/*
 * rcar_du_crtc.c  --  R-Car Display Unit CRTCs
 *
 * Copyright (C) 2013-2015 Renesas Electronics Corporation
 *
 * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 */

#include <linux/clk.h>
#include <linux/mutex.h>
#include <linux/sys_soc.h>

#include <drm/drmP.h>
#include <drm/drm_atomic.h>
#include <drm/drm_atomic_helper.h>
#include <drm/drm_crtc.h>
#include <drm/drm_crtc_helper.h>
#include <drm/drm_fb_cma_helper.h>
#include <drm/drm_gem_cma_helper.h>
#include <drm/drm_plane_helper.h>

#include "rcar_du_crtc.h"
#include "rcar_du_drv.h"
#include "rcar_du_kms.h"
#include "rcar_du_plane.h"
#include "rcar_du_regs.h"
#include "rcar_du_vsp.h"

static u32 rcar_du_crtc_read(struct rcar_du_crtc *rcrtc, u32 reg)
{
        struct rcar_du_device *rcdu = rcrtc->group->dev;

        return rcar_du_read(rcdu, rcrtc->mmio_offset + reg);
}

static void rcar_du_crtc_write(struct rcar_du_crtc *rcrtc, u32 reg, u32 data)
{
        struct rcar_du_device *rcdu = rcrtc->group->dev;

        rcar_du_write(rcdu, rcrtc->mmio_offset + reg, data);
}

static void rcar_du_crtc_clr(struct rcar_du_crtc *rcrtc, u32 reg, u32 clr)
{
        struct rcar_du_device *rcdu = rcrtc->group->dev;

        rcar_du_write(rcdu, rcrtc->mmio_offset + reg,
                      rcar_du_read(rcdu, rcrtc->mmio_offset + reg) & ~clr);
}

static void rcar_du_crtc_set(struct rcar_du_crtc *rcrtc, u32 reg, u32 set)
{
        struct rcar_du_device *rcdu = rcrtc->group->dev;

        rcar_du_write(rcdu, rcrtc->mmio_offset + reg,
                      rcar_du_read(rcdu, rcrtc->mmio_offset + reg) | set);
}

static void rcar_du_crtc_clr_set(struct rcar_du_crtc *rcrtc, u32 reg,
                                 u32 clr, u32 set)
{
        struct rcar_du_device *rcdu = rcrtc->group->dev;
        u32 value = rcar_du_read(rcdu, rcrtc->mmio_offset + reg);

        rcar_du_write(rcdu, rcrtc->mmio_offset + reg, (value & ~clr) | set);
}

static int rcar_du_crtc_get(struct rcar_du_crtc *rcrtc)
{
        int ret;

        ret = clk_prepare_enable(rcrtc->clock);
        if (ret < 0)
                return ret;

        ret = clk_prepare_enable(rcrtc->extclock);
        if (ret < 0)
                goto error_clock;

        ret = rcar_du_group_get(rcrtc->group);
        if (ret < 0)
                goto error_group;

        return 0;

error_group:
        clk_disable_unprepare(rcrtc->extclock);
error_clock:
        clk_disable_unprepare(rcrtc->clock);
        return ret;
}

static void rcar_du_crtc_put(struct rcar_du_crtc *rcrtc)
{
        rcar_du_group_put(rcrtc->group);

        clk_disable_unprepare(rcrtc->extclock);
        clk_disable_unprepare(rcrtc->clock);
}

/* -----------------------------------------------------------------------------
 * Hardware Setup
 */

struct dpll_info {
        unsigned int output;
        unsigned int fdpll;
        unsigned int n;
        unsigned int m;
};

static void rcar_du_dpll_divider(struct rcar_du_crtc *rcrtc,
                                 struct dpll_info *dpll,
                                 unsigned long input,
                                 unsigned long target)
{
        unsigned long best_diff = (unsigned long)-1;
        unsigned long diff;
        unsigned int fdpll;
        unsigned int m;
        unsigned int n;

