1 files changed, 991 insertions, 348 deletions

diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
index 5d450332c2fd..ad4ab894391a 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
@@ -1,74 +1,76 @@
+// SPDX-License-Identifier: GPL-2.0-only
 /*
- * Copyright (C) Fuzhou Rockchip Electronics Co.Ltd
+ * Copyright (C) Rockchip Electronics Co., Ltd.
  * Author:Mark Yao <mark.yao@rock-chips.com>
- *
- * This software is licensed under the terms of the GNU General Public
- * License version 2, as published by the Free Software Foundation, and
- * may be copied, distributed, and modified under those terms.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
  */
 
+#include <linux/clk.h>
+#include <linux/component.h>
+#include <linux/delay.h>
+#include <linux/iopoll.h>
+#include <linux/kernel.h>
+#include <linux/log2.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/overflow.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/reset.h>
+
 #include <drm/drm.h>
-#include <drm/drmP.h>
 #include <drm/drm_atomic.h>
+#include <drm/drm_atomic_uapi.h>
+#include <drm/drm_blend.h>
 #include <drm/drm_crtc.h>
-#include <drm/drm_crtc_helper.h>
 #include <drm/drm_flip_work.h>
-#include <drm/drm_plane_helper.h>
+#include <drm/drm_fourcc.h>
+#include <drm/drm_framebuffer.h>
+#include <drm/drm_gem_atomic_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
+#include <drm/drm_print.h>
+#include <drm/drm_probe_helper.h>
+#include <drm/drm_self_refresh_helper.h>
+#include <drm/drm_vblank.h>
+
 #ifdef CONFIG_DRM_ANALOGIX_DP
 #include <drm/bridge/analogix_dp.h>
 #endif
 
-#include <linux/kernel.h>
-#include <linux/module.h>
-#include <linux/platform_device.h>
-#include <linux/clk.h>
-#include <linux/iopoll.h>
-#include <linux/of.h>
-#include <linux/of_device.h>
-#include <linux/pm_runtime.h>
-#include <linux/component.h>
-
-#include <linux/reset.h>
-#include <linux/delay.h>
-
 #include "rockchip_drm_drv.h"
 #include "rockchip_drm_gem.h"
 #include "rockchip_drm_fb.h"
-#include "rockchip_drm_psr.h"
 #include "rockchip_drm_vop.h"
+#include "rockchip_rgb.h"
+
+#define VOP_WIN_SET(vop, win, name, v) \
+		vop_reg_set(vop, &win->phy->name, win->base, ~0, v, #name)
+#define VOP_SCL_SET(vop, win, name, v) \
+		vop_reg_set(vop, &win->phy->scl->name, win->base, ~0, v, #name)
+#define VOP_SCL_SET_EXT(vop, win, name, v) \
+		vop_reg_set(vop, &win->phy->scl->ext->name, \
+			    win->base, ~0, v, #name)
+
+#define VOP_WIN_YUV2YUV_SET(vop, win_yuv2yuv, name, v) \
+	do { \
+		if (win_yuv2yuv && win_yuv2yuv->name.mask) \
+			vop_reg_set(vop, &win_yuv2yuv->name, 0, ~0, v, #name); \
+	} while (0)
+
+#define VOP_WIN_YUV2YUV_COEFFICIENT_SET(vop, win_yuv2yuv, name, v) \
+	do { \
+		if (win_yuv2yuv && win_yuv2yuv->phy->name.mask) \
+			vop_reg_set(vop, &win_yuv2yuv->phy->name, win_yuv2yuv->base, ~0, v, #name); \
+	} while (0)
+
+#define VOP_INTR_SET_MASK(vop, name, mask, v) \
+		vop_reg_set(vop, &vop->data->intr->name, 0, mask, v, #name)
+
+#define VOP_REG_SET(vop, group, name, v) \
+		    vop_reg_set(vop, &vop->data->group->name, 0, ~0, v, #name)
+
+#define VOP_HAS_REG(vop, group, name) \
+		(!!(vop->data->group->name.mask))
 
-#define __REG_SET_RELAXED(x, off, mask, shift, v, write_mask) \
-		vop_mask_write(x, off, mask, shift, v, write_mask, true)
-
-#define __REG_SET_NORMAL(x, off, mask, shift, v, write_mask) \
-		vop_mask_write(x, off, mask, shift, v, write_mask, false)
-
-#define REG_SET(x, base, reg, v, mode) \
-		__REG_SET_##mode(x, base + reg.offset, \
-				 reg.mask, reg.shift, v, reg.write_mask)
-#define REG_SET_MASK(x, base, reg, mask, v, mode) \
-		__REG_SET_##mode(x, base + reg.offset, \
-				 mask, reg.shift, v, reg.write_mask)
-
-#define VOP_WIN_SET(x, win, name, v) \
-		REG_SET(x, win->base, win->phy->name, v, RELAXED)
-#define VOP_SCL_SET(x, win, name, v) \
-		REG_SET(x, win->base, win->phy->scl->name, v, RELAXED)
-#define VOP_SCL_SET_EXT(x, win, name, v) \
-		REG_SET(x, win->base, win->phy->scl->ext->name, v, RELAXED)
-#define VOP_CTRL_SET(x, name, v) \
-		REG_SET(x, 0, (x)->data->ctrl->name, v, NORMAL)
-
-#define VOP_INTR_GET(vop, name) \
-		vop_read_reg(vop, 0, &vop->data->ctrl->name)
-
-#define VOP_INTR_SET(vop, name, mask, v) \
-		REG_SET_MASK(vop, 0, vop->data->intr->name, mask, v, NORMAL)
 #define VOP_INTR_SET_TYPE(vop, name, type, v) \
 	do { \
 		int i, reg = 0, mask = 0; \
@@ -78,20 +80,51 @@
 				mask |= 1 << i; \
 			} \
 		} \
-		VOP_INTR_SET(vop, name, mask, reg); \
+		VOP_INTR_SET_MASK(vop, name, mask, reg); \
 	} while (0)
 #define VOP_INTR_GET_TYPE(vop, name, type) \
 		vop_get_intr_type(vop, &vop->data->intr->name, type)
 
-#define VOP_WIN_GET(x, win, name) \
-		vop_read_reg(x, win->base, &win->phy->name)
+#define VOP_WIN_GET(vop, win, name) \
+		vop_read_reg(vop, win->base, &win->phy->name)
+
+#define VOP_WIN_HAS_REG(win, name) \
+	(!!(win->phy->name.mask))
 
 #define VOP_WIN_GET_YRGBADDR(vop, win) \
 		vop_readl(vop, win->base + win->phy->yrgb_mst.offset)
 
+#define VOP_WIN_TO_INDEX(vop_win) \
+	((vop_win) - (vop_win)->vop->win)
+
+#define VOP_AFBC_SET(vop, name, v) \
+	do { \
+		if ((vop)->data->afbc) \
+			vop_reg_set((vop), &(vop)->data->afbc->name, \
+				    0, ~0, v, #name); \
+	} while (0)
+
 #define to_vop(x) container_of(x, struct vop, crtc)
 #define to_vop_win(x) container_of(x, struct vop_win, base)
 
+#define AFBC_FMT_RGB565		0x0
+#define AFBC_FMT_U8U8U8U8	0x5
+#define AFBC_FMT_U8U8U8		0x4
+
+#define AFBC_TILE_16x16		BIT(4)
+
+/*
+ * The coefficients of the following matrix are all fixed points.
+ * The format is S2.10 for the 3x3 part of the matrix, and S9.12 for the offsets.
+ * They are all represented in two's complement.
+ */
+static const uint32_t bt601_yuv2rgb[] = {
+	0x4A8, 0x0,    0x662,
+	0x4A8, 0x1E6F, 0x1CBF,
+	0x4A8, 0x812,  0x0,
+	0x321168, 0x0877CF, 0x2EB127
+};
+
 enum vop_pending {
 	VOP_PENDING_FB_UNREF,
 };
@@ -99,19 +132,19 @@ enum vop_pending {
 struct vop_win {
 	struct drm_plane base;
 	const struct vop_win_data *data;
+	const struct vop_win_yuv2yuv_data *yuv2yuv_data;
 	struct vop *vop;
 };
 
+struct rockchip_rgb;
 struct vop {
 	struct drm_crtc crtc;
 	struct device *dev;
 	struct drm_device *drm_dev;
 	bool is_enabled;
 
-	/* mutex vsync_ work */
-	struct mutex vsync_mutex;
-	bool vsync_work_pending;
 	struct completion dsp_hold_completion;
+	unsigned int win_enabled;
 
 	/* protected by dev->event_lock */
 	struct drm_pending_vblank_event *event;
@@ -125,6 +158,7 @@ struct vop {
 
 	uint32_t *regsbak;
 	void __iomem *regs;
+	void __iomem *lut_regs;
 
 	/* physical map length of vop register */
 	uint32_t len;
@@ -133,6 +167,8 @@ struct vop {
 	spinlock_t reg_lock;
 	/* lock vop irq reg */
 	spinlock_t irq_lock;
+	/* protects crtc enable/disable */
+	struct mutex vop_lock;
 
 	unsigned int irq;
 
@@ -146,15 +182,12 @@ struct vop {
 	/* vop dclk reset */
 	struct reset_control *dclk_rst;
 
+	/* optional internal rgb encoder */
+	struct rockchip_rgb *rgb;
+
 	struct vop_win win[];
 };
 
-static inline void vop_writel(struct vop *vop, uint32_t offset, uint32_t v)
-{
-	writel(v, vop->regs + offset);
-	vop->regsbak[offset >> 2] = v;
-}
-
 static inline uint32_t vop_readl(struct vop *vop, uint32_t offset)
 {
 	return readl(vop->regs + offset);
@@ -166,14 +199,22 @@ static inline uint32_t vop_read_reg(struct vop *vop, uint32_t base,
 	return (vop_readl(vop, base + reg->offset) >> reg->shift) & reg->mask;
 }
 
-static inline void vop_mask_write(struct vop *vop, uint32_t offset,
-				  uint32_t mask, uint32_t shift, uint32_t v,
-				  bool write_mask, bool relaxed)
+static void vop_reg_set(struct vop *vop, const struct vop_reg *reg,
+			uint32_t _offset, uint32_t _mask, uint32_t v,
+			const char *reg_name)
 {
-	if (!mask)
+	int offset, mask, shift;
+
+	if (!reg || !reg->mask) {
+		DRM_DEV_DEBUG(vop->dev, "Warning: not support %s\n", reg_name);
 		return;
+	}
+
+	offset = reg->offset + _offset;
+	mask = reg->mask & _mask;
+	shift = reg->shift;
 
-	if (write_mask) {
+	if (reg->write_mask) {
 		v = ((v << shift) & 0xffff) | (mask << (shift + 16));
 	} else {
 		uint32_t cached_val = vop->regsbak[offset >> 2];
@@ -182,7 +223,7 @@ static inline void vop_mask_write(struct vop *vop, uint32_t offset,
 		vop->regsbak[offset >> 2] = v;
 	}
 
