph_get_endpoint_by_id(dev_fwnode(csi2->dev), 0, 0, 0); > + if (!ep) { > + dev_err(csi2->dev, "Not connected to subdevice\n"); > + return -EINVAL; > + } > + > + ret = v4l2_fwnode_endpoint_parse(ep, &v4l2_ep); > + if (ret) { > + dev_err(csi2->dev, "Could not parse v4l2 endpoint\n"); > + fwnode_handle_put(ep); > + return -EINVAL; > + } > + > + ret = rzg2l_csi2_parse_v4l2(csi2, &v4l2_ep); > + if (ret) { > + fwnode_handle_put(ep); > + return ret; > + } > + > + fwnode = fwnode_graph_get_remote_endpoint(ep); > + fwnode_handle_put(ep); > + > + dev_dbg(csi2->dev, "Found '%pOF'\n", to_of_node(fwnode)); > + > + v4l2_async_nf_init(&csi2->notifier); > + csi2->notifier.ops = &rzg2l_csi2_notify_ops; > + > + asd = v4l2_async_nf_add_fwnode(&csi2->notifier, fwnode, > + struct v4l2_async_subdev); > + fwnode_handle_put(fwnode); > + if (IS_ERR(asd)) > + return PTR_ERR(asd); > + > + ret = v4l2_async_subdev_nf_register(&csi2->subdev, &csi2->notifier); > + if (ret) > + v4l2_async_nf_cleanup(&csi2->notifier); > + > + return ret; > +} > + > +/* ----------------------------------------------------------------------------- > + * Platform Device Driver. > + */ > + > +static const struct media_entity_operations rzg2l_csi2_entity_ops = { > + .link_validate = v4l2_subdev_link_validate, > +}; > + > +static int rzg2l_csi2_probe(struct platform_device *pdev) > +{ > + struct rzg2l_csi2 *csi2; > + struct clk *vclk; > + unsigned int i; > + int ret; > + > + csi2 = devm_kzalloc(&pdev->dev, sizeof(*csi2), GFP_KERNEL); > + if (!csi2) > + return -ENOMEM; > + > + csi2->base = devm_platform_ioremap_resource(pdev, 0); > + if (IS_ERR(csi2->base)) > + return PTR_ERR(csi2->base); > + > + csi2->rstc = devm_reset_control_get(&pdev->dev, "cmn-rstb"); > + if (IS_ERR(csi2->rstc)) > + return dev_err_probe(&pdev->dev, PTR_ERR(csi2->rstc), > + "failed to get cpg cmn-rstb\n"); > + > + vclk = devm_clk_get(&pdev->dev, "vclk"); > + if (IS_ERR(vclk)) > + return dev_err_probe(&pdev->dev, PTR_ERR(vclk), > + "failed to get vclk clock\n"); > + > + csi2->vclk_rate = clk_get_rate(vclk); > + devm_clk_put(&pdev->dev, vclk); Is there no need to ensure at runtime at the clock is on ? > + > + csi2->dev = &pdev->dev; > + mutex_init(&csi2->lock); > + csi2->stream_count = 0; > + > + platform_set_drvdata(pdev, csi2); > + > + ret = rzg2l_csi2_parse_dt(csi2); > + if (ret) > + goto error_mutex; > + > + csi2->subdev.owner = THIS_MODULE; You don't need this, v4l2_subdev_init() handles it. > + csi2->subdev.dev = &pdev->dev; > + v4l2_subdev_init(&csi2->subdev, &rzg2l_csi2_subdev_ops); > + v4l2_set_subdevdata(&csi2->subdev, &pdev->dev); > + snprintf(csi2->subdev.name, V4L2_SUBDEV_NAME_SIZE, "%s %s", sizeof(csi2->subdev.name) instead of V4L2_SUBDEV_NAME_SIZE, it's less error-prone. > + KBUILD_MODNAME, dev_name(&pdev->dev)); Don't use KBUILD_MODNAME but set a device name here, for instance use "csi-%s" as the format string. The subdev name shouldn't change if the module is renamed. > + csi2->subdev.flags = V4L2_SUBDEV_FL_HAS_DEVNODE; > + > + csi2->subdev.entity.function = MEDIA_ENT_F_PROC_VIDEO_PIXEL_FORMATTER; I think MEDIA_ENT_F_VID_IF_BRIDGE would be more appropriate. The pixel formatter reads or writes pixels from/to memory. > + csi2->subdev.entity.ops = &rzg2l_csi2_entity_ops; > + > + csi2->pads[RZG2L_CSI2_SINK].flags = MEDIA_PAD_FL_SINK; > + for (i = RZG2L_CSI2_SOURCE_VC0; i < NR_OF_RZG2L_CSI2_PAD; i++) > + csi2->pads[i].flags = MEDIA_PAD_FL_SOURCE; One source pad only. Support for multiple