Adding upstream version 5.10.209.upstream/5.10.209upstream

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

44 files changed, 15054 insertions, 0 deletions

diff --git a/drivers/gpu/drm/arm/Kconfig b/drivers/gpu/drm/arm/Kconfig
new file mode 100644
index 000000000..3a9e966e0
--- /dev/null
+++ b/drivers/gpu/drm/arm/Kconfig
@@ -0,0 +1,42 @@
+# SPDX-License-Identifier: GPL-2.0
+menu "ARM devices"
+
+config DRM_HDLCD
+	tristate "ARM HDLCD"
+	depends on DRM && OF && (ARM || ARM64 || COMPILE_TEST)
+	depends on COMMON_CLK
+	select DRM_KMS_HELPER
+	select DRM_KMS_CMA_HELPER
+	help
+	  Choose this option if you have an ARM High Definition Colour LCD
+	  controller.
+
+	  If M is selected the module will be called hdlcd.
+
+config DRM_HDLCD_SHOW_UNDERRUN
+	bool "Show underrun conditions"
+	depends on DRM_HDLCD
+	default n
+	help
+	  Enable this option to show in red colour the pixels that the
+	  HDLCD device did not fetch from framebuffer due to underrun
+	  conditions.
+
+config DRM_MALI_DISPLAY
+	tristate "ARM Mali Display Processor"
+	depends on DRM && OF && (ARM || ARM64 || COMPILE_TEST)
+	depends on COMMON_CLK
+	select DRM_KMS_HELPER
+	select DRM_KMS_CMA_HELPER
+	select DRM_GEM_CMA_HELPER
+	select VIDEOMODE_HELPERS
+	help
+	  Choose this option if you want to compile the ARM Mali Display
+	  Processor driver. It supports the DP500, DP550 and DP650 variants
+	  of the hardware.
+
+	  If compiled as a module it will be called mali-dp.
+
+source "drivers/gpu/drm/arm/display/Kconfig"
+
+endmenu
diff --git a/drivers/gpu/drm/arm/Makefile b/drivers/gpu/drm/arm/Makefile
new file mode 100644
index 000000000..3ced6fc9e
--- /dev/null
+++ b/drivers/gpu/drm/arm/Makefile
@@ -0,0 +1,7 @@
+# SPDX-License-Identifier: GPL-2.0-only
+hdlcd-y := hdlcd_drv.o hdlcd_crtc.o
+obj-$(CONFIG_DRM_HDLCD)	+= hdlcd.o
+mali-dp-y := malidp_drv.o malidp_hw.o malidp_planes.o malidp_crtc.o
+mali-dp-y += malidp_mw.o
+obj-$(CONFIG_DRM_MALI_DISPLAY)	+= mali-dp.o
+obj-$(CONFIG_DRM_KOMEDA) += display/
diff --git a/drivers/gpu/drm/arm/display/Kbuild b/drivers/gpu/drm/arm/display/Kbuild
new file mode 100644
index 000000000..382f1ca83
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/Kbuild
@@ -0,0 +1,3 @@
+# SPDX-License-Identifier: GPL-2.0
+
+obj-$(CONFIG_DRM_KOMEDA) += komeda/
diff --git a/drivers/gpu/drm/arm/display/Kconfig b/drivers/gpu/drm/arm/display/Kconfig
new file mode 100644
index 000000000..cec0639e3
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/Kconfig
@@ -0,0 +1,14 @@
+# SPDX-License-Identifier: GPL-2.0
+config DRM_KOMEDA
+	tristate "ARM Komeda display driver"
+	depends on DRM && OF
+	depends on COMMON_CLK
+	select DRM_KMS_HELPER
+	select DRM_KMS_CMA_HELPER
+	select DRM_GEM_CMA_HELPER
+	select VIDEOMODE_HELPERS
+	help
+	  Choose this option if you want to compile the ARM Komeda display
+	  Processor driver. It supports the D71 variants of the hardware.
+
+	  If compiled as a module it will be called komeda.
diff --git a/drivers/gpu/drm/arm/display/include/malidp_io.h b/drivers/gpu/drm/arm/display/include/malidp_io.h
new file mode 100644
index 000000000..9440dff94
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/include/malidp_io.h
@@ -0,0 +1,49 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#ifndef _MALIDP_IO_H_
+#define _MALIDP_IO_H_
+
+#include <linux/io.h>
+
+static inline u32
+malidp_read32(u32 __iomem *base, u32 offset)
+{
+	return readl((base + (offset >> 2)));
+}
+
+static inline void
+malidp_write32(u32 __iomem *base, u32 offset, u32 v)
+{
+	writel(v, (base + (offset >> 2)));
+}
+
+static inline void
+malidp_write64(u32 __iomem *base, u32 offset, u64 v)
+{
+	writel(lower_32_bits(v), (base + (offset >> 2)));
+	writel(upper_32_bits(v), (base + (offset >> 2) + 1));
+}
+
+static inline void
+malidp_write32_mask(u32 __iomem *base, u32 offset, u32 m, u32 v)
+{
+	u32 tmp = malidp_read32(base, offset);
+
+	tmp &= (~m);
+	malidp_write32(base, offset, v | tmp);
+}
+
+static inline void
+malidp_write_group(u32 __iomem *base, u32 offset, int num, const u32 *values)
+{
+	int i;
+
+	for (i = 0; i < num; i++)
+		malidp_write32(base, offset + i * 4, values[i]);
+}
+
+#endif /*_MALIDP_IO_H_*/
diff --git a/drivers/gpu/drm/arm/display/include/malidp_product.h b/drivers/gpu/drm/arm/display/include/malidp_product.h
new file mode 100644
index 000000000..16a8a2c22
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/include/malidp_product.h
@@ -0,0 +1,36 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#ifndef _MALIDP_PRODUCT_H_
+#define _MALIDP_PRODUCT_H_
+
+/* Product identification */
+#define MALIDP_CORE_ID(__product, __major, __minor, __status) \
+	((((__product) & 0xFFFF) << 16) | (((__major) & 0xF) << 12) | \
+	(((__minor) & 0xF) << 8) | ((__status) & 0xFF))
+
+#define MALIDP_CORE_ID_PRODUCT_ID(__core_id) ((__u32)(__core_id) >> 16)
+#define MALIDP_CORE_ID_MAJOR(__core_id)      (((__u32)(__core_id) >> 12) & 0xF)
+#define MALIDP_CORE_ID_MINOR(__core_id)      (((__u32)(__core_id) >> 8) & 0xF)
+#define MALIDP_CORE_ID_STATUS(__core_id)     (((__u32)(__core_id)) & 0xFF)
+
+/* Mali-display product IDs */
+#define MALIDP_D71_PRODUCT_ID	0x0071
+#define MALIDP_D32_PRODUCT_ID	0x0032
+
+union komeda_config_id {
+	struct {
+		__u32	max_line_sz:16,
+			n_pipelines:2,
+			n_scalers:2, /* number of scalers per pipeline */
+			n_layers:3, /* number of layers per pipeline */
+			n_richs:3, /* number of rich layers per pipeline */
+			reserved_bits:6;
+	};
+	__u32 value;
+};
+
+#endif /* _MALIDP_PRODUCT_H_ */
diff --git a/drivers/gpu/drm/arm/display/include/malidp_utils.h b/drivers/gpu/drm/arm/display/include/malidp_utils.h
new file mode 100644
index 000000000..49a1d7f35
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/include/malidp_utils.h
@@ -0,0 +1,43 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#ifndef _MALIDP_UTILS_
+#define _MALIDP_UTILS_
+
+#include <linux/delay.h>
+#include <linux/errno.h>
+
+#define has_bit(nr, mask)	(BIT(nr) & (mask))
+#define has_bits(bits, mask)	(((bits) & (mask)) == (bits))
+
+#define dp_wait_cond(__cond, __tries, __min_range, __max_range)	\
+({							\
+	int num_tries = __tries;			\
+	while (!__cond && (num_tries > 0)) {		\
+		usleep_range(__min_range, __max_range);	\
+		num_tries--;				\
+	}						\
+	(__cond) ? 0 : -ETIMEDOUT;			\
+})
+
+/* the restriction of range is [start, end] */
+struct malidp_range {
+	u32 start;
+	u32 end;
+};
+
+static inline void set_range(struct malidp_range *rg, u32 start, u32 end)
+{
+	rg->start = start;
+	rg->end   = end;
+}
+
+static inline bool in_range(struct malidp_range *rg, u32 v)
+{
+	return (v >= rg->start) && (v <= rg->end);
+}
+
+#endif /* _MALIDP_UTILS_ */
diff --git a/drivers/gpu/drm/arm/display/komeda/Makefile b/drivers/gpu/drm/arm/display/komeda/Makefile
new file mode 100644
index 000000000..1931a7fa1
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/Makefile
@@ -0,0 +1,26 @@
+# SPDX-License-Identifier: GPL-2.0
+
+ccflags-y := \
+	-I $(srctree)/$(src)/../include \
+	-I $(srctree)/$(src)
+
+komeda-y := \
+	komeda_drv.o \
+	komeda_dev.o \
+	komeda_format_caps.o \
+	komeda_color_mgmt.o \
+	komeda_pipeline.o \
+	komeda_pipeline_state.o \
+	komeda_framebuffer.o \
+	komeda_kms.o \
+	komeda_crtc.o \
+	komeda_plane.o \
+	komeda_wb_connector.o \
+	komeda_private_obj.o \
+	komeda_event.o
+
+komeda-y += \
+	d71/d71_dev.o \
+	d71/d71_component.o
+
+obj-$(CONFIG_DRM_KOMEDA) += komeda.o
diff --git a/drivers/gpu/drm/arm/display/komeda/d71/d71_component.c b/drivers/gpu/drm/arm/display/komeda/d71/d71_component.c
new file mode 100644
index 000000000..8a02ade36
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/d71/d71_component.c
@@ -0,0 +1,1442 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#include "d71_dev.h"
+#include "komeda_kms.h"
+#include "malidp_io.h"
+#include "komeda_framebuffer.h"
+#include "komeda_color_mgmt.h"
+
+static void get_resources_id(u32 hw_id, u32 *pipe_id, u32 *comp_id)
+{
+	u32 id = BLOCK_INFO_BLK_ID(hw_id);
+	u32 pipe = id;
+
+	switch (BLOCK_INFO_BLK_TYPE(hw_id)) {
+	case D71_BLK_TYPE_LPU_WB_LAYER:
+		id = KOMEDA_COMPONENT_WB_LAYER;
+		break;
+	case D71_BLK_TYPE_CU_SPLITTER:
+		id = KOMEDA_COMPONENT_SPLITTER;
+		break;
+	case D71_BLK_TYPE_CU_SCALER:
+		pipe = id / D71_PIPELINE_MAX_SCALERS;
+		id %= D71_PIPELINE_MAX_SCALERS;
+		id += KOMEDA_COMPONENT_SCALER0;
+		break;
+	case D71_BLK_TYPE_CU:
+		id += KOMEDA_COMPONENT_COMPIZ0;
+		break;
+	case D71_BLK_TYPE_LPU_LAYER:
+		pipe = id / D71_PIPELINE_MAX_LAYERS;
+		id %= D71_PIPELINE_MAX_LAYERS;
+		id += KOMEDA_COMPONENT_LAYER0;
+		break;
+	case D71_BLK_TYPE_DOU_IPS:
+		id += KOMEDA_COMPONENT_IPS0;
+		break;
+	case D71_BLK_TYPE_CU_MERGER:
+		id = KOMEDA_COMPONENT_MERGER;
+		break;
+	case D71_BLK_TYPE_DOU:
+		id = KOMEDA_COMPONENT_TIMING_CTRLR;
+		break;
+	default:
+		id = 0xFFFFFFFF;
+	}
+
+	if (comp_id)
+		*comp_id = id;
+
+	if (pipe_id)
+		*pipe_id = pipe;
+}
+
+static u32 get_valid_inputs(struct block_header *blk)
+{
+	u32 valid_inputs = 0, comp_id;
+	int i;
+
+	for (i = 0; i < PIPELINE_INFO_N_VALID_INPUTS(blk->pipeline_info); i++) {
+		get_resources_id(blk->input_ids[i], NULL, &comp_id);
+		if (comp_id == 0xFFFFFFFF)
+			continue;
+		valid_inputs |= BIT(comp_id);
+	}
+
+	return valid_inputs;
+}
+
+static void get_values_from_reg(void __iomem *reg, u32 offset,
+				u32 count, u32 *val)
+{
+	u32 i, addr;
+
+	for (i = 0; i < count; i++) {
+		addr = offset + (i << 2);
+		/* 0xA4 is WO register */
+		if (addr != 0xA4)
+			val[i] = malidp_read32(reg, addr);
+		else
+			val[i] = 0xDEADDEAD;
+	}
+}
+
+static void dump_block_header(struct seq_file *sf, void __iomem *reg)
+{
+	struct block_header hdr;
+	u32 i, n_input, n_output;
+
+	d71_read_block_header(reg, &hdr);
+	seq_printf(sf, "BLOCK_INFO:\t\t0x%X\n", hdr.block_info);
+	seq_printf(sf, "PIPELINE_INFO:\t\t0x%X\n", hdr.pipeline_info);
+
+	n_output = PIPELINE_INFO_N_OUTPUTS(hdr.pipeline_info);
+	n_input  = PIPELINE_INFO_N_VALID_INPUTS(hdr.pipeline_info);
+
+	for (i = 0; i < n_input; i++)
+		seq_printf(sf, "VALID_INPUT_ID%u:\t0x%X\n",
+			   i, hdr.input_ids[i]);
+
+	for (i = 0; i < n_output; i++)
+		seq_printf(sf, "OUTPUT_ID%u:\t\t0x%X\n",
+			   i, hdr.output_ids[i]);
+}
+
+/* On D71, we are using the global line size. From D32, every component have
+ * a line size register to indicate the fifo size.
+ */
+static u32 __get_blk_line_size(struct d71_dev *d71, u32 __iomem *reg,
+			       u32 max_default)
+{
+	if (!d71->periph_addr)
+		max_default = malidp_read32(reg, BLK_MAX_LINE_SIZE);
+
+	return max_default;
+}
+
+static u32 get_blk_line_size(struct d71_dev *d71, u32 __iomem *reg)
+{
+	return __get_blk_line_size(d71, reg, d71->max_line_size);
+}
+
+static u32 to_rot_ctrl(u32 rot)
+{
+	u32 lr_ctrl = 0;
+
+	switch (rot & DRM_MODE_ROTATE_MASK) {
+	case DRM_MODE_ROTATE_0:
+		lr_ctrl |= L_ROT(L_ROT_R0);
+		break;
+	case DRM_MODE_ROTATE_90:
+		lr_ctrl |= L_ROT(L_ROT_R90);
+		break;
+	case DRM_MODE_ROTATE_180:
+		lr_ctrl |= L_ROT(L_ROT_R180);
+		break;
+	case DRM_MODE_ROTATE_270:
+		lr_ctrl |= L_ROT(L_ROT_R270);
+		break;
+	}
+
+	if (rot & DRM_MODE_REFLECT_X)
+		lr_ctrl |= L_HFLIP;
+	if (rot & DRM_MODE_REFLECT_Y)
+		lr_ctrl |= L_VFLIP;
+
+	return lr_ctrl;
+}
+
+static u32 to_ad_ctrl(u64 modifier)
+{
+	u32 afbc_ctrl = AD_AEN;
+
+	if (!modifier)
+		return 0;
+
+	if ((modifier & AFBC_FORMAT_MOD_BLOCK_SIZE_MASK) ==
+	    AFBC_FORMAT_MOD_BLOCK_SIZE_32x8)
+		afbc_ctrl |= AD_WB;
+
+	if (modifier & AFBC_FORMAT_MOD_YTR)
+		afbc_ctrl |= AD_YT;
+	if (modifier & AFBC_FORMAT_MOD_SPLIT)
+		afbc_ctrl |= AD_BS;
+	if (modifier & AFBC_FORMAT_MOD_TILED)
+		afbc_ctrl |= AD_TH;
+
+	return afbc_ctrl;
+}
+
+static inline u32 to_d71_input_id(struct komeda_component_state *st, int idx)
+{
+	struct komeda_component_output *input = &st->inputs[idx];
+
+	/* if input is not active, set hw input_id(0) to disable it */
+	if (has_bit(idx, st->active_inputs))
+		return input->component->hw_id + input->output_port;
+	else
+		return 0;
+}
+
+static void d71_layer_update_fb(struct komeda_component *c,
+				struct komeda_fb *kfb,
+				dma_addr_t *addr)
+{
+	struct drm_framebuffer *fb = &kfb->base;
+	const struct drm_format_info *info = fb->format;
+	u32 __iomem *reg = c->reg;
+	int block_h;
+
+	if (info->num_planes > 2)
+		malidp_write64(reg, BLK_P2_PTR_LOW, addr[2]);
+
+	if (info->num_planes > 1) {
+		block_h = drm_format_info_block_height(info, 1);
+		malidp_write32(reg, BLK_P1_STRIDE, fb->pitches[1] * block_h);
+		malidp_write64(reg, BLK_P1_PTR_LOW, addr[1]);
+	}
+
+	block_h = drm_format_info_block_height(info, 0);
+	malidp_write32(reg, BLK_P0_STRIDE, fb->pitches[0] * block_h);
+	malidp_write64(reg, BLK_P0_PTR_LOW, addr[0]);
+	malidp_write32(reg, LAYER_FMT, kfb->format_caps->hw_id);
+}
+
+static void d71_layer_disable(struct komeda_component *c)
+{
+	malidp_write32_mask(c->reg, BLK_CONTROL, L_EN, 0);
+}
+
+static void d71_layer_update(struct komeda_component *c,
+			     struct komeda_component_state *state)
+{
+	struct komeda_layer_state *st = to_layer_st(state);
+	struct drm_plane_state *plane_st = state->plane->state;
+	struct drm_framebuffer *fb = plane_st->fb;
+	struct komeda_fb *kfb = to_kfb(fb);
+	u32 __iomem *reg = c->reg;
+	u32 ctrl_mask = L_EN | L_ROT(L_ROT_R270) | L_HFLIP | L_VFLIP | L_TBU_EN;
+	u32 ctrl = L_EN | to_rot_ctrl(st->rot);
+
+	d71_layer_update_fb(c, kfb, st->addr);
+
+	malidp_write32(reg, AD_CONTROL, to_ad_ctrl(fb->modifier));
+	if (fb->modifier) {
+		u64 addr;
+
+		malidp_write32(reg, LAYER_AD_H_CROP, HV_CROP(st->afbc_crop_l,
+							     st->afbc_crop_r));
+		malidp_write32(reg, LAYER_AD_V_CROP, HV_CROP(st->afbc_crop_t,
+							     st->afbc_crop_b));
+		/* afbc 1.2 wants payload, afbc 1.0/1.1 wants end_addr */
+		if (fb->modifier & AFBC_FORMAT_MOD_TILED)
+			addr = st->addr[0] + kfb->offset_payload;
+		else
+			addr = st->addr[0] + kfb->afbc_size - 1;
+
+		malidp_write32(reg, BLK_P1_PTR_LOW, lower_32_bits(addr));
+		malidp_write32(reg, BLK_P1_PTR_HIGH, upper_32_bits(addr));
+	}
+
+	if (fb->format->is_yuv) {
+		u32 upsampling = 0;
+
+		switch (kfb->format_caps->fourcc) {
+		case DRM_FORMAT_YUYV:
+			upsampling = fb->modifier ? LR_CHI422_BILINEAR :
+				     LR_CHI422_REPLICATION;
+			break;
+		case DRM_FORMAT_UYVY:
+			upsampling = LR_CHI422_REPLICATION;
+			break;
+		case DRM_FORMAT_NV12:
+		case DRM_FORMAT_YUV420_8BIT:
+		case DRM_FORMAT_YUV420_10BIT:
+		case DRM_FORMAT_YUV420:
+		case DRM_FORMAT_P010:
+		/* these fmt support MPGE/JPEG both, here perfer JPEG*/
+			upsampling = LR_CHI420_JPEG;
+			break;
+		case DRM_FORMAT_X0L2:
+			upsampling = LR_CHI420_JPEG;
+			break;
+		default:
+			break;
+		}
+
+		malidp_write32(reg, LAYER_R_CONTROL, upsampling);
+		malidp_write_group(reg, LAYER_YUV_RGB_COEFF0,
+				   KOMEDA_N_YUV2RGB_COEFFS,
+				   komeda_select_yuv2rgb_coeffs(
+					plane_st->color_encoding,
+					plane_st->color_range));
+	}
+
+	malidp_write32(reg, BLK_IN_SIZE, HV_SIZE(st->hsize, st->vsize));
+
+	if (kfb->is_va)
+		ctrl |= L_TBU_EN;
+	malidp_write32_mask(reg, BLK_CONTROL, ctrl_mask, ctrl);
+}
+
+static void d71_layer_dump(struct komeda_component *c, struct seq_file *sf)
+{
+	u32 v[15], i;
+	bool rich, rgb2rgb;
+	char *prefix;
+
+	get_values_from_reg(c->reg, LAYER_INFO, 1, &v[14]);
+	if (v[14] & 0x1) {
+		rich = true;
+		prefix = "LR_";
+	} else {
+		rich = false;
+		prefix = "LS_";
+	}
+
+	rgb2rgb = !!(v[14] & L_INFO_CM);
+
+	dump_block_header(sf, c->reg);
+
+	seq_printf(sf, "%sLAYER_INFO:\t\t0x%X\n", prefix, v[14]);
+
+	get_values_from_reg(c->reg, 0xD0, 1, v);
+	seq_printf(sf, "%sCONTROL:\t\t0x%X\n", prefix, v[0]);
+	if (rich) {
+		get_values_from_reg(c->reg, 0xD4, 1, v);
+		seq_printf(sf, "LR_RICH_CONTROL:\t0x%X\n", v[0]);
+	}
+	get_values_from_reg(c->reg, 0xD8, 4, v);
+	seq_printf(sf, "%sFORMAT:\t\t0x%X\n", prefix, v[0]);
+	seq_printf(sf, "%sIT_COEFFTAB:\t\t0x%X\n", prefix, v[1]);
+	seq_printf(sf, "%sIN_SIZE:\t\t0x%X\n", prefix, v[2]);
+	seq_printf(sf, "%sPALPHA:\t\t0x%X\n", prefix, v[3]);
+
+	get_values_from_reg(c->reg, 0x100, 3, v);
+	seq_printf(sf, "%sP0_PTR_LOW:\t\t0x%X\n", prefix, v[0]);
+	seq_printf(sf, "%sP0_PTR_HIGH:\t\t0x%X\n", prefix, v[1]);
+	seq_printf(sf, "%sP0_STRIDE:\t\t0x%X\n", prefix, v[2]);
+
+	get_values_from_reg(c->reg, 0x110, 2, v);
+	seq_printf(sf, "%sP1_PTR_LOW:\t\t0x%X\n", prefix, v[0]);
+	seq_printf(sf, "%sP1_PTR_HIGH:\t\t0x%X\n", prefix, v[1]);
+	if (rich) {
+		get_values_from_reg(c->reg, 0x118, 1, v);
+		seq_printf(sf, "LR_P1_STRIDE:\t\t0x%X\n", v[0]);
+
+		get_values_from_reg(c->reg, 0x120, 2, v);
+		seq_printf(sf, "LR_P2_PTR_LOW:\t\t0x%X\n", v[0]);
+		seq_printf(sf, "LR_P2_PTR_HIGH:\t\t0x%X\n", v[1]);
+
+		get_values_from_reg(c->reg, 0x130, 12, v);
+		for (i = 0; i < 12; i++)
+			seq_printf(sf, "LR_YUV_RGB_COEFF%u:\t0x%X\n", i, v[i]);
+	}
+
+	if (rgb2rgb) {
+		get_values_from_reg(c->reg, LAYER_RGB_RGB_COEFF0, 12, v);
+		for (i = 0; i < 12; i++)
+			seq_printf(sf, "LS_RGB_RGB_COEFF%u:\t0x%X\n", i, v[i]);
+	}
+
+	get_values_from_reg(c->reg, 0x160, 3, v);
+	seq_printf(sf, "%sAD_CONTROL:\t\t0x%X\n", prefix, v[0]);
+	seq_printf(sf, "%sAD_H_CROP:\t\t0x%X\n", prefix, v[1]);
+	seq_printf(sf, "%sAD_V_CROP:\t\t0x%X\n", prefix, v[2]);
+}
+
+static int d71_layer_validate(struct komeda_component *c,
+			      struct komeda_component_state *state)
+{
+	struct komeda_layer_state *st = to_layer_st(state);
+	struct komeda_layer *layer = to_layer(c);
+	struct drm_plane_state *plane_st;
+	struct drm_framebuffer *fb;
+	u32 fourcc, line_sz, max_line_sz;
+
+	plane_st = drm_atomic_get_new_plane_state(state->obj.state,
+						  state->plane);
+	fb = plane_st->fb;
+	fourcc = fb->format->format;
+
+	if (drm_rotation_90_or_270(st->rot))
+		line_sz = st->vsize - st->afbc_crop_t - st->afbc_crop_b;
+	else
+		line_sz = st->hsize - st->afbc_crop_l - st->afbc_crop_r;
+
+	if (fb->modifier) {
+		if ((fb->modifier & AFBC_FORMAT_MOD_BLOCK_SIZE_MASK) ==
+			AFBC_FORMAT_MOD_BLOCK_SIZE_32x8)
+			max_line_sz = layer->line_sz;
+		else
+			max_line_sz = layer->line_sz / 2;
+
+		if (line_sz > max_line_sz) {
+			DRM_DEBUG_ATOMIC("afbc request line_sz: %d exceed the max afbc line_sz: %d.\n",
+					 line_sz, max_line_sz);
+			return -EINVAL;
+		}
+	}
+
+	if (fourcc == DRM_FORMAT_YUV420_10BIT && line_sz > 2046 && (st->afbc_crop_l % 4)) {
+		DRM_DEBUG_ATOMIC("YUV420_10BIT input_hsize: %d exceed the max size 2046.\n",
+				 line_sz);
+		return -EINVAL;
+	}
+
+	if (fourcc == DRM_FORMAT_X0L2 && line_sz > 2046 && (st->addr[0] % 16)) {
+		DRM_DEBUG_ATOMIC("X0L2 input_hsize: %d exceed the max size 2046.\n",
+				 line_sz);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static const struct komeda_component_funcs d71_layer_funcs = {
+	.validate	= d71_layer_validate,
+	.update		= d71_layer_update,
+	.disable	= d71_layer_disable,
+	.dump_register	= d71_layer_dump,
+};
+
+static int d71_layer_init(struct d71_dev *d71,
+			  struct block_header *blk, u32 __iomem *reg)
+{
+	struct komeda_component *c;
+	struct komeda_layer *layer;
+	u32 pipe_id, layer_id, layer_info;
+
+	get_resources_id(blk->block_info, &pipe_id, &layer_id);
+	c = komeda_component_add(&d71->pipes[pipe_id]->base, sizeof(*layer),
+				 layer_id,
+				 BLOCK_INFO_INPUT_ID(blk->block_info),
+				 &d71_layer_funcs, 0,
+				 get_valid_inputs(blk),
+				 1, reg, "LPU%d_LAYER%d", pipe_id, layer_id);
+	if (IS_ERR(c)) {
+		DRM_ERROR("Failed to add layer component\n");
+		return PTR_ERR(c);
+	}
+
+	layer = to_layer(c);
+	layer_info = malidp_read32(reg, LAYER_INFO);
+
+	if (layer_info & L_INFO_RF)
+		layer->layer_type = KOMEDA_FMT_RICH_LAYER;
+	else
+		layer->layer_type = KOMEDA_FMT_SIMPLE_LAYER;
+
+	if (!d71->periph_addr) {
+		/* D32 or newer product */
+		layer->line_sz = malidp_read32(reg, BLK_MAX_LINE_SIZE);
+		layer->yuv_line_sz = L_INFO_YUV_MAX_LINESZ(layer_info);
+	} else if (d71->max_line_size > 2048) {
+		/* D71 4K */
+		layer->line_sz = d71->max_line_size;
+		layer->yuv_line_sz = layer->line_sz / 2;
+	} else	{
+		/* D71 2K */
+		if (layer->layer_type == KOMEDA_FMT_RICH_LAYER) {
+			/* rich layer is 4K configuration */
+			layer->line_sz = d71->max_line_size * 2;
+			layer->yuv_line_sz = layer->line_sz / 2;
+		} else {
+			layer->line_sz = d71->max_line_size;
+			layer->yuv_line_sz = 0;
+		}
+	}
+
+	set_range(&layer->hsize_in, 4, layer->line_sz);
+
+	set_range(&layer->vsize_in, 4, d71->max_vsize);
+
+	malidp_write32(reg, LAYER_PALPHA, D71_PALPHA_DEF_MAP);
+
+	layer->supported_rots = DRM_MODE_ROTATE_MASK | DRM_MODE_REFLECT_MASK;
+
+	return 0;
+}
+
+static void d71_wb_layer_update(struct komeda_component *c,
+				struct komeda_component_state *state)
+{
+	struct komeda_layer_state *st = to_layer_st(state);
+	struct drm_connector_state *conn_st = state->wb_conn->state;
+	struct komeda_fb *kfb = to_kfb(conn_st->writeback_job->fb);
+	u32 ctrl = L_EN | LW_OFM, mask = L_EN | LW_OFM | LW_TBU_EN;
+	u32 __iomem *reg = c->reg;
+
+	d71_layer_update_fb(c, kfb, st->addr);
+
+	if (kfb->is_va)
+		ctrl |= LW_TBU_EN;
+
+	malidp_write32(reg, BLK_IN_SIZE, HV_SIZE(st->hsize, st->vsize));
+	malidp_write32(reg, BLK_INPUT_ID0, to_d71_input_id(state, 0));
+	malidp_write32_mask(reg, BLK_CONTROL, mask, ctrl);
+}
+
+static void d71_wb_layer_dump(struct komeda_component *c, struct seq_file *sf)
+{
+	u32 v[12], i;
+
+	dump_block_header(sf, c->reg);
+
+	get_values_from_reg(c->reg, 0x80, 1, v);
+	seq_printf(sf, "LW_INPUT_ID0:\t\t0x%X\n", v[0]);
+
+	get_values_from_reg(c->reg, 0xD0, 3, v);
+	seq_printf(sf, "LW_CONTROL:\t\t0x%X\n", v[0]);
+	seq_printf(sf, "LW_PROG_LINE:\t\t0x%X\n", v[1]);
+	seq_printf(sf, "LW_FORMAT:\t\t0x%X\n", v[2]);
+
+	get_values_from_reg(c->reg, 0xE0, 1, v);
+	seq_printf(sf, "LW_IN_SIZE:\t\t0x%X\n", v[0]);
+
+	for (i = 0; i < 2; i++) {
+		get_values_from_reg(c->reg, 0x100 + i * 0x10, 3, v);
+		seq_printf(sf, "LW_P%u_PTR_LOW:\t\t0x%X\n", i, v[0]);
+		seq_printf(sf, "LW_P%u_PTR_HIGH:\t\t0x%X\n", i, v[1]);
+		seq_printf(sf, "LW_P%u_STRIDE:\t\t0x%X\n", i, v[2]);
+	}
+
+	get_values_from_reg(c->reg, 0x130, 12, v);
+	for (i = 0; i < 12; i++)
+		seq_printf(sf, "LW_RGB_YUV_COEFF%u:\t0x%X\n", i, v[i]);
+}
+
+static void d71_wb_layer_disable(struct komeda_component *c)
+{
+	malidp_write32(c->reg, BLK_INPUT_ID0, 0);
+	malidp_write32_mask(c->reg, BLK_CONTROL, L_EN, 0);
+}
+
+static const struct komeda_component_funcs d71_wb_layer_funcs = {
+	.update		= d71_wb_layer_update,
+	.disable	= d71_wb_layer_disable,
+	.dump_register	= d71_wb_layer_dump,
+};
+
+static int d71_wb_layer_init(struct d71_dev *d71,
+			     struct block_header *blk, u32 __iomem *reg)
+{
+	struct komeda_component *c;
+	struct komeda_layer *wb_layer;
+	u32 pipe_id, layer_id;
+
+	get_resources_id(blk->block_info, &pipe_id, &layer_id);
+
+	c = komeda_component_add(&d71->pipes[pipe_id]->base, sizeof(*wb_layer),
+				 layer_id, BLOCK_INFO_INPUT_ID(blk->block_info),
+				 &d71_wb_layer_funcs,
+				 1, get_valid_inputs(blk), 0, reg,
+				 "LPU%d_LAYER_WR", pipe_id);
+	if (IS_ERR(c)) {
+		DRM_ERROR("Failed to add wb_layer component\n");
+		return PTR_ERR(c);
+	}
+
+	wb_layer = to_layer(c);
+	wb_layer->layer_type = KOMEDA_FMT_WB_LAYER;
+	wb_layer->line_sz = get_blk_line_size(d71, reg);
+	wb_layer->yuv_line_sz = wb_layer->line_sz;
+
+	set_range(&wb_layer->hsize_in, 64, wb_layer->line_sz);
+	set_range(&wb_layer->vsize_in, 64, d71->max_vsize);
+
+	return 0;
+}
+
+static void d71_component_disable(struct komeda_component *c)
+{
+	u32 __iomem *reg = c->reg;
+	u32 i;
+
+	malidp_write32(reg, BLK_CONTROL, 0);
+
+	for (i = 0; i < c->max_active_inputs; i++) {
+		malidp_write32(reg, BLK_INPUT_ID0 + (i << 2), 0);
+
+		/* Besides clearing the input ID to zero, D71 compiz also has
+		 * input enable bit in CU_INPUTx_CONTROL which need to be
+		 * cleared.