-	if (relaxed)
+	if (reg->relaxed)
 		writel_relaxed(v, vop->regs + offset);
 	else
 		writel(v, vop->regs + offset);
@@ -204,17 +245,49 @@ static inline uint32_t vop_get_intr_type(struct vop *vop,
 
 static inline void vop_cfg_done(struct vop *vop)
 {
-	VOP_CTRL_SET(vop, cfg_done, 1);
+	VOP_REG_SET(vop, common, cfg_done, 1);
 }
 
-static bool has_rb_swapped(uint32_t format)
+static bool has_rb_swapped(uint32_t version, uint32_t format)
 {
 	switch (format) {
 	case DRM_FORMAT_XBGR8888:
 	case DRM_FORMAT_ABGR8888:
-	case DRM_FORMAT_BGR888:
 	case DRM_FORMAT_BGR565:
 		return true;
+	/*
+	 * full framework (IP version 3.x) only need rb swapped for RGB888 and
+	 * little framework (IP version 2.x) only need rb swapped for BGR888,
+	 * check for 3.x to also only rb swap BGR888 for unknown vop version
+	 */
+	case DRM_FORMAT_RGB888:
+		return VOP_MAJOR(version) == 3;
+	case DRM_FORMAT_BGR888:
+		return VOP_MAJOR(version) != 3;
+	default:
+		return false;
+	}
+}
+
+static bool has_uv_swapped(uint32_t format)
+{
+	switch (format) {
+	case DRM_FORMAT_NV21:
+	case DRM_FORMAT_NV61:
+	case DRM_FORMAT_NV42:
+		return true;
+	default:
+		return false;
+	}
+}
+
+static bool is_fmt_10(uint32_t format)
+{
+	switch (format) {
+	case DRM_FORMAT_NV15:
+	case DRM_FORMAT_NV20:
+	case DRM_FORMAT_NV30:
+		return true;
 	default:
 		return false;
 	}
@@ -235,10 +308,16 @@ static enum vop_data_format vop_convert_format(uint32_t format)
 	case DRM_FORMAT_BGR565:
 		return VOP_FMT_RGB565;
 	case DRM_FORMAT_NV12:
+	case DRM_FORMAT_NV15:
+	case DRM_FORMAT_NV21:
 		return VOP_FMT_YUV420SP;
 	case DRM_FORMAT_NV16:
+	case DRM_FORMAT_NV20:
+	case DRM_FORMAT_NV61:
 		return VOP_FMT_YUV422SP;
 	case DRM_FORMAT_NV24:
+	case DRM_FORMAT_NV30:
+	case DRM_FORMAT_NV42:
 		return VOP_FMT_YUV444SP;
 	default:
 		DRM_ERROR("unsupported format[%08x]\n", format);
@@ -246,26 +325,23 @@ static enum vop_data_format vop_convert_format(uint32_t format)
 	}
 }
 
-static bool is_yuv_support(uint32_t format)
-{
-	switch (format) {
-	case DRM_FORMAT_NV12:
-	case DRM_FORMAT_NV16:
-	case DRM_FORMAT_NV24:
-		return true;
-	default:
-		return false;
-	}
-}
-
-static bool is_alpha_support(uint32_t format)
+static int vop_convert_afbc_format(uint32_t format)
 {
 	switch (format) {
+	case DRM_FORMAT_XRGB8888:
 	case DRM_FORMAT_ARGB8888:
+	case DRM_FORMAT_XBGR8888:
 	case DRM_FORMAT_ABGR8888:
-		return true;
+		return AFBC_FMT_U8U8U8U8;
+	case DRM_FORMAT_RGB888:
+	case DRM_FORMAT_BGR888:
+		return AFBC_FMT_U8U8U8;
+	case DRM_FORMAT_RGB565:
+	case DRM_FORMAT_BGR565:
+		return AFBC_FMT_RGB565;
 	default:
-		return false;
+		DRM_DEBUG_KMS("unsupported AFBC format[%08x]\n", format);
+		return -EINVAL;
 	}
 }
 
@@ -275,6 +351,9 @@ static uint16_t scl_vop_cal_scale(enum scale_mode mode, uint32_t src,
 {
 	uint16_t val = 1 << SCL_FT_DEFAULT_FIXPOINT_SHIFT;
 
+	if (vskiplines)
+		*vskiplines = 0;
+
 	if (is_horizontal) {
 		if (mode == SCALE_UP)
 			val = GET_SCL_FT_BIC(src, dst);
@@ -302,23 +381,24 @@ static uint16_t scl_vop_cal_scale(enum scale_mode mode, uint32_t src,
 
 static void scl_vop_cal_scl_fac(struct vop *vop, const struct vop_win_data *win,
 			     uint32_t src_w, uint32_t src_h, uint32_t dst_w,
-			     uint32_t dst_h, uint32_t pixel_format)
+			     uint32_t dst_h, const struct drm_format_info *info)
 {
 	uint16_t yrgb_hor_scl_mode, yrgb_ver_scl_mode;
 	uint16_t cbcr_hor_scl_mode = SCALE_NONE;
 	uint16_t cbcr_ver_scl_mode = SCALE_NONE;
-	int hsub = drm_format_horz_chroma_subsampling(pixel_format);
-	int vsub = drm_format_vert_chroma_subsampling(pixel_format);
-	bool is_yuv = is_yuv_support(pixel_format);
-	uint16_t cbcr_src_w = src_w / hsub;
-	uint16_t cbcr_src_h = src_h / vsub;
+	bool is_yuv = false;
+	uint16_t cbcr_src_w = src_w / info->hsub;
+	uint16_t cbcr_src_h = src_h / info->vsub;
 	uint16_t vsu_mode;
 	uint16_t lb_mode;
 	uint32_t val;
-	int vskiplines = 0;
+	int vskiplines;
+
+	if (info->is_yuv)
+		is_yuv = true;
 
-	if (dst_w > 3840) {
-		DRM_DEV_ERROR(vop->dev, "Maximum dst width (3840) exceeded\n");
+	if (dst_w > 4096) {
+		DRM_DEV_ERROR(vop->dev, "Maximum dst width (4096) exceeded\n");
 		return;
 	}
 
@@ -497,28 +577,64 @@ static void vop_line_flag_irq_disable(struct vop *vop)
 	spin_unlock_irqrestore(&vop->irq_lock, flags);
 }
 
-static int vop_enable(struct drm_crtc *crtc)
+static int vop_core_clks_enable(struct vop *vop)
 {
-	struct vop *vop = to_vop(crtc);
 	int ret;
 
-	ret = pm_runtime_get_sync(vop->dev);
+	ret = clk_enable(vop->hclk);
+	if (ret < 0)
+		return ret;
+
+	ret = clk_enable(vop->aclk);
+	if (ret < 0)
+		goto err_disable_hclk;
+
+	return 0;
+
+err_disable_hclk:
+	clk_disable(vop->hclk);
+	return ret;
+}
+
+static void vop_core_clks_disable(struct vop *vop)
+{
+	clk_disable(vop->aclk);
+	clk_disable(vop->hclk);
+}
+
+static void vop_win_disable(struct vop *vop, const struct vop_win *vop_win)
+{
+	const struct vop_win_data *win = vop_win->data;
+
+	if (win->phy->scl && win->phy->scl->ext) {
+		VOP_SCL_SET_EXT(vop, win, yrgb_hor_scl_mode, SCALE_NONE);
+		VOP_SCL_SET_EXT(vop, win, yrgb_ver_scl_mode, SCALE_NONE);
+		VOP_SCL_SET_EXT(vop, win, cbcr_hor_scl_mode, SCALE_NONE);
+		VOP_SCL_SET_EXT(vop, win, cbcr_ver_scl_mode, SCALE_NONE);
+	}
+
+	VOP_WIN_SET(vop, win, enable, 0);
+	vop->win_enabled &= ~BIT(VOP_WIN_TO_INDEX(vop_win));
+}
+
+static int vop_enable(struct drm_crtc *crtc, struct drm_crtc_state *old_state)
+{
+	struct vop *vop = to_vop(crtc);
+	int ret, i;
+
+	ret = pm_runtime_resume_and_get(vop->dev);
 	if (ret < 0) {
-		dev_err(vop->dev, "failed to get pm runtime: %d\n", ret);
+		DRM_DEV_ERROR(vop->dev, "failed to get pm runtime: %d\n", ret);
 		return ret;
 	}
 
-	ret = clk_enable(vop->hclk);
+	ret = vop_core_clks_enable(vop);
 	if (WARN_ON(ret < 0))
 		goto err_put_pm_runtime;
 
 	ret = clk_enable(vop->dclk);
 	if (WARN_ON(ret < 0))
-		goto err_disable_hclk;
-
-	ret = clk_enable(vop->aclk);
-	if (WARN_ON(ret < 0))
-		goto err_disable_dclk;
+		goto err_disable_core;
 
 	/*
 	 * Slave iommu shares power, irq and clock with vop.  It was associated
@@ -528,13 +644,46 @@ static int vop_enable(struct drm_crtc *crtc)
 	 */
 	ret = rockchip_drm_dma_attach_device(vop->drm_dev, vop->dev);
 	if (ret) {
-		dev_err(vop->dev, "failed to attach dma mapping, %d\n", ret);
-		goto err_disable_aclk;
+		DRM_DEV_ERROR(vop->dev,
+			      "failed to attach dma mapping, %d\n", ret);
+		goto err_disable_dclk;
+	}
+
+	spin_lock(&vop->reg_lock);
+	for (i = 0; i < vop->len; i += 4)
+		writel_relaxed(vop->regsbak[i / 4], vop->regs + i);
+
+	/*
+	 * We need to make sure that all windows are disabled before we
+	 * enable the crtc. Otherwise we might try to scan from a destroyed
+	 * buffer later.
+	 *
+	 * In the case of enable-after-PSR, we don't need to worry about this
+	 * case since the buffer is guaranteed to be valid and disabling the
+	 * window will result in screen glitches on PSR exit.
+	 */
+	if (!old_state || !old_state->self_refresh_active) {
+		for (i = 0; i < vop->data->win_size; i++) {
+			struct vop_win *vop_win = &vop->win[i];
+
+			vop_win_disable(vop, vop_win);
+		}
+	}
+
+	if (vop->data->afbc) {
+		struct rockchip_crtc_state *s;
+		/*
+		 * Disable AFBC and forget there was a vop window with AFBC
+		 */
+		VOP_AFBC_SET(vop, enable, 0);
+		s = to_rockchip_crtc_state(crtc->state);
+		s->enable_afbc = false;
 	}
 
-	memcpy(vop->regs, vop->regsbak, vop->len);
 	vop_cfg_done(vop);
 
+	spin_unlock(&vop->reg_lock);
+
 	/*
 	 * At here, vop clock & iommu is enable, R/W vop regs would be safe.
 	 */
@@ -542,51 +691,55 @@ static int vop_enable(struct drm_crtc *crtc)
 
 	spin_lock(&vop->reg_lock);
 
-	VOP_CTRL_SET(vop, standby, 0);
+	VOP_REG_SET(vop, common, standby, 1);
 
 	spin_unlock(&vop->reg_lock);
 
-	enable_irq(vop->irq);
-
 	drm_crtc_vblank_on(crtc);
 
 	return 0;
 
-err_disable_aclk:
-	clk_disable(vop->aclk);
 err_disable_dclk:
 	clk_disable(vop->dclk);
-err_disable_hclk:
-	clk_disable(vop->hclk);
+err_disable_core:
+	vop_core_clks_disable(vop);
 err_put_pm_runtime:
 	pm_runtime_put_sync(vop->dev);
 	return ret;
 }
 