        /*
         *   fin                                 fvco        fout       fclkout
         * in --> [1/M] --> |PD| -> [LPF] -> [VCO] -> [1/P] -+-> [1/FDPLL] -> out
         *              +-> |  |                             |
         *              |                                    |
         *              +---------------- [1/N] <------------+
         *
         *      fclkout = fvco / P / FDPLL -- (1)
         *
         * fin/M = fvco/P/N
         *
         *      fvco = fin * P *  N / M -- (2)
         *
         * (1) + (2) indicates
         *
         *      fclkout = fin * N / M / FDPLL
         *
         * NOTES
         *      N       : (n + 1)
         *      M       : (m + 1)
         *      FDPLL   : (fdpll + 1)
         *      P       : 2
         *      2kHz < fvco < 4096MHz
         *
         * To minimize the jitter,
         * N : as large as possible
         * M : as small as possible
         */
        for (m = 0; m < 4; m++) {
                for (n = 119; n > 38; n--) {
                        /*
                         * This code only runs on 64-bit architectures, the
                         * unsigned long type can thus be used for 64-bit
                         * computation. It will still compile without any
                         * warning on 32-bit architectures.
                         *
                         * To optimize calculations, use fout instead of fvco
                         * to verify the VCO frequency constraint.
                         */
                        unsigned long fout = input * (n + 1) / (m + 1);

                        if (fout < 1000 || fout > 2048 * 1000 * 1000U)
                                continue;

                        for (fdpll = 1; fdpll < 32; fdpll++) {
                                unsigned long output;

                                output = fout / (fdpll + 1);
                                if (output >= 400 * 1000 * 1000)
                                        continue;

                                diff = abs((long)output - (long)target);
                                if (best_diff > diff) {
                                        best_diff = diff;
                                        dpll->n = n;
                                        dpll->m = m;
                                        dpll->fdpll = fdpll;
                                        dpll->output = output;
                                }

                                if (diff == 0)
                                        goto done;
                        }
                }
        }

done:
        dev_dbg(rcrtc->group->dev->dev,
                "output:%u, fdpll:%u, n:%u, m:%u, diff:%lu\n",
                 dpll->output, dpll->fdpll, dpll->n, dpll->m,
                 best_diff);
}

static const struct soc_device_attribute rcar_du_r8a7795_es1[] = {
        { .soc_id = "r8a7795", .revision = "ES1.*" },
        { /* sentinel */ }
};

static void rcar_du_crtc_set_display_timing(struct rcar_du_crtc *rcrtc)
{
        const struct drm_display_mode *mode = &rcrtc->crtc.state->adjusted_mode;
        struct rcar_du_device *rcdu = rcrtc->group->dev;
        unsigned long mode_clock = mode->clock * 1000;
        unsigned long clk;
        u32 value;
        u32 escr;
        u32 div;

        /*
         * Compute the clock divisor and select the internal or external dot
         * clock based on the requested frequency.
         */
        clk = clk_get_rate(rcrtc->clock);
        div = DIV_ROUND_CLOSEST(clk, mode_clock);
        div = clamp(div, 1U, 64U) - 1;
        escr = div | ESCR_DCLKSEL_CLKS;

        if (rcrtc->extclock) {
                struct dpll_info dpll = { 0 };
                unsigned long extclk;
                unsigned long extrate;
                unsigned long rate;
                u32 extdiv;

                extclk = clk_get_rate(rcrtc->extclock);
                if (rcdu->info->dpll_ch & (1 << rcrtc->index)) {
                        unsigned long target = mode_clock;

                        /*
                         * The H3 ES1.x exhibits dot clock duty cycle stability
                         * issues. We can work around them by configuring the
                         * DPLL to twice the desired frequency, coupled with a
                         * /2 post-divider. This isn't needed on other SoCs and
                         * breaks HDMI output on M3-W for a currently unknown
                         * reason, so restrict the workaround to H3 ES1.x.
                         */
                        if (soc_device_match(rcar_du_r8a7795_es1))
                                target *= 2;

                        rcar_du_dpll_divider(rcrtc, &dpll, extclk, target);
                        extclk = dpll.output;
                }

                extdiv = DIV_ROUND_CLOSEST(extclk, mode_clock);
                extdiv = clamp(extdiv, 1U, 64U) - 1;

                rate = clk / (div + 1);
                extrate = extclk / (extdiv + 1);

                if (abs((long)extrate - (long)mode_clock) <
                    abs((long)rate - (long)mode_clock)) {

                        if (rcdu->info->dpll_ch & (1 << rcrtc->index)) {
                                u32 dpllcr = DPLLCR_CODE | DPLLCR_CLKE
                                           | DPLLCR_FDPLL(dpll.fdpll)
                                           | DPLLCR_N(dpll.n) | DPLLCR_M(dpll.m)
                                           | DPLLCR_STBY;

                                if (rcrtc->index == 1)
                                        dpllcr |= DPLLCR_PLCS1
                                               |  DPLLCR_INCS_DOTCLKIN1;
                                else
                                        dpllcr |= DPLLCR_PLCS0
                                               |  DPLLCR_INCS_DOTCLKIN0;