+		 */
+		if (has_bit(c->id, KOMEDA_PIPELINE_COMPIZS))
+			malidp_write32(reg, CU_INPUT0_CONTROL +
+				       i * CU_PER_INPUT_REGS * 4,
+				       CU_INPUT_CTRL_ALPHA(0xFF));
+	}
+}
+
+static void compiz_enable_input(u32 __iomem *id_reg,
+				u32 __iomem *cfg_reg,
+				u32 input_hw_id,
+				struct komeda_compiz_input_cfg *cin)
+{
+	u32 ctrl = CU_INPUT_CTRL_EN;
+	u8 blend = cin->pixel_blend_mode;
+
+	if (blend == DRM_MODE_BLEND_PIXEL_NONE)
+		ctrl |= CU_INPUT_CTRL_PAD;
+	else if (blend == DRM_MODE_BLEND_PREMULTI)
+		ctrl |= CU_INPUT_CTRL_PMUL;
+
+	ctrl |= CU_INPUT_CTRL_ALPHA(cin->layer_alpha);
+
+	malidp_write32(id_reg, BLK_INPUT_ID0, input_hw_id);
+
+	malidp_write32(cfg_reg, CU_INPUT0_SIZE,
+		       HV_SIZE(cin->hsize, cin->vsize));
+	malidp_write32(cfg_reg, CU_INPUT0_OFFSET,
+		       HV_OFFSET(cin->hoffset, cin->voffset));
+	malidp_write32(cfg_reg, CU_INPUT0_CONTROL, ctrl);
+}
+
+static void d71_compiz_update(struct komeda_component *c,
+			      struct komeda_component_state *state)
+{
+	struct komeda_compiz_state *st = to_compiz_st(state);
+	u32 __iomem *reg = c->reg;
+	u32 __iomem *id_reg, *cfg_reg;
+	u32 index;
+
+	for_each_changed_input(state, index) {
+		id_reg = reg + index;
+		cfg_reg = reg + index * CU_PER_INPUT_REGS;
+		if (state->active_inputs & BIT(index)) {
+			compiz_enable_input(id_reg, cfg_reg,
+					    to_d71_input_id(state, index),
+					    &st->cins[index]);
+		} else {
+			malidp_write32(id_reg, BLK_INPUT_ID0, 0);
+			malidp_write32(cfg_reg, CU_INPUT0_CONTROL, 0);
+		}
+	}
+
+	malidp_write32(reg, BLK_SIZE, HV_SIZE(st->hsize, st->vsize));
+}
+
+static void d71_compiz_dump(struct komeda_component *c, struct seq_file *sf)
+{
+	u32 v[8], i;
+
+	dump_block_header(sf, c->reg);
+
+	get_values_from_reg(c->reg, 0x80, 5, v);
+	for (i = 0; i < 5; i++)
+		seq_printf(sf, "CU_INPUT_ID%u:\t\t0x%X\n", i, v[i]);
+
+	get_values_from_reg(c->reg, 0xA0, 5, v);
+	seq_printf(sf, "CU_IRQ_RAW_STATUS:\t0x%X\n", v[0]);
+	seq_printf(sf, "CU_IRQ_CLEAR:\t\t0x%X\n", v[1]);
+	seq_printf(sf, "CU_IRQ_MASK:\t\t0x%X\n", v[2]);
+	seq_printf(sf, "CU_IRQ_STATUS:\t\t0x%X\n", v[3]);
+	seq_printf(sf, "CU_STATUS:\t\t0x%X\n", v[4]);
+
+	get_values_from_reg(c->reg, 0xD0, 2, v);
+	seq_printf(sf, "CU_CONTROL:\t\t0x%X\n", v[0]);
+	seq_printf(sf, "CU_SIZE:\t\t0x%X\n", v[1]);
+
+	get_values_from_reg(c->reg, 0xDC, 1, v);
+	seq_printf(sf, "CU_BG_COLOR:\t\t0x%X\n", v[0]);
+
+	for (i = 0, v[4] = 0xE0; i < 5; i++, v[4] += 0x10) {
+		get_values_from_reg(c->reg, v[4], 3, v);
+		seq_printf(sf, "CU_INPUT%u_SIZE:\t\t0x%X\n", i, v[0]);
+		seq_printf(sf, "CU_INPUT%u_OFFSET:\t0x%X\n", i, v[1]);
+		seq_printf(sf, "CU_INPUT%u_CONTROL:\t0x%X\n", i, v[2]);
+	}
+
+	get_values_from_reg(c->reg, 0x130, 2, v);
+	seq_printf(sf, "CU_USER_LOW:\t\t0x%X\n", v[0]);
+	seq_printf(sf, "CU_USER_HIGH:\t\t0x%X\n", v[1]);
+}
+
+static const struct komeda_component_funcs d71_compiz_funcs = {
+	.update		= d71_compiz_update,
+	.disable	= d71_component_disable,
+	.dump_register	= d71_compiz_dump,
+};
+
+static int d71_compiz_init(struct d71_dev *d71,
+			   struct block_header *blk, u32 __iomem *reg)
+{
+	struct komeda_component *c;
+	struct komeda_compiz *compiz;
+	u32 pipe_id, comp_id;
+
+	get_resources_id(blk->block_info, &pipe_id, &comp_id);
+
+	c = komeda_component_add(&d71->pipes[pipe_id]->base, sizeof(*compiz),
+				 comp_id,
+				 BLOCK_INFO_INPUT_ID(blk->block_info),
+				 &d71_compiz_funcs,
+				 CU_NUM_INPUT_IDS, get_valid_inputs(blk),
+				 CU_NUM_OUTPUT_IDS, reg,
+				 "CU%d", pipe_id);
+	if (IS_ERR(c))
+		return PTR_ERR(c);
+
+	compiz = to_compiz(c);
+
+	set_range(&compiz->hsize, 64, get_blk_line_size(d71, reg));
+	set_range(&compiz->vsize, 64, d71->max_vsize);
+
+	return 0;
+}
+
+static void d71_scaler_update_filter_lut(u32 __iomem *reg, u32 hsize_in,
+					 u32 vsize_in, u32 hsize_out,
+					 u32 vsize_out)
+{
+	u32 val = 0;
+
+	if (hsize_in <= hsize_out)
+		val  |= 0x62;
+	else if (hsize_in <= (hsize_out + hsize_out / 2))
+		val |= 0x63;
+	else if (hsize_in <= hsize_out * 2)
+		val |= 0x64;
+	else if (hsize_in <= hsize_out * 2 + (hsize_out * 3) / 4)
+		val |= 0x65;
+	else
+		val |= 0x66;
+
+	if (vsize_in <= vsize_out)
+		val  |= SC_VTSEL(0x6A);
+	else if (vsize_in <= (vsize_out + vsize_out / 2))
+		val |= SC_VTSEL(0x6B);
+	else if (vsize_in <= vsize_out * 2)
+		val |= SC_VTSEL(0x6C);
+	else if (vsize_in <= vsize_out * 2 + vsize_out * 3 / 4)
+		val |= SC_VTSEL(0x6D);
+	else
+		val |= SC_VTSEL(0x6E);
+
+	malidp_write32(reg, SC_COEFFTAB, val);
+}
+
+static void d71_scaler_update(struct komeda_component *c,
+			      struct komeda_component_state *state)
+{
+	struct komeda_scaler_state *st = to_scaler_st(state);
+	u32 __iomem *reg = c->reg;
+	u32 init_ph, delta_ph, ctrl;
+
+	d71_scaler_update_filter_lut(reg, st->hsize_in, st->vsize_in,
+				     st->hsize_out, st->vsize_out);
+
+	malidp_write32(reg, BLK_IN_SIZE, HV_SIZE(st->hsize_in, st->vsize_in));
+	malidp_write32(reg, SC_OUT_SIZE, HV_SIZE(st->hsize_out, st->vsize_out));
+	malidp_write32(reg, SC_H_CROP, HV_CROP(st->left_crop, st->right_crop));
+
+	/* for right part, HW only sample the valid pixel which means the pixels
+	 * in left_crop will be jumpped, and the first sample pixel is:
+	 *
+	 * dst_a = st->total_hsize_out - st->hsize_out + st->left_crop + 0.5;
+	 *
+	 * Then the corresponding texel in src is:
+	 *
+	 * h_delta_phase = st->total_hsize_in / st->total_hsize_out;
+	 * src_a = dst_A * h_delta_phase;
+	 *
+	 * and h_init_phase is src_a deduct the real source start src_S;
+	 *
+	 * src_S = st->total_hsize_in - st->hsize_in;
+	 * h_init_phase = src_a - src_S;
+	 *
+	 * And HW precision for the initial/delta_phase is 16:16 fixed point,
+	 * the following is the simplified formula
+	 */
+	if (st->right_part) {
+		u32 dst_a = st->total_hsize_out - st->hsize_out + st->left_crop;
+
+		if (st->en_img_enhancement)
+			dst_a -= 1;
+
+		init_ph = ((st->total_hsize_in * (2 * dst_a + 1) -
+			    2 * st->total_hsize_out * (st->total_hsize_in -
+			    st->hsize_in)) << 15) / st->total_hsize_out;
+	} else {
+		init_ph = (st->total_hsize_in << 15) / st->total_hsize_out;
+	}
+
+	malidp_write32(reg, SC_H_INIT_PH, init_ph);
+
+	delta_ph = (st->total_hsize_in << 16) / st->total_hsize_out;
+	malidp_write32(reg, SC_H_DELTA_PH, delta_ph);
+
+	init_ph = (st->total_vsize_in << 15) / st->vsize_out;
+	malidp_write32(reg, SC_V_INIT_PH, init_ph);
+
+	delta_ph = (st->total_vsize_in << 16) / st->vsize_out;
+	malidp_write32(reg, SC_V_DELTA_PH, delta_ph);
+
+	ctrl = 0;
+	ctrl |= st->en_scaling ? SC_CTRL_SCL : 0;
+	ctrl |= st->en_alpha ? SC_CTRL_AP : 0;
+	ctrl |= st->en_img_enhancement ? SC_CTRL_IENH : 0;
+	/* If we use the hardware splitter we shouldn't set SC_CTRL_LS */
+	if (st->en_split &&
+	    state->inputs[0].component->id != KOMEDA_COMPONENT_SPLITTER)
+		ctrl |= SC_CTRL_LS;
+
+	malidp_write32(reg, BLK_CONTROL, ctrl);
+	malidp_write32(reg, BLK_INPUT_ID0, to_d71_input_id(state, 0));
+}
+
+static void d71_scaler_dump(struct komeda_component *c, struct seq_file *sf)
+{
+	u32 v[10];
+
+	dump_block_header(sf, c->reg);
+
+	get_values_from_reg(c->reg, 0x80, 1, v);
+	seq_printf(sf, "SC_INPUT_ID0:\t\t0x%X\n", v[0]);
+
+	get_values_from_reg(c->reg, 0xD0, 1, v);
+	seq_printf(sf, "SC_CONTROL:\t\t0x%X\n", v[0]);
+
+	get_values_from_reg(c->reg, 0xDC, 9, v);
+	seq_printf(sf, "SC_COEFFTAB:\t\t0x%X\n", v[0]);
+	seq_printf(sf, "SC_IN_SIZE:\t\t0x%X\n", v[1]);
+	seq_printf(sf, "SC_OUT_SIZE:\t\t0x%X\n", v[2]);
+	seq_printf(sf, "SC_H_CROP:\t\t0x%X\n", v[3]);
+	seq_printf(sf, "SC_V_CROP:\t\t0x%X\n", v[4]);
+	seq_printf(sf, "SC_H_INIT_PH:\t\t0x%X\n", v[5]);
+	seq_printf(sf, "SC_H_DELTA_PH:\t\t0x%X\n", v[6]);
+	seq_printf(sf, "SC_V_INIT_PH:\t\t0x%X\n", v[7]);
+	seq_printf(sf, "SC_V_DELTA_PH:\t\t0x%X\n", v[8]);
+
+	get_values_from_reg(c->reg, 0x130, 10, v);
+	seq_printf(sf, "SC_ENH_LIMITS:\t\t0x%X\n", v[0]);
+	seq_printf(sf, "SC_ENH_COEFF0:\t\t0x%X\n", v[1]);
+	seq_printf(sf, "SC_ENH_COEFF1:\t\t0x%X\n", v[2]);
+	seq_printf(sf, "SC_ENH_COEFF2:\t\t0x%X\n", v[3]);
+	seq_printf(sf, "SC_ENH_COEFF3:\t\t0x%X\n", v[4]);
+	seq_printf(sf, "SC_ENH_COEFF4:\t\t0x%X\n", v[5]);
+	seq_printf(sf, "SC_ENH_COEFF5:\t\t0x%X\n", v[6]);
+	seq_printf(sf, "SC_ENH_COEFF6:\t\t0x%X\n", v[7]);
+	seq_printf(sf, "SC_ENH_COEFF7:\t\t0x%X\n", v[8]);
+	seq_printf(sf, "SC_ENH_COEFF8:\t\t0x%X\n", v[9]);
+}
+
+static const struct komeda_component_funcs d71_scaler_funcs = {
+	.update		= d71_scaler_update,
+	.disable	= d71_component_disable,
+	.dump_register	= d71_scaler_dump,
+};
+
+static int d71_scaler_init(struct d71_dev *d71,
+			   struct block_header *blk, u32 __iomem *reg)
+{
+	struct komeda_component *c;
+	struct komeda_scaler *scaler;
+	u32 pipe_id, comp_id;
+
+	get_resources_id(blk->block_info, &pipe_id, &comp_id);
+
+	c = komeda_component_add(&d71->pipes[pipe_id]->base, sizeof(*scaler),
+				 comp_id, BLOCK_INFO_INPUT_ID(blk->block_info),
+				 &d71_scaler_funcs,
+				 1, get_valid_inputs(blk), 1, reg,
+				 "CU%d_SCALER%d",
+				 pipe_id, BLOCK_INFO_BLK_ID(blk->block_info));
+
+	if (IS_ERR(c)) {
+		DRM_ERROR("Failed to initialize scaler");
+		return PTR_ERR(c);
+	}
+
+	scaler = to_scaler(c);
+	set_range(&scaler->hsize, 4, __get_blk_line_size(d71, reg, 2048));
+	set_range(&scaler->vsize, 4, 4096);
+	scaler->max_downscaling = 6;
+	scaler->max_upscaling = 64;
+	scaler->scaling_split_overlap = 8;
+	scaler->enh_split_overlap = 1;
+
+	malidp_write32(c->reg, BLK_CONTROL, 0);
+
+	return 0;
+}
+
+static int d71_downscaling_clk_check(struct komeda_pipeline *pipe,
+				     struct drm_display_mode *mode,
+				     unsigned long aclk_rate,
+				     struct komeda_data_flow_cfg *dflow)
+{
+	u32 h_in = dflow->in_w;
+	u32 v_in = dflow->in_h;
+	u32 v_out = dflow->out_h;
+	u64 fraction, denominator;
+
+	/* D71 downscaling must satisfy the following equation
+	 *
+	 *   ACLK                   h_in * v_in
+	 * ------- >= ---------------------------------------------
+	 *  PXLCLK     (h_total - (1 + 2 * v_in / v_out)) * v_out
+	 *
+	 * In only horizontal downscaling situation, the right side should be
+	 * multiplied by (h_total - 3) / (h_active - 3), then equation becomes
+	 *
+	 *   ACLK          h_in
+	 * ------- >= ----------------
+	 *  PXLCLK     (h_active - 3)
+	 *
+	 * To avoid precision lost the equation 1 will be convert to:
+	 *
+	 *   ACLK             h_in * v_in
+	 * ------- >= -----------------------------------
+	 *  PXLCLK     (h_total -1 ) * v_out -  2 * v_in
+	 */
+	if (v_in == v_out) {
+		fraction = h_in;
+		denominator = mode->hdisplay - 3;
+	} else {
+		fraction = h_in * v_in;
+		denominator = (mode->htotal - 1) * v_out -  2 * v_in;
+	}
+
+	return aclk_rate * denominator >= mode->crtc_clock * 1000 * fraction ?
+	       0 : -EINVAL;
+}
+
+static void d71_splitter_update(struct komeda_component *c,
+				struct komeda_component_state *state)
+{
+	struct komeda_splitter_state *st = to_splitter_st(state);
+	u32 __iomem *reg = c->reg;
+
+	malidp_write32(reg, BLK_INPUT_ID0, to_d71_input_id(state, 0));
+	malidp_write32(reg, BLK_SIZE, HV_SIZE(st->hsize, st->vsize));
+	malidp_write32(reg, SP_OVERLAP_SIZE, st->overlap & 0x1FFF);
+	malidp_write32(reg, BLK_CONTROL, BLK_CTRL_EN);
+}
+
+static void d71_splitter_dump(struct komeda_component *c, struct seq_file *sf)
+{
+	u32 v[3];
+
+	dump_block_header(sf, c->reg);
+
+	get_values_from_reg(c->reg, BLK_INPUT_ID0, 1, v);
+	seq_printf(sf, "SP_INPUT_ID0:\t\t0x%X\n", v[0]);
+
+	get_values_from_reg(c->reg, BLK_CONTROL, 3, v);
+	seq_printf(sf, "SP_CONTROL:\t\t0x%X\n", v[0]);
+	seq_printf(sf, "SP_SIZE:\t\t0x%X\n", v[1]);
+	seq_printf(sf, "SP_OVERLAP_SIZE:\t0x%X\n", v[2]);
+}
+
+static const struct komeda_component_funcs d71_splitter_funcs = {
+	.update		= d71_splitter_update,
+	.disable	= d71_component_disable,
+	.dump_register	= d71_splitter_dump,
+};
+
+static int d71_splitter_init(struct d71_dev *d71,
+			     struct block_header *blk, u32 __iomem *reg)
+{
+	struct komeda_component *c;
+	struct komeda_splitter *splitter;
+	u32 pipe_id, comp_id;
+
+	get_resources_id(blk->block_info, &pipe_id, &comp_id);
+
+	c = komeda_component_add(&d71->pipes[pipe_id]->base, sizeof(*splitter),
+				 comp_id,
+				 BLOCK_INFO_INPUT_ID(blk->block_info),
+				 &d71_splitter_funcs,
+				 1, get_valid_inputs(blk), 2, reg,
+				 "CU%d_SPLITTER", pipe_id);
+
+	if (IS_ERR(c)) {
+		DRM_ERROR("Failed to initialize splitter");
+		return -1;
+	}
+
+	splitter = to_splitter(c);
+
+	set_range(&splitter->hsize, 4, get_blk_line_size(d71, reg));
+	set_range(&splitter->vsize, 4, d71->max_vsize);
+
+	return 0;
+}
+
+static void d71_merger_update(struct komeda_component *c,
+			      struct komeda_component_state *state)
+{
+	struct komeda_merger_state *st = to_merger_st(state);
+	u32 __iomem *reg = c->reg;
+	u32 index;
+
+	for_each_changed_input(state, index)
+		malidp_write32(reg, MG_INPUT_ID0 + index * 4,
+			       to_d71_input_id(state, index));
+
+	malidp_write32(reg, MG_SIZE, HV_SIZE(st->hsize_merged,
+					     st->vsize_merged));
+	malidp_write32(reg, BLK_CONTROL, BLK_CTRL_EN);
+}
+
+static void d71_merger_dump(struct komeda_component *c, struct seq_file *sf)
+{
+	u32 v;
+
+	dump_block_header(sf, c->reg);
+
+	get_values_from_reg(c->reg, MG_INPUT_ID0, 1, &v);
+	seq_printf(sf, "MG_INPUT_ID0:\t\t0x%X\n", v);
+
+	get_values_from_reg(c->reg, MG_INPUT_ID1, 1, &v);
+	seq_printf(sf, "MG_INPUT_ID1:\t\t0x%X\n", v);
+
+	get_values_from_reg(c->reg, BLK_CONTROL, 1, &v);
+	seq_printf(sf, "MG_CONTROL:\t\t0x%X\n", v);
+
+	get_values_from_reg(c->reg, MG_SIZE, 1, &v);
+	seq_printf(sf, "MG_SIZE:\t\t0x%X\n", v);
+}
+
+static const struct komeda_component_funcs d71_merger_funcs = {
+	.update		= d71_merger_update,
+	.disable	= d71_component_disable,
+	.dump_register	= d71_merger_dump,
+};
+
+static int d71_merger_init(struct d71_dev *d71,
+			   struct block_header *blk, u32 __iomem *reg)
+{
+	struct komeda_component *c;
+	struct komeda_merger *merger;
+	u32 pipe_id, comp_id;
+
+	get_resources_id(blk->block_info, &pipe_id, &comp_id);
+
+	c = komeda_component_add(&d71->pipes[pipe_id]->base, sizeof(*merger),
+				 comp_id,
+				 BLOCK_INFO_INPUT_ID(blk->block_info),
+				 &d71_merger_funcs,
+				 MG_NUM_INPUTS_IDS, get_valid_inputs(blk),
+				 MG_NUM_OUTPUTS_IDS, reg,
+				 "CU%d_MERGER", pipe_id);
+
+	if (IS_ERR(c)) {
+		DRM_ERROR("Failed to initialize merger.\n");
+		return PTR_ERR(c);
+	}
+
+	merger = to_merger(c);
+
+	set_range(&merger->hsize_merged, 4,
+		  __get_blk_line_size(d71, reg, 4032));
+	set_range(&merger->vsize_merged, 4, 4096);
+
+	return 0;
+}
+
+static void d71_improc_update(struct komeda_component *c,
+			      struct komeda_component_state *state)
+{
+	struct drm_crtc_state *crtc_st = state->crtc->state;
+	struct komeda_improc_state *st = to_improc_st(state);
+	struct d71_pipeline *pipe = to_d71_pipeline(c->pipeline);
+	u32 __iomem *reg = c->reg;
+	u32 index, mask = 0, ctrl = 0;
+
+	for_each_changed_input(state, index)
+		malidp_write32(reg, BLK_INPUT_ID0 + index * 4,
+			       to_d71_input_id(state, index));
+
+	malidp_write32(reg, BLK_SIZE, HV_SIZE(st->hsize, st->vsize));
+	malidp_write32(reg, IPS_DEPTH, st->color_depth);
+
+	if (crtc_st->color_mgmt_changed) {
+		mask |= IPS_CTRL_FT | IPS_CTRL_RGB;
+
+		if (crtc_st->gamma_lut) {
+			malidp_write_group(pipe->dou_ft_coeff_addr, FT_COEFF0,
+					   KOMEDA_N_GAMMA_COEFFS,
+					   st->fgamma_coeffs);
+			ctrl |= IPS_CTRL_FT; /* enable gamma */
+		}
+
+		if (crtc_st->ctm) {
+			malidp_write_group(reg, IPS_RGB_RGB_COEFF0,
+					   KOMEDA_N_CTM_COEFFS,
+					   st->ctm_coeffs);
+			ctrl |= IPS_CTRL_RGB; /* enable gamut */
+		}
+	}
+
+	mask |= IPS_CTRL_YUV | IPS_CTRL_CHD422 | IPS_CTRL_CHD420;
+
+	/* config color format */
+	if (st->color_format == DRM_COLOR_FORMAT_YCRCB420)
+		ctrl |= IPS_CTRL_YUV | IPS_CTRL_CHD422 | IPS_CTRL_CHD420;
+	else if (st->color_format == DRM_COLOR_FORMAT_YCRCB422)
+		ctrl |= IPS_CTRL_YUV | IPS_CTRL_CHD422;
+	else if (st->color_format == DRM_COLOR_FORMAT_YCRCB444)
+		ctrl |= IPS_CTRL_YUV;
+
+	malidp_write32_mask(reg, BLK_CONTROL, mask, ctrl);
+}
+
+static void d71_improc_dump(struct komeda_component *c, struct seq_file *sf)
+{
+	u32 v[12], i;
+
+	dump_block_header(sf, c->reg);
+
+	get_values_from_reg(c->reg, 0x80, 2, v);
+	seq_printf(sf, "IPS_INPUT_ID0:\t\t0x%X\n", v[0]);
+	seq_printf(sf, "IPS_INPUT_ID1:\t\t0x%X\n", v[1]);
+
+	get_values_from_reg(c->reg, 0xC0, 1, v);
+	seq_printf(sf, "IPS_INFO:\t\t0x%X\n", v[0]);
+
+	get_values_from_reg(c->reg, 0xD0, 3, v);
+	seq_printf(sf, "IPS_CONTROL:\t\t0x%X\n", v[0]);
+	seq_printf(sf, "IPS_SIZE:\t\t0x%X\n", v[1]);
+	seq_printf(sf, "IPS_DEPTH:\t\t0x%X\n", v[2]);
+
+	get_values_from_reg(c->reg, 0x130, 12, v);
+	for (i = 0; i < 12; i++)
+		seq_printf(sf, "IPS_RGB_RGB_COEFF%u:\t0x%X\n", i, v[i]);
+
+	get_values_from_reg(c->reg, 0x170, 12, v);
+	for (i = 0; i < 12; i++)
+		seq_printf(sf, "IPS_RGB_YUV_COEFF%u:\t0x%X\n", i, v[i]);
+}
+
+static const struct komeda_component_funcs d71_improc_funcs = {
+	.update		= d71_improc_update,
+	.disable	= d71_component_disable,
+	.dump_register	= d71_improc_dump,
+};
+
+static int d71_improc_init(struct d71_dev *d71,
+			   struct block_header *blk, u32 __iomem *reg)
+{
+	struct komeda_component *c;
+	struct komeda_improc *improc;
+	u32 pipe_id, comp_id, value;
+
+	get_resources_id(blk->block_info, &pipe_id, &comp_id);
+
+	c = komeda_component_add(&d71->pipes[pipe_id]->base, sizeof(*improc),
+				 comp_id,
+				 BLOCK_INFO_INPUT_ID(blk->block_info),
+				 &d71_improc_funcs, IPS_NUM_INPUT_IDS,
+				 get_valid_inputs(blk),
+				 IPS_NUM_OUTPUT_IDS, reg, "DOU%d_IPS", pipe_id);
+	if (IS_ERR(c)) {
+		DRM_ERROR("Failed to add improc component\n");
+		return PTR_ERR(c);
+	}
+
+	improc = to_improc(c);
+	improc->supported_color_depths = BIT(8) | BIT(10);
+	improc->supported_color_formats = DRM_COLOR_FORMAT_RGB444 |
+					  DRM_COLOR_FORMAT_YCRCB444 |
+					  DRM_COLOR_FORMAT_YCRCB422;
+	value = malidp_read32(reg, BLK_INFO);
+	if (value & IPS_INFO_CHD420)
+		improc->supported_color_formats |= DRM_COLOR_FORMAT_YCRCB420;
+
+	improc->supports_csc = true;
+	improc->supports_gamma = true;
+
+	return 0;
+}
+
+static void d71_timing_ctrlr_disable(struct komeda_component *c)
+{
+	malidp_write32_mask(c->reg, BLK_CONTROL, BS_CTRL_EN, 0);
+}
+
+static void d71_timing_ctrlr_update(struct komeda_component *c,
+				    struct komeda_component_state *state)
+{
+	struct drm_crtc_state *crtc_st = state->crtc->state;
+	struct drm_display_mode *mode = &crtc_st->adjusted_mode;
+	u32 __iomem *reg = c->reg;
+	u32 hactive, hfront_porch, hback_porch, hsync_len;
+	u32 vactive, vfront_porch, vback_porch, vsync_len;
+	u32 value;
+
+	hactive = mode->crtc_hdisplay;
+	hfront_porch = mode->crtc_hsync_start - mode->crtc_hdisplay;
+	hsync_len = mode->crtc_hsync_end - mode->crtc_hsync_start;
+	hback_porch = mode->crtc_htotal - mode->crtc_hsync_end;
+
+	vactive = mode->crtc_vdisplay;
+	vfront_porch = mode->crtc_vsync_start - mode->crtc_vdisplay;
+	vsync_len = mode->crtc_vsync_end - mode->crtc_vsync_start;
+	vback_porch = mode->crtc_vtotal - mode->crtc_vsync_end;
+
+	malidp_write32(reg, BS_ACTIVESIZE, HV_SIZE(hactive, vactive));
+	malidp_write32(reg, BS_HINTERVALS, BS_H_INTVALS(hfront_porch,
+							hback_porch));
+	malidp_write32(reg, BS_VINTERVALS, BS_V_INTVALS(vfront_porch,
+							vback_porch));
+
+	value = BS_SYNC_VSW(vsync_len) | BS_SYNC_HSW(hsync_len);
+	value |= mode->flags & DRM_MODE_FLAG_PVSYNC ? BS_SYNC_VSP : 0;
+	value |= mode->flags & DRM_MODE_FLAG_PHSYNC ? BS_SYNC_HSP : 0;
+	malidp_write32(reg, BS_SYNC, value);
+
+	malidp_write32(reg, BS_PROG_LINE, D71_DEFAULT_PREPRETCH_LINE - 1);
+	malidp_write32(reg, BS_PREFETCH_LINE, D71_DEFAULT_PREPRETCH_LINE);
+
+	/* configure bs control register */
+	value = BS_CTRL_EN | BS_CTRL_VM;
+	if (c->pipeline->dual_link) {
+		malidp_write32(reg, BS_DRIFT_TO, hfront_porch + 16);
+		value |= BS_CTRL_DL;
+	}
+
+	malidp_write32(reg, BLK_CONTROL, value);
+}
+
+static void d71_timing_ctrlr_dump(struct komeda_component *c,
+				  struct seq_file *sf)
+{
+	u32 v[8], i;
+
+	dump_block_header(sf, c->reg);
+
+	get_values_from_reg(c->reg, 0xC0, 1, v);
+	seq_printf(sf, "BS_INFO:\t\t0x%X\n", v[0]);
+
+	get_values_from_reg(c->reg, 0xD0, 8, v);
+	seq_printf(sf, "BS_CONTROL:\t\t0x%X\n", v[0]);
+	seq_printf(sf, "BS_PROG_LINE:\t\t0x%X\n", v[1]);
+	seq_printf(sf, "BS_PREFETCH_LINE:\t0x%X\n", v[2]);
+	seq_printf(sf, "BS_BG_COLOR:\t\t0x%X\n", v[3]);
+	seq_printf(sf, "BS_ACTIVESIZE:\t\t0x%X\n", v[4]);
+	seq_printf(sf, "BS_HINTERVALS:\t\t0x%X\n", v[5]);
+	seq_printf(sf, "BS_VINTERVALS:\t\t0x%X\n", v[6]);
+	seq_printf(sf, "BS_SYNC:\t\t0x%X\n", v[7]);
+
+	get_values_from_reg(c->reg, 0x100, 3, v);
+	seq_printf(sf, "BS_DRIFT_TO:\t\t0x%X\n", v[0]);
+	seq_printf(sf, "BS_FRAME_TO:\t\t0x%X\n", v[1]);
+	seq_printf(sf, "BS_TE_TO:\t\t0x%X\n", v[2]);
+
+	get_values_from_reg(c->reg, 0x110, 3, v);
+	for (i = 0; i < 3; i++)
+		seq_printf(sf, "BS_T%u_INTERVAL:\t\t0x%X\n", i, v[i]);
+
+	get_values_from_reg(c->reg, 0x120, 5, v);
+	for (i = 0; i < 2; i++) {
+		seq_printf(sf, "BS_CRC%u_LOW:\t\t0x%X\n", i, v[i << 1]);
+		seq_printf(sf, "BS_CRC%u_HIGH:\t\t0x%X\n", i, v[(i << 1) + 1]);
+	}
+	seq_printf(sf, "BS_USER:\t\t0x%X\n", v[4]);
+}
+
+static const struct komeda_component_funcs d71_timing_ctrlr_funcs = {
+	.update		= d71_timing_ctrlr_update,
+	.disable	= d71_timing_ctrlr_disable,
+	.dump_register	= d71_timing_ctrlr_dump,
+};
+
+static int d71_timing_ctrlr_init(struct d71_dev *d71,
+				 struct block_header *blk, u32 __iomem *reg)
+{
+	struct komeda_component *c;
+	struct komeda_timing_ctrlr *ctrlr;
+	u32 pipe_id, comp_id;
+
+	get_resources_id(blk->block_info, &pipe_id, &comp_id);
+
+	c = komeda_component_add(&d71->pipes[pipe_id]->base, sizeof(*ctrlr),
+				 KOMEDA_COMPONENT_TIMING_CTRLR,
+				 BLOCK_INFO_INPUT_ID(blk->block_info),
+				 &d71_timing_ctrlr_funcs,
+				 1, BIT(KOMEDA_COMPONENT_IPS0 + pipe_id),
+				 BS_NUM_OUTPUT_IDS, reg, "DOU%d_BS", pipe_id);
+	if (IS_ERR(c)) {
+		DRM_ERROR("Failed to add display_ctrl component\n");
+		return PTR_ERR(c);
+	}
+
+	ctrlr = to_ctrlr(c);
+
+	ctrlr->supports_dual_link = d71->supports_dual_link;
+
+	return 0;
+}
+
+int d71_probe_block(struct d71_dev *d71,
+		    struct block_header *blk, u32 __iomem *reg)
+{
+	struct d71_pipeline *pipe;