-static void vop_crtc_disable(struct drm_crtc *crtc)
+static void rockchip_drm_set_win_enabled(struct drm_crtc *crtc, bool enabled)
 {
-	struct vop *vop = to_vop(crtc);
-	int i;
+        struct vop *vop = to_vop(crtc);
+        int i;
 
-	WARN_ON(vop->event);
+        spin_lock(&vop->reg_lock);
 
-	rockchip_drm_psr_deactivate(&vop->crtc);
+        for (i = 0; i < vop->data->win_size; i++) {
+                struct vop_win *vop_win = &vop->win[i];
+                const struct vop_win_data *win = vop_win->data;
 
-	/*
-	 * We need to make sure that all windows are disabled before we
-	 * disable that crtc. Otherwise we might try to scan from a destroyed
-	 * buffer later.
-	 */
-	for (i = 0; i < vop->data->win_size; i++) {
-		struct vop_win *vop_win = &vop->win[i];
-		const struct vop_win_data *win = vop_win->data;
+                VOP_WIN_SET(vop, win, enable,
+                            enabled && (vop->win_enabled & BIT(i)));
+        }
+        vop_cfg_done(vop);
 
-		spin_lock(&vop->reg_lock);
-		VOP_WIN_SET(vop, win, enable, 0);
-		spin_unlock(&vop->reg_lock);
+        spin_unlock(&vop->reg_lock);
+}
+
+static void vop_crtc_atomic_disable(struct drm_crtc *crtc,
+				    struct drm_atomic_state *state)
+{
+	struct vop *vop = to_vop(crtc);
+
+	WARN_ON(vop->event);
+
+	if (crtc->state->self_refresh_active) {
+		rockchip_drm_set_win_enabled(crtc, false);
+		goto out;
 	}
 
-	vop_cfg_done(vop);
+	mutex_lock(&vop->vop_lock);
 
 	drm_crtc_vblank_off(crtc);
 
@@ -602,16 +755,16 @@ static void vop_crtc_disable(struct drm_crtc *crtc)
 
 	spin_lock(&vop->reg_lock);
 
-	VOP_CTRL_SET(vop, standby, 1);
+	VOP_REG_SET(vop, common, standby, 1);
 
 	spin_unlock(&vop->reg_lock);
 
-	wait_for_completion(&vop->dsp_hold_completion);
+	if (!wait_for_completion_timeout(&vop->dsp_hold_completion,
+					 msecs_to_jiffies(200)))
+		WARN(1, "%s: timed out waiting for DSP hold", crtc->name);
 
 	vop_dsp_hold_valid_irq_disable(vop);
 
-	disable_irq(vop->irq);
-
 	vop->is_enabled = false;
 
 	/*
@@ -620,10 +773,12 @@ static void vop_crtc_disable(struct drm_crtc *crtc)
 	rockchip_drm_dma_detach_device(vop->drm_dev, vop->dev);
 
 	clk_disable(vop->dclk);
-	clk_disable(vop->aclk);
-	clk_disable(vop->hclk);
+	vop_core_clks_disable(vop);
 	pm_runtime_put(vop->dev);
 
+	mutex_unlock(&vop->vop_lock);
+
+out:
 	if (crtc->state->event && !crtc->state->active) {
 		spin_lock_irq(&crtc->dev->event_lock);
 		drm_crtc_send_vblank_event(crtc, crtc->state->event);
@@ -633,45 +788,56 @@ static void vop_crtc_disable(struct drm_crtc *crtc)
 	}
 }
 
-static void vop_plane_destroy(struct drm_plane *plane)
+static inline bool rockchip_afbc(u64 modifier)
 {
-	drm_plane_cleanup(plane);
+	return modifier == ROCKCHIP_AFBC_MOD;
+}
+
+static bool rockchip_mod_supported(struct drm_plane *plane,
+				   u32 format, u64 modifier)
+{
+	if (modifier == DRM_FORMAT_MOD_LINEAR)
+		return true;
+
+	if (!rockchip_afbc(modifier)) {
+		DRM_DEBUG_KMS("Unsupported format modifier 0x%llx\n", modifier);
+
+		return false;
+	}
+
+	return vop_convert_afbc_format(format) >= 0;
 }
 
 static int vop_plane_atomic_check(struct drm_plane *plane,
-			   struct drm_plane_state *state)
+			   struct drm_atomic_state *state)
 {
-	struct drm_crtc *crtc = state->crtc;
+	struct drm_plane_state *new_plane_state = drm_atomic_get_new_plane_state(state,
+										 plane);
+	struct drm_crtc *crtc = new_plane_state->crtc;
 	struct drm_crtc_state *crtc_state;
-	struct drm_framebuffer *fb = state->fb;
+	struct drm_framebuffer *fb = new_plane_state->fb;
 	struct vop_win *vop_win = to_vop_win(plane);
 	const struct vop_win_data *win = vop_win->data;
 	int ret;
-	struct drm_rect clip;
 	int min_scale = win->phy->scl ? FRAC_16_16(1, 8) :
-					DRM_PLANE_HELPER_NO_SCALING;
+					DRM_PLANE_NO_SCALING;
 	int max_scale = win->phy->scl ? FRAC_16_16(8, 1) :
-					DRM_PLANE_HELPER_NO_SCALING;
+					DRM_PLANE_NO_SCALING;
 
-	if (!crtc || !fb)
+	if (!crtc || WARN_ON(!fb))
 		return 0;
 
-	crtc_state = drm_atomic_get_existing_crtc_state(state->state, crtc);
+	crtc_state = drm_atomic_get_new_crtc_state(state, crtc);
 	if (WARN_ON(!crtc_state))
 		return -EINVAL;
 
-	clip.x1 = 0;
-	clip.y1 = 0;
-	clip.x2 = crtc_state->adjusted_mode.hdisplay;
-	clip.y2 = crtc_state->adjusted_mode.vdisplay;
-
-	ret = drm_plane_helper_check_state(state, &clip,
-					   min_scale, max_scale,
-					   true, true);
+	ret = drm_atomic_helper_check_plane_state(new_plane_state, crtc_state,
+						  min_scale, max_scale,
+						  true, true);
 	if (ret)
 		return ret;
 
-	if (!state->visible)
+	if (!new_plane_state->visible)
 		return 0;
 
 	ret = vop_convert_format(fb->format->format);
@@ -682,17 +848,52 @@ static int vop_plane_atomic_check(struct drm_plane *plane,
 	 * Src.x1 can be odd when do clip, but yuv plane start point
 	 * need align with 2 pixel.
 	 */
-	if (is_yuv_support(fb->format->format) && ((state->src.x1 >> 16) % 2))
+	if (fb->format->is_yuv && ((new_plane_state->src.x1 >> 16) % 2)) {
+		DRM_DEBUG_KMS("Invalid Source: Yuv format not support odd xpos\n");
+		return -EINVAL;
+	}
+
+	if (fb->format->is_yuv && new_plane_state->rotation & DRM_MODE_REFLECT_Y) {
+		DRM_DEBUG_KMS("Invalid Source: Yuv format does not support this rotation\n");
 		return -EINVAL;
+	}
+
+	if (rockchip_afbc(fb->modifier)) {
+		struct vop *vop = to_vop(crtc);
+
+		if (!vop->data->afbc) {
+			DRM_DEBUG_KMS("vop does not support AFBC\n");
+			return -EINVAL;
+		}
+
+		ret = vop_convert_afbc_format(fb->format->format);
+		if (ret < 0)
+			return ret;
+
+		if (new_plane_state->src.x1 || new_plane_state->src.y1) {
+			DRM_DEBUG_KMS("AFBC does not support offset display, " \
+				      "xpos=%d, ypos=%d, offset=%d\n",
+				      new_plane_state->src.x1, new_plane_state->src.y1,
+				      fb->offsets[0]);
+			return -EINVAL;
+		}
+
+		if (new_plane_state->rotation && new_plane_state->rotation != DRM_MODE_ROTATE_0) {
+			DRM_DEBUG_KMS("No rotation support in AFBC, rotation=%d\n",
+				      new_plane_state->rotation);
+			return -EINVAL;
+		}
+	}
 
 	return 0;
 }
 
 static void vop_plane_atomic_disable(struct drm_plane *plane,
-				     struct drm_plane_state *old_state)
+				     struct drm_atomic_state *state)
 {
+	struct drm_plane_state *old_state = drm_atomic_get_old_plane_state(state,
+									   plane);
 	struct vop_win *vop_win = to_vop_win(plane);
-	const struct vop_win_data *win = vop_win->data;
 	struct vop *vop = to_vop(old_state->crtc);
 
 	if (!old_state->crtc)
@@ -700,32 +901,37 @@ static void vop_plane_atomic_disable(struct drm_plane *plane,
 
 	spin_lock(&vop->reg_lock);
 
-	VOP_WIN_SET(vop, win, enable, 0);
+	vop_win_disable(vop, vop_win);
 
 	spin_unlock(&vop->reg_lock);
 }
 
 static void vop_plane_atomic_update(struct drm_plane *plane,
-		struct drm_plane_state *old_state)
+		struct drm_atomic_state *state)
 {
-	struct drm_plane_state *state = plane->state;
-	struct drm_crtc *crtc = state->crtc;
+	struct drm_plane_state *new_state = drm_atomic_get_new_plane_state(state,
+									   plane);
+	struct drm_crtc *crtc = new_state->crtc;
 	struct vop_win *vop_win = to_vop_win(plane);
 	const struct vop_win_data *win = vop_win->data;
-	struct vop *vop = to_vop(state->crtc);
-	struct drm_framebuffer *fb = state->fb;
+	const struct vop_win_yuv2yuv_data *win_yuv2yuv = vop_win->yuv2yuv_data;
+	struct vop *vop = to_vop(new_state->crtc);
+	struct drm_framebuffer *fb = new_state->fb;
 	unsigned int actual_w, actual_h;
 	unsigned int dsp_stx, dsp_sty;
 	uint32_t act_info, dsp_info, dsp_st;
-	struct drm_rect *src = &state->src;
-	struct drm_rect *dest = &state->dst;
+	struct drm_rect *src = &new_state->src;
+	struct drm_rect *dest = &new_state->dst;
 	struct drm_gem_object *obj, *uv_obj;
 	struct rockchip_gem_object *rk_obj, *rk_uv_obj;
 	unsigned long offset;
 	dma_addr_t dma_addr;
 	uint32_t val;
-	bool rb_swap;
+	bool rb_swap, uv_swap;
+	int win_index = VOP_WIN_TO_INDEX(vop_win);
 	int format;
+	int is_yuv = fb->format->is_yuv;
+	int i;
 
 	/*
 	 * can't update plane when vop is disabled.
@@ -736,12 +942,12 @@ static void vop_plane_atomic_update(struct drm_plane *plane,
 	if (WARN_ON(!vop->is_enabled))
 		return;
 
-	if (!state->visible) {
-		vop_plane_atomic_disable(plane, old_state);
+	if (!new_state->visible) {
+		vop_plane_atomic_disable(plane, state);
 		return;
 	}
 