                                rcar_du_group_write(rcrtc->group, DPLLCR,
                                                    dpllcr);
                        }

                        escr = ESCR_DCLKSEL_DCLKIN | extdiv;
                }

                dev_dbg(rcrtc->group->dev->dev,
                        "mode clock %lu extrate %lu rate %lu ESCR 0x%08x\n",
                        mode_clock, extrate, rate, escr);
        }

        rcar_du_group_write(rcrtc->group, rcrtc->index % 2 ? ESCR2 : ESCR,
                            escr);
        rcar_du_group_write(rcrtc->group, rcrtc->index % 2 ? OTAR2 : OTAR, 0);

        /* Signal polarities */
        value = ((mode->flags & DRM_MODE_FLAG_PVSYNC) ? DSMR_VSL : 0)
              | ((mode->flags & DRM_MODE_FLAG_PHSYNC) ? DSMR_HSL : 0)
              | DSMR_DIPM_DISP | DSMR_CSPM;
        rcar_du_crtc_write(rcrtc, DSMR, value);

        /* Display timings */
        rcar_du_crtc_write(rcrtc, HDSR, mode->htotal - mode->hsync_start - 19);
        rcar_du_crtc_write(rcrtc, HDER, mode->htotal - mode->hsync_start +
                                        mode->hdisplay - 19);
        rcar_du_crtc_write(rcrtc, HSWR, mode->hsync_end -
                                        mode->hsync_start - 1);
        rcar_du_crtc_write(rcrtc, HCR,  mode->htotal - 1);

        rcar_du_crtc_write(rcrtc, VDSR, mode->crtc_vtotal -
                                        mode->crtc_vsync_end - 2);
        rcar_du_crtc_write(rcrtc, VDER, mode->crtc_vtotal -
                                        mode->crtc_vsync_end +
                                        mode->crtc_vdisplay - 2);
        rcar_du_crtc_write(rcrtc, VSPR, mode->crtc_vtotal -
                                        mode->crtc_vsync_end +
                                        mode->crtc_vsync_start - 1);
        rcar_du_crtc_write(rcrtc, VCR,  mode->crtc_vtotal - 1);

        rcar_du_crtc_write(rcrtc, DESR,  mode->htotal - mode->hsync_start - 1);
        rcar_du_crtc_write(rcrtc, DEWR,  mode->hdisplay);
}

void rcar_du_crtc_route_output(struct drm_crtc *crtc,
                               enum rcar_du_output output)
{
        struct rcar_du_crtc *rcrtc = to_rcar_crtc(crtc);
        struct rcar_du_device *rcdu = rcrtc->group->dev;

        /*
         * Store the route from the CRTC output to the DU output. The DU will be
         * configured when starting the CRTC.
         */
        rcrtc->outputs |= BIT(output);

        /*
         * Store RGB routing to DPAD0, the hardware will be configured when
         * starting the CRTC.
         */
        if (output == RCAR_DU_OUTPUT_DPAD0)
                rcdu->dpad0_source = rcrtc->index;
}

static unsigned int plane_zpos(struct rcar_du_plane *plane)
{
        return plane->plane.state->normalized_zpos;
}

static const struct rcar_du_format_info *
plane_format(struct rcar_du_plane *plane)
{
        return to_rcar_plane_state(plane->plane.state)->format;
}

static void rcar_du_crtc_update_planes(struct rcar_du_crtc *rcrtc)
{
        struct rcar_du_plane *planes[RCAR_DU_NUM_HW_PLANES];
        struct rcar_du_device *rcdu = rcrtc->group->dev;
        unsigned int num_planes = 0;
        unsigned int dptsr_planes;
        unsigned int hwplanes = 0;
        unsigned int prio = 0;
        unsigned int i;
        u32 dspr = 0;

        for (i = 0; i < rcrtc->group->num_planes; ++i) {
                struct rcar_du_plane *plane = &rcrtc->group->planes[i];
                unsigned int j;

                if (plane->plane.state->crtc != &rcrtc->crtc ||
                    !plane->plane.state->visible)
                        continue;

                /* Insert the plane in the sorted planes array. */
                for (j = num_planes++; j > 0; --j) {
                        if (plane_zpos(planes[j-1]) <= plane_zpos(plane))
                                break;
                        planes[j] = planes[j-1];
                }

                planes[j] = plane;
                prio += plane_format(plane)->planes * 4;
        }

        for (i = 0; i < num_planes; ++i) {
                struct rcar_du_plane *plane = planes[i];
                struct drm_plane_state *state = plane->plane.state;
                unsigned int index = to_rcar_plane_state(state)->hwindex;

                prio -= 4;
                dspr |= (index + 1) << prio;
                hwplanes |= 1 << index;

                if (plane_format(plane)->planes == 2) {
                        index = (index + 1) % 8;