+	int blk_id = BLOCK_INFO_BLK_ID(blk->block_info);
+
+	int err = 0;
+
+	switch (BLOCK_INFO_BLK_TYPE(blk->block_info)) {
+	case D71_BLK_TYPE_GCU:
+		break;
+
+	case D71_BLK_TYPE_LPU:
+		pipe = d71->pipes[blk_id];
+		pipe->lpu_addr = reg;
+		break;
+
+	case D71_BLK_TYPE_LPU_LAYER:
+		err = d71_layer_init(d71, blk, reg);
+		break;
+
+	case D71_BLK_TYPE_LPU_WB_LAYER:
+		err = d71_wb_layer_init(d71, blk, reg);
+		break;
+
+	case D71_BLK_TYPE_CU:
+		pipe = d71->pipes[blk_id];
+		pipe->cu_addr = reg;
+		err = d71_compiz_init(d71, blk, reg);
+		break;
+
+	case D71_BLK_TYPE_CU_SCALER:
+		err = d71_scaler_init(d71, blk, reg);
+		break;
+
+	case D71_BLK_TYPE_CU_SPLITTER:
+		err = d71_splitter_init(d71, blk, reg);
+		break;
+
+	case D71_BLK_TYPE_CU_MERGER:
+		err = d71_merger_init(d71, blk, reg);
+		break;
+
+	case D71_BLK_TYPE_DOU:
+		pipe = d71->pipes[blk_id];
+		pipe->dou_addr = reg;
+		break;
+
+	case D71_BLK_TYPE_DOU_IPS:
+		err = d71_improc_init(d71, blk, reg);
+		break;
+
+	case D71_BLK_TYPE_DOU_FT_COEFF:
+		pipe = d71->pipes[blk_id];
+		pipe->dou_ft_coeff_addr = reg;
+		break;
+
+	case D71_BLK_TYPE_DOU_BS:
+		err = d71_timing_ctrlr_init(d71, blk, reg);
+		break;
+
+	case D71_BLK_TYPE_GLB_LT_COEFF:
+		break;
+
+	case D71_BLK_TYPE_GLB_SCL_COEFF:
+		d71->glb_scl_coeff_addr[blk_id] = reg;
+		break;
+
+	default:
+		DRM_ERROR("Unknown block (block_info: 0x%x) is found\n",
+			  blk->block_info);
+		err = -EINVAL;
+		break;
+	}
+
+	return err;
+}
+
+static void d71_gcu_dump(struct d71_dev *d71, struct seq_file *sf)
+{
+	u32 v[5];
+
+	seq_puts(sf, "\n------ GCU ------\n");
+
+	get_values_from_reg(d71->gcu_addr, 0, 3, v);
+	seq_printf(sf, "GLB_ARCH_ID:\t\t0x%X\n", v[0]);
+	seq_printf(sf, "GLB_CORE_ID:\t\t0x%X\n", v[1]);
+	seq_printf(sf, "GLB_CORE_INFO:\t\t0x%X\n", v[2]);
+
+	get_values_from_reg(d71->gcu_addr, 0x10, 1, v);
+	seq_printf(sf, "GLB_IRQ_STATUS:\t\t0x%X\n", v[0]);
+
+	get_values_from_reg(d71->gcu_addr, 0xA0, 5, v);
+	seq_printf(sf, "GCU_IRQ_RAW_STATUS:\t0x%X\n", v[0]);
+	seq_printf(sf, "GCU_IRQ_CLEAR:\t\t0x%X\n", v[1]);
+	seq_printf(sf, "GCU_IRQ_MASK:\t\t0x%X\n", v[2]);
+	seq_printf(sf, "GCU_IRQ_STATUS:\t\t0x%X\n", v[3]);
+	seq_printf(sf, "GCU_STATUS:\t\t0x%X\n", v[4]);
+
+	get_values_from_reg(d71->gcu_addr, 0xD0, 3, v);
+	seq_printf(sf, "GCU_CONTROL:\t\t0x%X\n", v[0]);
+	seq_printf(sf, "GCU_CONFIG_VALID0:\t0x%X\n", v[1]);
+	seq_printf(sf, "GCU_CONFIG_VALID1:\t0x%X\n", v[2]);
+}
+
+static void d71_lpu_dump(struct d71_pipeline *pipe, struct seq_file *sf)
+{
+	u32 v[6];
+
+	seq_printf(sf, "\n------ LPU%d ------\n", pipe->base.id);
+
+	dump_block_header(sf, pipe->lpu_addr);
+
+	get_values_from_reg(pipe->lpu_addr, 0xA0, 6, v);
+	seq_printf(sf, "LPU_IRQ_RAW_STATUS:\t0x%X\n", v[0]);
+	seq_printf(sf, "LPU_IRQ_CLEAR:\t\t0x%X\n", v[1]);
+	seq_printf(sf, "LPU_IRQ_MASK:\t\t0x%X\n", v[2]);
+	seq_printf(sf, "LPU_IRQ_STATUS:\t\t0x%X\n", v[3]);
+	seq_printf(sf, "LPU_STATUS:\t\t0x%X\n", v[4]);
+	seq_printf(sf, "LPU_TBU_STATUS:\t\t0x%X\n", v[5]);
+
+	get_values_from_reg(pipe->lpu_addr, 0xC0, 1, v);
+	seq_printf(sf, "LPU_INFO:\t\t0x%X\n", v[0]);
+
+	get_values_from_reg(pipe->lpu_addr, 0xD0, 3, v);
+	seq_printf(sf, "LPU_RAXI_CONTROL:\t0x%X\n", v[0]);
+	seq_printf(sf, "LPU_WAXI_CONTROL:\t0x%X\n", v[1]);
+	seq_printf(sf, "LPU_TBU_CONTROL:\t0x%X\n", v[2]);
+}
+
+static void d71_dou_dump(struct d71_pipeline *pipe, struct seq_file *sf)
+{
+	u32 v[5];
+
+	seq_printf(sf, "\n------ DOU%d ------\n", pipe->base.id);
+
+	dump_block_header(sf, pipe->dou_addr);
+
+	get_values_from_reg(pipe->dou_addr, 0xA0, 5, v);
+	seq_printf(sf, "DOU_IRQ_RAW_STATUS:\t0x%X\n", v[0]);
+	seq_printf(sf, "DOU_IRQ_CLEAR:\t\t0x%X\n", v[1]);
+	seq_printf(sf, "DOU_IRQ_MASK:\t\t0x%X\n", v[2]);
+	seq_printf(sf, "DOU_IRQ_STATUS:\t\t0x%X\n", v[3]);
+	seq_printf(sf, "DOU_STATUS:\t\t0x%X\n", v[4]);
+}
+
+static void d71_pipeline_dump(struct komeda_pipeline *pipe, struct seq_file *sf)
+{
+	struct d71_pipeline *d71_pipe = to_d71_pipeline(pipe);
+
+	d71_lpu_dump(d71_pipe, sf);
+	d71_dou_dump(d71_pipe, sf);
+}
+
+const struct komeda_pipeline_funcs d71_pipeline_funcs = {
+	.downscaling_clk_check	= d71_downscaling_clk_check,
+	.dump_register		= d71_pipeline_dump,
+};
+
+void d71_dump(struct komeda_dev *mdev, struct seq_file *sf)
+{
+	struct d71_dev *d71 = mdev->chip_data;
+
+	d71_gcu_dump(d71, sf);
+}
diff --git a/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.c b/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.c
new file mode 100644
index 000000000..00fa56c29
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.c
@@ -0,0 +1,643 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+
+#include <drm/drm_print.h>
+#include "d71_dev.h"
+#include "malidp_io.h"
+
+static u64 get_lpu_event(struct d71_pipeline *d71_pipeline)
+{
+	u32 __iomem *reg = d71_pipeline->lpu_addr;
+	u32 status, raw_status;
+	u64 evts = 0ULL;
+
+	raw_status = malidp_read32(reg, BLK_IRQ_RAW_STATUS);
+	if (raw_status & LPU_IRQ_IBSY)
+		evts |= KOMEDA_EVENT_IBSY;
+	if (raw_status & LPU_IRQ_EOW)
+		evts |= KOMEDA_EVENT_EOW;
+	if (raw_status & LPU_IRQ_OVR)
+		evts |= KOMEDA_EVENT_OVR;
+
+	if (raw_status & (LPU_IRQ_ERR | LPU_IRQ_IBSY | LPU_IRQ_OVR)) {
+		u32 restore = 0, tbu_status;
+		/* Check error of LPU status */
+		status = malidp_read32(reg, BLK_STATUS);
+		if (status & LPU_STATUS_AXIE) {
+			restore |= LPU_STATUS_AXIE;
+			evts |= KOMEDA_ERR_AXIE;
+		}
+		if (status & LPU_STATUS_ACE0) {
+			restore |= LPU_STATUS_ACE0;
+			evts |= KOMEDA_ERR_ACE0;
+		}
+		if (status & LPU_STATUS_ACE1) {
+			restore |= LPU_STATUS_ACE1;
+			evts |= KOMEDA_ERR_ACE1;
+		}
+		if (status & LPU_STATUS_ACE2) {
+			restore |= LPU_STATUS_ACE2;
+			evts |= KOMEDA_ERR_ACE2;
+		}
+		if (status & LPU_STATUS_ACE3) {
+			restore |= LPU_STATUS_ACE3;
+			evts |= KOMEDA_ERR_ACE3;
+		}
+		if (status & LPU_STATUS_FEMPTY) {
+			restore |= LPU_STATUS_FEMPTY;
+			evts |= KOMEDA_EVENT_EMPTY;
+		}
+		if (status & LPU_STATUS_FFULL) {
+			restore |= LPU_STATUS_FFULL;
+			evts |= KOMEDA_EVENT_FULL;
+		}
+
+		if (restore != 0)
+			malidp_write32_mask(reg, BLK_STATUS, restore, 0);
+
+		restore = 0;
+		/* Check errors of TBU status */
+		tbu_status = malidp_read32(reg, LPU_TBU_STATUS);
+		if (tbu_status & LPU_TBU_STATUS_TCF) {
+			restore |= LPU_TBU_STATUS_TCF;
+			evts |= KOMEDA_ERR_TCF;
+		}
+		if (tbu_status & LPU_TBU_STATUS_TTNG) {
+			restore |= LPU_TBU_STATUS_TTNG;
+			evts |= KOMEDA_ERR_TTNG;
+		}
+		if (tbu_status & LPU_TBU_STATUS_TITR) {
+			restore |= LPU_TBU_STATUS_TITR;
+			evts |= KOMEDA_ERR_TITR;
+		}
+		if (tbu_status & LPU_TBU_STATUS_TEMR) {
+			restore |= LPU_TBU_STATUS_TEMR;
+			evts |= KOMEDA_ERR_TEMR;
+		}
+		if (tbu_status & LPU_TBU_STATUS_TTF) {
+			restore |= LPU_TBU_STATUS_TTF;
+			evts |= KOMEDA_ERR_TTF;
+		}
+		if (restore != 0)
+			malidp_write32_mask(reg, LPU_TBU_STATUS, restore, 0);
+	}
+
+	malidp_write32(reg, BLK_IRQ_CLEAR, raw_status);
+	return evts;
+}
+
+static u64 get_cu_event(struct d71_pipeline *d71_pipeline)
+{
+	u32 __iomem *reg = d71_pipeline->cu_addr;
+	u32 status, raw_status;
+	u64 evts = 0ULL;
+
+	raw_status = malidp_read32(reg, BLK_IRQ_RAW_STATUS);
+	if (raw_status & CU_IRQ_OVR)
+		evts |= KOMEDA_EVENT_OVR;
+
+	if (raw_status & (CU_IRQ_ERR | CU_IRQ_OVR)) {
+		status = malidp_read32(reg, BLK_STATUS) & 0x7FFFFFFF;
+		if (status & CU_STATUS_CPE)
+			evts |= KOMEDA_ERR_CPE;
+		if (status & CU_STATUS_ZME)
+			evts |= KOMEDA_ERR_ZME;
+		if (status & CU_STATUS_CFGE)
+			evts |= KOMEDA_ERR_CFGE;
+		if (status)
+			malidp_write32_mask(reg, BLK_STATUS, status, 0);
+	}
+
+	malidp_write32(reg, BLK_IRQ_CLEAR, raw_status);
+
+	return evts;
+}
+
+static u64 get_dou_event(struct d71_pipeline *d71_pipeline)
+{
+	u32 __iomem *reg = d71_pipeline->dou_addr;
+	u32 status, raw_status;
+	u64 evts = 0ULL;
+
+	raw_status = malidp_read32(reg, BLK_IRQ_RAW_STATUS);
+	if (raw_status & DOU_IRQ_PL0)
+		evts |= KOMEDA_EVENT_VSYNC;
+	if (raw_status & DOU_IRQ_UND)
+		evts |= KOMEDA_EVENT_URUN;
+
+	if (raw_status & (DOU_IRQ_ERR | DOU_IRQ_UND)) {
+		u32 restore  = 0;
+
+		status = malidp_read32(reg, BLK_STATUS);
+		if (status & DOU_STATUS_DRIFTTO) {
+			restore |= DOU_STATUS_DRIFTTO;
+			evts |= KOMEDA_ERR_DRIFTTO;
+		}
+		if (status & DOU_STATUS_FRAMETO) {
+			restore |= DOU_STATUS_FRAMETO;
+			evts |= KOMEDA_ERR_FRAMETO;
+		}
+		if (status & DOU_STATUS_TETO) {
+			restore |= DOU_STATUS_TETO;
+			evts |= KOMEDA_ERR_TETO;
+		}
+		if (status & DOU_STATUS_CSCE) {
+			restore |= DOU_STATUS_CSCE;
+			evts |= KOMEDA_ERR_CSCE;
+		}
+
+		if (restore != 0)
+			malidp_write32_mask(reg, BLK_STATUS, restore, 0);
+	}
+
+	malidp_write32(reg, BLK_IRQ_CLEAR, raw_status);
+	return evts;
+}
+
+static u64 get_pipeline_event(struct d71_pipeline *d71_pipeline, u32 gcu_status)
+{
+	u32 evts = 0ULL;
+
+	if (gcu_status & (GLB_IRQ_STATUS_LPU0 | GLB_IRQ_STATUS_LPU1))
+		evts |= get_lpu_event(d71_pipeline);
+
+	if (gcu_status & (GLB_IRQ_STATUS_CU0 | GLB_IRQ_STATUS_CU1))
+		evts |= get_cu_event(d71_pipeline);
+
+	if (gcu_status & (GLB_IRQ_STATUS_DOU0 | GLB_IRQ_STATUS_DOU1))
+		evts |= get_dou_event(d71_pipeline);
+
+	return evts;
+}
+
+static irqreturn_t
+d71_irq_handler(struct komeda_dev *mdev, struct komeda_events *evts)
+{
+	struct d71_dev *d71 = mdev->chip_data;
+	u32 status, gcu_status, raw_status;
+
+	gcu_status = malidp_read32(d71->gcu_addr, GLB_IRQ_STATUS);
+
+	if (gcu_status & GLB_IRQ_STATUS_GCU) {
+		raw_status = malidp_read32(d71->gcu_addr, BLK_IRQ_RAW_STATUS);
+		if (raw_status & GCU_IRQ_CVAL0)
+			evts->pipes[0] |= KOMEDA_EVENT_FLIP;
+		if (raw_status & GCU_IRQ_CVAL1)
+			evts->pipes[1] |= KOMEDA_EVENT_FLIP;
+		if (raw_status & GCU_IRQ_ERR) {
+			status = malidp_read32(d71->gcu_addr, BLK_STATUS);
+			if (status & GCU_STATUS_MERR) {
+				evts->global |= KOMEDA_ERR_MERR;
+				malidp_write32_mask(d71->gcu_addr, BLK_STATUS,
+						    GCU_STATUS_MERR, 0);
+			}
+		}
+
+		malidp_write32(d71->gcu_addr, BLK_IRQ_CLEAR, raw_status);
+	}
+
+	if (gcu_status & GLB_IRQ_STATUS_PIPE0)
+		evts->pipes[0] |= get_pipeline_event(d71->pipes[0], gcu_status);
+
+	if (gcu_status & GLB_IRQ_STATUS_PIPE1)
+		evts->pipes[1] |= get_pipeline_event(d71->pipes[1], gcu_status);
+
+	return IRQ_RETVAL(gcu_status);
+}
+
+#define ENABLED_GCU_IRQS	(GCU_IRQ_CVAL0 | GCU_IRQ_CVAL1 | \
+				 GCU_IRQ_MODE | GCU_IRQ_ERR)
+#define ENABLED_LPU_IRQS	(LPU_IRQ_IBSY | LPU_IRQ_ERR | LPU_IRQ_EOW)
+#define ENABLED_CU_IRQS		(CU_IRQ_OVR | CU_IRQ_ERR)
+#define ENABLED_DOU_IRQS	(DOU_IRQ_UND | DOU_IRQ_ERR)
+
+static int d71_enable_irq(struct komeda_dev *mdev)
+{
+	struct d71_dev *d71 = mdev->chip_data;
+	struct d71_pipeline *pipe;
+	u32 i;
+
+	malidp_write32_mask(d71->gcu_addr, BLK_IRQ_MASK,
+			    ENABLED_GCU_IRQS, ENABLED_GCU_IRQS);
+	for (i = 0; i < d71->num_pipelines; i++) {
+		pipe = d71->pipes[i];
+		malidp_write32_mask(pipe->cu_addr,  BLK_IRQ_MASK,
+				    ENABLED_CU_IRQS, ENABLED_CU_IRQS);
+		malidp_write32_mask(pipe->lpu_addr, BLK_IRQ_MASK,
+				    ENABLED_LPU_IRQS, ENABLED_LPU_IRQS);
+		malidp_write32_mask(pipe->dou_addr, BLK_IRQ_MASK,
+				    ENABLED_DOU_IRQS, ENABLED_DOU_IRQS);
+	}
+	return 0;
+}
+
+static int d71_disable_irq(struct komeda_dev *mdev)
+{
+	struct d71_dev *d71 = mdev->chip_data;
+	struct d71_pipeline *pipe;
+	u32 i;
+
+	malidp_write32_mask(d71->gcu_addr, BLK_IRQ_MASK, ENABLED_GCU_IRQS, 0);
+	for (i = 0; i < d71->num_pipelines; i++) {
+		pipe = d71->pipes[i];
+		malidp_write32_mask(pipe->cu_addr,  BLK_IRQ_MASK,
+				    ENABLED_CU_IRQS, 0);
+		malidp_write32_mask(pipe->lpu_addr, BLK_IRQ_MASK,
+				    ENABLED_LPU_IRQS, 0);
+		malidp_write32_mask(pipe->dou_addr, BLK_IRQ_MASK,
+				    ENABLED_DOU_IRQS, 0);
+	}
+	return 0;
+}
+
+static void d71_on_off_vblank(struct komeda_dev *mdev, int master_pipe, bool on)
+{
+	struct d71_dev *d71 = mdev->chip_data;
+	struct d71_pipeline *pipe = d71->pipes[master_pipe];
+
+	malidp_write32_mask(pipe->dou_addr, BLK_IRQ_MASK,
+			    DOU_IRQ_PL0, on ? DOU_IRQ_PL0 : 0);
+}
+
+static int to_d71_opmode(int core_mode)
+{
+	switch (core_mode) {
+	case KOMEDA_MODE_DISP0:
+		return DO0_ACTIVE_MODE;
+	case KOMEDA_MODE_DISP1:
+		return DO1_ACTIVE_MODE;
+	case KOMEDA_MODE_DUAL_DISP:
+		return DO01_ACTIVE_MODE;
+	case KOMEDA_MODE_INACTIVE:
+		return INACTIVE_MODE;
+	default:
+		WARN(1, "Unknown operation mode");
+		return INACTIVE_MODE;
+	}
+}
+
+static int d71_change_opmode(struct komeda_dev *mdev, int new_mode)
+{
+	struct d71_dev *d71 = mdev->chip_data;
+	u32 opmode = to_d71_opmode(new_mode);
+	int ret;
+
+	malidp_write32_mask(d71->gcu_addr, BLK_CONTROL, 0x7, opmode);
+
+	ret = dp_wait_cond(((malidp_read32(d71->gcu_addr, BLK_CONTROL) & 0x7) == opmode),
+			   100, 1000, 10000);
+
+	return ret;
+}
+
+static void d71_flush(struct komeda_dev *mdev,
+		      int master_pipe, u32 active_pipes)
+{
+	struct d71_dev *d71 = mdev->chip_data;
+	u32 reg_offset = (master_pipe == 0) ?
+			 GCU_CONFIG_VALID0 : GCU_CONFIG_VALID1;
+
+	malidp_write32(d71->gcu_addr, reg_offset, GCU_CONFIG_CVAL);
+}
+
+static int d71_reset(struct d71_dev *d71)
+{
+	u32 __iomem *gcu = d71->gcu_addr;
+	int ret;
+
+	malidp_write32_mask(gcu, BLK_CONTROL,
+			    GCU_CONTROL_SRST, GCU_CONTROL_SRST);
+
+	ret = dp_wait_cond(!(malidp_read32(gcu, BLK_CONTROL) & GCU_CONTROL_SRST),
+			   100, 1000, 10000);
+
+	return ret;
+}
+
+void d71_read_block_header(u32 __iomem *reg, struct block_header *blk)
+{
+	int i;
+
+	blk->block_info = malidp_read32(reg, BLK_BLOCK_INFO);
+	if (BLOCK_INFO_BLK_TYPE(blk->block_info) == D71_BLK_TYPE_RESERVED)
+		return;
+
+	blk->pipeline_info = malidp_read32(reg, BLK_PIPELINE_INFO);
+
+	/* get valid input and output ids */
+	for (i = 0; i < PIPELINE_INFO_N_VALID_INPUTS(blk->pipeline_info); i++)
+		blk->input_ids[i] = malidp_read32(reg + i, BLK_VALID_INPUT_ID0);
+	for (i = 0; i < PIPELINE_INFO_N_OUTPUTS(blk->pipeline_info); i++)
+		blk->output_ids[i] = malidp_read32(reg + i, BLK_OUTPUT_ID0);
+}
+
+static void d71_cleanup(struct komeda_dev *mdev)
+{
+	struct d71_dev *d71 = mdev->chip_data;
+
+	if (!d71)
+		return;
+
+	devm_kfree(mdev->dev, d71);
+	mdev->chip_data = NULL;
+}
+
+static int d71_enum_resources(struct komeda_dev *mdev)
+{
+	struct d71_dev *d71;
+	struct komeda_pipeline *pipe;
+	struct block_header blk;
+	u32 __iomem *blk_base;
+	u32 i, value, offset;
+	int err;
+
+	d71 = devm_kzalloc(mdev->dev, sizeof(*d71), GFP_KERNEL);
+	if (!d71)
+		return -ENOMEM;
+
+	mdev->chip_data = d71;
+	d71->mdev = mdev;
+	d71->gcu_addr = mdev->reg_base;
+	d71->periph_addr = mdev->reg_base + (D71_BLOCK_OFFSET_PERIPH >> 2);
+
+	err = d71_reset(d71);
+	if (err) {
+		DRM_ERROR("Fail to reset d71 device.\n");
+		goto err_cleanup;
+	}
+
+	/* probe GCU */
+	value = malidp_read32(d71->gcu_addr, GLB_CORE_INFO);
+	d71->num_blocks = value & 0xFF;
+	d71->num_pipelines = (value >> 8) & 0x7;
+
+	if (d71->num_pipelines > D71_MAX_PIPELINE) {
+		DRM_ERROR("d71 supports %d pipelines, but got: %d.\n",
+			  D71_MAX_PIPELINE, d71->num_pipelines);
+		err = -EINVAL;
+		goto err_cleanup;
+	}
+
+	/* Only the legacy HW has the periph block, the newer merges the periph
+	 * into GCU
+	 */
+	value = malidp_read32(d71->periph_addr, BLK_BLOCK_INFO);
+	if (BLOCK_INFO_BLK_TYPE(value) != D71_BLK_TYPE_PERIPH)
+		d71->periph_addr = NULL;
+
+	if (d71->periph_addr) {
+		/* probe PERIPHERAL in legacy HW */
+		value = malidp_read32(d71->periph_addr, PERIPH_CONFIGURATION_ID);
+
+		d71->max_line_size	= value & PERIPH_MAX_LINE_SIZE ? 4096 : 2048;
+		d71->max_vsize		= 4096;
+		d71->num_rich_layers	= value & PERIPH_NUM_RICH_LAYERS ? 2 : 1;
+		d71->supports_dual_link	= !!(value & PERIPH_SPLIT_EN);
+		d71->integrates_tbu	= !!(value & PERIPH_TBU_EN);
+	} else {
+		value = malidp_read32(d71->gcu_addr, GCU_CONFIGURATION_ID0);
+		d71->max_line_size	= GCU_MAX_LINE_SIZE(value);
+		d71->max_vsize		= GCU_MAX_NUM_LINES(value);
+
+		value = malidp_read32(d71->gcu_addr, GCU_CONFIGURATION_ID1);
+		d71->num_rich_layers	= GCU_NUM_RICH_LAYERS(value);
+		d71->supports_dual_link	= GCU_DISPLAY_SPLIT_EN(value);
+		d71->integrates_tbu	= GCU_DISPLAY_TBU_EN(value);
+	}
+
+	for (i = 0; i < d71->num_pipelines; i++) {
+		pipe = komeda_pipeline_add(mdev, sizeof(struct d71_pipeline),
+					   &d71_pipeline_funcs);
+		if (IS_ERR(pipe)) {
+			err = PTR_ERR(pipe);
+			goto err_cleanup;
+		}
+
+		/* D71 HW doesn't update shadow registers when display output
+		 * is turning off, so when we disable all pipeline components
+		 * together with display output disable by one flush or one
+		 * operation, the disable operation updated registers will not
+		 * be flush to or valid in HW, which may leads problem.
+		 * To workaround this problem, introduce a two phase disable.
+		 * Phase1: Disabling components with display is on to make sure
+		 *	   the disable can be flushed to HW.
+		 * Phase2: Only turn-off display output.
+		 */
+		value = KOMEDA_PIPELINE_IMPROCS |
+			BIT(KOMEDA_COMPONENT_TIMING_CTRLR);
+
+		pipe->standalone_disabled_comps = value;
+
+		d71->pipes[i] = to_d71_pipeline(pipe);
+	}
+
+	/* loop the register blks and probe.
+	 * NOTE: d71->num_blocks includes reserved blocks.
+	 * d71->num_blocks = GCU + valid blocks + reserved blocks
+	 */
+	i = 1; /* exclude GCU */
+	offset = D71_BLOCK_SIZE; /* skip GCU */
+	while (i < d71->num_blocks) {
+		blk_base = mdev->reg_base + (offset >> 2);
+
+		d71_read_block_header(blk_base, &blk);
+		if (BLOCK_INFO_BLK_TYPE(blk.block_info) != D71_BLK_TYPE_RESERVED) {
+			err = d71_probe_block(d71, &blk, blk_base);
+			if (err)
+				goto err_cleanup;
+		}
+
+		i++;
+		offset += D71_BLOCK_SIZE;
+	}
+
+	DRM_DEBUG("total %d (out of %d) blocks are found.\n",
+		  i, d71->num_blocks);
+
+	return 0;
+
+err_cleanup:
+	d71_cleanup(mdev);
+	return err;
+}
+
+#define __HW_ID(__group, __format) \
+	((((__group) & 0x7) << 3) | ((__format) & 0x7))
+
+#define RICH		KOMEDA_FMT_RICH_LAYER
+#define SIMPLE		KOMEDA_FMT_SIMPLE_LAYER
+#define RICH_SIMPLE	(KOMEDA_FMT_RICH_LAYER | KOMEDA_FMT_SIMPLE_LAYER)
+#define RICH_WB		(KOMEDA_FMT_RICH_LAYER | KOMEDA_FMT_WB_LAYER)
+#define RICH_SIMPLE_WB	(RICH_SIMPLE | KOMEDA_FMT_WB_LAYER)
+
+#define Rot_0		DRM_MODE_ROTATE_0
+#define Flip_H_V	(DRM_MODE_REFLECT_X | DRM_MODE_REFLECT_Y | Rot_0)
+#define Rot_ALL_H_V	(DRM_MODE_ROTATE_MASK | Flip_H_V)
+
+#define LYT_NM		BIT(AFBC_FORMAT_MOD_BLOCK_SIZE_16x16)
+#define LYT_WB		BIT(AFBC_FORMAT_MOD_BLOCK_SIZE_32x8)
+#define LYT_NM_WB	(LYT_NM | LYT_WB)
+
+#define AFB_TH		AFBC(_TILED | _SPARSE)
+#define AFB_TH_SC_YTR	AFBC(_TILED | _SC | _SPARSE | _YTR)
+#define AFB_TH_SC_YTR_BS AFBC(_TILED | _SC | _SPARSE | _YTR | _SPLIT)
+
+static struct komeda_format_caps d71_format_caps_table[] = {
+	/*   HW_ID    |        fourcc         |   layer_types |   rots    | afbc_layouts | afbc_features */
+	/* ABGR_2101010*/
+	{__HW_ID(0, 0),	DRM_FORMAT_ARGB2101010,	RICH_SIMPLE_WB,	Flip_H_V,		0, 0},
+	{__HW_ID(0, 1),	DRM_FORMAT_ABGR2101010,	RICH_SIMPLE_WB,	Flip_H_V,		0, 0},
+	{__HW_ID(0, 1),	DRM_FORMAT_ABGR2101010,	RICH_SIMPLE,	Rot_ALL_H_V,	LYT_NM_WB, AFB_TH_SC_YTR_BS}, /* afbc */
+	{__HW_ID(0, 2),	DRM_FORMAT_RGBA1010102,	RICH_SIMPLE_WB,	Flip_H_V,		0, 0},
+	{__HW_ID(0, 3),	DRM_FORMAT_BGRA1010102,	RICH_SIMPLE_WB,	Flip_H_V,		0, 0},
+	/* ABGR_8888*/
+	{__HW_ID(1, 0),	DRM_FORMAT_ARGB8888,	RICH_SIMPLE_WB,	Flip_H_V,		0, 0},
+	{__HW_ID(1, 1),	DRM_FORMAT_ABGR8888,	RICH_SIMPLE_WB,	Flip_H_V,		0, 0},
+	{__HW_ID(1, 1),	DRM_FORMAT_ABGR8888,	RICH_SIMPLE,	Rot_ALL_H_V,	LYT_NM_WB, AFB_TH_SC_YTR_BS}, /* afbc */
+	{__HW_ID(1, 2),	DRM_FORMAT_RGBA8888,	RICH_SIMPLE_WB,	Flip_H_V,		0, 0},
+	{__HW_ID(1, 3),	DRM_FORMAT_BGRA8888,	RICH_SIMPLE_WB,	Flip_H_V,		0, 0},
+	/* XBGB_8888 */
+	{__HW_ID(2, 0),	DRM_FORMAT_XRGB8888,	RICH_SIMPLE_WB,	Flip_H_V,		0, 0},
+	{__HW_ID(2, 1),	DRM_FORMAT_XBGR8888,	RICH_SIMPLE_WB,	Flip_H_V,		0, 0},
+	{__HW_ID(2, 2),	DRM_FORMAT_RGBX8888,	RICH_SIMPLE_WB,	Flip_H_V,		0, 0},
+	{__HW_ID(2, 3),	DRM_FORMAT_BGRX8888,	RICH_SIMPLE_WB,	Flip_H_V,		0, 0},
+	/* BGR_888 */ /* none-afbc RGB888 doesn't support rotation and flip */
+	{__HW_ID(3, 0),	DRM_FORMAT_RGB888,	RICH_SIMPLE_WB,	Rot_0,			0, 0},
+	{__HW_ID(3, 1),	DRM_FORMAT_BGR888,	RICH_SIMPLE_WB,	Rot_0,			0, 0},
+	{__HW_ID(3, 1),	DRM_FORMAT_BGR888,	RICH_SIMPLE,	Rot_ALL_H_V,	LYT_NM_WB, AFB_TH_SC_YTR_BS}, /* afbc */
+	/* BGR 16bpp */
+	{__HW_ID(4, 0),	DRM_FORMAT_RGBA5551,	RICH_SIMPLE,	Flip_H_V,		0, 0},
+	{__HW_ID(4, 1),	DRM_FORMAT_ABGR1555,	RICH_SIMPLE,	Flip_H_V,		0, 0},
+	{__HW_ID(4, 1),	DRM_FORMAT_ABGR1555,	RICH_SIMPLE,	Rot_ALL_H_V,	LYT_NM_WB, AFB_TH_SC_YTR}, /* afbc */
+	{__HW_ID(4, 2),	DRM_FORMAT_RGB565,	RICH_SIMPLE,	Flip_H_V,		0, 0},
+	{__HW_ID(4, 3),	DRM_FORMAT_BGR565,	RICH_SIMPLE,	Flip_H_V,		0, 0},
+	{__HW_ID(4, 3),	DRM_FORMAT_BGR565,	RICH_SIMPLE,	Rot_ALL_H_V,	LYT_NM_WB, AFB_TH_SC_YTR}, /* afbc */
+	{__HW_ID(4, 4), DRM_FORMAT_R8,		SIMPLE,		Rot_0,			0, 0},
+	/* YUV 444/422/420 8bit  */
+	{__HW_ID(5, 1),	DRM_FORMAT_YUYV,	RICH,		Rot_ALL_H_V,	LYT_NM, AFB_TH}, /* afbc */
+	{__HW_ID(5, 2),	DRM_FORMAT_YUYV,	RICH,		Flip_H_V,		0, 0},
+	{__HW_ID(5, 3),	DRM_FORMAT_UYVY,	RICH,		Flip_H_V,		0, 0},
+	{__HW_ID(5, 6),	DRM_FORMAT_NV12,	RICH,		Flip_H_V,		0, 0},
+	{__HW_ID(5, 6),	DRM_FORMAT_YUV420_8BIT,	RICH,		Rot_ALL_H_V,	LYT_NM, AFB_TH}, /* afbc */
+	{__HW_ID(5, 7),	DRM_FORMAT_YUV420,	RICH,		Flip_H_V,		0, 0},
+	/* YUV 10bit*/
+	{__HW_ID(6, 6),	DRM_FORMAT_X0L2,	RICH,		Flip_H_V,		0, 0},
+	{__HW_ID(6, 7),	DRM_FORMAT_P010,	RICH,		Flip_H_V,		0, 0},
+	{__HW_ID(6, 7),	DRM_FORMAT_YUV420_10BIT, RICH,		Rot_ALL_H_V,	LYT_NM, AFB_TH},
+};
+
+static bool d71_format_mod_supported(const struct komeda_format_caps *caps,
+				     u32 layer_type, u64 modifier, u32 rot)
+{
+	uint64_t layout = modifier & AFBC_FORMAT_MOD_BLOCK_SIZE_MASK;
+
+	if ((layout == AFBC_FORMAT_MOD_BLOCK_SIZE_32x8) &&
+	    drm_rotation_90_or_270(rot)) {
+		DRM_DEBUG_ATOMIC("D71 doesn't support ROT90 for WB-AFBC.\n");
+		return false;
+	}
+
+	return true;
+}
+
+static void d71_init_fmt_tbl(struct komeda_dev *mdev)
+{
+	struct komeda_format_caps_table *table = &mdev->fmt_tbl;
+
+	table->format_caps = d71_format_caps_table;
+	table->format_mod_supported = d71_format_mod_supported;
+	table->n_formats = ARRAY_SIZE(d71_format_caps_table);
+}
+
+static int d71_connect_iommu(struct komeda_dev *mdev)
+{
+	struct d71_dev *d71 = mdev->chip_data;
+	u32 __iomem *reg = d71->gcu_addr;
+	u32 check_bits = (d71->num_pipelines == 2) ?