-	obj = rockchip_fb_get_gem_obj(fb, 0);
+	obj = fb->obj[0];
 	rk_obj = to_rockchip_obj(obj);
 
 	actual_w = drm_rect_width(src) >> 16;
@@ -755,46 +961,93 @@ static void vop_plane_atomic_update(struct drm_plane *plane,
 	dsp_sty = dest->y1 + crtc->mode.vtotal - crtc->mode.vsync_start;
 	dsp_st = dsp_sty << 16 | (dsp_stx & 0xffff);
 
-	offset = (src->x1 >> 16) * fb->format->cpp[0];
+	if (fb->format->char_per_block[0])
+		offset = drm_format_info_min_pitch(fb->format, 0,
+						   src->x1 >> 16);
+	else
+		offset = (src->x1 >> 16) * fb->format->cpp[0];
+
 	offset += (src->y1 >> 16) * fb->pitches[0];
 	dma_addr = rk_obj->dma_addr + offset + fb->offsets[0];
 
+	/*
+	 * For y-mirroring we need to move address
+	 * to the beginning of the last line.
+	 */
+	if (new_state->rotation & DRM_MODE_REFLECT_Y)
+		dma_addr += (actual_h - 1) * fb->pitches[0];
+
 	format = vop_convert_format(fb->format->format);
 
 	spin_lock(&vop->reg_lock);
 
+	if (rockchip_afbc(fb->modifier)) {
+		int afbc_format = vop_convert_afbc_format(fb->format->format);
+
+		VOP_AFBC_SET(vop, format, afbc_format | AFBC_TILE_16x16);
+		VOP_AFBC_SET(vop, hreg_block_split, 0);
+		VOP_AFBC_SET(vop, win_sel, VOP_WIN_TO_INDEX(vop_win));
+		VOP_AFBC_SET(vop, hdr_ptr, dma_addr);
+		VOP_AFBC_SET(vop, pic_size, act_info);
+	}
+
 	VOP_WIN_SET(vop, win, format, format);
-	VOP_WIN_SET(vop, win, yrgb_vir, fb->pitches[0] >> 2);
+	VOP_WIN_SET(vop, win, fmt_10, is_fmt_10(fb->format->format));
+	VOP_WIN_SET(vop, win, yrgb_vir, DIV_ROUND_UP(fb->pitches[0], 4));
 	VOP_WIN_SET(vop, win, yrgb_mst, dma_addr);
-	if (is_yuv_support(fb->format->format)) {
-		int hsub = drm_format_horz_chroma_subsampling(fb->format->format);
-		int vsub = drm_format_vert_chroma_subsampling(fb->format->format);
-		int bpp = fb->format->cpp[1];
+	VOP_WIN_YUV2YUV_SET(vop, win_yuv2yuv, y2r_en, is_yuv);
+	VOP_WIN_SET(vop, win, y_mir_en,
+		    (new_state->rotation & DRM_MODE_REFLECT_Y) ? 1 : 0);
+	VOP_WIN_SET(vop, win, x_mir_en,
+		    (new_state->rotation & DRM_MODE_REFLECT_X) ? 1 : 0);
 
-		uv_obj = rockchip_fb_get_gem_obj(fb, 1);
+	if (is_yuv) {
+		uv_obj = fb->obj[1];
 		rk_uv_obj = to_rockchip_obj(uv_obj);
 
-		offset = (src->x1 >> 16) * bpp / hsub;
-		offset += (src->y1 >> 16) * fb->pitches[1] / vsub;
+		if (fb->format->char_per_block[1])
+			offset = drm_format_info_min_pitch(fb->format, 1,
+							   src->x1 >> 16);
+		else
+			offset = (src->x1 >> 16) * fb->format->cpp[1];
+		offset /= fb->format->hsub;
+		offset += (src->y1 >> 16) * fb->pitches[1] / fb->format->vsub;
 
 		dma_addr = rk_uv_obj->dma_addr + offset + fb->offsets[1];
-		VOP_WIN_SET(vop, win, uv_vir, fb->pitches[1] >> 2);
+		VOP_WIN_SET(vop, win, uv_vir, DIV_ROUND_UP(fb->pitches[1], 4));
 		VOP_WIN_SET(vop, win, uv_mst, dma_addr);
+
+		for (i = 0; i < NUM_YUV2YUV_COEFFICIENTS; i++) {
+			VOP_WIN_YUV2YUV_COEFFICIENT_SET(vop,
+							win_yuv2yuv,
+							y2r_coefficients[i],
+							bt601_yuv2rgb[i]);
+		}
+
+		uv_swap = has_uv_swapped(fb->format->format);
+		VOP_WIN_SET(vop, win, uv_swap, uv_swap);
 	}
 
 	if (win->phy->scl)
 		scl_vop_cal_scl_fac(vop, win, actual_w, actual_h,
 				    drm_rect_width(dest), drm_rect_height(dest),
-				    fb->format->format);
+				    fb->format);
 
 	VOP_WIN_SET(vop, win, act_info, act_info);
 	VOP_WIN_SET(vop, win, dsp_info, dsp_info);
 	VOP_WIN_SET(vop, win, dsp_st, dsp_st);
 
-	rb_swap = has_rb_swapped(fb->format->format);
+	rb_swap = has_rb_swapped(vop->data->version, fb->format->format);
 	VOP_WIN_SET(vop, win, rb_swap, rb_swap);
 
-	if (is_alpha_support(fb->format->format)) {
+	/*
+	 * Blending win0 with the background color doesn't seem to work
+	 * correctly. We only get the background color, no matter the contents
+	 * of the win0 framebuffer.  However, blending pre-multiplied color
+	 * with the default opaque black default background color is a no-op,
+	 * so we can just disable blending to get the correct result.
+	 */
+	if (fb->format->has_alpha && win_index > 0) {
 		VOP_WIN_SET(vop, win, dst_alpha_ctl,
 			    DST_FACTOR_M0(ALPHA_SRC_INVERSE));
 		val = SRC_ALPHA_EN(1) | SRC_COLOR_M0(ALPHA_SRC_PRE_MUL) |
@@ -803,27 +1056,111 @@ static void vop_plane_atomic_update(struct drm_plane *plane,
 			SRC_ALPHA_CAL_M0(ALPHA_NO_SATURATION) |
 			SRC_FACTOR_M0(ALPHA_ONE);
 		VOP_WIN_SET(vop, win, src_alpha_ctl, val);
+
+		VOP_WIN_SET(vop, win, alpha_pre_mul, ALPHA_SRC_PRE_MUL);
+		VOP_WIN_SET(vop, win, alpha_mode, ALPHA_PER_PIX);
+		VOP_WIN_SET(vop, win, alpha_en, 1);
 	} else {
 		VOP_WIN_SET(vop, win, src_alpha_ctl, SRC_ALPHA_EN(0));
+		VOP_WIN_SET(vop, win, alpha_en, 0);
 	}
 
 	VOP_WIN_SET(vop, win, enable, 1);
+	vop->win_enabled |= BIT(win_index);
 	spin_unlock(&vop->reg_lock);
 }
 
+static int vop_plane_atomic_async_check(struct drm_plane *plane,
+					struct drm_atomic_state *state, bool flip)
+{
+	struct drm_plane_state *new_plane_state = drm_atomic_get_new_plane_state(state,
+										 plane);
+	struct vop_win *vop_win = to_vop_win(plane);
+	const struct vop_win_data *win = vop_win->data;
+	int min_scale = win->phy->scl ? FRAC_16_16(1, 8) :
+					DRM_PLANE_NO_SCALING;
+	int max_scale = win->phy->scl ? FRAC_16_16(8, 1) :
+					DRM_PLANE_NO_SCALING;
+	struct drm_crtc_state *crtc_state;
+
+	if (plane != new_plane_state->crtc->cursor)
+		return -EINVAL;
+
+	if (!plane->state)
+		return -EINVAL;
+
+	if (!plane->state->fb)
+		return -EINVAL;
+
+	crtc_state = drm_atomic_get_new_crtc_state(state,
+						   new_plane_state->crtc);
+
+	/* Special case for asynchronous cursor updates. */
+	if (!crtc_state)
+		crtc_state = plane->crtc->state;
+
+	return drm_atomic_helper_check_plane_state(plane->state, crtc_state,
+						   min_scale, max_scale,
+						   true, true);
+}
+
+static void vop_plane_atomic_async_update(struct drm_plane *plane,
+					  struct drm_atomic_state *state)
+{
+	struct drm_plane_state *new_state = drm_atomic_get_new_plane_state(state,
+									   plane);
+	struct vop *vop = to_vop(plane->state->crtc);
+	struct drm_framebuffer *old_fb = plane->state->fb;
+
+	plane->state->crtc_x = new_state->crtc_x;
+	plane->state->crtc_y = new_state->crtc_y;
+	plane->state->crtc_h = new_state->crtc_h;
+	plane->state->crtc_w = new_state->crtc_w;
+	plane->state->src_x = new_state->src_x;
+	plane->state->src_y = new_state->src_y;
+	plane->state->src_h = new_state->src_h;
+	plane->state->src_w = new_state->src_w;
+	swap(plane->state->fb, new_state->fb);
+
+	if (vop->is_enabled) {
+		vop_plane_atomic_update(plane, state);
+		spin_lock(&vop->reg_lock);
+		vop_cfg_done(vop);
+		spin_unlock(&vop->reg_lock);
+
+		/*
+		 * A scanout can still be occurring, so we can't drop the
+		 * reference to the old framebuffer. To solve this we get a
+		 * reference to old_fb and set a worker to release it later.
+		 * FIXME: if we perform 500 async_update calls before the
+		 * vblank, then we can have 500 different framebuffers waiting
+		 * to be released.
+		 */
+		if (old_fb && plane->state->fb != old_fb) {
+			drm_framebuffer_get(old_fb);
+			WARN_ON(drm_crtc_vblank_get(plane->state->crtc) != 0);
+			drm_flip_work_queue(&vop->fb_unref_work, old_fb);
+			set_bit(VOP_PENDING_FB_UNREF, &vop->pending);
+		}
+	}
+}
+
 static const struct drm_plane_helper_funcs plane_helper_funcs = {
 	.atomic_check = vop_plane_atomic_check,
 	.atomic_update = vop_plane_atomic_update,
 	.atomic_disable = vop_plane_atomic_disable,
+	.atomic_async_check = vop_plane_atomic_async_check,
+	.atomic_async_update = vop_plane_atomic_async_update,
 };
 
 static const struct drm_plane_funcs vop_plane_funcs = {
 	.update_plane	= drm_atomic_helper_update_plane,
 	.disable_plane	= drm_atomic_helper_disable_plane,
-	.destroy = vop_plane_destroy,
+	.destroy = drm_plane_cleanup,
 	.reset = drm_atomic_helper_plane_reset,
 	.atomic_duplicate_state = drm_atomic_helper_plane_duplicate_state,
 	.atomic_destroy_state = drm_atomic_helper_plane_destroy_state,
+	.format_mod_supported = rockchip_mod_supported,
 };
 
 static int vop_crtc_enable_vblank(struct drm_crtc *crtc)
@@ -859,20 +1196,174 @@ static void vop_crtc_disable_vblank(struct drm_crtc *crtc)
 	spin_unlock_irqrestore(&vop->irq_lock, flags);
 }
 