                        prio -= 4;
                        dspr |= (index + 1) << prio;
                        hwplanes |= 1 << index;
                }
        }

        /* If VSP+DU integration is enabled the plane assignment is fixed. */
        if (rcar_du_has(rcdu, RCAR_DU_FEATURE_VSP1_SOURCE)) {
                if (rcdu->info->gen < 3) {
                        dspr = (rcrtc->index % 2) + 1;
                        hwplanes = 1 << (rcrtc->index % 2);
                } else {
                        dspr = (rcrtc->index % 2) ? 3 : 1;
                        hwplanes = 1 << ((rcrtc->index % 2) ? 2 : 0);
                }
        }

        /*
         * Update the planes to display timing and dot clock generator
         * associations.
         *
         * Updating the DPTSR register requires restarting the CRTC group,
         * resulting in visible flicker. To mitigate the issue only update the
         * association if needed by enabled planes. Planes being disabled will
         * keep their current association.
         */
        mutex_lock(&rcrtc->group->lock);

        dptsr_planes = rcrtc->index % 2 ? rcrtc->group->dptsr_planes | hwplanes
                     : rcrtc->group->dptsr_planes & ~hwplanes;

        if (dptsr_planes != rcrtc->group->dptsr_planes) {
                rcar_du_group_write(rcrtc->group, DPTSR,
                                    (dptsr_planes << 16) | dptsr_planes);
                rcrtc->group->dptsr_planes = dptsr_planes;

                if (rcrtc->group->used_crtcs)
                        rcar_du_group_restart(rcrtc->group);
        }

        /* Restart the group if plane sources have changed. */
        if (rcrtc->group->need_restart)
                rcar_du_group_restart(rcrtc->group);

        mutex_unlock(&rcrtc->group->lock);

        rcar_du_group_write(rcrtc->group, rcrtc->index % 2 ? DS2PR : DS1PR,
                            dspr);
}

/* -----------------------------------------------------------------------------
 * Page Flip
 */

void rcar_du_crtc_finish_page_flip(struct rcar_du_crtc *rcrtc)
{
        struct drm_pending_vblank_event *event;
        struct drm_device *dev = rcrtc->crtc.dev;
        unsigned long flags;

        spin_lock_irqsave(&dev->event_lock, flags);
        event = rcrtc->event;
        rcrtc->event = NULL;
        spin_unlock_irqrestore(&dev->event_lock, flags);

        if (event == NULL)
                return;

        spin_lock_irqsave(&dev->event_lock, flags);
        drm_crtc_send_vblank_event(&rcrtc->crtc, event);
        wake_up(&rcrtc->flip_wait);
        spin_unlock_irqrestore(&dev->event_lock, flags);

        drm_crtc_vblank_put(&rcrtc->crtc);
}

static bool rcar_du_crtc_page_flip_pending(struct rcar_du_crtc *rcrtc)
{
        struct drm_device *dev = rcrtc->crtc.dev;
        unsigned long flags;
        bool pending;

        spin_lock_irqsave(&dev->event_lock, flags);
        pending = rcrtc->event != NULL;
        spin_unlock_irqrestore(&dev->event_lock, flags);

        return pending;
}

static void rcar_du_crtc_wait_page_flip(struct rcar_du_crtc *rcrtc)
{
        struct rcar_du_device *rcdu = rcrtc->group->dev;

        if (wait_event_timeout(rcrtc->flip_wait,
                               !rcar_du_crtc_page_flip_pending(rcrtc),
                               msecs_to_jiffies(50)))
                return;

        dev_warn(rcdu->dev, "page flip timeout\n");

        rcar_du_crtc_finish_page_flip(rcrtc);
}

/* -----------------------------------------------------------------------------
 * Start/Stop and Suspend/Resume
 */

static void rcar_du_crtc_setup(struct rcar_du_crtc *rcrtc)
{
        /* Set display off and background to black */
        rcar_du_crtc_write(rcrtc, DOOR, DOOR_RGB(0, 0, 0));
        rcar_du_crtc_write(rcrtc, BPOR, BPOR_RGB(0, 0, 0));

        /* Configure display timings and output routing */
        rcar_du_crtc_set_display_timing(rcrtc);
        rcar_du_group_set_routing(rcrtc->group);

        /* Start with all planes disabled. */
        rcar_du_group_write(rcrtc->group, rcrtc->index % 2 ? DS2PR : DS1PR, 0);