+			 GCU_STATUS_TCS0 | GCU_STATUS_TCS1 : GCU_STATUS_TCS0;
+	int i, ret;
+
+	if (!d71->integrates_tbu)
+		return -1;
+
+	malidp_write32_mask(reg, BLK_CONTROL, 0x7, TBU_CONNECT_MODE);
+
+	ret = dp_wait_cond(has_bits(check_bits, malidp_read32(reg, BLK_STATUS)),
+			100, 1000, 1000);
+	if (ret < 0) {
+		DRM_ERROR("timed out connecting to TCU!\n");
+		malidp_write32_mask(reg, BLK_CONTROL, 0x7, INACTIVE_MODE);
+		return ret;
+	}
+
+	for (i = 0; i < d71->num_pipelines; i++)
+		malidp_write32_mask(d71->pipes[i]->lpu_addr, LPU_TBU_CONTROL,
+				    LPU_TBU_CTRL_TLBPEN, LPU_TBU_CTRL_TLBPEN);
+	return 0;
+}
+
+static int d71_disconnect_iommu(struct komeda_dev *mdev)
+{
+	struct d71_dev *d71 = mdev->chip_data;
+	u32 __iomem *reg = d71->gcu_addr;
+	u32 check_bits = (d71->num_pipelines == 2) ?
+			 GCU_STATUS_TCS0 | GCU_STATUS_TCS1 : GCU_STATUS_TCS0;
+	int ret;
+
+	malidp_write32_mask(reg, BLK_CONTROL, 0x7, TBU_DISCONNECT_MODE);
+
+	ret = dp_wait_cond(((malidp_read32(reg, BLK_STATUS) & check_bits) == 0),
+			100, 1000, 1000);
+	if (ret < 0) {
+		DRM_ERROR("timed out disconnecting from TCU!\n");
+		malidp_write32_mask(reg, BLK_CONTROL, 0x7, INACTIVE_MODE);
+	}
+
+	return ret;
+}
+
+static const struct komeda_dev_funcs d71_chip_funcs = {
+	.init_format_table	= d71_init_fmt_tbl,
+	.enum_resources		= d71_enum_resources,
+	.cleanup		= d71_cleanup,
+	.irq_handler		= d71_irq_handler,
+	.enable_irq		= d71_enable_irq,
+	.disable_irq		= d71_disable_irq,
+	.on_off_vblank		= d71_on_off_vblank,
+	.change_opmode		= d71_change_opmode,
+	.flush			= d71_flush,
+	.connect_iommu		= d71_connect_iommu,
+	.disconnect_iommu	= d71_disconnect_iommu,
+	.dump_register		= d71_dump,
+};
+
+const struct komeda_dev_funcs *
+d71_identify(u32 __iomem *reg_base, struct komeda_chip_info *chip)
+{
+	const struct komeda_dev_funcs *funcs;
+	u32 product_id;
+
+	chip->core_id = malidp_read32(reg_base, GLB_CORE_ID);
+
+	product_id = MALIDP_CORE_ID_PRODUCT_ID(chip->core_id);
+
+	switch (product_id) {
+	case MALIDP_D71_PRODUCT_ID:
+	case MALIDP_D32_PRODUCT_ID:
+		funcs = &d71_chip_funcs;
+		break;
+	default:
+		DRM_ERROR("Unsupported product: 0x%x\n", product_id);
+		return NULL;
+	}
+
+	chip->arch_id	= malidp_read32(reg_base, GLB_ARCH_ID);
+	chip->core_info	= malidp_read32(reg_base, GLB_CORE_INFO);
+	chip->bus_width	= D71_BUS_WIDTH_16_BYTES;
+
+	return funcs;
+}
diff --git a/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.h b/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.h
new file mode 100644
index 000000000..c7357c2b9
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/d71/d71_dev.h
@@ -0,0 +1,54 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#ifndef _D71_DEV_H_
+#define _D71_DEV_H_
+
+#include "komeda_dev.h"
+#include "komeda_pipeline.h"
+#include "d71_regs.h"
+
+struct d71_pipeline {
+	struct komeda_pipeline base;
+
+	/* d71 private pipeline blocks */
+	u32 __iomem	*lpu_addr;
+	u32 __iomem	*cu_addr;
+	u32 __iomem	*dou_addr;
+	u32 __iomem	*dou_ft_coeff_addr; /* forward transform coeffs table */
+};
+
+struct d71_dev {
+	struct komeda_dev *mdev;
+
+	int	num_blocks;
+	int	num_pipelines;
+	int	num_rich_layers;
+	u32	max_line_size;
+	u32	max_vsize;
+	u32	supports_dual_link : 1;
+	u32	integrates_tbu : 1;
+
+	/* global register blocks */
+	u32 __iomem	*gcu_addr;
+	/* scaling coeffs table */
+	u32 __iomem	*glb_scl_coeff_addr[D71_MAX_GLB_SCL_COEFF];
+	u32 __iomem	*periph_addr;
+
+	struct d71_pipeline *pipes[D71_MAX_PIPELINE];
+};
+
+#define to_d71_pipeline(x)	container_of(x, struct d71_pipeline, base)
+
+extern const struct komeda_pipeline_funcs d71_pipeline_funcs;
+
+int d71_probe_block(struct d71_dev *d71,
+		    struct block_header *blk, u32 __iomem *reg);
+void d71_read_block_header(u32 __iomem *reg, struct block_header *blk);
+
+void d71_dump(struct komeda_dev *mdev, struct seq_file *sf);
+
+#endif /* !_D71_DEV_H_ */
diff --git a/drivers/gpu/drm/arm/display/komeda/d71/d71_regs.h b/drivers/gpu/drm/arm/display/komeda/d71/d71_regs.h
new file mode 100644
index 000000000..e80172a0b
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/d71/d71_regs.h
@@ -0,0 +1,541 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#ifndef _D71_REG_H_
+#define _D71_REG_H_
+
+/* Common block registers offset */
+#define BLK_BLOCK_INFO		0x000
+#define BLK_PIPELINE_INFO	0x004
+#define BLK_MAX_LINE_SIZE	0x008
+#define BLK_VALID_INPUT_ID0	0x020
+#define BLK_OUTPUT_ID0		0x060
+#define BLK_INPUT_ID0		0x080
+#define BLK_IRQ_RAW_STATUS	0x0A0
+#define BLK_IRQ_CLEAR		0x0A4
+#define BLK_IRQ_MASK		0x0A8
+#define BLK_IRQ_STATUS		0x0AC
+#define BLK_STATUS		0x0B0
+#define BLK_INFO		0x0C0
+#define BLK_CONTROL		0x0D0
+#define BLK_SIZE		0x0D4
+#define BLK_IN_SIZE		0x0E0
+
+#define BLK_P0_PTR_LOW		0x100
+#define BLK_P0_PTR_HIGH		0x104
+#define BLK_P0_STRIDE		0x108
+#define BLK_P1_PTR_LOW		0x110
+#define BLK_P1_PTR_HIGH		0x114
+#define BLK_P1_STRIDE		0x118
+#define BLK_P2_PTR_LOW		0x120
+#define BLK_P2_PTR_HIGH		0x124
+
+#define BLOCK_INFO_N_SUBBLKS(x)	((x) & 0x000F)
+#define BLOCK_INFO_BLK_ID(x)	(((x) & 0x00F0) >> 4)
+#define BLOCK_INFO_BLK_TYPE(x)	(((x) & 0xFF00) >> 8)
+#define BLOCK_INFO_INPUT_ID(x)	((x) & 0xFFF0)
+#define BLOCK_INFO_TYPE_ID(x)	(((x) & 0x0FF0) >> 4)
+
+#define PIPELINE_INFO_N_OUTPUTS(x)	((x) & 0x000F)
+#define PIPELINE_INFO_N_VALID_INPUTS(x)	(((x) & 0x0F00) >> 8)
+
+/* Common block control register bits */
+#define BLK_CTRL_EN		BIT(0)
+/* Common size macro */
+#define HV_SIZE(h, v)		(((h) & 0x1FFF) + (((v) & 0x1FFF) << 16))
+#define HV_OFFSET(h, v)		(((h) & 0xFFF) + (((v) & 0xFFF) << 16))
+#define HV_CROP(h, v)		(((h) & 0xFFF) + (((v) & 0xFFF) << 16))
+
+/* AD_CONTROL register */
+#define AD_CONTROL		0x160
+
+/* AD_CONTROL register bits */
+#define AD_AEN			BIT(0)
+#define AD_YT			BIT(1)
+#define AD_BS			BIT(2)
+#define AD_WB			BIT(3)
+#define AD_TH			BIT(4)
+
+/* Global Control Unit */
+#define GLB_ARCH_ID		0x000
+#define GLB_CORE_ID		0x004
+#define GLB_CORE_INFO		0x008
+#define GLB_IRQ_STATUS		0x010
+
+#define GCU_CONFIG_VALID0	0x0D4
+#define GCU_CONFIG_VALID1	0x0D8
+
+/* GCU_CONTROL_BITS */
+#define GCU_CONTROL_MODE(x)	((x) & 0x7)
+#define GCU_CONTROL_SRST	BIT(16)
+
+/* GCU_CONFIGURATION registers */
+#define GCU_CONFIGURATION_ID0	0x100
+#define GCU_CONFIGURATION_ID1	0x104
+
+/* GCU configuration */
+#define GCU_MAX_LINE_SIZE(x)	((x) & 0xFFFF)
+#define GCU_MAX_NUM_LINES(x)	((x) >> 16)
+#define GCU_NUM_RICH_LAYERS(x)	((x) & 0x7)
+#define GCU_NUM_PIPELINES(x)	(((x) >> 3) & 0x7)
+#define GCU_NUM_SCALERS(x)	(((x) >> 6) & 0x7)
+#define GCU_DISPLAY_SPLIT_EN(x)	(((x) >> 16) & 0x1)
+#define GCU_DISPLAY_TBU_EN(x)	(((x) >> 17) & 0x1)
+
+/* GCU opmode */
+#define INACTIVE_MODE		0
+#define TBU_CONNECT_MODE	1
+#define TBU_DISCONNECT_MODE	2
+#define DO0_ACTIVE_MODE		3
+#define DO1_ACTIVE_MODE		4
+#define DO01_ACTIVE_MODE	5
+
+/* GLB_IRQ_STATUS bits */
+#define GLB_IRQ_STATUS_GCU	BIT(0)
+#define GLB_IRQ_STATUS_LPU0	BIT(8)
+#define GLB_IRQ_STATUS_LPU1	BIT(9)
+#define GLB_IRQ_STATUS_ATU0	BIT(10)
+#define GLB_IRQ_STATUS_ATU1	BIT(11)
+#define GLB_IRQ_STATUS_ATU2	BIT(12)
+#define GLB_IRQ_STATUS_ATU3	BIT(13)
+#define GLB_IRQ_STATUS_CU0	BIT(16)
+#define GLB_IRQ_STATUS_CU1	BIT(17)
+#define GLB_IRQ_STATUS_DOU0	BIT(24)
+#define GLB_IRQ_STATUS_DOU1	BIT(25)
+
+#define GLB_IRQ_STATUS_PIPE0	(GLB_IRQ_STATUS_LPU0 |\
+				 GLB_IRQ_STATUS_ATU0 |\
+				 GLB_IRQ_STATUS_ATU1 |\
+				 GLB_IRQ_STATUS_CU0 |\
+				 GLB_IRQ_STATUS_DOU0)
+
+#define GLB_IRQ_STATUS_PIPE1	(GLB_IRQ_STATUS_LPU1 |\
+				 GLB_IRQ_STATUS_ATU2 |\
+				 GLB_IRQ_STATUS_ATU3 |\
+				 GLB_IRQ_STATUS_CU1 |\
+				 GLB_IRQ_STATUS_DOU1)
+
+#define GLB_IRQ_STATUS_ATU	(GLB_IRQ_STATUS_ATU0 |\
+				 GLB_IRQ_STATUS_ATU1 |\
+				 GLB_IRQ_STATUS_ATU2 |\
+				 GLB_IRQ_STATUS_ATU3)
+
+/* GCU_IRQ_BITS */
+#define GCU_IRQ_CVAL0		BIT(0)
+#define GCU_IRQ_CVAL1		BIT(1)
+#define GCU_IRQ_MODE		BIT(4)
+#define GCU_IRQ_ERR		BIT(11)
+
+/* GCU_STATUS_BITS */
+#define GCU_STATUS_MODE(x)	((x) & 0x7)
+#define GCU_STATUS_MERR		BIT(4)
+#define GCU_STATUS_TCS0		BIT(8)
+#define GCU_STATUS_TCS1		BIT(9)
+#define GCU_STATUS_ACTIVE	BIT(31)
+
+/* GCU_CONFIG_VALIDx BITS */
+#define GCU_CONFIG_CVAL		BIT(0)
+
+/* PERIPHERAL registers */
+#define PERIPH_MAX_LINE_SIZE	BIT(0)
+#define PERIPH_NUM_RICH_LAYERS	BIT(4)
+#define PERIPH_SPLIT_EN		BIT(8)
+#define PERIPH_TBU_EN		BIT(12)
+#define PERIPH_AFBC_DMA_EN	BIT(16)
+#define PERIPH_CONFIGURATION_ID	0x1D4
+
+/* LPU register */
+#define LPU_TBU_STATUS		0x0B4
+#define LPU_RAXI_CONTROL	0x0D0
+#define LPU_WAXI_CONTROL	0x0D4
+#define LPU_TBU_CONTROL		0x0D8
+
+/* LPU_xAXI_CONTROL_BITS */
+#define TO_RAXI_AOUTSTDCAPB(x)	(x)
+#define TO_RAXI_BOUTSTDCAPB(x)	((x) << 8)
+#define TO_RAXI_BEN(x)		((x) << 15)
+#define TO_xAXI_BURSTLEN(x)	((x) << 16)
+#define TO_xAXI_AxQOS(x)	((x) << 24)
+#define TO_xAXI_ORD(x)		((x) << 31)
+#define TO_WAXI_OUTSTDCAPB(x)	(x)
+
+#define RAXI_AOUTSTDCAPB_MASK	0x7F
+#define RAXI_BOUTSTDCAPB_MASK	0x7F00
+#define RAXI_BEN_MASK		BIT(15)
+#define xAXI_BURSTLEN_MASK	0x3F0000
+#define xAXI_AxQOS_MASK		0xF000000
+#define xAXI_ORD_MASK		BIT(31)
+#define WAXI_OUTSTDCAPB_MASK	0x3F
+
+/* LPU_TBU_CONTROL BITS */
+#define TO_TBU_DOUTSTDCAPB(x)	(x)
+#define TBU_DOUTSTDCAPB_MASK	0x3F
+
+/* LPU_IRQ_BITS */
+#define LPU_IRQ_OVR		BIT(9)
+#define LPU_IRQ_IBSY		BIT(10)
+#define LPU_IRQ_ERR		BIT(11)
+#define LPU_IRQ_EOW		BIT(12)
+#define LPU_IRQ_PL0		BIT(13)
+
+/* LPU_STATUS_BITS */
+#define LPU_STATUS_AXIED(x)	((x) & 0xF)
+#define LPU_STATUS_AXIE		BIT(4)
+#define LPU_STATUS_AXIRP	BIT(5)
+#define LPU_STATUS_AXIWP	BIT(6)
+#define LPU_STATUS_FEMPTY	BIT(11)
+#define LPU_STATUS_FFULL	BIT(14)
+#define LPU_STATUS_ACE0		BIT(16)
+#define LPU_STATUS_ACE1		BIT(17)
+#define LPU_STATUS_ACE2		BIT(18)
+#define LPU_STATUS_ACE3		BIT(19)
+#define LPU_STATUS_ACTIVE	BIT(31)
+
+#define AXIEID_MASK		0xF
+#define AXIE_MASK		LPU_STATUS_AXIE
+#define AXIRP_MASK		LPU_STATUS_AXIRP
+#define AXIWP_MASK		LPU_STATUS_AXIWP
+
+#define FROM_AXIEID(reg)	((reg) & AXIEID_MASK)
+#define TO_AXIE(x)		((x) << 4)
+#define FROM_AXIRP(reg)		(((reg) & AXIRP_MASK) >> 5)
+#define FROM_AXIWP(reg)		(((reg) & AXIWP_MASK) >> 6)
+
+/* LPU_TBU_STATUS_BITS */
+#define LPU_TBU_STATUS_TCF	BIT(1)
+#define LPU_TBU_STATUS_TTNG	BIT(2)
+#define LPU_TBU_STATUS_TITR	BIT(8)
+#define LPU_TBU_STATUS_TEMR	BIT(16)
+#define LPU_TBU_STATUS_TTF	BIT(31)
+
+/* LPU_TBU_CONTROL BITS */
+#define LPU_TBU_CTRL_TLBPEN	BIT(16)
+
+/* CROSSBAR CONTROL BITS */
+#define CBU_INPUT_CTRL_EN	BIT(0)
+#define CBU_NUM_INPUT_IDS	5
+#define CBU_NUM_OUTPUT_IDS	5
+
+/* CU register */
+#define CU_BG_COLOR		0x0DC
+#define CU_INPUT0_SIZE		0x0E0
+#define CU_INPUT0_OFFSET	0x0E4
+#define CU_INPUT0_CONTROL	0x0E8
+#define CU_INPUT1_SIZE		0x0F0
+#define CU_INPUT1_OFFSET	0x0F4
+#define CU_INPUT1_CONTROL	0x0F8
+#define CU_INPUT2_SIZE		0x100
+#define CU_INPUT2_OFFSET	0x104
+#define CU_INPUT2_CONTROL	0x108
+#define CU_INPUT3_SIZE		0x110
+#define CU_INPUT3_OFFSET	0x114
+#define CU_INPUT3_CONTROL	0x118
+#define CU_INPUT4_SIZE		0x120
+#define CU_INPUT4_OFFSET	0x124
+#define CU_INPUT4_CONTROL	0x128
+
+#define CU_PER_INPUT_REGS	4
+
+#define CU_NUM_INPUT_IDS	5
+#define CU_NUM_OUTPUT_IDS	1
+
+/* CU control register bits */
+#define CU_CTRL_COPROC		BIT(0)
+
+/* CU_IRQ_BITS */
+#define CU_IRQ_OVR		BIT(9)
+#define CU_IRQ_ERR		BIT(11)
+
+/* CU_STATUS_BITS */
+#define CU_STATUS_CPE		BIT(0)
+#define CU_STATUS_ZME		BIT(1)
+#define CU_STATUS_CFGE		BIT(2)
+#define CU_STATUS_ACTIVE	BIT(31)
+
+/* CU input control register bits */
+#define CU_INPUT_CTRL_EN	BIT(0)
+#define CU_INPUT_CTRL_PAD	BIT(1)
+#define CU_INPUT_CTRL_PMUL	BIT(2)
+#define CU_INPUT_CTRL_ALPHA(x)	(((x) & 0xFF) << 8)
+
+/* DOU register */
+
+/* DOU_IRQ_BITS */
+#define DOU_IRQ_UND		BIT(8)
+#define DOU_IRQ_ERR		BIT(11)
+#define DOU_IRQ_PL0		BIT(13)
+#define DOU_IRQ_PL1		BIT(14)
+
+/* DOU_STATUS_BITS */
+#define DOU_STATUS_DRIFTTO	BIT(0)
+#define DOU_STATUS_FRAMETO	BIT(1)
+#define DOU_STATUS_TETO		BIT(2)
+#define DOU_STATUS_CSCE		BIT(8)
+#define DOU_STATUS_ACTIVE	BIT(31)
+
+/* Layer registers */
+#define LAYER_INFO		0x0C0
+#define LAYER_R_CONTROL		0x0D4
+#define LAYER_FMT		0x0D8
+#define LAYER_LT_COEFFTAB	0x0DC
+#define LAYER_PALPHA		0x0E4
+
+#define LAYER_YUV_RGB_COEFF0	0x130
+
+#define LAYER_AD_H_CROP		0x164
+#define LAYER_AD_V_CROP		0x168
+
+#define LAYER_RGB_RGB_COEFF0	0x170
+
+/* L_CONTROL_BITS */
+#define L_EN			BIT(0)
+#define L_IT			BIT(4)
+#define L_R2R			BIT(5)
+#define L_FT			BIT(6)
+#define L_ROT(x)		(((x) & 3) << 8)
+#define L_HFLIP			BIT(10)
+#define L_VFLIP			BIT(11)
+#define L_TBU_EN		BIT(16)
+#define L_A_RCACHE(x)		(((x) & 0xF) << 28)
+#define L_ROT_R0		0
+#define L_ROT_R90		1
+#define L_ROT_R180		2
+#define L_ROT_R270		3
+
+/* LAYER_R_CONTROL BITS */
+#define LR_CHI422_BILINEAR	0
+#define LR_CHI422_REPLICATION	1
+#define LR_CHI420_JPEG		(0 << 2)
+#define LR_CHI420_MPEG		(1 << 2)
+
+#define L_ITSEL(x)		((x) & 0xFFF)
+#define L_FTSEL(x)		(((x) & 0xFFF) << 16)
+
+#define LAYER_PER_PLANE_REGS	4
+
+/* Layer_WR registers */
+#define LAYER_WR_PROG_LINE	0x0D4
+#define LAYER_WR_FORMAT		0x0D8
+
+/* Layer_WR control bits */
+#define LW_OFM			BIT(4)
+#define LW_LALPHA(x)		(((x) & 0xFF) << 8)
+#define LW_A_WCACHE(x)		(((x) & 0xF) << 28)
+#define LW_TBU_EN		BIT(16)
+
+#define AxCACHE_MASK		0xF0000000
+
+/* Layer AXI R/W cache setting */
+#define AxCACHE_B		BIT(0)	/* Bufferable */
+#define AxCACHE_M		BIT(1)	/* Modifiable */
+#define AxCACHE_RA		BIT(2)	/* Read-Allocate */
+#define AxCACHE_WA		BIT(3)	/* Write-Allocate */
+
+/* Layer info bits */
+#define L_INFO_RF		BIT(0)
+#define L_INFO_CM		BIT(1)
+#define L_INFO_ABUF_SIZE(x)	(((x) >> 4) & 0x7)
+#define L_INFO_YUV_MAX_LINESZ(x)	(((x) >> 16) & 0xFFFF)
+
+/* Scaler registers */
+#define SC_COEFFTAB		0x0DC
+#define SC_OUT_SIZE		0x0E4
+#define SC_H_CROP		0x0E8
+#define SC_V_CROP		0x0EC
+#define SC_H_INIT_PH		0x0F0
+#define SC_H_DELTA_PH		0x0F4
+#define SC_V_INIT_PH		0x0F8
+#define SC_V_DELTA_PH		0x0FC
+#define SC_ENH_LIMITS		0x130
+#define SC_ENH_COEFF0		0x134
+
+#define SC_MAX_ENH_COEFF	9
+
+/* SC_CTRL_BITS */
+#define SC_CTRL_SCL		BIT(0)
+#define SC_CTRL_LS		BIT(1)
+#define SC_CTRL_AP		BIT(4)
+#define SC_CTRL_IENH		BIT(8)
+#define SC_CTRL_RGBSM		BIT(16)
+#define SC_CTRL_ASM		BIT(17)
+
+#define SC_VTSEL(vtal)		((vtal) << 16)
+
+#define SC_NUM_INPUTS_IDS	1
+#define SC_NUM_OUTPUTS_IDS	1
+
+#define MG_NUM_INPUTS_IDS	2
+#define MG_NUM_OUTPUTS_IDS	1
+
+/* Merger registers */
+#define MG_INPUT_ID0		BLK_INPUT_ID0
+#define MG_INPUT_ID1		(MG_INPUT_ID0 + 4)
+#define MG_SIZE			BLK_SIZE
+
+/* Splitter registers */
+#define SP_OVERLAP_SIZE		0xD8
+
+/* Backend registers */
+#define BS_INFO			0x0C0
+#define BS_PROG_LINE		0x0D4
+#define BS_PREFETCH_LINE	0x0D8
+#define BS_BG_COLOR		0x0DC
+#define BS_ACTIVESIZE		0x0E0
+#define BS_HINTERVALS		0x0E4
+#define BS_VINTERVALS		0x0E8
+#define BS_SYNC			0x0EC
+#define BS_DRIFT_TO		0x100
+#define BS_FRAME_TO		0x104
+#define BS_TE_TO		0x108
+#define BS_T0_INTERVAL		0x110
+#define BS_T1_INTERVAL		0x114
+#define BS_T2_INTERVAL		0x118
+#define BS_CRC0_LOW		0x120
+#define BS_CRC0_HIGH		0x124
+#define BS_CRC1_LOW		0x128
+#define BS_CRC1_HIGH		0x12C
+#define BS_USER			0x130
+
+/* BS control register bits */
+#define BS_CTRL_EN		BIT(0)
+#define BS_CTRL_VM		BIT(1)
+#define BS_CTRL_BM		BIT(2)
+#define BS_CTRL_HMASK		BIT(4)
+#define BS_CTRL_VD		BIT(5)
+#define BS_CTRL_TE		BIT(8)
+#define BS_CTRL_TS		BIT(9)
+#define BS_CTRL_TM		BIT(12)
+#define BS_CTRL_DL		BIT(16)
+#define BS_CTRL_SBS		BIT(17)
+#define BS_CTRL_CRC		BIT(18)
+#define BS_CTRL_PM		BIT(20)
+
+/* BS active size/intervals */
+#define BS_H_INTVALS(hfp, hbp)	(((hfp) & 0xFFF) + (((hbp) & 0x3FF) << 16))
+#define BS_V_INTVALS(vfp, vbp)  (((vfp) & 0x3FFF) + (((vbp) & 0xFF) << 16))
+
+/* BS_SYNC bits */
+#define BS_SYNC_HSW(x)		((x) & 0x3FF)
+#define BS_SYNC_HSP		BIT(12)
+#define BS_SYNC_VSW(x)		(((x) & 0xFF) << 16)
+#define BS_SYNC_VSP		BIT(28)
+
+#define BS_NUM_INPUT_IDS	0
+#define BS_NUM_OUTPUT_IDS	0
+
+/* Image process registers */
+#define IPS_DEPTH		0x0D8
+#define IPS_RGB_RGB_COEFF0	0x130
+#define IPS_RGB_YUV_COEFF0	0x170
+
+#define IPS_DEPTH_MARK		0xF
+
+/* IPS control register bits */
+#define IPS_CTRL_RGB		BIT(0)
+#define IPS_CTRL_FT		BIT(4)
+#define IPS_CTRL_YUV		BIT(8)
+#define IPS_CTRL_CHD422		BIT(9)
+#define IPS_CTRL_CHD420		BIT(10)
+#define IPS_CTRL_LPF		BIT(11)
+#define IPS_CTRL_DITH		BIT(12)
+#define IPS_CTRL_CLAMP		BIT(16)
+#define IPS_CTRL_SBS		BIT(17)
+
+/* IPS info register bits */
+#define IPS_INFO_CHD420		BIT(10)
+
+#define IPS_NUM_INPUT_IDS	2
+#define IPS_NUM_OUTPUT_IDS	1
+
+/* FT_COEFF block registers */
+#define FT_COEFF0		0x80
+#define GLB_IT_COEFF		0x80
+
+/* GLB_SC_COEFF registers */
+#define GLB_SC_COEFF_ADDR	0x0080
+#define GLB_SC_COEFF_DATA	0x0084
+#define GLB_LT_COEFF_DATA	0x0080
+
+#define GLB_SC_COEFF_MAX_NUM	1024
+#define GLB_LT_COEFF_NUM	65
+/* GLB_SC_ADDR */
+#define SC_COEFF_R_ADDR		BIT(18)
+#define SC_COEFF_G_ADDR		BIT(17)
+#define SC_COEFF_B_ADDR		BIT(16)
+
+#define SC_COEFF_DATA(x, y)	(((y) & 0xFFFF) | (((x) & 0xFFFF) << 16))
+
+enum d71_blk_type {
+	D71_BLK_TYPE_GCU		= 0x00,
+	D71_BLK_TYPE_LPU		= 0x01,
+	D71_BLK_TYPE_CU			= 0x02,
+	D71_BLK_TYPE_DOU		= 0x03,
+	D71_BLK_TYPE_AEU		= 0x04,
+	D71_BLK_TYPE_GLB_LT_COEFF	= 0x05,
+	D71_BLK_TYPE_GLB_SCL_COEFF	= 0x06, /* SH/SV scaler coeff */
+	D71_BLK_TYPE_GLB_SC_COEFF	= 0x07,
+	D71_BLK_TYPE_PERIPH		= 0x08,
+	D71_BLK_TYPE_LPU_TRUSTED	= 0x09,
+	D71_BLK_TYPE_AEU_TRUSTED	= 0x0A,
+	D71_BLK_TYPE_LPU_LAYER		= 0x10,
+	D71_BLK_TYPE_LPU_WB_LAYER	= 0x11,
+	D71_BLK_TYPE_CU_SPLITTER	= 0x20,
+	D71_BLK_TYPE_CU_SCALER		= 0x21,
+	D71_BLK_TYPE_CU_MERGER		= 0x22,
+	D71_BLK_TYPE_DOU_IPS		= 0x30,
+	D71_BLK_TYPE_DOU_BS		= 0x31,
+	D71_BLK_TYPE_DOU_FT_COEFF	= 0x32,
+	D71_BLK_TYPE_AEU_DS		= 0x40,
+	D71_BLK_TYPE_AEU_AES		= 0x41,
+	D71_BLK_TYPE_RESERVED		= 0xFF
+};
+
+/* Constant of components */
+#define D71_MAX_PIPELINE		2
+#define D71_PIPELINE_MAX_SCALERS	2
+#define D71_PIPELINE_MAX_LAYERS		4
+
+#define D71_MAX_GLB_IT_COEFF		3
+#define D71_MAX_GLB_SCL_COEFF		4
+
+#define D71_MAX_LAYERS_PER_LPU		4
+#define D71_BLOCK_MAX_INPUT		9
+#define D71_BLOCK_MAX_OUTPUT		5
+#define D71_MAX_SC_PER_CU		2
+
+#define D71_BLOCK_OFFSET_PERIPH		0xFE00
+#define D71_BLOCK_SIZE			0x0200
+
+#define D71_DEFAULT_PREPRETCH_LINE	5
+#define D71_BUS_WIDTH_16_BYTES		16
+
+#define D71_SC_MAX_UPSCALING		64
+#define D71_SC_MAX_DOWNSCALING		6
+#define D71_SC_SPLIT_OVERLAP		8
+#define D71_SC_ENH_SPLIT_OVERLAP	1
+
+#define D71_MG_MIN_MERGED_SIZE		4
+#define D71_MG_MAX_MERGED_HSIZE		4032
+#define D71_MG_MAX_MERGED_VSIZE		4096
+
+#define D71_PALPHA_DEF_MAP		0xFFAA5500
+#define D71_LAYER_CONTROL_DEFAULT	0x30000000
+#define D71_WB_LAYER_CONTROL_DEFAULT	0x3000FF00
+#define D71_BS_CONTROL_DEFAULT		0x00000002
+
+struct block_header {
+	u32 block_info;
+	u32 pipeline_info;
+	u32 input_ids[D71_BLOCK_MAX_INPUT];
+	u32 output_ids[D71_BLOCK_MAX_OUTPUT];
+};
+
+static inline u32 get_block_type(struct block_header *blk)
+{
+	return BLOCK_INFO_BLK_TYPE(blk->block_info);
+}
+
+#endif /* !_D71_REG_H_ */
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_color_mgmt.c b/drivers/gpu/drm/arm/display/komeda/komeda_color_mgmt.c
new file mode 100644
index 000000000..d8e449e6e
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_color_mgmt.c
@@ -0,0 +1,133 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2019 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+
+#include "komeda_color_mgmt.h"
+
+/* 10bit precision YUV2RGB matrix */
+static const s32 yuv2rgb_bt601_narrow[KOMEDA_N_YUV2RGB_COEFFS] = {
+	1192,    0, 1634,
+	1192, -401, -832,
+	1192, 2066,    0,
+	  64,  512,  512
+};
+
+static const s32 yuv2rgb_bt601_wide[KOMEDA_N_YUV2RGB_COEFFS] = {
+	1024,    0, 1436,
+	1024, -352, -731,
+	1024, 1815,    0,
+	   0,  512,  512
+};
+
+static const s32 yuv2rgb_bt709_narrow[KOMEDA_N_YUV2RGB_COEFFS] = {
+	1192,    0, 1836,
+	1192, -218, -546,
+	1192, 2163,    0,
+	  64,  512,  512
+};
+
+static const s32 yuv2rgb_bt709_wide[KOMEDA_N_YUV2RGB_COEFFS] = {
+	1024,    0, 1613,
+	1024, -192, -479,
+	1024, 1900,    0,
+	   0,  512,  512
+};
+
+static const s32 yuv2rgb_bt2020[KOMEDA_N_YUV2RGB_COEFFS] = {
+	1024,    0, 1476,
+	1024, -165, -572,
+	1024, 1884,    0,
+	   0,  512,  512
+};
+
+const s32 *komeda_select_yuv2rgb_coeffs(u32 color_encoding, u32 color_range)
+{
+	bool narrow = color_range == DRM_COLOR_YCBCR_LIMITED_RANGE;
+	const s32 *coeffs;
+
+	switch (color_encoding) {
+	case DRM_COLOR_YCBCR_BT709:
+		coeffs = narrow ? yuv2rgb_bt709_narrow : yuv2rgb_bt709_wide;
+		break;
+	case DRM_COLOR_YCBCR_BT601:
+		coeffs = narrow ? yuv2rgb_bt601_narrow : yuv2rgb_bt601_wide;
+		break;
+	case DRM_COLOR_YCBCR_BT2020:
+		coeffs = yuv2rgb_bt2020;
+		break;
+	default:
+		coeffs = NULL;
+		break;
+	}
+
+	return coeffs;
+}
+
+struct gamma_curve_sector {
+	u32 boundary_start;
+	u32 num_of_segments;
+	u32 segment_width;
+};
+
+struct gamma_curve_segment {
+	u32 start;
+	u32 end;
+};
+
+static struct gamma_curve_sector sector_tbl[] = {
+	{ 0,    4,  4   },
+	{ 16,   4,  4   },
+	{ 32,   4,  8   },
+	{ 64,   4,  16  },
+	{ 128,  4,  32  },
+	{ 256,  4,  64  },
+	{ 512,  16, 32  },
+	{ 1024, 24, 128 },
+};
+
+static void
+drm_lut_to_coeffs(struct drm_property_blob *lut_blob, u32 *coeffs,
+		  struct gamma_curve_sector *sector_tbl, u32 num_sectors)
+{
+	struct drm_color_lut *lut;
+	u32 i, j, in, num = 0;
+
+	if (!lut_blob)
+		return;
+
+	lut = lut_blob->data;
+
+	for (i = 0; i < num_sectors; i++) {
+		for (j = 0; j < sector_tbl[i].num_of_segments; j++) {
+			in = sector_tbl[i].boundary_start +
+			     j * sector_tbl[i].segment_width;
+
+			coeffs[num++] = drm_color_lut_extract(lut[in].red,
+						KOMEDA_COLOR_PRECISION);
+		}
+	}
+
+	coeffs[num] = BIT(KOMEDA_COLOR_PRECISION);
+}
+
+void drm_lut_to_fgamma_coeffs(struct drm_property_blob *lut_blob, u32 *coeffs)
+{
+	drm_lut_to_coeffs(lut_blob, coeffs, sector_tbl, ARRAY_SIZE(sector_tbl));
+}
+
+void drm_ctm_to_coeffs(struct drm_property_blob *ctm_blob, u32 *coeffs)
+{
+	struct drm_color_ctm *ctm;
+	u32 i;
+
+	if (!ctm_blob)
+		return;
+
+	ctm = ctm_blob->data;
+
+	for (i = 0; i < KOMEDA_N_CTM_COEFFS; i++)
+		coeffs[i] = drm_color_ctm_s31_32_to_qm_n(ctm->matrix[i], 3, 12);
+}
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_color_mgmt.h b/drivers/gpu/drm/arm/display/komeda/komeda_color_mgmt.h
new file mode 100644
index 000000000..2f4668466
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_color_mgmt.h
@@ -0,0 +1,25 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * (C) COPYRIGHT 2019 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+
+#ifndef _KOMEDA_COLOR_MGMT_H_
+#define _KOMEDA_COLOR_MGMT_H_
+
+#include <drm/drm_color_mgmt.h>
+
+#define KOMEDA_N_YUV2RGB_COEFFS		12
+#define KOMEDA_N_RGB2YUV_COEFFS		12
+#define KOMEDA_COLOR_PRECISION		12
+#define KOMEDA_N_GAMMA_COEFFS		65
+#define KOMEDA_COLOR_LUT_SIZE		BIT(KOMEDA_COLOR_PRECISION)
+#define KOMEDA_N_CTM_COEFFS		9
+
+void drm_lut_to_fgamma_coeffs(struct drm_property_blob *lut_blob, u32 *coeffs);
+void drm_ctm_to_coeffs(struct drm_property_blob *ctm_blob, u32 *coeffs);
+
+const s32 *komeda_select_yuv2rgb_coeffs(u32 color_encoding, u32 color_range);
+
+#endif /*_KOMEDA_COLOR_MGMT_H_*/
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_crtc.c b/drivers/gpu/drm/arm/display/komeda/komeda_crtc.c
new file mode 100644
index 000000000..f33418d6e
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_crtc.c
@@ -0,0 +1,636 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#include <linux/clk.h>
+#include <linux/pm_runtime.h>
+#include <linux/spinlock.h>
+
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_plane_helper.h>
+#include <drm/drm_print.h>
+#include <drm/drm_vblank.h>
+
+#include "komeda_dev.h"
+#include "komeda_kms.h"
+
+void komeda_crtc_get_color_config(struct drm_crtc_state *crtc_st,
+				  u32 *color_depths, u32 *color_formats)
+{
+	struct drm_connector *conn;
+	struct drm_connector_state *conn_st;
+	u32 conn_color_formats = ~0u;
+	int i, min_bpc = 31, conn_bpc = 0;
+
+	for_each_new_connector_in_state(crtc_st->state, conn, conn_st, i) {
+		if (conn_st->crtc != crtc_st->crtc)
+			continue;
+
+		conn_bpc = conn->display_info.bpc ? conn->display_info.bpc : 8;
+		conn_color_formats &= conn->display_info.color_formats;
+
+		if (conn_bpc < min_bpc)
+			min_bpc = conn_bpc;
+	}
+
+	/* connector doesn't config any color_format, use RGB444 as default */
+	if (!conn_color_formats)
+		conn_color_formats = DRM_COLOR_FORMAT_RGB444;
+
+	*color_depths = GENMASK(min_bpc, 0);
+	*color_formats = conn_color_formats;
+}
+
+static void komeda_crtc_update_clock_ratio(struct komeda_crtc_state *kcrtc_st)
+{
+	u64 pxlclk, aclk;
+
+	if (!kcrtc_st->base.active) {
+		kcrtc_st->clock_ratio = 0;
+		return;
+	}
+
+	pxlclk = kcrtc_st->base.adjusted_mode.crtc_clock * 1000ULL;
+	aclk = komeda_crtc_get_aclk(kcrtc_st);
+
+	kcrtc_st->clock_ratio = div64_u64(aclk << 32, pxlclk);
+}
+
+/**
+ * komeda_crtc_atomic_check - build display output data flow
+ * @crtc: DRM crtc
+ * @state: the crtc state object
+ *
+ * crtc_atomic_check is the final check stage, so beside build a display data
+ * pipeline according to the crtc_state, but still needs to release or disable
+ * the unclaimed pipeline resources.