+static enum drm_mode_status vop_crtc_mode_valid(struct drm_crtc *crtc,
+						const struct drm_display_mode *mode)
+{
+	struct vop *vop = to_vop(crtc);
+
+	if (vop->data->max_output.width && mode->hdisplay > vop->data->max_output.width)
+		return MODE_BAD_HVALUE;
+
+	return MODE_OK;
+}
+
 static bool vop_crtc_mode_fixup(struct drm_crtc *crtc,
 				const struct drm_display_mode *mode,
 				struct drm_display_mode *adjusted_mode)
 {
 	struct vop *vop = to_vop(crtc);
+	unsigned long rate;
 
-	adjusted_mode->clock =
-		clk_round_rate(vop->dclk, mode->clock * 1000) / 1000;
+	/*
+	 * Clock craziness.
+	 *
+	 * Key points:
+	 *
+	 * - DRM works in kHz.
+	 * - Clock framework works in Hz.
+	 * - Rockchip's clock driver picks the clock rate that is the
+	 *   same _OR LOWER_ than the one requested.
+	 *
+	 * Action plan:
+	 *
+	 * 1. Try to set the exact rate first, and confirm the clock framework
+	 *    can provide it.
+	 *
+	 * 2. If the clock framework cannot provide the exact rate, we should
+	 *    add 999 Hz to the requested rate.  That way if the clock we need
+	 *    is 60000001 Hz (~60 MHz) and DRM tells us to make 60000 kHz then
+	 *    the clock framework will actually give us the right clock.
+	 *
+	 * 3. Get the clock framework to round the rate for us to tell us
+	 *    what it will actually make.
+	 *
+	 * 4. Store the rounded up rate so that we don't need to worry about
+	 *    this in the actual clk_set_rate().
+	 */
+	rate = clk_round_rate(vop->dclk, adjusted_mode->clock * 1000);
+	if (rate / 1000 != adjusted_mode->clock)
+		rate = clk_round_rate(vop->dclk,
+				      adjusted_mode->clock * 1000 + 999);
+	adjusted_mode->clock = DIV_ROUND_UP(rate, 1000);
 
 	return true;
 }
 
-static void vop_crtc_enable(struct drm_crtc *crtc)
+static bool vop_dsp_lut_is_enabled(struct vop *vop)
+{
+	return vop_read_reg(vop, 0, &vop->data->common->dsp_lut_en);
+}
+
+static u32 vop_lut_buffer_index(struct vop *vop)
+{
+	return vop_read_reg(vop, 0, &vop->data->common->lut_buffer_index);
+}
+
+static void vop_crtc_write_gamma_lut(struct vop *vop, struct drm_crtc *crtc)
 {
+	struct drm_color_lut *lut = crtc->state->gamma_lut->data;
+	unsigned int i, bpc = ilog2(vop->data->lut_size);
+
+	for (i = 0; i < crtc->gamma_size; i++) {
+		u32 word;
+
+		word = (drm_color_lut_extract(lut[i].red, bpc) << (2 * bpc)) |
+		       (drm_color_lut_extract(lut[i].green, bpc) << bpc) |
+			drm_color_lut_extract(lut[i].blue, bpc);
+		writel(word, vop->lut_regs + i * 4);
+	}
+}
+
+static void vop_crtc_gamma_set(struct vop *vop, struct drm_crtc *crtc,
+			       struct drm_crtc_state *old_state)
+{
+	struct drm_crtc_state *state = crtc->state;
+	unsigned int idle;
+	u32 lut_idx, old_idx;
+	int ret;
+
+	if (!vop->lut_regs)
+		return;
+
+	if (!state->gamma_lut || !VOP_HAS_REG(vop, common, update_gamma_lut)) {
+		/*
+		 * To disable gamma (gamma_lut is null) or to write
+		 * an update to the LUT, clear dsp_lut_en.
+		 */
+		spin_lock(&vop->reg_lock);
+		VOP_REG_SET(vop, common, dsp_lut_en, 0);
+		vop_cfg_done(vop);
+		spin_unlock(&vop->reg_lock);
+
+		/*
+		 * In order to write the LUT to the internal memory,
+		 * we need to first make sure the dsp_lut_en bit is cleared.
+		 */
+		ret = readx_poll_timeout(vop_dsp_lut_is_enabled, vop,
+					 idle, !idle, 5, 30 * 1000);
+		if (ret) {
+			DRM_DEV_ERROR(vop->dev, "display LUT RAM enable timeout!\n");
+			return;
+		}
+
+		if (!state->gamma_lut)
+			return;
+	} else {
+		/*
+		 * On RK3399 the gamma LUT can updated without clearing dsp_lut_en,
+		 * by setting update_gamma_lut then waiting for lut_buffer_index change
+		 */
+		old_idx = vop_lut_buffer_index(vop);
+	}
+
+	spin_lock(&vop->reg_lock);
+	vop_crtc_write_gamma_lut(vop, crtc);
+	VOP_REG_SET(vop, common, dsp_lut_en, 1);
+	VOP_REG_SET(vop, common, update_gamma_lut, 1);
+	vop_cfg_done(vop);
+	spin_unlock(&vop->reg_lock);
+
+	if (VOP_HAS_REG(vop, common, update_gamma_lut)) {
+		ret = readx_poll_timeout(vop_lut_buffer_index, vop,
+					 lut_idx, lut_idx != old_idx, 5, 30 * 1000);
+		if (ret) {
+			DRM_DEV_ERROR(vop->dev, "gamma LUT update timeout!\n");
+			return;
+		}
+
+		/*
+		 * update_gamma_lut is auto cleared by HW, but write 0 to clear the bit
+		 * in our backup of the regs.
+		 */
+		spin_lock(&vop->reg_lock);
+		VOP_REG_SET(vop, common, update_gamma_lut, 0);
+		spin_unlock(&vop->reg_lock);
+	}
+}
+
+static void vop_crtc_atomic_begin(struct drm_crtc *crtc,
+				  struct drm_atomic_state *state)
+{
+	struct drm_crtc_state *crtc_state = drm_atomic_get_new_crtc_state(state,
+									  crtc);
+	struct drm_crtc_state *old_crtc_state = drm_atomic_get_old_crtc_state(state,
+									      crtc);
+	struct vop *vop = to_vop(crtc);
+
+	/*
+	 * Only update GAMMA if the 'active' flag is not changed,
+	 * otherwise it's updated by .atomic_enable.
+	 */
+	if (crtc_state->color_mgmt_changed &&
+	    !crtc_state->active_changed)
+		vop_crtc_gamma_set(vop, crtc, old_crtc_state);
+}
+
+static void vop_crtc_atomic_enable(struct drm_crtc *crtc,
+				   struct drm_atomic_state *state)
+{
+	struct drm_crtc_state *old_state = drm_atomic_get_old_crtc_state(state,
+									 crtc);
 	struct vop *vop = to_vop(crtc);
 	const struct vop_data *vop_data = vop->data;
 	struct rockchip_crtc_state *s = to_rockchip_crtc_state(crtc->state);
@@ -888,80 +1379,59 @@ static void vop_crtc_enable(struct drm_crtc *crtc)
 	u16 vact_st = adjusted_mode->vtotal - adjusted_mode->vsync_start;
 	u16 vact_end = vact_st + vdisplay;
 	uint32_t pin_pol, val;
+	int dither_bpc = s->output_bpc ? s->output_bpc : 10;
 	int ret;
 
-	WARN_ON(vop->event);
-
-	ret = vop_enable(crtc);
-	if (ret) {
-		DRM_DEV_ERROR(vop->dev, "Failed to enable vop (%d)\n", ret);
+	if (old_state && old_state->self_refresh_active) {
+		drm_crtc_vblank_on(crtc);
+		rockchip_drm_set_win_enabled(crtc, true);
 		return;
 	}
 
-	/*
-	 * If dclk rate is zero, mean that scanout is stop,
-	 * we don't need wait any more.
-	 */
-	if (clk_get_rate(vop->dclk)) {
-		/*
-		 * Rk3288 vop timing register is immediately, when configure
-		 * display timing on display time, may cause tearing.
-		 *
-		 * Vop standby will take effect at end of current frame,
-		 * if dsp hold valid irq happen, it means standby complete.
-		 *
-		 * mode set:
-		 *    standby and wait complete --> |----
-		 *                                  | display time
-		 *                                  |----
-		 *                                  |---> dsp hold irq
-		 *     configure display timing --> |
-		 *         standby exit             |
-		 *                                  | new frame start.
-		 */
-
-		reinit_completion(&vop->dsp_hold_completion);
-		vop_dsp_hold_valid_irq_enable(vop);
-
-		spin_lock(&vop->reg_lock);
-
-		VOP_CTRL_SET(vop, standby, 1);
-
-		spin_unlock(&vop->reg_lock);
+	mutex_lock(&vop->vop_lock);
 
-		wait_for_completion(&vop->dsp_hold_completion);
+	WARN_ON(vop->event);
 
-		vop_dsp_hold_valid_irq_disable(vop);
+	ret = vop_enable(crtc, old_state);
+	if (ret) {
+		mutex_unlock(&vop->vop_lock);
+		DRM_DEV_ERROR(vop->dev, "Failed to enable vop (%d)\n", ret);
+		return;
 	}
-
-	pin_pol = BIT(DCLK_INVERT);
-	pin_pol |= (adjusted_mode->flags & DRM_MODE_FLAG_PHSYNC) ?
+	pin_pol = (adjusted_mode->flags & DRM_MODE_FLAG_PHSYNC) ?
 		   BIT(HSYNC_POSITIVE) : 0;
 	pin_pol |= (adjusted_mode->flags & DRM_MODE_FLAG_PVSYNC) ?
 		   BIT(VSYNC_POSITIVE) : 0;
-	VOP_CTRL_SET(vop, pin_pol, pin_pol);
+	VOP_REG_SET(vop, output, pin_pol, pin_pol);
+	VOP_REG_SET(vop, output, mipi_dual_channel_en, 0);
 