        /* Enable the VSP compositor. */
        if (rcar_du_has(rcrtc->group->dev, RCAR_DU_FEATURE_VSP1_SOURCE))
                rcar_du_vsp_enable(rcrtc);

        /* Turn vertical blanking interrupt reporting on. */
        drm_crtc_vblank_on(&rcrtc->crtc);
}

static void rcar_du_crtc_start(struct rcar_du_crtc *rcrtc)
{
        bool interlaced;

        /*
         * Select master sync mode. This enables display operation in master
         * sync mode (with the HSYNC and VSYNC signals configured as outputs and
         * actively driven).
         */
        interlaced = rcrtc->crtc.mode.flags & DRM_MODE_FLAG_INTERLACE;
        rcar_du_crtc_clr_set(rcrtc, DSYSR, DSYSR_TVM_MASK | DSYSR_SCM_MASK,
                             (interlaced ? DSYSR_SCM_INT_VIDEO : 0) |
                             DSYSR_TVM_MASTER);

        rcar_du_group_start_stop(rcrtc->group, true);
}

static void rcar_du_crtc_disable_planes(struct rcar_du_crtc *rcrtc)
{
        struct rcar_du_device *rcdu = rcrtc->group->dev;
        struct drm_crtc *crtc = &rcrtc->crtc;
        u32 status;

        /* Make sure vblank interrupts are enabled. */
        drm_crtc_vblank_get(crtc);

        /*
         * Disable planes and calculate how many vertical blanking interrupts we
         * have to wait for. If a vertical blanking interrupt has been triggered
         * but not processed yet, we don't know whether it occurred before or
         * after the planes got disabled. We thus have to wait for two vblank
         * interrupts in that case.
         */
        spin_lock_irq(&rcrtc->vblank_lock);
        rcar_du_group_write(rcrtc->group, rcrtc->index % 2 ? DS2PR : DS1PR, 0);
        status = rcar_du_crtc_read(rcrtc, DSSR);
        rcrtc->vblank_count = status & DSSR_VBK ? 2 : 1;
        spin_unlock_irq(&rcrtc->vblank_lock);

        if (!wait_event_timeout(rcrtc->vblank_wait, rcrtc->vblank_count == 0,
                                msecs_to_jiffies(100)))
                dev_warn(rcdu->dev, "vertical blanking timeout\n");

        drm_crtc_vblank_put(crtc);
}

static void rcar_du_crtc_stop(struct rcar_du_crtc *rcrtc)
{
        struct drm_crtc *crtc = &rcrtc->crtc;

        /*
         * Disable all planes and wait for the change to take effect. This is
         * required as the plane enable registers are updated on vblank, and no
         * vblank will occur once the CRTC is stopped. Disabling planes when
         * starting the CRTC thus wouldn't be enough as it would start scanning
         * out immediately from old frame buffers until the next vblank.
         *
         * This increases the CRTC stop delay, especially when multiple CRTCs
         * are stopped in one operation as we now wait for one vblank per CRTC.
         * Whether this can be improved needs to be researched.
         */
        rcar_du_crtc_disable_planes(rcrtc);

        /*
         * Disable vertical blanking interrupt reporting. We first need to wait
         * for page flip completion before stopping the CRTC as userspace
         * expects page flips to eventually complete.
         */
        rcar_du_crtc_wait_page_flip(rcrtc);
        drm_crtc_vblank_off(crtc);

        /* Disable the VSP compositor. */
        if (rcar_du_has(rcrtc->group->dev, RCAR_DU_FEATURE_VSP1_SOURCE))
                rcar_du_vsp_disable(rcrtc);

        /*
         * Select switch sync mode. This stops display operation and configures
         * the HSYNC and VSYNC signals as inputs.
         */
        rcar_du_crtc_clr_set(rcrtc, DSYSR, DSYSR_TVM_MASK, DSYSR_TVM_SWITCH);

        rcar_du_group_start_stop(rcrtc->group, false);
}

/* -----------------------------------------------------------------------------
 * CRTC Functions
 */

static void rcar_du_crtc_atomic_enable(struct drm_crtc *crtc,
                                       struct drm_crtc_state *old_state)
{
        struct rcar_du_crtc *rcrtc = to_rcar_crtc(crtc);