+ *
+ * RETURNS:
+ * Zero for success or -errno
+ */
+static int
+komeda_crtc_atomic_check(struct drm_crtc *crtc,
+			 struct drm_crtc_state *state)
+{
+	struct komeda_crtc *kcrtc = to_kcrtc(crtc);
+	struct komeda_crtc_state *kcrtc_st = to_kcrtc_st(state);
+	int err;
+
+	if (drm_atomic_crtc_needs_modeset(state))
+		komeda_crtc_update_clock_ratio(kcrtc_st);
+
+	if (state->active) {
+		err = komeda_build_display_data_flow(kcrtc, kcrtc_st);
+		if (err)
+			return err;
+	}
+
+	/* release unclaimed pipeline resources */
+	err = komeda_release_unclaimed_resources(kcrtc->slave, kcrtc_st);
+	if (err)
+		return err;
+
+	err = komeda_release_unclaimed_resources(kcrtc->master, kcrtc_st);
+	if (err)
+		return err;
+
+	return 0;
+}
+
+/* For active a crtc, mainly need two parts of preparation
+ * 1. adjust display operation mode.
+ * 2. enable needed clk
+ */
+static int
+komeda_crtc_prepare(struct komeda_crtc *kcrtc)
+{
+	struct komeda_dev *mdev = kcrtc->base.dev->dev_private;
+	struct komeda_pipeline *master = kcrtc->master;
+	struct komeda_crtc_state *kcrtc_st = to_kcrtc_st(kcrtc->base.state);
+	struct drm_display_mode *mode = &kcrtc_st->base.adjusted_mode;
+	u32 new_mode;
+	int err;
+
+	mutex_lock(&mdev->lock);
+
+	new_mode = mdev->dpmode | BIT(master->id);
+	if (WARN_ON(new_mode == mdev->dpmode)) {
+		err = 0;
+		goto unlock;
+	}
+
+	err = mdev->funcs->change_opmode(mdev, new_mode);
+	if (err) {
+		DRM_ERROR("failed to change opmode: 0x%x -> 0x%x.\n,",
+			  mdev->dpmode, new_mode);
+		goto unlock;
+	}
+
+	mdev->dpmode = new_mode;
+	/* Only need to enable aclk on single display mode, but no need to
+	 * enable aclk it on dual display mode, since the dual mode always
+	 * switch from single display mode, the aclk already enabled, no need
+	 * to enable it again.
+	 */
+	if (new_mode != KOMEDA_MODE_DUAL_DISP) {
+		err = clk_set_rate(mdev->aclk, komeda_crtc_get_aclk(kcrtc_st));
+		if (err)
+			DRM_ERROR("failed to set aclk.\n");
+		err = clk_prepare_enable(mdev->aclk);
+		if (err)
+			DRM_ERROR("failed to enable aclk.\n");
+	}
+
+	err = clk_set_rate(master->pxlclk, mode->crtc_clock * 1000);
+	if (err)
+		DRM_ERROR("failed to set pxlclk for pipe%d\n", master->id);
+	err = clk_prepare_enable(master->pxlclk);
+	if (err)
+		DRM_ERROR("failed to enable pxl clk for pipe%d.\n", master->id);
+
+unlock:
+	mutex_unlock(&mdev->lock);
+
+	return err;
+}
+
+static int
+komeda_crtc_unprepare(struct komeda_crtc *kcrtc)
+{
+	struct komeda_dev *mdev = kcrtc->base.dev->dev_private;
+	struct komeda_pipeline *master = kcrtc->master;
+	u32 new_mode;
+	int err;
+
+	mutex_lock(&mdev->lock);
+
+	new_mode = mdev->dpmode & (~BIT(master->id));
+
+	if (WARN_ON(new_mode == mdev->dpmode)) {
+		err = 0;
+		goto unlock;
+	}
+
+	err = mdev->funcs->change_opmode(mdev, new_mode);
+	if (err) {
+		DRM_ERROR("failed to change opmode: 0x%x -> 0x%x.\n,",
+			  mdev->dpmode, new_mode);
+		goto unlock;
+	}
+
+	mdev->dpmode = new_mode;
+
+	clk_disable_unprepare(master->pxlclk);
+	if (new_mode == KOMEDA_MODE_INACTIVE)
+		clk_disable_unprepare(mdev->aclk);
+
+unlock:
+	mutex_unlock(&mdev->lock);
+
+	return err;
+}
+
+void komeda_crtc_handle_event(struct komeda_crtc   *kcrtc,
+			      struct komeda_events *evts)
+{
+	struct drm_crtc *crtc = &kcrtc->base;
+	u32 events = evts->pipes[kcrtc->master->id];
+
+	if (events & KOMEDA_EVENT_VSYNC)
+		drm_crtc_handle_vblank(crtc);
+
+	if (events & KOMEDA_EVENT_EOW) {
+		struct komeda_wb_connector *wb_conn = kcrtc->wb_conn;
+
+		if (wb_conn)
+			drm_writeback_signal_completion(&wb_conn->base, 0);
+		else
+			DRM_WARN("CRTC[%d]: EOW happen but no wb_connector.\n",
+				 drm_crtc_index(&kcrtc->base));
+	}
+	/* will handle it together with the write back support */
+	if (events & KOMEDA_EVENT_EOW)
+		DRM_DEBUG("EOW.\n");
+
+	if (events & KOMEDA_EVENT_FLIP) {
+		unsigned long flags;
+		struct drm_pending_vblank_event *event;
+
+		spin_lock_irqsave(&crtc->dev->event_lock, flags);
+		if (kcrtc->disable_done) {
+			complete_all(kcrtc->disable_done);
+			kcrtc->disable_done = NULL;
+		} else if (crtc->state->event) {
+			event = crtc->state->event;
+			/*
+			 * Consume event before notifying drm core that flip
+			 * happened.
+			 */
+			crtc->state->event = NULL;
+			drm_crtc_send_vblank_event(crtc, event);
+		} else {
+			DRM_WARN("CRTC[%d]: FLIP happen but no pending commit.\n",
+				 drm_crtc_index(&kcrtc->base));
+		}
+		spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
+	}
+}
+
+static void
+komeda_crtc_do_flush(struct drm_crtc *crtc,
+		     struct drm_crtc_state *old)
+{
+	struct komeda_crtc *kcrtc = to_kcrtc(crtc);
+	struct komeda_crtc_state *kcrtc_st = to_kcrtc_st(crtc->state);
+	struct komeda_dev *mdev = kcrtc->base.dev->dev_private;
+	struct komeda_pipeline *master = kcrtc->master;
+	struct komeda_pipeline *slave = kcrtc->slave;
+	struct komeda_wb_connector *wb_conn = kcrtc->wb_conn;
+	struct drm_connector_state *conn_st;
+
+	DRM_DEBUG_ATOMIC("CRTC%d_FLUSH: active_pipes: 0x%x, affected: 0x%x.\n",
+			 drm_crtc_index(crtc),
+			 kcrtc_st->active_pipes, kcrtc_st->affected_pipes);
+
+	/* step 1: update the pipeline/component state to HW */
+	if (has_bit(master->id, kcrtc_st->affected_pipes))
+		komeda_pipeline_update(master, old->state);
+
+	if (slave && has_bit(slave->id, kcrtc_st->affected_pipes))
+		komeda_pipeline_update(slave, old->state);
+
+	conn_st = wb_conn ? wb_conn->base.base.state : NULL;
+	if (conn_st && conn_st->writeback_job)
+		drm_writeback_queue_job(&wb_conn->base, conn_st);
+
+	/* step 2: notify the HW to kickoff the update */
+	mdev->funcs->flush(mdev, master->id, kcrtc_st->active_pipes);
+}
+
+static void
+komeda_crtc_atomic_enable(struct drm_crtc *crtc,
+			  struct drm_crtc_state *old)
+{
+	pm_runtime_get_sync(crtc->dev->dev);
+	komeda_crtc_prepare(to_kcrtc(crtc));
+	drm_crtc_vblank_on(crtc);
+	WARN_ON(drm_crtc_vblank_get(crtc));
+	komeda_crtc_do_flush(crtc, old);
+}
+
+static void
+komeda_crtc_flush_and_wait_for_flip_done(struct komeda_crtc *kcrtc,
+					 struct completion *input_flip_done)
+{
+	struct drm_device *drm = kcrtc->base.dev;
+	struct komeda_dev *mdev = kcrtc->master->mdev;
+	struct completion *flip_done;
+	struct completion temp;
+	int timeout;
+
+	/* if caller doesn't send a flip_done, use a private flip_done */
+	if (input_flip_done) {
+		flip_done = input_flip_done;
+	} else {
+		init_completion(&temp);
+		kcrtc->disable_done = &temp;
+		flip_done = &temp;
+	}
+
+	mdev->funcs->flush(mdev, kcrtc->master->id, 0);
+
+	/* wait the flip take affect.*/
+	timeout = wait_for_completion_timeout(flip_done, HZ);
+	if (timeout == 0) {
+		DRM_ERROR("wait pipe%d flip done timeout\n", kcrtc->master->id);
+		if (!input_flip_done) {
+			unsigned long flags;
+
+			spin_lock_irqsave(&drm->event_lock, flags);
+			kcrtc->disable_done = NULL;
+			spin_unlock_irqrestore(&drm->event_lock, flags);
+		}
+	}
+}
+
+static void
+komeda_crtc_atomic_disable(struct drm_crtc *crtc,
+			   struct drm_crtc_state *old)
+{
+	struct komeda_crtc *kcrtc = to_kcrtc(crtc);
+	struct komeda_crtc_state *old_st = to_kcrtc_st(old);
+	struct komeda_pipeline *master = kcrtc->master;
+	struct komeda_pipeline *slave  = kcrtc->slave;
+	struct completion *disable_done;
+	bool needs_phase2 = false;
+
+	DRM_DEBUG_ATOMIC("CRTC%d_DISABLE: active_pipes: 0x%x, affected: 0x%x\n",
+			 drm_crtc_index(crtc),
+			 old_st->active_pipes, old_st->affected_pipes);
+
+	if (slave && has_bit(slave->id, old_st->active_pipes))
+		komeda_pipeline_disable(slave, old->state);
+
+	if (has_bit(master->id, old_st->active_pipes))
+		needs_phase2 = komeda_pipeline_disable(master, old->state);
+
+	/* crtc_disable has two scenarios according to the state->active switch.
+	 * 1. active -> inactive
+	 *    this commit is a disable commit. and the commit will be finished
+	 *    or done after the disable operation. on this case we can directly
+	 *    use the crtc->state->event to tracking the HW disable operation.
+	 * 2. active -> active
+	 *    the crtc->commit is not for disable, but a modeset operation when
+	 *    crtc is active, such commit actually has been completed by 3
+	 *    DRM operations:
+	 *    crtc_disable, update_planes(crtc_flush), crtc_enable
+	 *    so on this case the crtc->commit is for the whole process.
+	 *    we can not use it for tracing the disable, we need a temporary
+	 *    flip_done for tracing the disable. and crtc->state->event for
+	 *    the crtc_enable operation.
+	 *    That's also the reason why skip modeset commit in
+	 *    komeda_crtc_atomic_flush()
+	 */
+	disable_done = (needs_phase2 || crtc->state->active) ?
+		       NULL : &crtc->state->commit->flip_done;
+
+	/* wait phase 1 disable done */
+	komeda_crtc_flush_and_wait_for_flip_done(kcrtc, disable_done);
+
+	/* phase 2 */
+	if (needs_phase2) {
+		komeda_pipeline_disable(kcrtc->master, old->state);
+
+		disable_done = crtc->state->active ?
+			       NULL : &crtc->state->commit->flip_done;
+
+		komeda_crtc_flush_and_wait_for_flip_done(kcrtc, disable_done);
+	}
+
+	drm_crtc_vblank_put(crtc);
+	drm_crtc_vblank_off(crtc);
+	komeda_crtc_unprepare(kcrtc);
+	pm_runtime_put(crtc->dev->dev);
+}
+
+static void
+komeda_crtc_atomic_flush(struct drm_crtc *crtc,
+			 struct drm_crtc_state *old)
+{
+	/* commit with modeset will be handled in enable/disable */
+	if (drm_atomic_crtc_needs_modeset(crtc->state))
+		return;
+
+	komeda_crtc_do_flush(crtc, old);
+}
+
+/* Returns the minimum frequency of the aclk rate (main engine clock) in Hz */
+static unsigned long
+komeda_calc_min_aclk_rate(struct komeda_crtc *kcrtc,
+			  unsigned long pxlclk)
+{
+	/* Once dual-link one display pipeline drives two display outputs,
+	 * the aclk needs run on the double rate of pxlclk
+	 */
+	if (kcrtc->master->dual_link)
+		return pxlclk * 2;
+	else
+		return pxlclk;
+}
+
+/* Get current aclk rate that specified by state */
+unsigned long komeda_crtc_get_aclk(struct komeda_crtc_state *kcrtc_st)
+{
+	struct drm_crtc *crtc = kcrtc_st->base.crtc;
+	struct komeda_dev *mdev = crtc->dev->dev_private;
+	unsigned long pxlclk = kcrtc_st->base.adjusted_mode.crtc_clock * 1000;
+	unsigned long min_aclk;
+
+	min_aclk = komeda_calc_min_aclk_rate(to_kcrtc(crtc), pxlclk);
+
+	return clk_round_rate(mdev->aclk, min_aclk);
+}
+
+static enum drm_mode_status
+komeda_crtc_mode_valid(struct drm_crtc *crtc, const struct drm_display_mode *m)
+{
+	struct komeda_dev *mdev = crtc->dev->dev_private;
+	struct komeda_crtc *kcrtc = to_kcrtc(crtc);
+	struct komeda_pipeline *master = kcrtc->master;
+	unsigned long min_pxlclk, min_aclk;
+
+	if (m->flags & DRM_MODE_FLAG_INTERLACE)
+		return MODE_NO_INTERLACE;
+
+	min_pxlclk = m->clock * 1000;
+	if (master->dual_link)
+		min_pxlclk /= 2;
+
+	if (min_pxlclk != clk_round_rate(master->pxlclk, min_pxlclk)) {
+		DRM_DEBUG_ATOMIC("pxlclk doesn't support %lu Hz\n", min_pxlclk);
+
+		return MODE_NOCLOCK;
+	}
+
+	min_aclk = komeda_calc_min_aclk_rate(to_kcrtc(crtc), min_pxlclk);
+	if (clk_round_rate(mdev->aclk, min_aclk) < min_aclk) {
+		DRM_DEBUG_ATOMIC("engine clk can't satisfy the requirement of %s-clk: %lu.\n",
+				 m->name, min_pxlclk);
+
+		return MODE_CLOCK_HIGH;
+	}
+
+	return MODE_OK;
+}
+
+static bool komeda_crtc_mode_fixup(struct drm_crtc *crtc,
+				   const struct drm_display_mode *m,
+				   struct drm_display_mode *adjusted_mode)
+{
+	struct komeda_crtc *kcrtc = to_kcrtc(crtc);
+	unsigned long clk_rate;
+
+	drm_mode_set_crtcinfo(adjusted_mode, 0);
+	/* In dual link half the horizontal settings */
+	if (kcrtc->master->dual_link) {
+		adjusted_mode->crtc_clock /= 2;
+		adjusted_mode->crtc_hdisplay /= 2;
+		adjusted_mode->crtc_hsync_start /= 2;
+		adjusted_mode->crtc_hsync_end /= 2;
+		adjusted_mode->crtc_htotal /= 2;
+	}
+
+	clk_rate = adjusted_mode->crtc_clock * 1000;
+	/* crtc_clock will be used as the komeda output pixel clock */
+	adjusted_mode->crtc_clock = clk_round_rate(kcrtc->master->pxlclk,
+						   clk_rate) / 1000;
+
+	return true;
+}
+
+static const struct drm_crtc_helper_funcs komeda_crtc_helper_funcs = {
+	.atomic_check	= komeda_crtc_atomic_check,
+	.atomic_flush	= komeda_crtc_atomic_flush,
+	.atomic_enable	= komeda_crtc_atomic_enable,
+	.atomic_disable	= komeda_crtc_atomic_disable,
+	.mode_valid	= komeda_crtc_mode_valid,
+	.mode_fixup	= komeda_crtc_mode_fixup,
+};
+
+static void komeda_crtc_reset(struct drm_crtc *crtc)
+{
+	struct komeda_crtc_state *state;
+
+	if (crtc->state)
+		__drm_atomic_helper_crtc_destroy_state(crtc->state);
+
+	kfree(to_kcrtc_st(crtc->state));
+	crtc->state = NULL;
+
+	state = kzalloc(sizeof(*state), GFP_KERNEL);
+	if (state)
+		__drm_atomic_helper_crtc_reset(crtc, &state->base);
+}
+
+static struct drm_crtc_state *
+komeda_crtc_atomic_duplicate_state(struct drm_crtc *crtc)
+{
+	struct komeda_crtc_state *old = to_kcrtc_st(crtc->state);
+	struct komeda_crtc_state *new;
+
+	new = kzalloc(sizeof(*new), GFP_KERNEL);
+	if (!new)
+		return NULL;
+
+	__drm_atomic_helper_crtc_duplicate_state(crtc, &new->base);
+
+	new->affected_pipes = old->active_pipes;
+	new->clock_ratio = old->clock_ratio;
+	new->max_slave_zorder = old->max_slave_zorder;
+
+	return &new->base;
+}
+
+static void komeda_crtc_atomic_destroy_state(struct drm_crtc *crtc,
+					     struct drm_crtc_state *state)
+{
+	__drm_atomic_helper_crtc_destroy_state(state);
+	kfree(to_kcrtc_st(state));
+}
+
+static int komeda_crtc_vblank_enable(struct drm_crtc *crtc)
+{
+	struct komeda_dev *mdev = crtc->dev->dev_private;
+	struct komeda_crtc *kcrtc = to_kcrtc(crtc);
+
+	mdev->funcs->on_off_vblank(mdev, kcrtc->master->id, true);
+	return 0;
+}
+
+static void komeda_crtc_vblank_disable(struct drm_crtc *crtc)
+{
+	struct komeda_dev *mdev = crtc->dev->dev_private;
+	struct komeda_crtc *kcrtc = to_kcrtc(crtc);
+
+	mdev->funcs->on_off_vblank(mdev, kcrtc->master->id, false);
+}
+
+static const struct drm_crtc_funcs komeda_crtc_funcs = {
+	.gamma_set		= drm_atomic_helper_legacy_gamma_set,
+	.destroy		= drm_crtc_cleanup,
+	.set_config		= drm_atomic_helper_set_config,
+	.page_flip		= drm_atomic_helper_page_flip,
+	.reset			= komeda_crtc_reset,
+	.atomic_duplicate_state	= komeda_crtc_atomic_duplicate_state,
+	.atomic_destroy_state	= komeda_crtc_atomic_destroy_state,
+	.enable_vblank		= komeda_crtc_vblank_enable,
+	.disable_vblank		= komeda_crtc_vblank_disable,
+};
+
+int komeda_kms_setup_crtcs(struct komeda_kms_dev *kms,
+			   struct komeda_dev *mdev)
+{
+	struct komeda_crtc *crtc;
+	struct komeda_pipeline *master;
+	char str[16];
+	int i;
+
+	kms->n_crtcs = 0;
+
+	for (i = 0; i < mdev->n_pipelines; i++) {
+		crtc = &kms->crtcs[kms->n_crtcs];
+		master = mdev->pipelines[i];
+
+		crtc->master = master;
+		crtc->slave  = komeda_pipeline_get_slave(master);
+
+		if (crtc->slave)
+			sprintf(str, "pipe-%d", crtc->slave->id);
+		else
+			sprintf(str, "None");
+
+		DRM_INFO("CRTC-%d: master(pipe-%d) slave(%s).\n",
+			 kms->n_crtcs, master->id, str);
+
+		kms->n_crtcs++;
+	}
+
+	return 0;
+}
+
+static struct drm_plane *
+get_crtc_primary(struct komeda_kms_dev *kms, struct komeda_crtc *crtc)
+{
+	struct komeda_plane *kplane;
+	struct drm_plane *plane;
+
+	drm_for_each_plane(plane, &kms->base) {
+		if (plane->type != DRM_PLANE_TYPE_PRIMARY)
+			continue;
+
+		kplane = to_kplane(plane);
+		/* only master can be primary */
+		if (kplane->layer->base.pipeline == crtc->master)
+			return plane;
+	}
+
+	return NULL;
+}
+
+static int komeda_crtc_add(struct komeda_kms_dev *kms,
+			   struct komeda_crtc *kcrtc)
+{
+	struct drm_crtc *crtc = &kcrtc->base;
+	int err;
+
+	err = drm_crtc_init_with_planes(&kms->base, crtc,
+					get_crtc_primary(kms, kcrtc), NULL,
+					&komeda_crtc_funcs, NULL);
+	if (err)
+		return err;
+
+	drm_crtc_helper_add(crtc, &komeda_crtc_helper_funcs);
+
+	crtc->port = kcrtc->master->of_output_port;
+
+	drm_crtc_enable_color_mgmt(crtc, 0, true, KOMEDA_COLOR_LUT_SIZE);
+
+	return err;
+}
+
+int komeda_kms_add_crtcs(struct komeda_kms_dev *kms, struct komeda_dev *mdev)
+{
+	int i, err;
+
+	for (i = 0; i < kms->n_crtcs; i++) {
+		err = komeda_crtc_add(kms, &kms->crtcs[i]);
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_dev.c b/drivers/gpu/drm/arm/display/komeda/komeda_dev.c
new file mode 100644
index 000000000..1d767473b
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_dev.c
@@ -0,0 +1,359 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#include <linux/io.h>
+#include <linux/iommu.h>
+#include <linux/of_device.h>
+#include <linux/of_graph.h>
+#include <linux/of_reserved_mem.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/dma-mapping.h>
+#ifdef CONFIG_DEBUG_FS
+#include <linux/debugfs.h>
+#include <linux/seq_file.h>
+#endif
+
+#include <drm/drm_print.h>
+
+#include "komeda_dev.h"
+
+static int komeda_register_show(struct seq_file *sf, void *x)
+{
+	struct komeda_dev *mdev = sf->private;
+	int i;
+
+	seq_puts(sf, "\n====== Komeda register dump =========\n");
+
+	pm_runtime_get_sync(mdev->dev);
+
+	if (mdev->funcs->dump_register)
+		mdev->funcs->dump_register(mdev, sf);
+
+	for (i = 0; i < mdev->n_pipelines; i++)
+		komeda_pipeline_dump_register(mdev->pipelines[i], sf);
+
+	pm_runtime_put(mdev->dev);
+
+	return 0;
+}
+
+static int komeda_register_open(struct inode *inode, struct file *filp)
+{
+	return single_open(filp, komeda_register_show, inode->i_private);
+}
+
+static const struct file_operations komeda_register_fops = {
+	.owner		= THIS_MODULE,
+	.open		= komeda_register_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+};
+
+#ifdef CONFIG_DEBUG_FS
+static void komeda_debugfs_init(struct komeda_dev *mdev)
+{
+	if (!debugfs_initialized())
+		return;
+
+	mdev->debugfs_root = debugfs_create_dir("komeda", NULL);
+	debugfs_create_file("register", 0444, mdev->debugfs_root,
+			    mdev, &komeda_register_fops);
+	debugfs_create_x16("err_verbosity", 0664, mdev->debugfs_root,
+			   &mdev->err_verbosity);
+}
+#endif
+
+static ssize_t
+core_id_show(struct device *dev, struct device_attribute *attr, char *buf)
+{
+	struct komeda_dev *mdev = dev_to_mdev(dev);
+
+	return snprintf(buf, PAGE_SIZE, "0x%08x\n", mdev->chip.core_id);
+}
+static DEVICE_ATTR_RO(core_id);
+
+static ssize_t
+config_id_show(struct device *dev, struct device_attribute *attr, char *buf)
+{
+	struct komeda_dev *mdev = dev_to_mdev(dev);
+	struct komeda_pipeline *pipe = mdev->pipelines[0];
+	union komeda_config_id config_id;
+	int i;
+
+	memset(&config_id, 0, sizeof(config_id));
+
+	config_id.max_line_sz = pipe->layers[0]->hsize_in.end;
+	config_id.n_pipelines = mdev->n_pipelines;
+	config_id.n_scalers = pipe->n_scalers;
+	config_id.n_layers = pipe->n_layers;
+	config_id.n_richs = 0;
+	for (i = 0; i < pipe->n_layers; i++) {
+		if (pipe->layers[i]->layer_type == KOMEDA_FMT_RICH_LAYER)
+			config_id.n_richs++;
+	}
+	return snprintf(buf, PAGE_SIZE, "0x%08x\n", config_id.value);
+}
+static DEVICE_ATTR_RO(config_id);
+
+static ssize_t
+aclk_hz_show(struct device *dev, struct device_attribute *attr, char *buf)
+{
+	struct komeda_dev *mdev = dev_to_mdev(dev);
+
+	return snprintf(buf, PAGE_SIZE, "%lu\n", clk_get_rate(mdev->aclk));
+}
+static DEVICE_ATTR_RO(aclk_hz);
+
+static struct attribute *komeda_sysfs_entries[] = {
+	&dev_attr_core_id.attr,
+	&dev_attr_config_id.attr,
+	&dev_attr_aclk_hz.attr,
+	NULL,
+};
+
+static struct attribute_group komeda_sysfs_attr_group = {
+	.attrs = komeda_sysfs_entries,
+};
+
+static int komeda_parse_pipe_dt(struct komeda_pipeline *pipe)
+{
+	struct device_node *np = pipe->of_node;
+	struct clk *clk;
+
+	clk = of_clk_get_by_name(np, "pxclk");
+	if (IS_ERR(clk)) {
+		DRM_ERROR("get pxclk for pipeline %d failed!\n", pipe->id);
+		return PTR_ERR(clk);
+	}
+	pipe->pxlclk = clk;
+
+	/* enum ports */
+	pipe->of_output_links[0] =
+		of_graph_get_remote_node(np, KOMEDA_OF_PORT_OUTPUT, 0);
+	pipe->of_output_links[1] =
+		of_graph_get_remote_node(np, KOMEDA_OF_PORT_OUTPUT, 1);
+	pipe->of_output_port =
+		of_graph_get_port_by_id(np, KOMEDA_OF_PORT_OUTPUT);
+
+	pipe->dual_link = pipe->of_output_links[0] && pipe->of_output_links[1];
+
+	return 0;
+}
+
+static int komeda_parse_dt(struct device *dev, struct komeda_dev *mdev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct device_node *child, *np = dev->of_node;
+	struct komeda_pipeline *pipe;
+	u32 pipe_id = U32_MAX;
+	int ret = -1;
+
+	mdev->irq  = platform_get_irq(pdev, 0);
+	if (mdev->irq < 0) {
+		DRM_ERROR("could not get IRQ number.\n");
+		return mdev->irq;
+	}
+
+	/* Get the optional framebuffer memory resource */
+	ret = of_reserved_mem_device_init(dev);
+	if (ret && ret != -ENODEV)
+		return ret;
+	ret = 0;
+
+	for_each_available_child_of_node(np, child) {
+		if (of_node_name_eq(child, "pipeline")) {
+			of_property_read_u32(child, "reg", &pipe_id);
+			if (pipe_id >= mdev->n_pipelines) {
+				DRM_WARN("Skip the redundant DT node: pipeline-%u.\n",
+					 pipe_id);
+				continue;
+			}
+			mdev->pipelines[pipe_id]->of_node = of_node_get(child);
+		}
+	}
+
+	for (pipe_id = 0; pipe_id < mdev->n_pipelines; pipe_id++) {
+		pipe = mdev->pipelines[pipe_id];
+
+		if (!pipe->of_node) {
+			DRM_ERROR("Pipeline-%d doesn't have a DT node.\n",
+				  pipe->id);
+			return -EINVAL;
+		}
+		ret = komeda_parse_pipe_dt(pipe);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+struct komeda_dev *komeda_dev_create(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	komeda_identify_func komeda_identify;
+	struct komeda_dev *mdev;
+	int err = 0;
+
+	komeda_identify = of_device_get_match_data(dev);
+	if (!komeda_identify)
+		return ERR_PTR(-ENODEV);
+
+	mdev = devm_kzalloc(dev, sizeof(*mdev), GFP_KERNEL);
+	if (!mdev)
+		return ERR_PTR(-ENOMEM);
+
+	mutex_init(&mdev->lock);
+
+	mdev->dev = dev;
+	mdev->reg_base = devm_platform_ioremap_resource(pdev, 0);
+	if (IS_ERR(mdev->reg_base)) {
+		DRM_ERROR("Map register space failed.\n");
+		err = PTR_ERR(mdev->reg_base);
+		mdev->reg_base = NULL;
+		goto err_cleanup;
+	}
+
+	mdev->aclk = devm_clk_get(dev, "aclk");
+	if (IS_ERR(mdev->aclk)) {
+		DRM_ERROR("Get engine clk failed.\n");
+		err = PTR_ERR(mdev->aclk);
+		mdev->aclk = NULL;
+		goto err_cleanup;
+	}
+
+	clk_prepare_enable(mdev->aclk);
+
+	mdev->funcs = komeda_identify(mdev->reg_base, &mdev->chip);
+	if (!mdev->funcs) {
+		DRM_ERROR("Failed to identify the HW.\n");
+		err = -ENODEV;
+		goto disable_clk;
+	}
+
+	DRM_INFO("Found ARM Mali-D%x version r%dp%d\n",
+		 MALIDP_CORE_ID_PRODUCT_ID(mdev->chip.core_id),
+		 MALIDP_CORE_ID_MAJOR(mdev->chip.core_id),
+		 MALIDP_CORE_ID_MINOR(mdev->chip.core_id));
+
+	mdev->funcs->init_format_table(mdev);
+
+	err = mdev->funcs->enum_resources(mdev);
+	if (err) {
+		DRM_ERROR("enumerate display resource failed.\n");
+		goto disable_clk;
+	}
+
+	err = komeda_parse_dt(dev, mdev);
+	if (err) {
+		DRM_ERROR("parse device tree failed.\n");
+		goto disable_clk;
+	}
+
+	err = komeda_assemble_pipelines(mdev);
+	if (err) {
+		DRM_ERROR("assemble display pipelines failed.\n");
+		goto disable_clk;
+	}
+
+	dev->dma_parms = &mdev->dma_parms;
+	dma_set_max_seg_size(dev, DMA_BIT_MASK(32));
+
+	mdev->iommu = iommu_get_domain_for_dev(mdev->dev);
+	if (!mdev->iommu)
+		DRM_INFO("continue without IOMMU support!\n");
+
+	clk_disable_unprepare(mdev->aclk);