virtual channels should be implemented using the stream API that is under development (see https://lore.kernel.org/linux-media/20220810121122.3149086-1-tomi.valkeinen@xxxxxxxxxxxxxxxx/T/#t). > + > + ret = media_entity_pads_init(&csi2->subdev.entity, NR_OF_RZG2L_CSI2_PAD, csi2->pads); > + if (ret) > + goto error_async; > + > + pm_runtime_enable(&pdev->dev); > + > + ret = v4l2_subdev_init_finalize(&csi2->subdev); > + if (ret < 0) > + goto error_pm; > + > + ret = v4l2_async_register_subdev(&csi2->subdev); > + if (ret < 0) > + goto error_subdev; > + > + dev_info(csi2->dev, "%d lanes found\n", csi2->lanes); > + > + return 0; > + > +error_subdev: > + v4l2_subdev_cleanup(&csi2->subdev); > +error_pm: > + pm_runtime_disable(&pdev->dev); > +error_async: > + v4l2_async_nf_unregister(&csi2->notifier); > + v4l2_async_nf_cleanup(&csi2->notifier); > +error_mutex: > + mutex_destroy(&csi2->lock); > + > + return ret; > +} > + > +static const struct of_device_id rzg2l_csi2_of_table[] = { > + { .compatible = "renesas,rzg2l-csi2", }, > + { /* sentinel */ } > +}; > + > +static int rzg2l_csi2_remove(struct platform_device *pdev) > +{ > + struct rzg2l_csi2 *csi2 = platform_get_drvdata(pdev); > + > + v4l2_async_nf_unregister(&csi2->notifier); > + v4l2_async_nf_cleanup(&csi2->notifier); > + v4l2_async_unregister_subdev(&csi2->subdev); > + v4l2_subdev_cleanup(&csi2->subdev); > + > + pm_runtime_disable(&pdev->dev); > + mutex_destroy(&csi2->lock); > + > + return 0; > +} > + > +static struct platform_driver rzg2l_csi2_pdrv = { > + .remove = rzg2l_csi2_remove, > + .probe = rzg2l_csi2_probe, > + .driver = { > + .name = "rzg2l-csi2", > + .of_match_table = rzg2l_csi2_of_table, > + }, > +}; > + > +module_platform_driver(rzg2l_csi2_pdrv); > + > +MODULE_AUTHOR("Lad Prabhakar <prabhakar.mahadev-lad.rj@xxxxxxxxxxxxxx>"); > +MODULE_DESCRIPTION("Renesas RZ/G2L MIPI CSI2 receiver driver"); > +MODULE_LICENSE("GPL"); > diff --git a/drivers/media/platform/renesas/rzg2l-cru/rzg2l-csi2.h b/drivers/media/platform/renesas/rzg2l-cru/rzg2l-csi2.h > new file mode 100644 > index 000000000000..7192923517f6 > --- /dev/null > +++ b/drivers/media/platform/renesas/rzg2l-cru/rzg2l-csi2.h > @@ -0,0 +1,49 @@ > +/* SPDX-License-Identifier: GPL-2.0+ */ > +/* > + * Copyright (C) 2022 Renesas Electronics Corp. > + */ > + > +#ifndef __RZG2L_CSI2__ > +#define __RZG2L_CSI2__ > + > +enum rzg2l_csi2_pads { > + RZG2L_CSI2_SINK = 0, > + RZG2L_CSI2_SOURCE_VC0, > + RZG2L_CSI2_SOURCE_VC1, > + RZG2L_CSI2_SOURCE_VC2, > + RZG2L_CSI2_SOURCE_VC3, > + NR_OF_RZG2L_CSI2_PAD, > +}; > + > +struct rzg2l_csi2 { > + struct device *dev; > + void __iomem *base; > + struct reset_control *rstc; > + unsigned long vclk_rate; > + > + struct v4l2_subdev subdev; > + struct media_pad pads[NR_OF_RZG2L_CSI2_PAD]; > + > + struct v4l2_async_notifier notifier; > + struct v4l2_subdev *remote; You can name this field source (or remote_source, up to you), it will be more explicit. > + > + struct v4l2_mbus_framefmt mf; > + > + struct mutex lock; > + int stream_count; > + > + unsigned short lanes; > + > + unsigned long hsfreq; > +}; > + > +static inline struct rzg2l_csi2 *sd_to_csi2(struct v4l2_subdev *sd) > +{ > + return container_of(sd, struct rzg2l_csi2, subdev); > +} > + > +int rzg2l_csi2_cmn_rstb_deassert(struct rzg2l_csi2 *csi2); > +int rzg2l_csi2_dphy_setting(struct rzg2l_csi2 *csi2, bool on); > +void rzg2l_csi2_mipi_link_setting(struct rzg2l_csi2 *csi2, bool on); > + > +#endif -- Regards, Laurent Pinchart