        /*
         * If the CRTC has already been setup by the .atomic_begin() handler we
         * can skip the setup stage.
         */
        if (!rcrtc->initialized) {
                rcar_du_crtc_get(rcrtc);
                rcar_du_crtc_setup(rcrtc);
                rcrtc->initialized = true;
        }

        rcar_du_crtc_start(rcrtc);
}

static void rcar_du_crtc_atomic_disable(struct drm_crtc *crtc,
                                        struct drm_crtc_state *old_state)
{
        struct rcar_du_crtc *rcrtc = to_rcar_crtc(crtc);

        rcar_du_crtc_stop(rcrtc);
        rcar_du_crtc_put(rcrtc);

        spin_lock_irq(&crtc->dev->event_lock);
        if (crtc->state->event) {
                drm_crtc_send_vblank_event(crtc, crtc->state->event);
                crtc->state->event = NULL;
        }
        spin_unlock_irq(&crtc->dev->event_lock);

        rcrtc->initialized = false;
        rcrtc->outputs = 0;
}

static void rcar_du_crtc_atomic_begin(struct drm_crtc *crtc,
                                      struct drm_crtc_state *old_crtc_state)
{
        struct rcar_du_crtc *rcrtc = to_rcar_crtc(crtc);

        WARN_ON(!crtc->state->enable);

        /*
         * If a mode set is in progress we can be called with the CRTC disabled.
         * We then need to first setup the CRTC in order to configure planes.
         * The .atomic_enable() handler will notice and skip the CRTC setup.
         */
        if (!rcrtc->initialized) {
                rcar_du_crtc_get(rcrtc);
                rcar_du_crtc_setup(rcrtc);
                rcrtc->initialized = true;
        }

        if (rcar_du_has(rcrtc->group->dev, RCAR_DU_FEATURE_VSP1_SOURCE))
                rcar_du_vsp_atomic_begin(rcrtc);
}

static void rcar_du_crtc_atomic_flush(struct drm_crtc *crtc,
                                      struct drm_crtc_state *old_crtc_state)
{
        struct rcar_du_crtc *rcrtc = to_rcar_crtc(crtc);
        struct drm_device *dev = rcrtc->crtc.dev;
        unsigned long flags;

        rcar_du_crtc_update_planes(rcrtc);

        if (crtc->state->event) {
                WARN_ON(drm_crtc_vblank_get(crtc) != 0);

                spin_lock_irqsave(&dev->event_lock, flags);
                rcrtc->event = crtc->state->event;
                crtc->state->event = NULL;
                spin_unlock_irqrestore(&dev->event_lock, flags);
        }

        if (rcar_du_has(rcrtc->group->dev, RCAR_DU_FEATURE_VSP1_SOURCE))
                rcar_du_vsp_atomic_flush(rcrtc);
}

static const struct drm_crtc_helper_funcs crtc_helper_funcs = {
        .atomic_begin = rcar_du_crtc_atomic_begin,
        .atomic_flush = rcar_du_crtc_atomic_flush,
        .atomic_enable = rcar_du_crtc_atomic_enable,
        .atomic_disable = rcar_du_crtc_atomic_disable,
};

static struct drm_crtc_state *
rcar_du_crtc_atomic_duplicate_state(struct drm_crtc *crtc)
{
        struct rcar_du_crtc_state *state;
        struct rcar_du_crtc_state *copy;

        if (WARN_ON(!crtc->state))
                return NULL;

        state = to_rcar_crtc_state(crtc->state);
        copy = kmemdup(state, sizeof(*state), GFP_KERNEL);
        if (copy == NULL)
                return NULL;

        __drm_atomic_helper_crtc_duplicate_state(crtc, &copy->state);

        return &copy->state;
}

static void rcar_du_crtc_atomic_destroy_state(struct drm_crtc *crtc,
                                              struct drm_crtc_state *state)
{
        __drm_atomic_helper_crtc_destroy_state(state);
        kfree(to_rcar_crtc_state(state));
}

static void rcar_du_crtc_reset(struct drm_crtc *crtc)
{
        struct rcar_du_crtc_state *state;

        if (crtc->state) {
                rcar_du_crtc_atomic_destroy_state(crtc, crtc->state);
                crtc->state = NULL;
        }

        state = kzalloc(sizeof(*state), GFP_KERNEL);
        if (state == NULL)
                return;

        state->crc.source = VSP1_DU_CRC_NONE;
        state->crc.index = 0;

        crtc->state = &state->state;
        crtc->state->crtc = crtc;
}

static int rcar_du_crtc_enable_vblank(struct drm_crtc *crtc)
{
        struct rcar_du_crtc *rcrtc = to_rcar_crtc(crtc);

        rcar_du_crtc_write(rcrtc, DSRCR, DSRCR_VBCL);
        rcar_du_crtc_set(rcrtc, DIER, DIER_VBE);
        rcrtc->vblank_enable = true;