+
+	err = sysfs_create_group(&dev->kobj, &komeda_sysfs_attr_group);
+	if (err) {
+		DRM_ERROR("create sysfs group failed.\n");
+		goto err_cleanup;
+	}
+
+	mdev->err_verbosity = KOMEDA_DEV_PRINT_ERR_EVENTS;
+
+#ifdef CONFIG_DEBUG_FS
+	komeda_debugfs_init(mdev);
+#endif
+
+	return mdev;
+
+disable_clk:
+	clk_disable_unprepare(mdev->aclk);
+err_cleanup:
+	komeda_dev_destroy(mdev);
+	return ERR_PTR(err);
+}
+
+void komeda_dev_destroy(struct komeda_dev *mdev)
+{
+	struct device *dev = mdev->dev;
+	const struct komeda_dev_funcs *funcs = mdev->funcs;
+	int i;
+
+	sysfs_remove_group(&dev->kobj, &komeda_sysfs_attr_group);
+
+#ifdef CONFIG_DEBUG_FS
+	debugfs_remove_recursive(mdev->debugfs_root);
+#endif
+
+	if (mdev->aclk)
+		clk_prepare_enable(mdev->aclk);
+
+	for (i = 0; i < mdev->n_pipelines; i++) {
+		komeda_pipeline_destroy(mdev, mdev->pipelines[i]);
+		mdev->pipelines[i] = NULL;
+	}
+
+	mdev->n_pipelines = 0;
+
+	of_reserved_mem_device_release(dev);
+
+	if (funcs && funcs->cleanup)
+		funcs->cleanup(mdev);
+
+	if (mdev->reg_base) {
+		devm_iounmap(dev, mdev->reg_base);
+		mdev->reg_base = NULL;
+	}
+
+	if (mdev->aclk) {
+		clk_disable_unprepare(mdev->aclk);
+		devm_clk_put(dev, mdev->aclk);
+		mdev->aclk = NULL;
+	}
+
+	devm_kfree(dev, mdev);
+}
+
+int komeda_dev_resume(struct komeda_dev *mdev)
+{
+	clk_prepare_enable(mdev->aclk);
+
+	mdev->funcs->enable_irq(mdev);
+
+	if (mdev->iommu && mdev->funcs->connect_iommu)
+		if (mdev->funcs->connect_iommu(mdev))
+			DRM_ERROR("connect iommu failed.\n");
+
+	return 0;
+}
+
+int komeda_dev_suspend(struct komeda_dev *mdev)
+{
+	if (mdev->iommu && mdev->funcs->disconnect_iommu)
+		if (mdev->funcs->disconnect_iommu(mdev))
+			DRM_ERROR("disconnect iommu failed.\n");
+
+	mdev->funcs->disable_irq(mdev);
+
+	clk_disable_unprepare(mdev->aclk);
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_dev.h b/drivers/gpu/drm/arm/display/komeda/komeda_dev.h
new file mode 100644
index 000000000..ce27f2f27
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_dev.h
@@ -0,0 +1,245 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#ifndef _KOMEDA_DEV_H_
+#define _KOMEDA_DEV_H_
+
+#include <linux/device.h>
+#include <linux/clk.h>
+#include "komeda_pipeline.h"
+#include "malidp_product.h"
+#include "komeda_format_caps.h"
+
+#define KOMEDA_EVENT_VSYNC		BIT_ULL(0)
+#define KOMEDA_EVENT_FLIP		BIT_ULL(1)
+#define KOMEDA_EVENT_URUN		BIT_ULL(2)
+#define KOMEDA_EVENT_IBSY		BIT_ULL(3)
+#define KOMEDA_EVENT_OVR		BIT_ULL(4)
+#define KOMEDA_EVENT_EOW		BIT_ULL(5)
+#define KOMEDA_EVENT_MODE		BIT_ULL(6)
+#define KOMEDA_EVENT_FULL		BIT_ULL(7)
+#define KOMEDA_EVENT_EMPTY		BIT_ULL(8)
+
+#define KOMEDA_ERR_TETO			BIT_ULL(14)
+#define KOMEDA_ERR_TEMR			BIT_ULL(15)
+#define KOMEDA_ERR_TITR			BIT_ULL(16)
+#define KOMEDA_ERR_CPE			BIT_ULL(17)
+#define KOMEDA_ERR_CFGE			BIT_ULL(18)
+#define KOMEDA_ERR_AXIE			BIT_ULL(19)
+#define KOMEDA_ERR_ACE0			BIT_ULL(20)
+#define KOMEDA_ERR_ACE1			BIT_ULL(21)
+#define KOMEDA_ERR_ACE2			BIT_ULL(22)
+#define KOMEDA_ERR_ACE3			BIT_ULL(23)
+#define KOMEDA_ERR_DRIFTTO		BIT_ULL(24)
+#define KOMEDA_ERR_FRAMETO		BIT_ULL(25)
+#define KOMEDA_ERR_CSCE			BIT_ULL(26)
+#define KOMEDA_ERR_ZME			BIT_ULL(27)
+#define KOMEDA_ERR_MERR			BIT_ULL(28)
+#define KOMEDA_ERR_TCF			BIT_ULL(29)
+#define KOMEDA_ERR_TTNG			BIT_ULL(30)
+#define KOMEDA_ERR_TTF			BIT_ULL(31)
+
+#define KOMEDA_ERR_EVENTS	\
+	(KOMEDA_EVENT_URUN	| KOMEDA_EVENT_IBSY	| KOMEDA_EVENT_OVR |\
+	KOMEDA_ERR_TETO		| KOMEDA_ERR_TEMR	| KOMEDA_ERR_TITR |\
+	KOMEDA_ERR_CPE		| KOMEDA_ERR_CFGE	| KOMEDA_ERR_AXIE |\
+	KOMEDA_ERR_ACE0		| KOMEDA_ERR_ACE1	| KOMEDA_ERR_ACE2 |\
+	KOMEDA_ERR_ACE3		| KOMEDA_ERR_DRIFTTO	| KOMEDA_ERR_FRAMETO |\
+	KOMEDA_ERR_ZME		| KOMEDA_ERR_MERR	| KOMEDA_ERR_TCF |\
+	KOMEDA_ERR_TTNG		| KOMEDA_ERR_TTF)
+
+#define KOMEDA_WARN_EVENTS	\
+	(KOMEDA_ERR_CSCE | KOMEDA_EVENT_FULL | KOMEDA_EVENT_EMPTY)
+
+#define KOMEDA_INFO_EVENTS (0 \
+			    | KOMEDA_EVENT_VSYNC \
+			    | KOMEDA_EVENT_FLIP \
+			    | KOMEDA_EVENT_EOW \
+			    | KOMEDA_EVENT_MODE \
+			    )
+
+/* pipeline DT ports */
+enum {
+	KOMEDA_OF_PORT_OUTPUT		= 0,
+	KOMEDA_OF_PORT_COPROC		= 1,
+};
+
+struct komeda_chip_info {
+	u32 arch_id;
+	u32 core_id;
+	u32 core_info;
+	u32 bus_width;
+};
+
+struct komeda_dev;
+
+struct komeda_events {
+	u64 global;
+	u64 pipes[KOMEDA_MAX_PIPELINES];
+};
+
+/**
+ * struct komeda_dev_funcs
+ *
+ * Supplied by chip level and returned by the chip entry function xxx_identify,
+ */
+struct komeda_dev_funcs {
+	/**
+	 * @init_format_table:
+	 *
+	 * initialize &komeda_dev->format_table, this function should be called
+	 * before the &enum_resource
+	 */
+	void (*init_format_table)(struct komeda_dev *mdev);
+	/**
+	 * @enum_resources:
+	 *
+	 * for CHIP to report or add pipeline and component resources to CORE
+	 */
+	int (*enum_resources)(struct komeda_dev *mdev);
+	/** @cleanup: call to chip to cleanup komeda_dev->chip data */
+	void (*cleanup)(struct komeda_dev *mdev);
+	/** @connect_iommu: Optional, connect to external iommu */
+	int (*connect_iommu)(struct komeda_dev *mdev);
+	/** @disconnect_iommu: Optional, disconnect to external iommu */
+	int (*disconnect_iommu)(struct komeda_dev *mdev);
+	/**
+	 * @irq_handler:
+	 *
+	 * for CORE to get the HW event from the CHIP when interrupt happened.
+	 */
+	irqreturn_t (*irq_handler)(struct komeda_dev *mdev,
+				   struct komeda_events *events);
+	/** @enable_irq: enable irq */
+	int (*enable_irq)(struct komeda_dev *mdev);
+	/** @disable_irq: disable irq */
+	int (*disable_irq)(struct komeda_dev *mdev);
+	/** @on_off_vblank: notify HW to on/off vblank */
+	void (*on_off_vblank)(struct komeda_dev *mdev,
+			      int master_pipe, bool on);
+
+	/** @dump_register: Optional, dump registers to seq_file */
+	void (*dump_register)(struct komeda_dev *mdev, struct seq_file *seq);
+	/**
+	 * @change_opmode:
+	 *
+	 * Notify HW to switch to a new display operation mode.
+	 */
+	int (*change_opmode)(struct komeda_dev *mdev, int new_mode);
+	/** @flush: Notify the HW to flush or kickoff the update */
+	void (*flush)(struct komeda_dev *mdev,
+		      int master_pipe, u32 active_pipes);
+};
+
+/*
+ * DISPLAY_MODE describes how many display been enabled, and which will be
+ * passed to CHIP by &komeda_dev_funcs->change_opmode(), then CHIP can do the
+ * pipeline resources assignment according to this usage hint.
+ * -   KOMEDA_MODE_DISP0: Only one display enabled, pipeline-0 work as master.
+ * -   KOMEDA_MODE_DISP1: Only one display enabled, pipeline-0 work as master.
+ * -   KOMEDA_MODE_DUAL_DISP: Dual display mode, both display has been enabled.
+ * And D71 supports assign two pipelines to one single display on mode
+ * KOMEDA_MODE_DISP0/DISP1
+ */
+enum {
+	KOMEDA_MODE_INACTIVE	= 0,
+	KOMEDA_MODE_DISP0	= BIT(0),
+	KOMEDA_MODE_DISP1	= BIT(1),
+	KOMEDA_MODE_DUAL_DISP	= KOMEDA_MODE_DISP0 | KOMEDA_MODE_DISP1,
+};
+
+/**
+ * struct komeda_dev
+ *
+ * Pipeline and component are used to describe how to handle the pixel data.
+ * komeda_device is for describing the whole view of the device, and the
+ * control-abilites of device.
+ */
+struct komeda_dev {
+	/** @dev: the base device structure */
+	struct device *dev;
+	/** @reg_base: the base address of komeda io space */
+	u32 __iomem   *reg_base;
+	/** @dma_parms: the dma parameters of komeda */
+	struct device_dma_parameters dma_parms;
+
+	/** @chip: the basic chip information */
+	struct komeda_chip_info chip;
+	/** @fmt_tbl: initialized by &komeda_dev_funcs->init_format_table */
+	struct komeda_format_caps_table fmt_tbl;
+	/** @aclk: HW main engine clk */
+	struct clk *aclk;
+
+	/** @irq: irq number */
+	int irq;
+
+	/** @lock: used to protect dpmode */
+	struct mutex lock;
+	/** @dpmode: current display mode */
+	u32 dpmode;
+
+	/** @n_pipelines: the number of pipe in @pipelines */
+	int n_pipelines;
+	/** @pipelines: the komeda pipelines */
+	struct komeda_pipeline *pipelines[KOMEDA_MAX_PIPELINES];
+
+	/** @funcs: chip funcs to access to HW */
+	const struct komeda_dev_funcs *funcs;
+	/**
+	 * @chip_data:
+	 *
+	 * chip data will be added by &komeda_dev_funcs.enum_resources() and
+	 * destroyed by &komeda_dev_funcs.cleanup()
+	 */
+	void *chip_data;
+
+	/** @iommu: iommu domain */
+	struct iommu_domain *iommu;
+
+	/** @debugfs_root: root directory of komeda debugfs */
+	struct dentry *debugfs_root;
+	/**
+	 * @err_verbosity: bitmask for how much extra info to print on error
+	 *
+	 * See KOMEDA_DEV_* macros for details. Low byte contains the debug
+	 * level categories, the high byte contains extra debug options.
+	 */
+	u16 err_verbosity;
+	/* Print a single line per error per frame with error events. */
+#define KOMEDA_DEV_PRINT_ERR_EVENTS BIT(0)
+	/* Print a single line per warning per frame with error events. */
+#define KOMEDA_DEV_PRINT_WARN_EVENTS BIT(1)
+	/* Print a single line per info event per frame with error events. */
+#define KOMEDA_DEV_PRINT_INFO_EVENTS BIT(2)
+	/* Dump DRM state on an error or warning event. */
+#define KOMEDA_DEV_PRINT_DUMP_STATE_ON_EVENT BIT(8)
+	/* Disable rate limiting of event prints (normally one per commit) */
+#define KOMEDA_DEV_PRINT_DISABLE_RATELIMIT BIT(12)
+};
+
+static inline bool
+komeda_product_match(struct komeda_dev *mdev, u32 target)
+{
+	return MALIDP_CORE_ID_PRODUCT_ID(mdev->chip.core_id) == target;
+}
+
+typedef const struct komeda_dev_funcs *
+(*komeda_identify_func)(u32 __iomem *reg, struct komeda_chip_info *chip);
+
+const struct komeda_dev_funcs *
+d71_identify(u32 __iomem *reg, struct komeda_chip_info *chip);
+
+struct komeda_dev *komeda_dev_create(struct device *dev);
+void komeda_dev_destroy(struct komeda_dev *mdev);
+
+struct komeda_dev *dev_to_mdev(struct device *dev);
+
+void komeda_print_events(struct komeda_events *evts, struct drm_device *dev);
+
+int komeda_dev_resume(struct komeda_dev *mdev);
+int komeda_dev_suspend(struct komeda_dev *mdev);
+
+#endif /*_KOMEDA_DEV_H_*/
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_drv.c b/drivers/gpu/drm/arm/display/komeda/komeda_drv.c
new file mode 100644
index 000000000..e7933930a
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_drv.c
@@ -0,0 +1,205 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/platform_device.h>
+#include <linux/component.h>
+#include <linux/pm_runtime.h>
+#include <drm/drm_of.h>
+#include "komeda_dev.h"
+#include "komeda_kms.h"
+
+struct komeda_drv {
+	struct komeda_dev *mdev;
+	struct komeda_kms_dev *kms;
+};
+
+struct komeda_dev *dev_to_mdev(struct device *dev)
+{
+	struct komeda_drv *mdrv = dev_get_drvdata(dev);
+
+	return mdrv ? mdrv->mdev : NULL;
+}
+
+static void komeda_unbind(struct device *dev)
+{
+	struct komeda_drv *mdrv = dev_get_drvdata(dev);
+
+	if (!mdrv)
+		return;
+
+	komeda_kms_detach(mdrv->kms);
+
+	if (pm_runtime_enabled(dev))
+		pm_runtime_disable(dev);
+	else
+		komeda_dev_suspend(mdrv->mdev);
+
+	komeda_dev_destroy(mdrv->mdev);
+
+	dev_set_drvdata(dev, NULL);
+	devm_kfree(dev, mdrv);
+}
+
+static int komeda_bind(struct device *dev)
+{
+	struct komeda_drv *mdrv;
+	int err;
+
+	mdrv = devm_kzalloc(dev, sizeof(*mdrv), GFP_KERNEL);
+	if (!mdrv)
+		return -ENOMEM;
+
+	mdrv->mdev = komeda_dev_create(dev);
+	if (IS_ERR(mdrv->mdev)) {
+		err = PTR_ERR(mdrv->mdev);
+		goto free_mdrv;
+	}
+
+	pm_runtime_enable(dev);
+	if (!pm_runtime_enabled(dev))
+		komeda_dev_resume(mdrv->mdev);
+
+	mdrv->kms = komeda_kms_attach(mdrv->mdev);
+	if (IS_ERR(mdrv->kms)) {
+		err = PTR_ERR(mdrv->kms);
+		goto destroy_mdev;
+	}
+
+	dev_set_drvdata(dev, mdrv);
+
+	return 0;
+
+destroy_mdev:
+	if (pm_runtime_enabled(dev))
+		pm_runtime_disable(dev);
+	else
+		komeda_dev_suspend(mdrv->mdev);
+
+	komeda_dev_destroy(mdrv->mdev);
+
+free_mdrv:
+	devm_kfree(dev, mdrv);
+	return err;
+}
+
+static const struct component_master_ops komeda_master_ops = {
+	.bind	= komeda_bind,
+	.unbind	= komeda_unbind,
+};
+
+static int compare_of(struct device *dev, void *data)
+{
+	return dev->of_node == data;
+}
+
+static void komeda_add_slave(struct device *master,
+			     struct component_match **match,
+			     struct device_node *np,
+			     u32 port, u32 endpoint)
+{
+	struct device_node *remote;
+
+	remote = of_graph_get_remote_node(np, port, endpoint);
+	if (remote) {
+		drm_of_component_match_add(master, match, compare_of, remote);
+		of_node_put(remote);
+	}
+}
+
+static int komeda_platform_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct component_match *match = NULL;
+	struct device_node *child;
+
+	if (!dev->of_node)
+		return -ENODEV;
+
+	for_each_available_child_of_node(dev->of_node, child) {
+		if (of_node_cmp(child->name, "pipeline") != 0)
+			continue;
+
+		/* add connector */
+		komeda_add_slave(dev, &match, child, KOMEDA_OF_PORT_OUTPUT, 0);
+		komeda_add_slave(dev, &match, child, KOMEDA_OF_PORT_OUTPUT, 1);
+	}
+
+	return component_master_add_with_match(dev, &komeda_master_ops, match);
+}
+
+static int komeda_platform_remove(struct platform_device *pdev)
+{
+	component_master_del(&pdev->dev, &komeda_master_ops);
+	return 0;
+}
+
+static const struct of_device_id komeda_of_match[] = {
+	{ .compatible = "arm,mali-d71", .data = d71_identify, },
+	{ .compatible = "arm,mali-d32", .data = d71_identify, },
+	{},
+};
+
+MODULE_DEVICE_TABLE(of, komeda_of_match);
+
+static int __maybe_unused komeda_rt_pm_suspend(struct device *dev)
+{
+	struct komeda_drv *mdrv = dev_get_drvdata(dev);
+
+	return komeda_dev_suspend(mdrv->mdev);
+}
+
+static int __maybe_unused komeda_rt_pm_resume(struct device *dev)
+{
+	struct komeda_drv *mdrv = dev_get_drvdata(dev);
+
+	return komeda_dev_resume(mdrv->mdev);
+}
+
+static int __maybe_unused komeda_pm_suspend(struct device *dev)
+{
+	struct komeda_drv *mdrv = dev_get_drvdata(dev);
+	int res;
+
+	res = drm_mode_config_helper_suspend(&mdrv->kms->base);
+
+	if (!pm_runtime_status_suspended(dev))
+		komeda_dev_suspend(mdrv->mdev);
+
+	return res;
+}
+
+static int __maybe_unused komeda_pm_resume(struct device *dev)
+{
+	struct komeda_drv *mdrv = dev_get_drvdata(dev);
+
+	if (!pm_runtime_status_suspended(dev))
+		komeda_dev_resume(mdrv->mdev);
+
+	return drm_mode_config_helper_resume(&mdrv->kms->base);
+}
+
+static const struct dev_pm_ops komeda_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(komeda_pm_suspend, komeda_pm_resume)
+	SET_RUNTIME_PM_OPS(komeda_rt_pm_suspend, komeda_rt_pm_resume, NULL)
+};
+
+static struct platform_driver komeda_platform_driver = {
+	.probe	= komeda_platform_probe,
+	.remove	= komeda_platform_remove,
+	.driver	= {
+		.name = "komeda",
+		.of_match_table	= komeda_of_match,
+		.pm = &komeda_pm_ops,
+	},
+};
+
+module_platform_driver(komeda_platform_driver);
+
+MODULE_AUTHOR("James.Qian.Wang <james.qian.wang@arm.com>");
+MODULE_DESCRIPTION("Komeda KMS driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_event.c b/drivers/gpu/drm/arm/display/komeda/komeda_event.c
new file mode 100644
index 000000000..53f944e66
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_event.c
@@ -0,0 +1,158 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2019 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#include <drm/drm_atomic.h>
+#include <drm/drm_print.h>
+
+#include "komeda_dev.h"
+
+struct komeda_str {
+	char *str;
+	u32 sz;
+	u32 len;
+};
+
+/* return 0 on success,  < 0 on no space.
+ */
+__printf(2, 3)
+static int komeda_sprintf(struct komeda_str *str, const char *fmt, ...)
+{
+	va_list args;
+	int num, free_sz;
+	int err;
+
+	free_sz = str->sz - str->len - 1;
+	if (free_sz <= 0)
+		return -ENOSPC;
+
+	va_start(args, fmt);
+
+	num = vsnprintf(str->str + str->len, free_sz, fmt, args);
+
+	va_end(args);
+
+	if (num < free_sz) {
+		str->len += num;
+		err = 0;
+	} else {
+		str->len = str->sz - 1;
+		err = -ENOSPC;
+	}
+
+	return err;
+}
+
+static void evt_sprintf(struct komeda_str *str, u64 evt, const char *msg)
+{
+	if (evt)
+		komeda_sprintf(str, msg);
+}
+
+static void evt_str(struct komeda_str *str, u64 events)
+{
+	if (events == 0ULL) {
+		komeda_sprintf(str, "None");
+		return;
+	}
+
+	evt_sprintf(str, events & KOMEDA_EVENT_VSYNC, "VSYNC|");
+	evt_sprintf(str, events & KOMEDA_EVENT_FLIP, "FLIP|");
+	evt_sprintf(str, events & KOMEDA_EVENT_EOW, "EOW|");
+	evt_sprintf(str, events & KOMEDA_EVENT_MODE, "OP-MODE|");
+
+	evt_sprintf(str, events & KOMEDA_EVENT_URUN, "UNDERRUN|");
+	evt_sprintf(str, events & KOMEDA_EVENT_OVR, "OVERRUN|");
+
+	/* GLB error */
+	evt_sprintf(str, events & KOMEDA_ERR_MERR, "MERR|");
+	evt_sprintf(str, events & KOMEDA_ERR_FRAMETO, "FRAMETO|");
+
+	/* DOU error */
+	evt_sprintf(str, events & KOMEDA_ERR_DRIFTTO, "DRIFTTO|");
+	evt_sprintf(str, events & KOMEDA_ERR_FRAMETO, "FRAMETO|");
+	evt_sprintf(str, events & KOMEDA_ERR_TETO, "TETO|");
+	evt_sprintf(str, events & KOMEDA_ERR_CSCE, "CSCE|");
+
+	/* LPU errors or events */
+	evt_sprintf(str, events & KOMEDA_EVENT_IBSY, "IBSY|");
+	evt_sprintf(str, events & KOMEDA_EVENT_EMPTY, "EMPTY|");
+	evt_sprintf(str, events & KOMEDA_EVENT_FULL, "FULL|");
+	evt_sprintf(str, events & KOMEDA_ERR_AXIE, "AXIE|");
+	evt_sprintf(str, events & KOMEDA_ERR_ACE0, "ACE0|");
+	evt_sprintf(str, events & KOMEDA_ERR_ACE1, "ACE1|");
+	evt_sprintf(str, events & KOMEDA_ERR_ACE2, "ACE2|");
+	evt_sprintf(str, events & KOMEDA_ERR_ACE3, "ACE3|");
+
+	/* LPU TBU errors*/
+	evt_sprintf(str, events & KOMEDA_ERR_TCF, "TCF|");
+	evt_sprintf(str, events & KOMEDA_ERR_TTNG, "TTNG|");
+	evt_sprintf(str, events & KOMEDA_ERR_TITR, "TITR|");
+	evt_sprintf(str, events & KOMEDA_ERR_TEMR, "TEMR|");
+	evt_sprintf(str, events & KOMEDA_ERR_TTF, "TTF|");
+
+	/* CU errors*/
+	evt_sprintf(str, events & KOMEDA_ERR_CPE, "COPROC|");
+	evt_sprintf(str, events & KOMEDA_ERR_ZME, "ZME|");
+	evt_sprintf(str, events & KOMEDA_ERR_CFGE, "CFGE|");
+	evt_sprintf(str, events & KOMEDA_ERR_TEMR, "TEMR|");
+
+	if (str->len > 0 && (str->str[str->len - 1] == '|')) {
+		str->str[str->len - 1] = 0;
+		str->len--;
+	}
+}
+
+static bool is_new_frame(struct komeda_events *a)
+{
+	return (a->pipes[0] | a->pipes[1]) &
+	       (KOMEDA_EVENT_FLIP | KOMEDA_EVENT_EOW);
+}
+
+void komeda_print_events(struct komeda_events *evts, struct drm_device *dev)
+{
+	u64 print_evts = 0;
+	static bool en_print = true;
+	struct komeda_dev *mdev = dev->dev_private;
+	u16 const err_verbosity = mdev->err_verbosity;
+	u64 evts_mask = evts->global | evts->pipes[0] | evts->pipes[1];
+
+	/* reduce the same msg print, only print the first evt for one frame */
+	if (evts->global || is_new_frame(evts))
+		en_print = true;
+	if (!(err_verbosity & KOMEDA_DEV_PRINT_DISABLE_RATELIMIT) && !en_print)
+		return;
+
+	if (err_verbosity & KOMEDA_DEV_PRINT_ERR_EVENTS)
+		print_evts |= KOMEDA_ERR_EVENTS;
+	if (err_verbosity & KOMEDA_DEV_PRINT_WARN_EVENTS)
+		print_evts |= KOMEDA_WARN_EVENTS;
+	if (err_verbosity & KOMEDA_DEV_PRINT_INFO_EVENTS)
+		print_evts |= KOMEDA_INFO_EVENTS;
+
+	if (evts_mask & print_evts) {
+		char msg[256];
+		struct komeda_str str;
+		struct drm_printer p = drm_info_printer(dev->dev);
+
+		str.str = msg;
+		str.sz  = sizeof(msg);
+		str.len = 0;
+
+		komeda_sprintf(&str, "gcu: ");
+		evt_str(&str, evts->global);
+		komeda_sprintf(&str, ", pipes[0]: ");
+		evt_str(&str, evts->pipes[0]);
+		komeda_sprintf(&str, ", pipes[1]: ");
+		evt_str(&str, evts->pipes[1]);
+
+		DRM_ERROR("err detect: %s\n", msg);
+		if ((err_verbosity & KOMEDA_DEV_PRINT_DUMP_STATE_ON_EVENT) &&
+		    (evts_mask & (KOMEDA_ERR_EVENTS | KOMEDA_WARN_EVENTS)))
+			drm_state_dump(dev, &p);
+
+		en_print = false;
+	}
+}
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_format_caps.c b/drivers/gpu/drm/arm/display/komeda/komeda_format_caps.c
new file mode 100644
index 000000000..c9a1edb9a
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_format_caps.c
@@ -0,0 +1,152 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+
+#include <linux/slab.h>
+#include "komeda_format_caps.h"
+#include "malidp_utils.h"
+
+const struct komeda_format_caps *
+komeda_get_format_caps(struct komeda_format_caps_table *table,
+		       u32 fourcc, u64 modifier)
+{
+	const struct komeda_format_caps *caps;
+	u64 afbc_features = modifier & ~(AFBC_FORMAT_MOD_BLOCK_SIZE_MASK);
+	u32 afbc_layout = modifier & AFBC_FORMAT_MOD_BLOCK_SIZE_MASK;
+	int id;
+
+	for (id = 0; id < table->n_formats; id++) {
+		caps = &table->format_caps[id];
+
+		if (fourcc != caps->fourcc)
+			continue;
+
+		if ((modifier == 0ULL) && (caps->supported_afbc_layouts == 0))
+			return caps;
+
+		if (has_bits(afbc_features, caps->supported_afbc_features) &&
+		    has_bit(afbc_layout, caps->supported_afbc_layouts))
+			return caps;
+	}
+
+	return NULL;
+}
+
+u32 komeda_get_afbc_format_bpp(const struct drm_format_info *info, u64 modifier)
+{
+	u32 bpp;
+
+	switch (info->format) {
+	case DRM_FORMAT_YUV420_8BIT:
+		bpp = 12;
+		break;
+	case DRM_FORMAT_YUV420_10BIT:
+		bpp = 15;
+		break;
+	default:
+		bpp = info->cpp[0] * 8;
+		break;
+	}
+
+	return bpp;
+}
+
+/* Two assumptions
+ * 1. RGB always has YTR
+ * 2. Tiled RGB always has SC
+ */
+u64 komeda_supported_modifiers[] = {
+	/* AFBC_16x16 + features: YUV+RGB both */
+	AFBC_16x16(0),
+	/* SPARSE */
+	AFBC_16x16(_SPARSE),
+	/* YTR + (SPARSE) */
+	AFBC_16x16(_YTR | _SPARSE),
+	AFBC_16x16(_YTR),
+	/* SPLIT + SPARSE + YTR RGB only */
+	/* split mode is only allowed for sparse mode */
+	AFBC_16x16(_SPLIT | _SPARSE | _YTR),
+	/* TILED + (SPARSE) */
+	/* TILED YUV format only */
+	AFBC_16x16(_TILED | _SPARSE),
+	AFBC_16x16(_TILED),
+	/* TILED + SC + (SPLIT+SPARSE | SPARSE) + (YTR) */
+	AFBC_16x16(_TILED | _SC | _SPLIT | _SPARSE | _YTR),
+	AFBC_16x16(_TILED | _SC | _SPARSE | _YTR),
+	AFBC_16x16(_TILED | _SC | _YTR),
+	/* AFBC_32x8 + features: which are RGB formats only */
+	/* YTR + (SPARSE) */
+	AFBC_32x8(_YTR | _SPARSE),
+	AFBC_32x8(_YTR),
+	/* SPLIT + SPARSE + (YTR) */
+	/* split mode is only allowed for sparse mode */
+	AFBC_32x8(_SPLIT | _SPARSE | _YTR),
+	/* TILED + SC + (SPLIT+SPARSE | SPARSE) + YTR */
+	AFBC_32x8(_TILED | _SC | _SPLIT | _SPARSE | _YTR),
+	AFBC_32x8(_TILED | _SC | _SPARSE | _YTR),
+	AFBC_32x8(_TILED | _SC | _YTR),
+	DRM_FORMAT_MOD_LINEAR,
+	DRM_FORMAT_MOD_INVALID
+};
+
+bool komeda_format_mod_supported(struct komeda_format_caps_table *table,
+				 u32 layer_type, u32 fourcc, u64 modifier,
+				 u32 rot)
+{
+	const struct komeda_format_caps *caps;
+
+	caps = komeda_get_format_caps(table, fourcc, modifier);
+	if (!caps)
+		return false;
+
+	if (!(caps->supported_layer_types & layer_type))
+		return false;
+
+	if (table->format_mod_supported)
+		return table->format_mod_supported(caps, layer_type, modifier,
+						   rot);
+
+	return true;
+}
+
+u32 *komeda_get_layer_fourcc_list(struct komeda_format_caps_table *table,
+				  u32 layer_type, u32 *n_fmts)
+{
+	const struct komeda_format_caps *cap;
+	u32 *fmts;
+	int i, j, n = 0;
+
+	fmts = kcalloc(table->n_formats, sizeof(u32), GFP_KERNEL);
+	if (!fmts)
+		return NULL;
+
+	for (i = 0; i < table->n_formats; i++) {
+		cap = &table->format_caps[i];
+		if (!(layer_type & cap->supported_layer_types) ||
+		    (cap->fourcc == 0))
+			continue;
+
+		/* one fourcc may has two caps items in table (afbc/none-afbc),
+		 * so check the existing list to avoid adding a duplicated one.