 	switch (s->output_type) {
 	case DRM_MODE_CONNECTOR_LVDS:
-		VOP_CTRL_SET(vop, rgb_en, 1);
-		VOP_CTRL_SET(vop, rgb_pin_pol, pin_pol);
+		VOP_REG_SET(vop, output, rgb_dclk_pol, 1);
+		VOP_REG_SET(vop, output, rgb_pin_pol, pin_pol);
+		VOP_REG_SET(vop, output, rgb_en, 1);
 		break;
 	case DRM_MODE_CONNECTOR_eDP:
-		VOP_CTRL_SET(vop, edp_pin_pol, pin_pol);
-		VOP_CTRL_SET(vop, edp_en, 1);
+		VOP_REG_SET(vop, output, edp_dclk_pol, 1);
+		VOP_REG_SET(vop, output, edp_pin_pol, pin_pol);
+		VOP_REG_SET(vop, output, edp_en, 1);
 		break;
 	case DRM_MODE_CONNECTOR_HDMIA:
-		VOP_CTRL_SET(vop, hdmi_pin_pol, pin_pol);
-		VOP_CTRL_SET(vop, hdmi_en, 1);
+		VOP_REG_SET(vop, output, hdmi_dclk_pol, 1);
+		VOP_REG_SET(vop, output, hdmi_pin_pol, pin_pol);
+		VOP_REG_SET(vop, output, hdmi_en, 1);
 		break;
 	case DRM_MODE_CONNECTOR_DSI:
-		VOP_CTRL_SET(vop, mipi_pin_pol, pin_pol);
-		VOP_CTRL_SET(vop, mipi_en, 1);
+		VOP_REG_SET(vop, output, mipi_dclk_pol, 1);
+		VOP_REG_SET(vop, output, mipi_pin_pol, pin_pol);
+		VOP_REG_SET(vop, output, mipi_en, 1);
+		VOP_REG_SET(vop, output, mipi_dual_channel_en,
+			    !!(s->output_flags & ROCKCHIP_OUTPUT_DSI_DUAL));
 		break;
 	case DRM_MODE_CONNECTOR_DisplayPort:
-		pin_pol &= ~BIT(DCLK_INVERT);
-		VOP_CTRL_SET(vop, dp_pin_pol, pin_pol);
-		VOP_CTRL_SET(vop, dp_en, 1);
+		VOP_REG_SET(vop, output, dp_dclk_pol, 0);
+		VOP_REG_SET(vop, output, dp_pin_pol, pin_pol);
+		VOP_REG_SET(vop, output, dp_en, 1);
 		break;
 	default:
 		DRM_DEV_ERROR(vop->dev, "unsupported connector_type [%d]\n",
@@ -974,27 +1444,48 @@ static void vop_crtc_enable(struct drm_crtc *crtc)
 	if (s->output_mode == ROCKCHIP_OUT_MODE_AAAA &&
 	    !(vop_data->feature & VOP_FEATURE_OUTPUT_RGB10))
 		s->output_mode = ROCKCHIP_OUT_MODE_P888;
-	VOP_CTRL_SET(vop, out_mode, s->output_mode);
 
-	VOP_CTRL_SET(vop, htotal_pw, (htotal << 16) | hsync_len);
+	if (s->output_mode == ROCKCHIP_OUT_MODE_AAAA && dither_bpc <= 8)
+		VOP_REG_SET(vop, common, pre_dither_down, 1);
+	else
+		VOP_REG_SET(vop, common, pre_dither_down, 0);
+
+	if (dither_bpc == 6) {
+		VOP_REG_SET(vop, common, dither_down_sel, DITHER_DOWN_ALLEGRO);
+		VOP_REG_SET(vop, common, dither_down_mode, RGB888_TO_RGB666);
+		VOP_REG_SET(vop, common, dither_down_en, 1);
+	} else {
+		VOP_REG_SET(vop, common, dither_down_en, 0);
+	}
+
+	VOP_REG_SET(vop, common, out_mode, s->output_mode);
+
+	VOP_REG_SET(vop, modeset, htotal_pw, (htotal << 16) | hsync_len);
 	val = hact_st << 16;
 	val |= hact_end;
-	VOP_CTRL_SET(vop, hact_st_end, val);
-	VOP_CTRL_SET(vop, hpost_st_end, val);
+	VOP_REG_SET(vop, modeset, hact_st_end, val);
+	VOP_REG_SET(vop, modeset, hpost_st_end, val);
 
-	VOP_CTRL_SET(vop, vtotal_pw, (vtotal << 16) | vsync_len);
+	VOP_REG_SET(vop, modeset, vtotal_pw, (vtotal << 16) | vsync_len);
 	val = vact_st << 16;
 	val |= vact_end;
-	VOP_CTRL_SET(vop, vact_st_end, val);
-	VOP_CTRL_SET(vop, vpost_st_end, val);
+	VOP_REG_SET(vop, modeset, vact_st_end, val);
+	VOP_REG_SET(vop, modeset, vpost_st_end, val);
 
-	VOP_CTRL_SET(vop, line_flag_num[0], vact_end);
+	VOP_REG_SET(vop, intr, line_flag_num[0], vact_end);
 
 	clk_set_rate(vop->dclk, adjusted_mode->clock * 1000);
 
-	VOP_CTRL_SET(vop, standby, 0);
+	VOP_REG_SET(vop, common, standby, 0);
+	mutex_unlock(&vop->vop_lock);
 
-	rockchip_drm_psr_activate(&vop->crtc);
+	/*
+	 * If we have a GAMMA LUT in the state, then let's make sure
+	 * it's updated. We might be coming out of suspend,
+	 * which means the LUT internal memory needs to be re-written.
+	 */
+	if (crtc->state->gamma_lut)
+		vop_crtc_gamma_set(vop, crtc, old_state);
 }
 
 static bool vop_fs_irq_is_pending(struct vop *vop)
@@ -1023,13 +1514,63 @@ static void vop_wait_for_irq_handler(struct vop *vop)
 	synchronize_irq(vop->irq);
 }
 
+static int vop_crtc_atomic_check(struct drm_crtc *crtc,
+				 struct drm_atomic_state *state)
+{
+	struct drm_crtc_state *crtc_state = drm_atomic_get_new_crtc_state(state,
+									  crtc);
+	struct vop *vop = to_vop(crtc);
+	struct drm_plane *plane;
+	struct drm_plane_state *plane_state;
+	struct rockchip_crtc_state *s;
+	int afbc_planes = 0;
+
+	if (vop->lut_regs && crtc_state->color_mgmt_changed &&
+	    crtc_state->gamma_lut) {
+		unsigned int len;
+
+		len = drm_color_lut_size(crtc_state->gamma_lut);
+		if (len != crtc->gamma_size) {
+			DRM_DEBUG_KMS("Invalid LUT size; got %d, expected %d\n",
+				      len, crtc->gamma_size);
+			return -EINVAL;
+		}
+	}
+
+	drm_atomic_crtc_state_for_each_plane(plane, crtc_state) {
+		plane_state =
+			drm_atomic_get_plane_state(crtc_state->state, plane);
+		if (IS_ERR(plane_state)) {
+			DRM_DEBUG_KMS("Cannot get plane state for plane %s\n",
+				      plane->name);
+			return PTR_ERR(plane_state);
+		}
+
+		if (drm_is_afbc(plane_state->fb->modifier))
+			++afbc_planes;
+	}
+
+	if (afbc_planes > 1) {
+		DRM_DEBUG_KMS("Invalid number of AFBC planes; got %d, expected at most 1\n", afbc_planes);
+		return -EINVAL;
+	}
+
+	s = to_rockchip_crtc_state(crtc_state);
+	s->enable_afbc = afbc_planes > 0;
+
+	return 0;
+}
+
 static void vop_crtc_atomic_flush(struct drm_crtc *crtc,
-				  struct drm_crtc_state *old_crtc_state)
+				  struct drm_atomic_state *state)
 {
+	struct drm_crtc_state *old_crtc_state = drm_atomic_get_old_crtc_state(state,
+									      crtc);
 	struct drm_atomic_state *old_state = old_crtc_state->state;
-	struct drm_plane_state *old_plane_state;
+	struct drm_plane_state *old_plane_state, *new_plane_state;
 	struct vop *vop = to_vop(crtc);
 	struct drm_plane *plane;
+	struct rockchip_crtc_state *s;
 	int i;
 
 	if (WARN_ON(!vop->is_enabled))
@@ -1037,8 +1578,15 @@ static void vop_crtc_atomic_flush(struct drm_crtc *crtc,
 
 	spin_lock(&vop->reg_lock);
 
+	/* Enable AFBC if there is some AFBC window, disable otherwise. */
+	s = to_rockchip_crtc_state(crtc->state);
+	VOP_AFBC_SET(vop, enable, s->enable_afbc);
 	vop_cfg_done(vop);
 
+	/* Ack the DMA transfer of the previous frame (RK3066). */
+	if (VOP_HAS_REG(vop, common, dma_stop))
+		VOP_REG_SET(vop, common, dma_stop, 0);
+
 	spin_unlock(&vop->reg_lock);
 
 	/*
@@ -1058,55 +1606,40 @@ static void vop_crtc_atomic_flush(struct drm_crtc *crtc,
 	}
 	spin_unlock_irq(&crtc->dev->event_lock);
 
-	for_each_plane_in_state(old_state, plane, old_plane_state, i) {
+	for_each_oldnew_plane_in_state(old_state, plane, old_plane_state,
+				       new_plane_state, i) {
 		if (!old_plane_state->fb)
 			continue;
 
-		if (old_plane_state->fb == plane->state->fb)
+		if (old_plane_state->fb == new_plane_state->fb)
 			continue;
 
-		drm_framebuffer_reference(old_plane_state->fb);
+		drm_framebuffer_get(old_plane_state->fb);
+		WARN_ON(drm_crtc_vblank_get(crtc) != 0);
 		drm_flip_work_queue(&vop->fb_unref_work, old_plane_state->fb);
 		set_bit(VOP_PENDING_FB_UNREF, &vop->pending);
-		WARN_ON(drm_crtc_vblank_get(crtc) != 0);
 	}
 }
 
-static void vop_crtc_atomic_begin(struct drm_crtc *crtc,
-				  struct drm_crtc_state *old_crtc_state)
-{
-	rockchip_drm_psr_flush(crtc);
-}
-
 static const struct drm_crtc_helper_funcs vop_crtc_helper_funcs = {
-	.enable = vop_crtc_enable,
-	.disable = vop_crtc_disable,
+	.mode_valid = vop_crtc_mode_valid,
 	.mode_fixup = vop_crtc_mode_fixup,
-	.atomic_flush = vop_crtc_atomic_flush,
+	.atomic_check = vop_crtc_atomic_check,
 	.atomic_begin = vop_crtc_atomic_begin,
+	.atomic_flush = vop_crtc_atomic_flush,
+	.atomic_enable = vop_crtc_atomic_enable,
+	.atomic_disable = vop_crtc_atomic_disable,
 };
 
-static void vop_crtc_destroy(struct drm_crtc *crtc)
-{
-	drm_crtc_cleanup(crtc);
-}
-
-static void vop_crtc_reset(struct drm_crtc *crtc)
-{
-	if (crtc->state)
-		__drm_atomic_helper_crtc_destroy_state(crtc->state);
-	kfree(crtc->state);
-
-	crtc->state = kzalloc(sizeof(struct rockchip_crtc_state), GFP_KERNEL);
-	if (crtc->state)
-		crtc->state->crtc = crtc;
-}
-
 static struct drm_crtc_state *vop_crtc_duplicate_state(struct drm_crtc *crtc)
 {
 	struct rockchip_crtc_state *rockchip_state;
 
-	rockchip_state = kzalloc(sizeof(*rockchip_state), GFP_KERNEL);
+	if (WARN_ON(!crtc->state))
+		return NULL;
+
+	rockchip_state = kmemdup(to_rockchip_crtc_state(crtc->state),
+				 sizeof(*rockchip_state), GFP_KERNEL);
 	if (!rockchip_state)
 		return NULL;
 
@@ -1123,6 +1656,20 @@ static void vop_crtc_destroy_state(struct drm_crtc *crtc,
 	kfree(s);
 }
 