        return 0;
}

static void rcar_du_crtc_disable_vblank(struct drm_crtc *crtc)
{
        struct rcar_du_crtc *rcrtc = to_rcar_crtc(crtc);

        rcar_du_crtc_clr(rcrtc, DIER, DIER_VBE);
        rcrtc->vblank_enable = false;
}

static int rcar_du_crtc_set_crc_source(struct drm_crtc *crtc,
                                       const char *source_name,
                                       size_t *values_cnt)
{
        struct rcar_du_crtc *rcrtc = to_rcar_crtc(crtc);
        struct drm_modeset_acquire_ctx ctx;
        struct drm_crtc_state *crtc_state;
        struct drm_atomic_state *state;
        enum vsp1_du_crc_source source;
        unsigned int index = 0;
        unsigned int i;
        int ret;

        /*
         * Parse the source name. Supported values are "plane%u" to compute the
         * CRC on an input plane (%u is the plane ID), and "auto" to compute the
         * CRC on the composer (VSP) output.
         */
        if (!source_name) {
                source = VSP1_DU_CRC_NONE;
        } else if (!strcmp(source_name, "auto")) {
                source = VSP1_DU_CRC_OUTPUT;
        } else if (strstarts(source_name, "plane")) {
                source = VSP1_DU_CRC_PLANE;

                ret = kstrtouint(source_name + strlen("plane"), 10, &index);
                if (ret < 0)
                        return ret;

                for (i = 0; i < rcrtc->vsp->num_planes; ++i) {
                        if (index == rcrtc->vsp->planes[i].plane.base.id) {
                                index = i;
                                break;
                        }
                }

                if (i >= rcrtc->vsp->num_planes)
                        return -EINVAL;
        } else {
                return -EINVAL;
        }

        *values_cnt = 1;

        /* Perform an atomic commit to set the CRC source. */
        drm_modeset_acquire_init(&ctx, 0);

        state = drm_atomic_state_alloc(crtc->dev);
        if (!state) {
                ret = -ENOMEM;
                goto unlock;
        }

        state->acquire_ctx = &ctx;

retry:
        crtc_state = drm_atomic_get_crtc_state(state, crtc);
        if (!IS_ERR(crtc_state)) {
                struct rcar_du_crtc_state *rcrtc_state;

                rcrtc_state = to_rcar_crtc_state(crtc_state);
                rcrtc_state->crc.source = source;
                rcrtc_state->crc.index = index;

                ret = drm_atomic_commit(state);
        } else {
                ret = PTR_ERR(crtc_state);
        }

        if (ret == -EDEADLK) {
                drm_atomic_state_clear(state);
                drm_modeset_backoff(&ctx);
                goto retry;
        }

        drm_atomic_state_put(state);

unlock:
        drm_modeset_drop_locks(&ctx);
        drm_modeset_acquire_fini(&ctx);

        return 0;
}

static const struct drm_crtc_funcs crtc_funcs_gen2 = {
        .reset = rcar_du_crtc_reset,
        .destroy = drm_crtc_cleanup,
        .set_config = drm_atomic_helper_set_config,
        .page_flip = drm_atomic_helper_page_flip,
        .atomic_duplicate_state = rcar_du_crtc_atomic_duplicate_state,
        .atomic_destroy_state = rcar_du_crtc_atomic_destroy_state,
        .enable_vblank = rcar_du_crtc_enable_vblank,
        .disable_vblank = rcar_du_crtc_disable_vblank,
};

static const struct drm_crtc_funcs crtc_funcs_gen3 = {
        .reset = rcar_du_crtc_reset,
        .destroy = drm_crtc_cleanup,
        .set_config = drm_atomic_helper_set_config,
        .page_flip = drm_atomic_helper_page_flip,
        .atomic_duplicate_state = rcar_du_crtc_atomic_duplicate_state,
        .atomic_destroy_state = rcar_du_crtc_atomic_destroy_state,
        .enable_vblank = rcar_du_crtc_enable_vblank,
        .disable_vblank = rcar_du_crtc_disable_vblank,
        .set_crc_source = rcar_du_crtc_set_crc_source,
};