+		 */
+		for (j = n - 1; j >= 0; j--)
+			if (fmts[j] == cap->fourcc)
+				break;
+
+		if (j < 0)
+			fmts[n++] = cap->fourcc;
+	}
+
+	if (n_fmts)
+		*n_fmts = n;
+
+	return fmts;
+}
+
+void komeda_put_fourcc_list(u32 *fourcc_list)
+{
+	kfree(fourcc_list);
+}
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_format_caps.h b/drivers/gpu/drm/arm/display/komeda/komeda_format_caps.h
new file mode 100644
index 000000000..32273cf18
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_format_caps.h
@@ -0,0 +1,112 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+
+#ifndef _KOMEDA_FORMAT_CAPS_H_
+#define _KOMEDA_FORMAT_CAPS_H_
+
+#include <linux/types.h>
+#include <uapi/drm/drm_fourcc.h>
+#include <drm/drm_fourcc.h>
+
+#define AFBC(x)		DRM_FORMAT_MOD_ARM_AFBC(x)
+
+/* afbc layerout */
+#define AFBC_16x16(x)	AFBC(AFBC_FORMAT_MOD_BLOCK_SIZE_16x16 | (x))
+#define AFBC_32x8(x)	AFBC(AFBC_FORMAT_MOD_BLOCK_SIZE_32x8 | (x))
+
+/* afbc features */
+#define _YTR		AFBC_FORMAT_MOD_YTR
+#define _SPLIT		AFBC_FORMAT_MOD_SPLIT
+#define _SPARSE		AFBC_FORMAT_MOD_SPARSE
+#define _CBR		AFBC_FORMAT_MOD_CBR
+#define _TILED		AFBC_FORMAT_MOD_TILED
+#define _SC		AFBC_FORMAT_MOD_SC
+
+/* layer_type */
+#define KOMEDA_FMT_RICH_LAYER		BIT(0)
+#define KOMEDA_FMT_SIMPLE_LAYER		BIT(1)
+#define KOMEDA_FMT_WB_LAYER		BIT(2)
+
+#define AFBC_TH_LAYOUT_ALIGNMENT	8
+#define AFBC_HEADER_SIZE		16
+#define AFBC_SUPERBLK_ALIGNMENT		128
+#define AFBC_SUPERBLK_PIXELS		256
+#define AFBC_BODY_START_ALIGNMENT	1024
+#define AFBC_TH_BODY_START_ALIGNMENT	4096
+
+/**
+ * struct komeda_format_caps
+ *
+ * komeda_format_caps is for describing ARM display specific features and
+ * limitations for a specific format, and format_caps will be linked into
+ * &komeda_framebuffer like a extension of &drm_format_info.
+ *
+ * NOTE: one fourcc may has two different format_caps items for fourcc and
+ * fourcc+modifier
+ *
+ * @hw_id: hw format id, hw specific value.
+ * @fourcc: drm fourcc format.
+ * @supported_layer_types: indicate which layer supports this format
+ * @supported_rots: allowed rotations for this format
+ * @supported_afbc_layouts: supported afbc layerout
+ * @supported_afbc_features: supported afbc features
+ */
+struct komeda_format_caps {
+	u32 hw_id;
+	u32 fourcc;
+	u32 supported_layer_types;
+	u32 supported_rots;
+	u32 supported_afbc_layouts;
+	u64 supported_afbc_features;
+};
+
+/**
+ * struct komeda_format_caps_table - format_caps mananger
+ *
+ * @n_formats: the size of format_caps list.
+ * @format_caps: format_caps list.
+ * @format_mod_supported: Optional. Some HW may have special requirements or
+ * limitations which can not be described by format_caps, this func supply HW
+ * the ability to do the further HW specific check.
+ */
+struct komeda_format_caps_table {
+	u32 n_formats;
+	const struct komeda_format_caps *format_caps;
+	bool (*format_mod_supported)(const struct komeda_format_caps *caps,
+				     u32 layer_type, u64 modifier, u32 rot);
+};
+
+extern u64 komeda_supported_modifiers[];
+
+static inline const char *komeda_get_format_name(u32 fourcc, u64 modifier)
+{
+	struct drm_format_name_buf buf;
+	static char name[64];
+
+	snprintf(name, sizeof(name), "%s with modifier: 0x%llx.",
+		 drm_get_format_name(fourcc, &buf), modifier);
+
+	return name;
+}
+
+const struct komeda_format_caps *
+komeda_get_format_caps(struct komeda_format_caps_table *table,
+		       u32 fourcc, u64 modifier);
+
+u32 komeda_get_afbc_format_bpp(const struct drm_format_info *info,
+			       u64 modifier);
+
+u32 *komeda_get_layer_fourcc_list(struct komeda_format_caps_table *table,
+				  u32 layer_type, u32 *n_fmts);
+
+void komeda_put_fourcc_list(u32 *fourcc_list);
+
+bool komeda_format_mod_supported(struct komeda_format_caps_table *table,
+				 u32 layer_type, u32 fourcc, u64 modifier,
+				 u32 rot);
+
+#endif
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_framebuffer.c b/drivers/gpu/drm/arm/display/komeda/komeda_framebuffer.c
new file mode 100644
index 000000000..170f9dc8e
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_framebuffer.c
@@ -0,0 +1,283 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#include <drm/drm_device.h>
+#include <drm/drm_fb_cma_helper.h>
+#include <drm/drm_gem.h>
+#include <drm/drm_gem_cma_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
+
+#include "komeda_framebuffer.h"
+#include "komeda_dev.h"
+
+static void komeda_fb_destroy(struct drm_framebuffer *fb)
+{
+	struct komeda_fb *kfb = to_kfb(fb);
+	u32 i;
+
+	for (i = 0; i < fb->format->num_planes; i++)
+		drm_gem_object_put(fb->obj[i]);
+
+	drm_framebuffer_cleanup(fb);
+	kfree(kfb);
+}
+
+static int komeda_fb_create_handle(struct drm_framebuffer *fb,
+				   struct drm_file *file, u32 *handle)
+{
+	return drm_gem_handle_create(file, fb->obj[0], handle);
+}
+
+static const struct drm_framebuffer_funcs komeda_fb_funcs = {
+	.destroy	= komeda_fb_destroy,
+	.create_handle	= komeda_fb_create_handle,
+};
+
+static int
+komeda_fb_afbc_size_check(struct komeda_fb *kfb, struct drm_file *file,
+			  const struct drm_mode_fb_cmd2 *mode_cmd)
+{
+	struct drm_framebuffer *fb = &kfb->base;
+	const struct drm_format_info *info = fb->format;
+	struct drm_gem_object *obj;
+	u32 alignment_w = 0, alignment_h = 0, alignment_header, n_blocks, bpp;
+	u64 min_size;
+
+	obj = drm_gem_object_lookup(file, mode_cmd->handles[0]);
+	if (!obj) {
+		DRM_DEBUG_KMS("Failed to lookup GEM object\n");
+		return -ENOENT;
+	}
+
+	switch (fb->modifier & AFBC_FORMAT_MOD_BLOCK_SIZE_MASK) {
+	case AFBC_FORMAT_MOD_BLOCK_SIZE_32x8:
+		alignment_w = 32;
+		alignment_h = 8;
+		break;
+	case AFBC_FORMAT_MOD_BLOCK_SIZE_16x16:
+		alignment_w = 16;
+		alignment_h = 16;
+		break;
+	default:
+		WARN(1, "Invalid AFBC_FORMAT_MOD_BLOCK_SIZE: %lld.\n",
+		     fb->modifier & AFBC_FORMAT_MOD_BLOCK_SIZE_MASK);
+		break;
+	}
+
+	/* tiled header afbc */
+	if (fb->modifier & AFBC_FORMAT_MOD_TILED) {
+		alignment_w *= AFBC_TH_LAYOUT_ALIGNMENT;
+		alignment_h *= AFBC_TH_LAYOUT_ALIGNMENT;
+		alignment_header = AFBC_TH_BODY_START_ALIGNMENT;
+	} else {
+		alignment_header = AFBC_BODY_START_ALIGNMENT;
+	}
+
+	kfb->aligned_w = ALIGN(fb->width, alignment_w);
+	kfb->aligned_h = ALIGN(fb->height, alignment_h);
+
+	if (fb->offsets[0] % alignment_header) {
+		DRM_DEBUG_KMS("afbc offset alignment check failed.\n");
+		goto check_failed;
+	}
+
+	n_blocks = (kfb->aligned_w * kfb->aligned_h) / AFBC_SUPERBLK_PIXELS;
+	kfb->offset_payload = ALIGN(n_blocks * AFBC_HEADER_SIZE,
+				    alignment_header);
+
+	bpp = komeda_get_afbc_format_bpp(info, fb->modifier);
+	kfb->afbc_size = kfb->offset_payload + n_blocks *
+			 ALIGN(bpp * AFBC_SUPERBLK_PIXELS / 8,
+			       AFBC_SUPERBLK_ALIGNMENT);
+	min_size = kfb->afbc_size + fb->offsets[0];
+	if (min_size > obj->size) {
+		DRM_DEBUG_KMS("afbc size check failed, obj_size: 0x%zx. min_size 0x%llx.\n",
+			      obj->size, min_size);
+		goto check_failed;
+	}
+
+	fb->obj[0] = obj;
+	return 0;
+
+check_failed:
+	drm_gem_object_put(obj);
+	return -EINVAL;
+}
+
+static int
+komeda_fb_none_afbc_size_check(struct komeda_dev *mdev, struct komeda_fb *kfb,
+			       struct drm_file *file,
+			       const struct drm_mode_fb_cmd2 *mode_cmd)
+{
+	struct drm_framebuffer *fb = &kfb->base;
+	const struct drm_format_info *info = fb->format;
+	struct drm_gem_object *obj;
+	u32 i, block_h;
+	u64 min_size;
+
+	if (komeda_fb_check_src_coords(kfb, 0, 0, fb->width, fb->height))
+		return -EINVAL;
+
+	for (i = 0; i < info->num_planes; i++) {
+		obj = drm_gem_object_lookup(file, mode_cmd->handles[i]);
+		if (!obj) {
+			DRM_DEBUG_KMS("Failed to lookup GEM object\n");
+			return -ENOENT;
+		}
+		fb->obj[i] = obj;
+
+		block_h = drm_format_info_block_height(info, i);
+		if ((fb->pitches[i] * block_h) % mdev->chip.bus_width) {
+			DRM_DEBUG_KMS("Pitch[%d]: 0x%x doesn't align to 0x%x\n",
+				      i, fb->pitches[i], mdev->chip.bus_width);
+			return -EINVAL;
+		}
+
+		min_size = komeda_fb_get_pixel_addr(kfb, 0, fb->height, i)
+			 - to_drm_gem_cma_obj(obj)->paddr;
+		if (obj->size < min_size) {
+			DRM_DEBUG_KMS("The fb->obj[%d] size: 0x%zx lower than the minimum requirement: 0x%llx.\n",
+				      i, obj->size, min_size);
+			return -EINVAL;
+		}
+	}
+
+	if (fb->format->num_planes == 3) {
+		if (fb->pitches[1] != fb->pitches[2]) {
+			DRM_DEBUG_KMS("The pitch[1] and [2] are not same\n");
+			return -EINVAL;
+		}
+	}
+
+	return 0;
+}
+
+struct drm_framebuffer *
+komeda_fb_create(struct drm_device *dev, struct drm_file *file,
+		 const struct drm_mode_fb_cmd2 *mode_cmd)
+{
+	struct komeda_dev *mdev = dev->dev_private;
+	struct komeda_fb *kfb;
+	int ret = 0, i;
+
+	kfb = kzalloc(sizeof(*kfb), GFP_KERNEL);
+	if (!kfb)
+		return ERR_PTR(-ENOMEM);
+
+	kfb->format_caps = komeda_get_format_caps(&mdev->fmt_tbl,
+						  mode_cmd->pixel_format,
+						  mode_cmd->modifier[0]);
+	if (!kfb->format_caps) {
+		DRM_DEBUG_KMS("FMT %x is not supported.\n",
+			      mode_cmd->pixel_format);
+		kfree(kfb);
+		return ERR_PTR(-EINVAL);
+	}
+
+	drm_helper_mode_fill_fb_struct(dev, &kfb->base, mode_cmd);
+
+	if (kfb->base.modifier)
+		ret = komeda_fb_afbc_size_check(kfb, file, mode_cmd);
+	else
+		ret = komeda_fb_none_afbc_size_check(mdev, kfb, file, mode_cmd);
+	if (ret < 0)
+		goto err_cleanup;
+
+	ret = drm_framebuffer_init(dev, &kfb->base, &komeda_fb_funcs);
+	if (ret < 0) {
+		DRM_DEBUG_KMS("failed to initialize fb\n");
+
+		goto err_cleanup;
+	}
+
+	kfb->is_va = mdev->iommu ? true : false;
+
+	return &kfb->base;
+
+err_cleanup:
+	for (i = 0; i < kfb->base.format->num_planes; i++)
+		drm_gem_object_put(kfb->base.obj[i]);
+
+	kfree(kfb);
+	return ERR_PTR(ret);
+}
+
+int komeda_fb_check_src_coords(const struct komeda_fb *kfb,
+			       u32 src_x, u32 src_y, u32 src_w, u32 src_h)
+{
+	const struct drm_framebuffer *fb = &kfb->base;
+	const struct drm_format_info *info = fb->format;
+	u32 block_w = drm_format_info_block_width(fb->format, 0);
+	u32 block_h = drm_format_info_block_height(fb->format, 0);
+
+	if ((src_x + src_w > fb->width) || (src_y + src_h > fb->height)) {
+		DRM_DEBUG_ATOMIC("Invalid source coordinate.\n");
+		return -EINVAL;
+	}
+
+	if ((src_x % info->hsub) || (src_w % info->hsub) ||
+	    (src_y % info->vsub) || (src_h % info->vsub)) {
+		DRM_DEBUG_ATOMIC("Wrong subsampling dimension x:%d, y:%d, w:%d, h:%d for format: %x.\n",
+				 src_x, src_y, src_w, src_h, info->format);
+		return -EINVAL;
+	}
+
+	if ((src_x % block_w) || (src_w % block_w) ||
+	    (src_y % block_h) || (src_h % block_h)) {
+		DRM_DEBUG_ATOMIC("x:%d, y:%d, w:%d, h:%d should be multiple of block_w/h for format: %x.\n",
+				 src_x, src_y, src_w, src_h, info->format);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+dma_addr_t
+komeda_fb_get_pixel_addr(struct komeda_fb *kfb, int x, int y, int plane)
+{
+	struct drm_framebuffer *fb = &kfb->base;
+	const struct drm_gem_cma_object *obj;
+	u32 offset, plane_x, plane_y, block_w, block_sz;
+
+	if (plane >= fb->format->num_planes) {
+		DRM_DEBUG_KMS("Out of max plane num.\n");
+		return -EINVAL;
+	}
+
+	obj = drm_fb_cma_get_gem_obj(fb, plane);
+
+	offset = fb->offsets[plane];
+	if (!fb->modifier) {
+		block_w = drm_format_info_block_width(fb->format, plane);
+		block_sz = fb->format->char_per_block[plane];
+		plane_x = x / (plane ? fb->format->hsub : 1);
+		plane_y = y / (plane ? fb->format->vsub : 1);
+
+		offset += (plane_x / block_w) * block_sz
+			+ plane_y * fb->pitches[plane];
+	}
+
+	return obj->paddr + offset;
+}
+
+/* if the fb can be supported by a specific layer */
+bool komeda_fb_is_layer_supported(struct komeda_fb *kfb, u32 layer_type,
+				  u32 rot)
+{
+	struct drm_framebuffer *fb = &kfb->base;
+	struct komeda_dev *mdev = fb->dev->dev_private;
+	u32 fourcc = fb->format->format;
+	u64 modifier = fb->modifier;
+	bool supported;
+
+	supported = komeda_format_mod_supported(&mdev->fmt_tbl, layer_type,
+						fourcc, modifier, rot);
+	if (!supported)
+		DRM_DEBUG_ATOMIC("Layer TYPE: %d doesn't support fb FMT: %s.\n",
+			layer_type, komeda_get_format_name(fourcc, modifier));
+
+	return supported;
+}
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_framebuffer.h b/drivers/gpu/drm/arm/display/komeda/komeda_framebuffer.h
new file mode 100644
index 000000000..c61ca98a3
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_framebuffer.h
@@ -0,0 +1,48 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#ifndef _KOMEDA_FRAMEBUFFER_H_
+#define _KOMEDA_FRAMEBUFFER_H_
+
+#include <drm/drm_framebuffer.h>
+#include "komeda_format_caps.h"
+
+/**
+ * struct komeda_fb - Entending drm_framebuffer with komeda attribute
+ */
+struct komeda_fb {
+	/** @base: &drm_framebuffer */
+	struct drm_framebuffer base;
+	/**
+	 * @format_caps:
+	 * extends drm_format_info for komeda specific information
+	 */
+	const struct komeda_format_caps *format_caps;
+	/** @is_va: if smmu is enabled, it will be true */
+	bool is_va;
+	/** @aligned_w: aligned frame buffer width */
+	u32 aligned_w;
+	/** @aligned_h: aligned frame buffer height */
+	u32 aligned_h;
+	/** @afbc_size: minimum size of afbc */
+	u32 afbc_size;
+	/** @offset_payload: start of afbc body buffer */
+	u32 offset_payload;
+};
+
+#define to_kfb(dfb)	container_of(dfb, struct komeda_fb, base)
+
+struct drm_framebuffer *
+komeda_fb_create(struct drm_device *dev, struct drm_file *file,
+		const struct drm_mode_fb_cmd2 *mode_cmd);
+int komeda_fb_check_src_coords(const struct komeda_fb *kfb,
+			       u32 src_x, u32 src_y, u32 src_w, u32 src_h);
+dma_addr_t
+komeda_fb_get_pixel_addr(struct komeda_fb *kfb, int x, int y, int plane);
+bool komeda_fb_is_layer_supported(struct komeda_fb *kfb, u32 layer_type,
+		u32 rot);
+
+#endif
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_kms.c b/drivers/gpu/drm/arm/display/komeda/komeda_kms.c
new file mode 100644
index 000000000..1f6682032
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_kms.c
@@ -0,0 +1,337 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#include <linux/component.h>
+#include <linux/interrupt.h>
+
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_drv.h>
+#include <drm/drm_fb_helper.h>
+#include <drm/drm_gem_cma_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
+#include <drm/drm_irq.h>
+#include <drm/drm_managed.h>
+#include <drm/drm_probe_helper.h>
+#include <drm/drm_vblank.h>
+
+#include "komeda_dev.h"
+#include "komeda_framebuffer.h"
+#include "komeda_kms.h"
+
+DEFINE_DRM_GEM_CMA_FOPS(komeda_cma_fops);
+
+static int komeda_gem_cma_dumb_create(struct drm_file *file,
+				      struct drm_device *dev,
+				      struct drm_mode_create_dumb *args)
+{
+	struct komeda_dev *mdev = dev->dev_private;
+	u32 pitch = DIV_ROUND_UP(args->width * args->bpp, 8);
+
+	args->pitch = ALIGN(pitch, mdev->chip.bus_width);
+
+	return drm_gem_cma_dumb_create_internal(file, dev, args);
+}
+
+static irqreturn_t komeda_kms_irq_handler(int irq, void *data)
+{
+	struct drm_device *drm = data;
+	struct komeda_dev *mdev = drm->dev_private;
+	struct komeda_kms_dev *kms = to_kdev(drm);
+	struct komeda_events evts;
+	irqreturn_t status;
+	u32 i;
+
+	/* Call into the CHIP to recognize events */
+	memset(&evts, 0, sizeof(evts));
+	status = mdev->funcs->irq_handler(mdev, &evts);
+
+	komeda_print_events(&evts, drm);
+
+	/* Notify the crtc to handle the events */
+	for (i = 0; i < kms->n_crtcs; i++)
+		komeda_crtc_handle_event(&kms->crtcs[i], &evts);
+
+	return status;
+}
+
+static struct drm_driver komeda_kms_driver = {
+	.driver_features = DRIVER_GEM | DRIVER_MODESET | DRIVER_ATOMIC,
+	.lastclose			= drm_fb_helper_lastclose,
+	DRM_GEM_CMA_DRIVER_OPS_WITH_DUMB_CREATE(komeda_gem_cma_dumb_create),
+	.fops = &komeda_cma_fops,
+	.name = "komeda",
+	.desc = "Arm Komeda Display Processor driver",
+	.date = "20181101",
+	.major = 0,
+	.minor = 1,
+};
+
+static void komeda_kms_commit_tail(struct drm_atomic_state *old_state)
+{
+	struct drm_device *dev = old_state->dev;
+
+	drm_atomic_helper_commit_modeset_disables(dev, old_state);
+
+	drm_atomic_helper_commit_planes(dev, old_state,
+					DRM_PLANE_COMMIT_ACTIVE_ONLY);
+
+	drm_atomic_helper_commit_modeset_enables(dev, old_state);
+
+	drm_atomic_helper_wait_for_flip_done(dev, old_state);
+
+	drm_atomic_helper_commit_hw_done(old_state);
+
+	drm_atomic_helper_cleanup_planes(dev, old_state);
+}
+
+static const struct drm_mode_config_helper_funcs komeda_mode_config_helpers = {
+	.atomic_commit_tail = komeda_kms_commit_tail,
+};
+
+static int komeda_plane_state_list_add(struct drm_plane_state *plane_st,
+				       struct list_head *zorder_list)
+{
+	struct komeda_plane_state *new = to_kplane_st(plane_st);
+	struct komeda_plane_state *node, *last;
+
+	last = list_empty(zorder_list) ?
+	       NULL : list_last_entry(zorder_list, typeof(*last), zlist_node);
+
+	/* Considering the list sequence is zpos increasing, so if list is empty
+	 * or the zpos of new node bigger than the last node in list, no need
+	 * loop and just insert the new one to the tail of the list.
+	 */
+	if (!last || (new->base.zpos > last->base.zpos)) {
+		list_add_tail(&new->zlist_node, zorder_list);
+		return 0;
+	}
+
+	/* Build the list by zpos increasing */
+	list_for_each_entry(node, zorder_list, zlist_node) {
+		if (new->base.zpos < node->base.zpos) {
+			list_add_tail(&new->zlist_node, &node->zlist_node);
+			break;
+		} else if (node->base.zpos == new->base.zpos) {
+			struct drm_plane *a = node->base.plane;
+			struct drm_plane *b = new->base.plane;
+
+			/* Komeda doesn't support setting a same zpos for
+			 * different planes.
+			 */
+			DRM_DEBUG_ATOMIC("PLANE: %s and PLANE: %s are configured same zpos: %d.\n",
+					 a->name, b->name, node->base.zpos);
+			return -EINVAL;
+		}
+	}
+
+	return 0;
+}
+
+static int komeda_crtc_normalize_zpos(struct drm_crtc *crtc,
+				      struct drm_crtc_state *crtc_st)
+{
+	struct drm_atomic_state *state = crtc_st->state;
+	struct komeda_crtc *kcrtc = to_kcrtc(crtc);
+	struct komeda_crtc_state *kcrtc_st = to_kcrtc_st(crtc_st);
+	struct komeda_plane_state *kplane_st;
+	struct drm_plane_state *plane_st;
+	struct drm_plane *plane;
+	struct list_head zorder_list;
+	int order = 0, err;
+
+	DRM_DEBUG_ATOMIC("[CRTC:%d:%s] calculating normalized zpos values\n",
+			 crtc->base.id, crtc->name);
+
+	INIT_LIST_HEAD(&zorder_list);
+
+	/* This loop also added all effected planes into the new state */
+	drm_for_each_plane_mask(plane, crtc->dev, crtc_st->plane_mask) {
+		plane_st = drm_atomic_get_plane_state(state, plane);
+		if (IS_ERR(plane_st))
+			return PTR_ERR(plane_st);
+
+		/* Build a list by zpos increasing */
+		err = komeda_plane_state_list_add(plane_st, &zorder_list);
+		if (err)
+			return err;
+	}
+
+	kcrtc_st->max_slave_zorder = 0;
+
+	list_for_each_entry(kplane_st, &zorder_list, zlist_node) {
+		plane_st = &kplane_st->base;
+		plane = plane_st->plane;
+
+		plane_st->normalized_zpos = order++;
+		/* When layer_split has been enabled, one plane will be handled
+		 * by two separated komeda layers (left/right), which may needs
+		 * two zorders.
+		 * - zorder: for left_layer for left display part.
+		 * - zorder + 1: will be reserved for right layer.
+		 */
+		if (to_kplane_st(plane_st)->layer_split)
+			order++;
+
+		DRM_DEBUG_ATOMIC("[PLANE:%d:%s] zpos:%d, normalized zpos: %d\n",
+				 plane->base.id, plane->name,
+				 plane_st->zpos, plane_st->normalized_zpos);
+
+		/* calculate max slave zorder */
+		if (has_bit(drm_plane_index(plane), kcrtc->slave_planes))
+			kcrtc_st->max_slave_zorder =
+				max(plane_st->normalized_zpos,
+				    kcrtc_st->max_slave_zorder);
+	}
+
+	crtc_st->zpos_changed = true;
+
+	return 0;
+}
+
+static int komeda_kms_check(struct drm_device *dev,
+			    struct drm_atomic_state *state)
+{
+	struct drm_crtc *crtc;
+	struct drm_crtc_state *new_crtc_st;
+	int i, err;
+
+	err = drm_atomic_helper_check_modeset(dev, state);
+	if (err)
+		return err;
+
+	/* Komeda need to re-calculate resource assumption in every commit
+	 * so need to add all affected_planes (even unchanged) to
+	 * drm_atomic_state.
+	 */
+	for_each_new_crtc_in_state(state, crtc, new_crtc_st, i) {
+		err = drm_atomic_add_affected_planes(state, crtc);
+		if (err)
+			return err;
+
+		err = komeda_crtc_normalize_zpos(crtc, new_crtc_st);
+		if (err)
+			return err;
+	}
+
+	err = drm_atomic_helper_check_planes(dev, state);
+	if (err)
+		return err;
+
+	return 0;
+}
+
+static const struct drm_mode_config_funcs komeda_mode_config_funcs = {
+	.fb_create		= komeda_fb_create,
+	.atomic_check		= komeda_kms_check,
+	.atomic_commit		= drm_atomic_helper_commit,
+};
+
+static void komeda_kms_mode_config_init(struct komeda_kms_dev *kms,
+					struct komeda_dev *mdev)
+{
+	struct drm_mode_config *config = &kms->base.mode_config;
+
+	drm_mode_config_init(&kms->base);
+
+	komeda_kms_setup_crtcs(kms, mdev);
+
+	/* Get value from dev */
+	config->min_width	= 0;
+	config->min_height	= 0;
+	config->max_width	= 4096;
+	config->max_height	= 4096;
+	config->allow_fb_modifiers = true;
+
+	config->funcs = &komeda_mode_config_funcs;
+	config->helper_private = &komeda_mode_config_helpers;
+}
+
+struct komeda_kms_dev *komeda_kms_attach(struct komeda_dev *mdev)
+{
+	struct komeda_kms_dev *kms;
+	struct drm_device *drm;
+	int err;
+
+	kms = devm_drm_dev_alloc(mdev->dev, &komeda_kms_driver,
+				 struct komeda_kms_dev, base);
+	if (IS_ERR(kms))
+		return kms;
+
+	drm = &kms->base;
+
+	drm->dev_private = mdev;
+
+	komeda_kms_mode_config_init(kms, mdev);
+
+	err = komeda_kms_add_private_objs(kms, mdev);
+	if (err)
+		goto cleanup_mode_config;
+
+	err = komeda_kms_add_planes(kms, mdev);
+	if (err)
+		goto cleanup_mode_config;
+
+	err = drm_vblank_init(drm, kms->n_crtcs);
+	if (err)
+		goto cleanup_mode_config;
+
+	err = komeda_kms_add_crtcs(kms, mdev);
+	if (err)
+		goto cleanup_mode_config;
+
+	err = komeda_kms_add_wb_connectors(kms, mdev);
+	if (err)
+		goto cleanup_mode_config;
+
+	err = component_bind_all(mdev->dev, kms);
+	if (err)
+		goto cleanup_mode_config;
+
+	drm_mode_config_reset(drm);
+
+	err = devm_request_irq(drm->dev, mdev->irq,
+			       komeda_kms_irq_handler, IRQF_SHARED,
+			       drm->driver->name, drm);
+	if (err)
+		goto free_component_binding;
+
+	drm->irq_enabled = true;
+
+	drm_kms_helper_poll_init(drm);
+
+	err = drm_dev_register(drm, 0);
+	if (err)
+		goto free_interrupts;
+
+	return kms;
+
+free_interrupts:
+	drm_kms_helper_poll_fini(drm);
+	drm->irq_enabled = false;
+free_component_binding:
+	component_unbind_all(mdev->dev, drm);
+cleanup_mode_config:
+	drm_mode_config_cleanup(drm);
+	komeda_kms_cleanup_private_objs(kms);
+	drm->dev_private = NULL;
+	return ERR_PTR(err);
+}
+
+void komeda_kms_detach(struct komeda_kms_dev *kms)
+{
+	struct drm_device *drm = &kms->base;
+	struct komeda_dev *mdev = drm->dev_private;
+
+	drm_dev_unregister(drm);
+	drm_kms_helper_poll_fini(drm);
+	drm_atomic_helper_shutdown(drm);
+	drm->irq_enabled = false;
+	component_unbind_all(mdev->dev, drm);
+	drm_mode_config_cleanup(drm);
+	komeda_kms_cleanup_private_objs(kms);
+	drm->dev_private = NULL;
+}
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_kms.h b/drivers/gpu/drm/arm/display/komeda/komeda_kms.h
new file mode 100644
index 000000000..456f3c435
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_kms.h
@@ -0,0 +1,189 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#ifndef _KOMEDA_KMS_H_
+#define _KOMEDA_KMS_H_
+
+#include <linux/list.h>
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_device.h>
+#include <drm/drm_writeback.h>
+#include <drm/drm_print.h>
+
+/**
+ * struct komeda_plane - komeda instance of drm_plane
+ */
+struct komeda_plane {
+	/** @base: &drm_plane */
+	struct drm_plane base;
+	/**
+	 * @layer:
+	 *
+	 * represents available layer input pipelines for this plane.