+static void vop_crtc_reset(struct drm_crtc *crtc)
+{
+	struct rockchip_crtc_state *crtc_state =
+		kzalloc(sizeof(*crtc_state), GFP_KERNEL);
+
+	if (crtc->state)
+		vop_crtc_destroy_state(crtc, crtc->state);
+
+	if (crtc_state)
+		__drm_atomic_helper_crtc_reset(crtc, &crtc_state->base);
+	else
+		__drm_atomic_helper_crtc_reset(crtc, NULL);
+}
+
 #ifdef CONFIG_DRM_ANALOGIX_DP
 static struct drm_connector *vop_get_edp_connector(struct vop *vop)
 {
@@ -1142,7 +1689,7 @@ static struct drm_connector *vop_get_edp_connector(struct vop *vop)
 }
 
 static int vop_crtc_set_crc_source(struct drm_crtc *crtc,
-				   const char *source_name, size_t *values_cnt)
+				   const char *source_name)
 {
 	struct vop *vop = to_vop(crtc);
 	struct drm_connector *connector;
@@ -1152,8 +1699,6 @@ static int vop_crtc_set_crc_source(struct drm_crtc *crtc,
 	if (!connector)
 		return -EINVAL;
 
-	*values_cnt = 3;
-
 	if (source_name && strcmp(source_name, "auto") == 0)
 		ret = analogix_dp_start_crc(connector);
 	else if (!source_name)
@@ -1163,9 +1708,28 @@ static int vop_crtc_set_crc_source(struct drm_crtc *crtc,
 
 	return ret;
 }
+
+static int
+vop_crtc_verify_crc_source(struct drm_crtc *crtc, const char *source_name,
+			   size_t *values_cnt)
+{
+	if (source_name && strcmp(source_name, "auto") != 0)
+		return -EINVAL;
+
+	*values_cnt = 3;
+	return 0;
+}
+
 #else
 static int vop_crtc_set_crc_source(struct drm_crtc *crtc,
-				   const char *source_name, size_t *values_cnt)
+				   const char *source_name)
+{
+	return -ENODEV;
+}
+
+static int
+vop_crtc_verify_crc_source(struct drm_crtc *crtc, const char *source_name,
+			   size_t *values_cnt)
 {
 	return -ENODEV;
 }
@@ -1174,13 +1738,14 @@ static int vop_crtc_set_crc_source(struct drm_crtc *crtc,
 static const struct drm_crtc_funcs vop_crtc_funcs = {
 	.set_config = drm_atomic_helper_set_config,
 	.page_flip = drm_atomic_helper_page_flip,
-	.destroy = vop_crtc_destroy,
+	.destroy = drm_crtc_cleanup,
 	.reset = vop_crtc_reset,
 	.atomic_duplicate_state = vop_crtc_duplicate_state,
 	.atomic_destroy_state = vop_crtc_destroy_state,
 	.enable_vblank = vop_crtc_enable_vblank,
 	.disable_vblank = vop_crtc_disable_vblank,
 	.set_crc_source = vop_crtc_set_crc_source,
+	.verify_crc_source = vop_crtc_verify_crc_source,
 };
 
 static void vop_fb_unref_worker(struct drm_flip_work *work, void *val)
@@ -1189,22 +1754,21 @@ static void vop_fb_unref_worker(struct drm_flip_work *work, void *val)
 	struct drm_framebuffer *fb = val;
 
 	drm_crtc_vblank_put(&vop->crtc);
-	drm_framebuffer_unreference(fb);
+	drm_framebuffer_put(fb);
 }
 
 static void vop_handle_vblank(struct vop *vop)
 {
 	struct drm_device *drm = vop->drm_dev;
 	struct drm_crtc *crtc = &vop->crtc;
-	unsigned long flags;
 
-	spin_lock_irqsave(&drm->event_lock, flags);
+	spin_lock(&drm->event_lock);
 	if (vop->event) {
 		drm_crtc_send_vblank_event(crtc, vop->event);
 		drm_crtc_vblank_put(crtc);
 		vop->event = NULL;
 	}
-	spin_unlock_irqrestore(&drm->event_lock, flags);
+	spin_unlock(&drm->event_lock);
 
 	if (test_and_clear_bit(VOP_PENDING_FB_UNREF, &vop->pending))
 		drm_flip_work_commit(&vop->fb_unref_work, system_unbound_wq);
@@ -1215,25 +1779,36 @@ static irqreturn_t vop_isr(int irq, void *data)
 	struct vop *vop = data;
 	struct drm_crtc *crtc = &vop->crtc;
 	uint32_t active_irqs;
-	unsigned long flags;
 	int ret = IRQ_NONE;
 
 	/*
+	 * The irq is shared with the iommu. If the runtime-pm state of the
+	 * vop-device is disabled the irq has to be targeted at the iommu.
+	 */
+	if (!pm_runtime_get_if_in_use(vop->dev))
+		return IRQ_NONE;
+
+	if (vop_core_clks_enable(vop)) {
+		DRM_DEV_ERROR_RATELIMITED(vop->dev, "couldn't enable clocks\n");
+		goto out;
+	}
+
+	/*
 	 * interrupt register has interrupt status, enable and clear bits, we
 	 * must hold irq_lock to avoid a race with enable/disable_vblank().
 	*/
-	spin_lock_irqsave(&vop->irq_lock, flags);
+	spin_lock(&vop->irq_lock);
 
 	active_irqs = VOP_INTR_GET_TYPE(vop, status, INTR_MASK);
 	/* Clear all active interrupt sources */
 	if (active_irqs)
 		VOP_INTR_SET_TYPE(vop, clear, active_irqs, 1);
 
-	spin_unlock_irqrestore(&vop->irq_lock, flags);
+	spin_unlock(&vop->irq_lock);
 
 	/* This is expected for vop iommu irqs, since the irq is shared */
 	if (!active_irqs)
-		return IRQ_NONE;
+		goto out_disable;
 
 	if (active_irqs & DSP_HOLD_VALID_INTR) {
 		complete(&vop->dsp_hold_completion);
@@ -1259,9 +1834,25 @@ static irqreturn_t vop_isr(int irq, void *data)
 		DRM_DEV_ERROR(vop->dev, "Unknown VOP IRQs: %#02x\n",
 			      active_irqs);
 
+out_disable:
+	vop_core_clks_disable(vop);
+out:
+	pm_runtime_put(vop->dev);
 	return ret;
 }
 
+static void vop_plane_add_properties(struct drm_plane *plane,
+				     const struct vop_win_data *win_data)
+{
+	unsigned int flags = 0;
+
+	flags |= VOP_WIN_HAS_REG(win_data, x_mir_en) ? DRM_MODE_REFLECT_X : 0;
+	flags |= VOP_WIN_HAS_REG(win_data, y_mir_en) ? DRM_MODE_REFLECT_Y : 0;
+	if (flags)
+		drm_plane_create_rotation_property(plane, DRM_MODE_ROTATE_0,
+						   DRM_MODE_ROTATE_0 | flags);
+}
+
 static int vop_create_crtc(struct vop *vop)
 {
 	const struct vop_data *vop_data = vop->data;
@@ -1290,6 +1881,7 @@ static int vop_create_crtc(struct vop *vop)
 					       0, &vop_plane_funcs,
 					       win_data->phy->data_formats,
 					       win_data->phy->nformats,
+					       win_data->phy->format_modifiers,
 					       win_data->type, NULL);
 		if (ret) {
 			DRM_DEV_ERROR(vop->dev, "failed to init plane %d\n",
@@ -1299,6 +1891,7 @@ static int vop_create_crtc(struct vop *vop)
 
 		plane = &vop_win->base;
 		drm_plane_helper_add(plane, &plane_helper_funcs);
+		vop_plane_add_properties(plane, win_data);
 		if (plane->type == DRM_PLANE_TYPE_PRIMARY)
 			primary = plane;
 		else if (plane->type == DRM_PLANE_TYPE_CURSOR)
@@ -1311,6 +1904,10 @@ static int vop_create_crtc(struct vop *vop)
 		goto err_cleanup_planes;
 
 	drm_crtc_helper_add(crtc, &vop_crtc_helper_funcs);
+	if (vop->lut_regs) {
+		drm_mode_crtc_set_gamma_size(crtc, vop_data->lut_size);
+		drm_crtc_enable_color_mgmt(crtc, 0, false, vop_data->lut_size);
+	}
 
 	/*
 	 * Create drm_planes for overlay windows with possible_crtcs restricted
@@ -1319,7 +1916,7 @@ static int vop_create_crtc(struct vop *vop)
 	for (i = 0; i < vop_data->win_size; i++) {
 		struct vop_win *vop_win = &vop->win[i];
 		const struct vop_win_data *win_data = vop_win->data;
-		unsigned long possible_crtcs = 1 << drm_crtc_index(crtc);
+		unsigned long possible_crtcs = drm_crtc_mask(crtc);
 
 		if (win_data->type != DRM_PLANE_TYPE_OVERLAY)
 			continue;
@@ -1329,6 +1926,7 @@ static int vop_create_crtc(struct vop *vop)
 					       &vop_plane_funcs,
 					       win_data->phy->data_formats,
 					       win_data->phy->nformats,
+					       win_data->phy->format_modifiers,
 					       win_data->type, NULL);
 		if (ret) {
 			DRM_DEV_ERROR(vop->dev, "failed to init overlay %d\n",
@@ -1336,12 +1934,13 @@ static int vop_create_crtc(struct vop *vop)
 			goto err_cleanup_crtc;
 		}
 		drm_plane_helper_add(&vop_win->base, &plane_helper_funcs);
+		vop_plane_add_properties(&vop_win->base, win_data);
 	}
 
 	port = of_get_child_by_name(dev->of_node, "port");
 	if (!port) {
-		DRM_DEV_ERROR(vop->dev, "no port node found in %s\n",
-			      dev->of_node->full_name);
+		DRM_DEV_ERROR(vop->dev, "no port node found in %pOF\n",
+			      dev->of_node);
 		ret = -ENOENT;
 		goto err_cleanup_crtc;
 	}
@@ -1353,6 +1952,12 @@ static int vop_create_crtc(struct vop *vop)
 	init_completion(&vop->line_flag_completion);
 	crtc->port = port;
 
+	ret = drm_self_refresh_helper_init(crtc);
+	if (ret)
+		DRM_DEV_DEBUG_KMS(vop->dev,
+			"Failed to init %s with SR helpers %d, ignoring\n",
+			crtc->name, ret);
+
 	return 0;
 
 err_cleanup_crtc:
@@ -1370,6 +1975,8 @@ static void vop_destroy_crtc(struct vop *vop)
 	struct drm_device *drm_dev = vop->drm_dev;
 	struct drm_plane *plane, *tmp;
 
+	drm_self_refresh_helper_cleanup(crtc);
+
 	of_node_put(crtc->port);
 
 	/*
@@ -1382,7 +1989,7 @@ static void vop_destroy_crtc(struct vop *vop)
 	 */
 	list_for_each_entry_safe(plane, tmp, &drm_dev->mode_config.plane_list,
 				 head)
-		vop_plane_destroy(plane);
+		drm_plane_cleanup(plane);
 