/* -----------------------------------------------------------------------------
 * Interrupt Handling
 */

static irqreturn_t rcar_du_crtc_irq(int irq, void *arg)
{
        struct rcar_du_crtc *rcrtc = arg;
        struct rcar_du_device *rcdu = rcrtc->group->dev;
        irqreturn_t ret = IRQ_NONE;
        u32 status;

        spin_lock(&rcrtc->vblank_lock);

        status = rcar_du_crtc_read(rcrtc, DSSR);
        rcar_du_crtc_write(rcrtc, DSRCR, status & DSRCR_MASK);

        if (status & DSSR_VBK) {
                /*
                 * Wake up the vblank wait if the counter reaches 0. This must
                 * be protected by the vblank_lock to avoid races in
                 * rcar_du_crtc_disable_planes().
                 */
                if (rcrtc->vblank_count) {
                        if (--rcrtc->vblank_count == 0)
                                wake_up(&rcrtc->vblank_wait);
                }
        }

        spin_unlock(&rcrtc->vblank_lock);

        if (status & DSSR_VBK) {
                if (rcdu->info->gen < 3) {
                        drm_crtc_handle_vblank(&rcrtc->crtc);
                        rcar_du_crtc_finish_page_flip(rcrtc);
                }

                ret = IRQ_HANDLED;
        }

        return ret;
}

/* -----------------------------------------------------------------------------
 * Initialization
 */

int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int swindex,
                        unsigned int hwindex)
{
        static const unsigned int mmio_offsets[] = {
                DU0_REG_OFFSET, DU1_REG_OFFSET, DU2_REG_OFFSET, DU3_REG_OFFSET
        };

        struct rcar_du_device *rcdu = rgrp->dev;
        struct platform_device *pdev = to_platform_device(rcdu->dev);
        struct rcar_du_crtc *rcrtc = &rcdu->crtcs[swindex];
        struct drm_crtc *crtc = &rcrtc->crtc;
        struct drm_plane *primary;
        unsigned int irqflags;
        struct clk *clk;
        char clk_name[9];
        char *name;
        int irq;
        int ret;

        /* Get the CRTC clock and the optional external clock. */
        if (rcar_du_has(rcdu, RCAR_DU_FEATURE_CRTC_IRQ_CLOCK)) {
                sprintf(clk_name, "du.%u", hwindex);
                name = clk_name;
        } else {
                name = NULL;
        }

        rcrtc->clock = devm_clk_get(rcdu->dev, name);
        if (IS_ERR(rcrtc->clock)) {
                dev_err(rcdu->dev, "no clock for DU channel %u\n", hwindex);
                return PTR_ERR(rcrtc->clock);
        }

        sprintf(clk_name, "dclkin.%u", hwindex);
        clk = devm_clk_get(rcdu->dev, clk_name);
        if (!IS_ERR(clk)) {
                rcrtc->extclock = clk;
        } else if (PTR_ERR(rcrtc->clock) == -EPROBE_DEFER) {
                dev_info(rcdu->dev, "can't get external clock %u\n", hwindex);
                return -EPROBE_DEFER;
        }

        init_waitqueue_head(&rcrtc->flip_wait);
        init_waitqueue_head(&rcrtc->vblank_wait);
        spin_lock_init(&rcrtc->vblank_lock);

        rcrtc->group = rgrp;
        rcrtc->mmio_offset = mmio_offsets[hwindex];
        rcrtc->index = hwindex;

        if (rcar_du_has(rcdu, RCAR_DU_FEATURE_VSP1_SOURCE))
                primary = &rcrtc->vsp->planes[rcrtc->vsp_pipe].plane;
        else
                primary = &rgrp->planes[swindex % 2].plane;

        ret = drm_crtc_init_with_planes(rcdu->ddev, crtc, primary, NULL,
                                        rcdu->info->gen <= 2 ?
                                        &crtc_funcs_gen2 : &crtc_funcs_gen3,
                                        NULL);
        if (ret < 0)
                return ret;

        drm_crtc_helper_add(crtc, &crtc_helper_funcs);

        /* Start with vertical blanking interrupt reporting disabled. */
        drm_crtc_vblank_off(crtc);

        /* Register the interrupt handler. */
        if (rcar_du_has(rcdu, RCAR_DU_FEATURE_CRTC_IRQ_CLOCK)) {
                /* The IRQ's are associated with the CRTC (sw)index. */
                irq = platform_get_irq(pdev, swindex);
                irqflags = 0;
        } else {
                irq = platform_get_irq(pdev, 0);
                irqflags = IRQF_SHARED;
        }

        if (irq < 0) {
                dev_err(rcdu->dev, "no IRQ for CRTC %u\n", swindex);
                return irq;
        }

        ret = devm_request_irq(rcdu->dev, irq, rcar_du_crtc_irq, irqflags,
                               dev_name(rcdu->dev), rcrtc);
        if (ret < 0) {
                dev_err(rcdu->dev,
                        "failed to register IRQ for CRTC %u\n", swindex);
                return ret;
        }

        return 0;
}