 	/*
 	 * Destroy CRTC after vop_plane_destroy() since vop_disable_plane()
@@ -1394,49 +2001,47 @@ static void vop_destroy_crtc(struct vop *vop)
 
 static int vop_initial(struct vop *vop)
 {
-	const struct vop_data *vop_data = vop->data;
-	const struct vop_reg_data *init_table = vop_data->init_table;
 	struct reset_control *ahb_rst;
 	int i, ret;
 
 	vop->hclk = devm_clk_get(vop->dev, "hclk_vop");
 	if (IS_ERR(vop->hclk)) {
-		dev_err(vop->dev, "failed to get hclk source\n");
+		DRM_DEV_ERROR(vop->dev, "failed to get hclk source\n");
 		return PTR_ERR(vop->hclk);
 	}
 	vop->aclk = devm_clk_get(vop->dev, "aclk_vop");
 	if (IS_ERR(vop->aclk)) {
-		dev_err(vop->dev, "failed to get aclk source\n");
+		DRM_DEV_ERROR(vop->dev, "failed to get aclk source\n");
 		return PTR_ERR(vop->aclk);
 	}
 	vop->dclk = devm_clk_get(vop->dev, "dclk_vop");
 	if (IS_ERR(vop->dclk)) {
-		dev_err(vop->dev, "failed to get dclk source\n");
+		DRM_DEV_ERROR(vop->dev, "failed to get dclk source\n");
 		return PTR_ERR(vop->dclk);
 	}
 
-	ret = pm_runtime_get_sync(vop->dev);
+	ret = pm_runtime_resume_and_get(vop->dev);
 	if (ret < 0) {
-		dev_err(vop->dev, "failed to get pm runtime: %d\n", ret);
+		DRM_DEV_ERROR(vop->dev, "failed to get pm runtime: %d\n", ret);
 		return ret;
 	}
 
 	ret = clk_prepare(vop->dclk);
 	if (ret < 0) {
-		dev_err(vop->dev, "failed to prepare dclk\n");
+		DRM_DEV_ERROR(vop->dev, "failed to prepare dclk\n");
 		goto err_put_pm_runtime;
 	}
 
 	/* Enable both the hclk and aclk to setup the vop */
 	ret = clk_prepare_enable(vop->hclk);
 	if (ret < 0) {
-		dev_err(vop->dev, "failed to prepare/enable hclk\n");
+		DRM_DEV_ERROR(vop->dev, "failed to prepare/enable hclk\n");
 		goto err_unprepare_dclk;
 	}
 
 	ret = clk_prepare_enable(vop->aclk);
 	if (ret < 0) {
-		dev_err(vop->dev, "failed to prepare/enable aclk\n");
+		DRM_DEV_ERROR(vop->dev, "failed to prepare/enable aclk\n");
 		goto err_disable_hclk;
 	}
 
@@ -1445,7 +2050,7 @@ static int vop_initial(struct vop *vop)
 	 */
 	ahb_rst = devm_reset_control_get(vop->dev, "ahb");
 	if (IS_ERR(ahb_rst)) {
-		dev_err(vop->dev, "failed to get ahb reset\n");
+		DRM_DEV_ERROR(vop->dev, "failed to get ahb reset\n");
 		ret = PTR_ERR(ahb_rst);
 		goto err_disable_aclk;
 	}
@@ -1453,15 +2058,23 @@ static int vop_initial(struct vop *vop)
 	usleep_range(10, 20);
 	reset_control_deassert(ahb_rst);
 
-	memcpy(vop->regsbak, vop->regs, vop->len);
+	VOP_INTR_SET_TYPE(vop, clear, INTR_MASK, 1);
+	VOP_INTR_SET_TYPE(vop, enable, INTR_MASK, 0);
 
-	for (i = 0; i < vop_data->table_size; i++)
-		vop_writel(vop, init_table[i].offset, init_table[i].value);
+	for (i = 0; i < vop->len; i += sizeof(u32))
+		vop->regsbak[i / 4] = readl_relaxed(vop->regs + i);
 
-	for (i = 0; i < vop_data->win_size; i++) {
-		const struct vop_win_data *win = &vop_data->win[i];
+	VOP_REG_SET(vop, misc, global_regdone_en, 1);
+	VOP_REG_SET(vop, common, dsp_blank, 0);
 
-		VOP_WIN_SET(vop, win, enable, 0);
+	for (i = 0; i < vop->data->win_size; i++) {
+		struct vop_win *vop_win = &vop->win[i];
+		const struct vop_win_data *win = vop_win->data;
+		int channel = i * 2 + 1;
+
+		VOP_WIN_SET(vop, win, channel, (channel + 1) << 4 | channel);
+		vop_win_disable(vop, vop_win);
+		VOP_WIN_SET(vop, win, gate, 1);
 	}
 
 	vop_cfg_done(vop);
@@ -1471,7 +2084,7 @@ static int vop_initial(struct vop *vop)
 	 */
 	vop->dclk_rst = devm_reset_control_get(vop->dev, "dclk");
 	if (IS_ERR(vop->dclk_rst)) {
-		dev_err(vop->dev, "failed to get dclk reset\n");
+		DRM_DEV_ERROR(vop->dev, "failed to get dclk reset\n");
 		ret = PTR_ERR(vop->dclk_rst);
 		goto err_disable_aclk;
 	}
@@ -1513,6 +2126,9 @@ static void vop_win_init(struct vop *vop)
 
 		vop_win->data = win_data;
 		vop_win->vop = vop;
+
+		if (vop_data->win_yuv2yuv)
+			vop_win->yuv2yuv_data = &vop_data->win_yuv2yuv[i];
 	}
 }
 
@@ -1530,15 +2146,21 @@ int rockchip_drm_wait_vact_end(struct drm_crtc *crtc, unsigned int mstimeout)
 {
 	struct vop *vop = to_vop(crtc);
 	unsigned long jiffies_left;
+	int ret = 0;
 
 	if (!crtc || !vop->is_enabled)
 		return -ENODEV;
 
-	if (mstimeout <= 0)
-		return -EINVAL;
+	mutex_lock(&vop->vop_lock);
+	if (mstimeout <= 0) {
+		ret = -EINVAL;
+		goto out;
+	}
 
-	if (vop_line_flag_irq_is_enabled(vop))
-		return -EBUSY;
+	if (vop_line_flag_irq_is_enabled(vop)) {
+		ret = -EBUSY;
+		goto out;
+	}
 
 	reinit_completion(&vop->line_flag_completion);
 	vop_line_flag_irq_enable(vop);
@@ -1548,11 +2170,14 @@ int rockchip_drm_wait_vact_end(struct drm_crtc *crtc, unsigned int mstimeout)
 	vop_line_flag_irq_disable(vop);
 
 	if (jiffies_left == 0) {
-		dev_err(vop->dev, "Timeout waiting for IRQ\n");
-		return -ETIMEDOUT;
+		DRM_DEV_ERROR(vop->dev, "Timeout waiting for IRQ\n");
+		ret = -ETIMEDOUT;
+		goto out;
 	}
 
-	return 0;
+out:
+	mutex_unlock(&vop->vop_lock);
+	return ret;
 }
 EXPORT_SYMBOL(rockchip_drm_wait_vact_end);
 
@@ -1563,7 +2188,6 @@ static int vop_bind(struct device *dev, struct device *master, void *data)
 	struct drm_device *drm_dev = data;
 	struct vop *vop;
 	struct resource *res;
-	size_t alloc_size;
 	int ret, irq;
 
 	vop_data = of_device_get_match_data(dev);
@@ -1571,8 +2195,8 @@ static int vop_bind(struct device *dev, struct device *master, void *data)
 		return -ENODEV;
 
 	/* Allocate vop struct and its vop_win array */
-	alloc_size = sizeof(*vop) + sizeof(*vop->win) * vop_data->win_size;
-	vop = devm_kzalloc(dev, alloc_size, GFP_KERNEL);
+	vop = devm_kzalloc(dev, struct_size(vop, win, vop_data->win_size),
+			   GFP_KERNEL);
 	if (!vop)
 		return -ENOMEM;
 
@@ -1584,10 +2208,21 @@ static int vop_bind(struct device *dev, struct device *master, void *data)
 	vop_win_init(vop);
 
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	vop->len = resource_size(res);
 	vop->regs = devm_ioremap_resource(dev, res);
 	if (IS_ERR(vop->regs))
 		return PTR_ERR(vop->regs);
+	vop->len = resource_size(res);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 1);
+	if (res) {
+		if (vop_data->lut_size != 1024 && vop_data->lut_size != 256) {
+			DRM_DEV_ERROR(dev, "unsupported gamma LUT size %d\n", vop_data->lut_size);
+			return -EINVAL;
+		}
+		vop->lut_regs = devm_ioremap_resource(dev, res);
+		if (IS_ERR(vop->lut_regs))
+			return PTR_ERR(vop->lut_regs);
+	}
 
 	vop->regsbak = devm_kzalloc(dev, vop->len, GFP_KERNEL);
 	if (!vop->regsbak)
@@ -1595,43 +2230,48 @@ static int vop_bind(struct device *dev, struct device *master, void *data)
 
 	irq = platform_get_irq(pdev, 0);
 	if (irq < 0) {
-		dev_err(dev, "cannot find irq for vop\n");
+		DRM_DEV_ERROR(dev, "cannot find irq for vop\n");
 		return irq;
 	}
 	vop->irq = (unsigned int)irq;
 
 	spin_lock_init(&vop->reg_lock);
 	spin_lock_init(&vop->irq_lock);
-
-	mutex_init(&vop->vsync_mutex);
-
-	ret = devm_request_irq(dev, vop->irq, vop_isr,
-			       IRQF_SHARED, dev_name(dev), vop);
-	if (ret)
-		return ret;
-
-	/* IRQ is initially disabled; it gets enabled in power_on */
-	disable_irq(vop->irq);
+	mutex_init(&vop->vop_lock);
 
 	ret = vop_create_crtc(vop);
 	if (ret)
-		goto err_enable_irq;
+		return ret;
 
 	pm_runtime_enable(&pdev->dev);
 
 	ret = vop_initial(vop);
 	if (ret < 0) {
-		dev_err(&pdev->dev, "cannot initial vop dev - err %d\n", ret);
+		DRM_DEV_ERROR(&pdev->dev,
+			      "cannot initial vop dev - err %d\n", ret);
+		goto err_disable_pm_runtime;
+	}
+
+	ret = devm_request_irq(dev, vop->irq, vop_isr,
+			       IRQF_SHARED, dev_name(dev), vop);
+	if (ret)
 		goto err_disable_pm_runtime;
+
+	if (vop->data->feature & VOP_FEATURE_INTERNAL_RGB) {
+		vop->rgb = rockchip_rgb_init(dev, &vop->crtc, vop->drm_dev, 0);
+		if (IS_ERR(vop->rgb)) {
+			ret = PTR_ERR(vop->rgb);
+			goto err_disable_pm_runtime;
+		}
 	}
 
+	rockchip_drm_dma_init_device(drm_dev, dev);
+
 	return 0;
 
 err_disable_pm_runtime:
 	pm_runtime_disable(&pdev->dev);
 	vop_destroy_crtc(vop);
-err_enable_irq:
-	enable_irq(vop->irq); /* To balance out the disable_irq above */
 	return ret;
 }
 
@@ -1639,6 +2279,9 @@ static void vop_unbind(struct device *dev, struct device *master, void *data)
 {
 	struct vop *vop = dev_get_drvdata(dev);
 
+	if (vop->rgb)
+		rockchip_rgb_fini(vop->rgb);
+
 	pm_runtime_disable(dev);
 	vop_destroy_crtc(vop);