+	 *
+	 * NOTE:
+	 * the layer is not for a specific Layer, but indicate a group of
+	 * Layers with same capabilities.
+	 */
+	struct komeda_layer *layer;
+};
+
+/**
+ * struct komeda_plane_state
+ *
+ * The plane_state can be split into two data flow (left/right) and handled
+ * by two layers &komeda_plane.layer and &komeda_plane.layer.right
+ */
+struct komeda_plane_state {
+	/** @base: &drm_plane_state */
+	struct drm_plane_state base;
+	/** @zlist_node: zorder list node */
+	struct list_head zlist_node;
+
+	/** @layer_split: on/off layer_split */
+	u8 layer_split : 1;
+};
+
+/**
+ * struct komeda_wb_connector
+ */
+struct komeda_wb_connector {
+	/** @base: &drm_writeback_connector */
+	struct drm_writeback_connector base;
+
+	/** @wb_layer: represents associated writeback pipeline of komeda */
+	struct komeda_layer *wb_layer;
+};
+
+/**
+ * struct komeda_crtc
+ */
+struct komeda_crtc {
+	/** @base: &drm_crtc */
+	struct drm_crtc base;
+	/** @master: only master has display output */
+	struct komeda_pipeline *master;
+	/**
+	 * @slave: optional
+	 *
+	 * Doesn't have its own display output, the handled data flow will
+	 * merge into the master.
+	 */
+	struct komeda_pipeline *slave;
+
+	/** @slave_planes: komeda slave planes mask */
+	u32 slave_planes;
+
+	/** @wb_conn: komeda write back connector */
+	struct komeda_wb_connector *wb_conn;
+
+	/** @disable_done: this flip_done is for tracing the disable */
+	struct completion *disable_done;
+};
+
+/**
+ * struct komeda_crtc_state
+ */
+struct komeda_crtc_state {
+	/** @base: &drm_crtc_state */
+	struct drm_crtc_state base;
+
+	/* private properties */
+
+	/* computed state which are used by validate/check */
+	/**
+	 * @affected_pipes:
+	 * the affected pipelines in once display instance
+	 */
+	u32 affected_pipes;
+	/**
+	 * @active_pipes:
+	 * the active pipelines in once display instance
+	 */
+	u32 active_pipes;
+
+	/** @clock_ratio: ratio of (aclk << 32)/pxlclk */
+	u64 clock_ratio;
+
+	/** @max_slave_zorder: the maximum of slave zorder */
+	u32 max_slave_zorder;
+};
+
+/** struct komeda_kms_dev - for gather KMS related things */
+struct komeda_kms_dev {
+	/** @base: &drm_device */
+	struct drm_device base;
+
+	/** @n_crtcs: valid numbers of crtcs in &komeda_kms_dev.crtcs */
+	int n_crtcs;
+	/** @crtcs: crtcs list */
+	struct komeda_crtc crtcs[KOMEDA_MAX_PIPELINES];
+};
+
+#define to_kplane(p)	container_of(p, struct komeda_plane, base)
+#define to_kplane_st(p)	container_of(p, struct komeda_plane_state, base)
+#define to_kconn(p)	container_of(p, struct komeda_wb_connector, base)
+#define to_kcrtc(p)	container_of(p, struct komeda_crtc, base)
+#define to_kcrtc_st(p)	container_of(p, struct komeda_crtc_state, base)
+#define to_kdev(p)	container_of(p, struct komeda_kms_dev, base)
+#define to_wb_conn(x)	container_of(x, struct drm_writeback_connector, base)
+
+static inline bool is_writeback_only(struct drm_crtc_state *st)
+{
+	struct komeda_wb_connector *wb_conn = to_kcrtc(st->crtc)->wb_conn;
+	struct drm_connector *conn = wb_conn ? &wb_conn->base.base : NULL;
+
+	return conn && (st->connector_mask == BIT(drm_connector_index(conn)));
+}
+
+static inline bool
+is_only_changed_connector(struct drm_crtc_state *st, struct drm_connector *conn)
+{
+	struct drm_crtc_state *old_st;
+	u32 changed_connectors;
+
+	old_st = drm_atomic_get_old_crtc_state(st->state, st->crtc);
+	changed_connectors = st->connector_mask ^ old_st->connector_mask;
+
+	return BIT(drm_connector_index(conn)) == changed_connectors;
+}
+
+static inline bool has_flip_h(u32 rot)
+{
+	u32 rotation = drm_rotation_simplify(rot,
+					     DRM_MODE_ROTATE_0 |
+					     DRM_MODE_ROTATE_90 |
+					     DRM_MODE_REFLECT_MASK);
+
+	if (rotation & DRM_MODE_ROTATE_90)
+		return !!(rotation & DRM_MODE_REFLECT_Y);
+	else
+		return !!(rotation & DRM_MODE_REFLECT_X);
+}
+
+void komeda_crtc_get_color_config(struct drm_crtc_state *crtc_st,
+				  u32 *color_depths, u32 *color_formats);
+unsigned long komeda_crtc_get_aclk(struct komeda_crtc_state *kcrtc_st);
+
+int komeda_kms_setup_crtcs(struct komeda_kms_dev *kms, struct komeda_dev *mdev);
+
+int komeda_kms_add_crtcs(struct komeda_kms_dev *kms, struct komeda_dev *mdev);
+int komeda_kms_add_planes(struct komeda_kms_dev *kms, struct komeda_dev *mdev);
+int komeda_kms_add_private_objs(struct komeda_kms_dev *kms,
+				struct komeda_dev *mdev);
+int komeda_kms_add_wb_connectors(struct komeda_kms_dev *kms,
+				 struct komeda_dev *mdev);
+void komeda_kms_cleanup_private_objs(struct komeda_kms_dev *kms);
+
+void komeda_crtc_handle_event(struct komeda_crtc   *kcrtc,
+			      struct komeda_events *evts);
+
+struct komeda_kms_dev *komeda_kms_attach(struct komeda_dev *mdev);
+void komeda_kms_detach(struct komeda_kms_dev *kms);
+
+#endif /*_KOMEDA_KMS_H_*/
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_pipeline.c b/drivers/gpu/drm/arm/display/komeda/komeda_pipeline.c
new file mode 100644
index 000000000..79a633984
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_pipeline.c
@@ -0,0 +1,376 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#include <drm/drm_print.h>
+
+#include "komeda_dev.h"
+#include "komeda_pipeline.h"
+
+/** komeda_pipeline_add - Add a pipeline to &komeda_dev */
+struct komeda_pipeline *
+komeda_pipeline_add(struct komeda_dev *mdev, size_t size,
+		    const struct komeda_pipeline_funcs *funcs)
+{
+	struct komeda_pipeline *pipe;
+
+	if (mdev->n_pipelines + 1 > KOMEDA_MAX_PIPELINES) {
+		DRM_ERROR("Exceed max support %d pipelines.\n",
+			  KOMEDA_MAX_PIPELINES);
+		return ERR_PTR(-ENOSPC);
+	}
+
+	if (size < sizeof(*pipe)) {
+		DRM_ERROR("Request pipeline size too small.\n");
+		return ERR_PTR(-EINVAL);
+	}
+
+	pipe = devm_kzalloc(mdev->dev, size, GFP_KERNEL);
+	if (!pipe)
+		return ERR_PTR(-ENOMEM);
+
+	pipe->mdev = mdev;
+	pipe->id   = mdev->n_pipelines;
+	pipe->funcs = funcs;
+
+	mdev->pipelines[mdev->n_pipelines] = pipe;
+	mdev->n_pipelines++;
+
+	return pipe;
+}
+
+void komeda_pipeline_destroy(struct komeda_dev *mdev,
+			     struct komeda_pipeline *pipe)
+{
+	struct komeda_component *c;
+	int i;
+	unsigned long avail_comps = pipe->avail_comps;
+
+	for_each_set_bit(i, &avail_comps, 32) {
+		c = komeda_pipeline_get_component(pipe, i);
+		komeda_component_destroy(mdev, c);
+	}
+
+	clk_put(pipe->pxlclk);
+
+	of_node_put(pipe->of_output_links[0]);
+	of_node_put(pipe->of_output_links[1]);
+	of_node_put(pipe->of_output_port);
+	of_node_put(pipe->of_node);
+
+	devm_kfree(mdev->dev, pipe);
+}
+
+static struct komeda_component **
+komeda_pipeline_get_component_pos(struct komeda_pipeline *pipe, int id)
+{
+	struct komeda_dev *mdev = pipe->mdev;
+	struct komeda_pipeline *temp = NULL;
+	struct komeda_component **pos = NULL;
+
+	switch (id) {
+	case KOMEDA_COMPONENT_LAYER0:
+	case KOMEDA_COMPONENT_LAYER1:
+	case KOMEDA_COMPONENT_LAYER2:
+	case KOMEDA_COMPONENT_LAYER3:
+		pos = to_cpos(pipe->layers[id - KOMEDA_COMPONENT_LAYER0]);
+		break;
+	case KOMEDA_COMPONENT_WB_LAYER:
+		pos = to_cpos(pipe->wb_layer);
+		break;
+	case KOMEDA_COMPONENT_COMPIZ0:
+	case KOMEDA_COMPONENT_COMPIZ1:
+		temp = mdev->pipelines[id - KOMEDA_COMPONENT_COMPIZ0];
+		if (!temp) {
+			DRM_ERROR("compiz-%d doesn't exist.\n", id);
+			return NULL;
+		}
+		pos = to_cpos(temp->compiz);
+		break;
+	case KOMEDA_COMPONENT_SCALER0:
+	case KOMEDA_COMPONENT_SCALER1:
+		pos = to_cpos(pipe->scalers[id - KOMEDA_COMPONENT_SCALER0]);
+		break;
+	case KOMEDA_COMPONENT_SPLITTER:
+		pos = to_cpos(pipe->splitter);
+		break;
+	case KOMEDA_COMPONENT_MERGER:
+		pos = to_cpos(pipe->merger);
+		break;
+	case KOMEDA_COMPONENT_IPS0:
+	case KOMEDA_COMPONENT_IPS1:
+		temp = mdev->pipelines[id - KOMEDA_COMPONENT_IPS0];
+		if (!temp) {
+			DRM_ERROR("ips-%d doesn't exist.\n", id);
+			return NULL;
+		}
+		pos = to_cpos(temp->improc);
+		break;
+	case KOMEDA_COMPONENT_TIMING_CTRLR:
+		pos = to_cpos(pipe->ctrlr);
+		break;
+	default:
+		pos = NULL;
+		DRM_ERROR("Unknown pipeline resource ID: %d.\n", id);
+		break;
+	}
+
+	return pos;
+}
+
+struct komeda_component *
+komeda_pipeline_get_component(struct komeda_pipeline *pipe, int id)
+{
+	struct komeda_component **pos = NULL;
+	struct komeda_component *c = NULL;
+
+	pos = komeda_pipeline_get_component_pos(pipe, id);
+	if (pos)
+		c = *pos;
+
+	return c;
+}
+
+struct komeda_component *
+komeda_pipeline_get_first_component(struct komeda_pipeline *pipe,
+				    u32 comp_mask)
+{
+	struct komeda_component *c = NULL;
+	int id;
+
+	id = find_first_bit((unsigned long *)&comp_mask, 32);
+	if (id < 32)
+		c = komeda_pipeline_get_component(pipe, id);
+
+	return c;
+}
+
+static struct komeda_component *
+komeda_component_pickup_input(struct komeda_component *c, u32 avail_comps)
+{
+	u32 avail_inputs = c->supported_inputs & (avail_comps);
+
+	return komeda_pipeline_get_first_component(c->pipeline, avail_inputs);
+}
+
+/** komeda_component_add - Add a component to &komeda_pipeline */
+struct komeda_component *
+komeda_component_add(struct komeda_pipeline *pipe,
+		     size_t comp_sz, u32 id, u32 hw_id,
+		     const struct komeda_component_funcs *funcs,
+		     u8 max_active_inputs, u32 supported_inputs,
+		     u8 max_active_outputs, u32 __iomem *reg,
+		     const char *name_fmt, ...)
+{
+	struct komeda_component **pos;
+	struct komeda_component *c;
+	int idx, *num = NULL;
+
+	if (max_active_inputs > KOMEDA_COMPONENT_N_INPUTS) {
+		WARN(1, "please large KOMEDA_COMPONENT_N_INPUTS to %d.\n",
+		     max_active_inputs);
+		return ERR_PTR(-ENOSPC);
+	}
+
+	pos = komeda_pipeline_get_component_pos(pipe, id);
+	if (!pos || (*pos))
+		return ERR_PTR(-EINVAL);
+
+	if (has_bit(id, KOMEDA_PIPELINE_LAYERS)) {
+		idx = id - KOMEDA_COMPONENT_LAYER0;
+		num = &pipe->n_layers;
+		if (idx != pipe->n_layers) {
+			DRM_ERROR("please add Layer by id sequence.\n");
+			return ERR_PTR(-EINVAL);
+		}
+	} else if (has_bit(id,  KOMEDA_PIPELINE_SCALERS)) {
+		idx = id - KOMEDA_COMPONENT_SCALER0;
+		num = &pipe->n_scalers;
+		if (idx != pipe->n_scalers) {
+			DRM_ERROR("please add Scaler by id sequence.\n");
+			return ERR_PTR(-EINVAL);
+		}
+	}
+
+	c = devm_kzalloc(pipe->mdev->dev, comp_sz, GFP_KERNEL);
+	if (!c)
+		return ERR_PTR(-ENOMEM);
+
+	c->id = id;
+	c->hw_id = hw_id;
+	c->reg = reg;
+	c->pipeline = pipe;
+	c->max_active_inputs = max_active_inputs;
+	c->max_active_outputs = max_active_outputs;
+	c->supported_inputs = supported_inputs;
+	c->funcs = funcs;
+
+	if (name_fmt) {
+		va_list args;
+
+		va_start(args, name_fmt);
+		vsnprintf(c->name, sizeof(c->name), name_fmt, args);
+		va_end(args);
+	}
+
+	if (num)
+		*num = *num + 1;
+
+	pipe->avail_comps |= BIT(c->id);
+	*pos = c;
+
+	return c;
+}
+
+void komeda_component_destroy(struct komeda_dev *mdev,
+			      struct komeda_component *c)
+{
+	devm_kfree(mdev->dev, c);
+}
+
+static void komeda_component_dump(struct komeda_component *c)
+{
+	if (!c)
+		return;
+
+	DRM_DEBUG("	%s: ID %d-0x%08lx.\n",
+		  c->name, c->id, BIT(c->id));
+	DRM_DEBUG("		max_active_inputs:%d, supported_inputs: 0x%08x.\n",
+		  c->max_active_inputs, c->supported_inputs);
+	DRM_DEBUG("		max_active_outputs:%d, supported_outputs: 0x%08x.\n",
+		  c->max_active_outputs, c->supported_outputs);
+}
+
+static void komeda_pipeline_dump(struct komeda_pipeline *pipe)
+{
+	struct komeda_component *c;
+	int id;
+	unsigned long avail_comps = pipe->avail_comps;
+
+	DRM_INFO("Pipeline-%d: n_layers: %d, n_scalers: %d, output: %s.\n",
+		 pipe->id, pipe->n_layers, pipe->n_scalers,
+		 pipe->dual_link ? "dual-link" : "single-link");
+	DRM_INFO("	output_link[0]: %s.\n",
+		 pipe->of_output_links[0] ?
+		 pipe->of_output_links[0]->full_name : "none");
+	DRM_INFO("	output_link[1]: %s.\n",
+		 pipe->of_output_links[1] ?
+		 pipe->of_output_links[1]->full_name : "none");
+
+	for_each_set_bit(id, &avail_comps, 32) {
+		c = komeda_pipeline_get_component(pipe, id);
+
+		komeda_component_dump(c);
+	}
+}
+
+static void komeda_component_verify_inputs(struct komeda_component *c)
+{
+	struct komeda_pipeline *pipe = c->pipeline;
+	struct komeda_component *input;
+	int id;
+	unsigned long supported_inputs = c->supported_inputs;
+
+	for_each_set_bit(id, &supported_inputs, 32) {
+		input = komeda_pipeline_get_component(pipe, id);
+		if (!input) {
+			c->supported_inputs &= ~(BIT(id));
+			DRM_WARN("Can not find input(ID-%d) for component: %s.\n",
+				 id, c->name);
+			continue;
+		}
+
+		input->supported_outputs |= BIT(c->id);
+	}
+}
+
+static struct komeda_layer *
+komeda_get_layer_split_right_layer(struct komeda_pipeline *pipe,
+				   struct komeda_layer *left)
+{
+	int index = left->base.id - KOMEDA_COMPONENT_LAYER0;
+	int i;
+
+	for (i = index + 1; i < pipe->n_layers; i++)
+		if (left->layer_type == pipe->layers[i]->layer_type)
+			return pipe->layers[i];
+	return NULL;
+}
+
+static void komeda_pipeline_assemble(struct komeda_pipeline *pipe)
+{
+	struct komeda_component *c;
+	struct komeda_layer *layer;
+	int i, id;
+	unsigned long avail_comps = pipe->avail_comps;
+
+	for_each_set_bit(id, &avail_comps, 32) {
+		c = komeda_pipeline_get_component(pipe, id);
+		komeda_component_verify_inputs(c);
+	}
+	/* calculate right layer for the layer split */
+	for (i = 0; i < pipe->n_layers; i++) {
+		layer = pipe->layers[i];
+
+		layer->right = komeda_get_layer_split_right_layer(pipe, layer);
+	}
+
+	if (pipe->dual_link && !pipe->ctrlr->supports_dual_link) {
+		pipe->dual_link = false;
+		DRM_WARN("PIPE-%d doesn't support dual-link, ignore DT dual-link configuration.\n",
+			 pipe->id);
+	}
+}
+
+/* if pipeline_A accept another pipeline_B's component as input, treat
+ * pipeline_B as slave of pipeline_A.
+ */
+struct komeda_pipeline *
+komeda_pipeline_get_slave(struct komeda_pipeline *master)
+{
+	struct komeda_component *slave;
+
+	slave = komeda_component_pickup_input(&master->compiz->base,
+					      KOMEDA_PIPELINE_COMPIZS);
+
+	return slave ? slave->pipeline : NULL;
+}
+
+int komeda_assemble_pipelines(struct komeda_dev *mdev)
+{
+	struct komeda_pipeline *pipe;
+	int i;
+
+	for (i = 0; i < mdev->n_pipelines; i++) {
+		pipe = mdev->pipelines[i];
+
+		komeda_pipeline_assemble(pipe);
+		komeda_pipeline_dump(pipe);
+	}
+
+	return 0;
+}
+
+void komeda_pipeline_dump_register(struct komeda_pipeline *pipe,
+				   struct seq_file *sf)
+{
+	struct komeda_component *c;
+	u32 id;
+	unsigned long avail_comps;
+
+	seq_printf(sf, "\n======== Pipeline-%d ==========\n", pipe->id);
+
+	if (pipe->funcs && pipe->funcs->dump_register)
+		pipe->funcs->dump_register(pipe, sf);
+
+	avail_comps = pipe->avail_comps;
+	for_each_set_bit(id, &avail_comps, 32) {
+		c = komeda_pipeline_get_component(pipe, id);
+
+		seq_printf(sf, "\n------%s------\n", c->name);
+		if (c->funcs->dump_register)
+			c->funcs->dump_register(c, sf);
+	}
+}
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_pipeline.h b/drivers/gpu/drm/arm/display/komeda/komeda_pipeline.h
new file mode 100644
index 000000000..ac8725e24
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_pipeline.h
@@ -0,0 +1,565 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#ifndef _KOMEDA_PIPELINE_H_
+#define _KOMEDA_PIPELINE_H_
+
+#include <linux/types.h>
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include "malidp_utils.h"
+#include "komeda_color_mgmt.h"
+
+#define KOMEDA_MAX_PIPELINES		2
+#define KOMEDA_PIPELINE_MAX_LAYERS	4
+#define KOMEDA_PIPELINE_MAX_SCALERS	2
+#define KOMEDA_COMPONENT_N_INPUTS	5
+
+/* pipeline component IDs */
+enum {
+	KOMEDA_COMPONENT_LAYER0		= 0,
+	KOMEDA_COMPONENT_LAYER1		= 1,
+	KOMEDA_COMPONENT_LAYER2		= 2,
+	KOMEDA_COMPONENT_LAYER3		= 3,
+	KOMEDA_COMPONENT_WB_LAYER	= 7, /* write back layer */
+	KOMEDA_COMPONENT_SCALER0	= 8,
+	KOMEDA_COMPONENT_SCALER1	= 9,
+	KOMEDA_COMPONENT_SPLITTER	= 12,
+	KOMEDA_COMPONENT_MERGER		= 14,
+	KOMEDA_COMPONENT_COMPIZ0	= 16, /* compositor */
+	KOMEDA_COMPONENT_COMPIZ1	= 17,
+	KOMEDA_COMPONENT_IPS0		= 20, /* post image processor */
+	KOMEDA_COMPONENT_IPS1		= 21,
+	KOMEDA_COMPONENT_TIMING_CTRLR	= 22, /* timing controller */
+};
+
+#define KOMEDA_PIPELINE_LAYERS		(BIT(KOMEDA_COMPONENT_LAYER0) |\
+					 BIT(KOMEDA_COMPONENT_LAYER1) |\
+					 BIT(KOMEDA_COMPONENT_LAYER2) |\
+					 BIT(KOMEDA_COMPONENT_LAYER3))
+
+#define KOMEDA_PIPELINE_SCALERS		(BIT(KOMEDA_COMPONENT_SCALER0) |\
+					 BIT(KOMEDA_COMPONENT_SCALER1))
+
+#define KOMEDA_PIPELINE_COMPIZS		(BIT(KOMEDA_COMPONENT_COMPIZ0) |\
+					 BIT(KOMEDA_COMPONENT_COMPIZ1))
+
+#define KOMEDA_PIPELINE_IMPROCS		(BIT(KOMEDA_COMPONENT_IPS0) |\
+					 BIT(KOMEDA_COMPONENT_IPS1))
+struct komeda_component;
+struct komeda_component_state;
+
+/** komeda_component_funcs - component control functions */
+struct komeda_component_funcs {
+	/** @validate: optional,
+	 * component may has special requirements or limitations, this function
+	 * supply HW the ability to do the further HW specific check.
+	 */
+	int (*validate)(struct komeda_component *c,
+			struct komeda_component_state *state);
+	/** @update: update is a active update */
+	void (*update)(struct komeda_component *c,
+		       struct komeda_component_state *state);
+	/** @disable: disable component */
+	void (*disable)(struct komeda_component *c);
+	/** @dump_register: Optional, dump registers to seq_file */
+	void (*dump_register)(struct komeda_component *c, struct seq_file *seq);
+};
+
+/**
+ * struct komeda_component
+ *
+ * struct komeda_component describe the data flow capabilities for how to link a
+ * component into the display pipeline.
+ * all specified components are subclass of this structure.
+ */
+struct komeda_component {
+	/** @obj: treat component as private obj */
+	struct drm_private_obj obj;
+	/** @pipeline: the komeda pipeline this component belongs to */
+	struct komeda_pipeline *pipeline;
+	/** @name: component name */
+	char name[32];
+	/**
+	 * @reg:
+	 * component register base,
+	 * which is initialized by chip and used by chip only
+	 */
+	u32 __iomem *reg;
+	/** @id: component id */
+	u32 id;
+	/**
+	 * @hw_id: component hw id,
+	 * which is initialized by chip and used by chip only
+	 */
+	u32 hw_id;
+
+	/**
+	 * @max_active_inputs:
+	 * @max_active_outputs:
+	 *
+	 * maximum number of inputs/outputs that can be active at the same time
+	 * Note:
+	 * the number isn't the bit number of @supported_inputs or
+	 * @supported_outputs, but may be less than it, since component may not
+	 * support enabling all @supported_inputs/outputs at the same time.
+	 */
+	u8 max_active_inputs;
+	/** @max_active_outputs: maximum number of outputs */
+	u8 max_active_outputs;
+	/**
+	 * @supported_inputs:
+	 * @supported_outputs:
+	 *
+	 * bitmask of BIT(component->id) for the supported inputs/outputs,
+	 * describes the possibilities of how a component is linked into a
+	 * pipeline.
+	 */
+	u32 supported_inputs;
+	/** @supported_outputs: bitmask of supported output componenet ids */
+	u32 supported_outputs;
+
+	/**
+	 * @funcs: chip functions to access HW
+	 */
+	const struct komeda_component_funcs *funcs;
+};
+
+/**
+ * struct komeda_component_output
+ *
+ * a component has multiple outputs, if want to know where the data
+ * comes from, only know the component is not enough, we still need to know
+ * its output port
+ */
+struct komeda_component_output {
+	/** @component: indicate which component the data comes from */
+	struct komeda_component *component;
+	/**
+	 * @output_port:
+	 * the output port of the &komeda_component_output.component
+	 */
+	u8 output_port;
+};
+
+/**
+ * struct komeda_component_state
+ *
+ * component_state is the data flow configuration of the component, and it's
+ * the superclass of all specific component_state like @komeda_layer_state,
+ * @komeda_scaler_state
+ */
+struct komeda_component_state {
+	/** @obj: tracking component_state by drm_atomic_state */
+	struct drm_private_state obj;
+	/** @component: backpointer to the component */
+	struct komeda_component *component;
+	/**
+	 * @binding_user:
+	 * currently bound user, the user can be @crtc, @plane or @wb_conn,
+	 * which is valid decided by @component and @inputs
+	 *
+	 * -  Layer: its user always is plane.
+	 * -  compiz/improc/timing_ctrlr: the user is crtc.
+	 * -  wb_layer: wb_conn;
+	 * -  scaler: plane when input is layer, wb_conn if input is compiz.
+	 */
+	union {
+		/** @crtc: backpointer for user crtc */
+		struct drm_crtc *crtc;
+		/** @plane: backpointer for user plane */
+		struct drm_plane *plane;
+		/** @wb_conn: backpointer for user wb_connector  */
+		struct drm_connector *wb_conn;
+		void *binding_user;
+	};
+
+	/**
+	 * @active_inputs:
+	 *
+	 * active_inputs is bitmask of @inputs index
+	 *
+	 * -  active_inputs = changed_active_inputs | unchanged_active_inputs
+	 * -  affected_inputs = old->active_inputs | new->active_inputs;
+	 * -  disabling_inputs = affected_inputs ^ active_inputs;
+	 * -  changed_inputs = disabling_inputs | changed_active_inputs;
+	 *
+	 * NOTE:
+	 * changed_inputs doesn't include all active_input but only
+	 * @changed_active_inputs, and this bitmask can be used in chip
+	 * level for dirty update.
+	 */
+	u16 active_inputs;
+	/** @changed_active_inputs: bitmask of the changed @active_inputs */
+	u16 changed_active_inputs;
+	/** @affected_inputs: bitmask for affected @inputs */
+	u16 affected_inputs;
+	/**
+	 * @inputs:
+	 *
+	 * the specific inputs[i] only valid on BIT(i) has been set in
+	 * @active_inputs, if not the inputs[i] is undefined.
+	 */
+	struct komeda_component_output inputs[KOMEDA_COMPONENT_N_INPUTS];
+};
+
+static inline u16 component_disabling_inputs(struct komeda_component_state *st)
+{
+	return st->affected_inputs ^ st->active_inputs;
+}
+
+static inline u16 component_changed_inputs(struct komeda_component_state *st)
+{
+	return component_disabling_inputs(st) | st->changed_active_inputs;
+}
+
+#define for_each_changed_input(st, i)	\
+	for ((i) = 0; (i) < (st)->component->max_active_inputs; (i)++)	\
+		if (has_bit((i), component_changed_inputs(st)))
+
+#define to_comp(__c)	(((__c) == NULL) ? NULL : &((__c)->base))
+#define to_cpos(__c)	((struct komeda_component **)&(__c))
+
+struct komeda_layer {
+	struct komeda_component base;
+	/* accepted h/v input range before rotation */
+	struct malidp_range hsize_in, vsize_in;
+	u32 layer_type; /* RICH, SIMPLE or WB */
+	u32 line_sz;
+	u32 yuv_line_sz; /* maximum line size for YUV422 and YUV420 */
+	u32 supported_rots;
+	/* komeda supports layer split which splits a whole image to two parts
+	 * left and right and handle them by two individual layer processors
+	 * Note: left/right are always according to the final display rect,
+	 * not the source buffer.
+	 */
+	struct komeda_layer *right;
+};
+
+struct komeda_layer_state {
+	struct komeda_component_state base;
+	/* layer specific configuration state */
+	u16 hsize, vsize;
+	u32 rot;
+	u16 afbc_crop_l;
+	u16 afbc_crop_r;
+	u16 afbc_crop_t;
+	u16 afbc_crop_b;
+	dma_addr_t addr[3];
+};
+
+struct komeda_scaler {
+	struct komeda_component base;
+	struct malidp_range hsize, vsize;
+	u32 max_upscaling;
+	u32 max_downscaling;
+	u8 scaling_split_overlap; /* split overlap for scaling */
+	u8 enh_split_overlap; /* split overlap for image enhancement */
+};
+
+struct komeda_scaler_state {
+	struct komeda_component_state base;
+	u16 hsize_in, vsize_in;
+	u16 hsize_out, vsize_out;
+	u16 total_hsize_in, total_vsize_in;
+	u16 total_hsize_out; /* total_xxxx are size before split */
+	u16 left_crop, right_crop;
+	u8 en_scaling : 1,
+	   en_alpha : 1, /* enable alpha processing */
+	   en_img_enhancement : 1,
+	   en_split : 1,
+	   right_part : 1; /* right part of split image */
+};
+
+struct komeda_compiz {
+	struct komeda_component base;
+	struct malidp_range hsize, vsize;
+};
+
+struct komeda_compiz_input_cfg {
+	u16 hsize, vsize;
+	u16 hoffset, voffset;
+	u8 pixel_blend_mode, layer_alpha;
+};
+
+struct komeda_compiz_state {
+	struct komeda_component_state base;
+	/* composition size */
+	u16 hsize, vsize;
+	struct komeda_compiz_input_cfg cins[KOMEDA_COMPONENT_N_INPUTS];
+};
+
+struct komeda_merger {
+	struct komeda_component base;
+	struct malidp_range hsize_merged;
+	struct malidp_range vsize_merged;
+};
+
+struct komeda_merger_state {
+	struct komeda_component_state base;
+	u16 hsize_merged;
+	u16 vsize_merged;
+};
+
+struct komeda_splitter {
+	struct komeda_component base;
+	struct malidp_range hsize, vsize;
+};
+
+struct komeda_splitter_state {
+	struct komeda_component_state base;
+	u16 hsize, vsize;
+	u16 overlap;
+};
+
+struct komeda_improc {
+	struct komeda_component base;
+	u32 supported_color_formats;  /* DRM_RGB/YUV444/YUV420*/
+	u32 supported_color_depths; /* BIT(8) | BIT(10)*/
+	u8 supports_degamma : 1;
+	u8 supports_csc : 1;
+	u8 supports_gamma : 1;
+};
+
+struct komeda_improc_state {
+	struct komeda_component_state base;
+	u8 color_format, color_depth;
+	u16 hsize, vsize;
+	u32 fgamma_coeffs[KOMEDA_N_GAMMA_COEFFS];
+	u32 ctm_coeffs[KOMEDA_N_CTM_COEFFS];
+};
+
+/* display timing controller */
+struct komeda_timing_ctrlr {
+	struct komeda_component base;
+	u8 supports_dual_link : 1;
+};
+
+struct komeda_timing_ctrlr_state {
+	struct komeda_component_state base;
+};
+
+/* Why define A separated structure but not use plane_state directly ?
+ * 1. Komeda supports layer_split which means a plane_state can be split and
+ *    handled by two layers, one layer only handle half of plane image.
+ * 2. Fix up the user properties according to HW's capabilities, like user
+ *    set rotation to R180, but HW only supports REFLECT_X+Y. the rot here is
+ *    after drm_rotation_simplify()
+ */
+struct komeda_data_flow_cfg {
+	struct komeda_component_output input;
+	u16 in_x, in_y, in_w, in_h;
+	u32 out_x, out_y, out_w, out_h;
+	u16 total_in_h, total_in_w;
+	u16 total_out_w;
+	u16 left_crop, right_crop, overlap;
+	u32 rot;
+	int blending_zorder;
+	u8 pixel_blend_mode, layer_alpha;
+	u8 en_scaling : 1,
+	   en_img_enhancement : 1,
+	   en_split : 1,
+	   is_yuv : 1,
+	   right_part : 1; /* right part of display image if split enabled */
+};
+
+struct komeda_pipeline_funcs {
+	/* check if the aclk (main engine clock) can satisfy the clock
+	 * requirements of the downscaling that specified by dflow
+	 */
+	int (*downscaling_clk_check)(struct komeda_pipeline *pipe,
+				     struct drm_display_mode *mode,
+				     unsigned long aclk_rate,
+				     struct komeda_data_flow_cfg *dflow);
+	/* dump_register: Optional, dump registers to seq_file */
+	void (*dump_register)(struct komeda_pipeline *pipe,
+			      struct seq_file *sf);
+};
+
+/**
+ * struct komeda_pipeline
+ *
+ * Represent a complete display pipeline and hold all functional components.
+ */
+struct komeda_pipeline {
+	/** @obj: link pipeline as private obj of drm_atomic_state */
+	struct drm_private_obj obj;
+	/** @mdev: the parent komeda_dev */
+	struct komeda_dev *mdev;
+	/** @pxlclk: pixel clock */
+	struct clk *pxlclk;
+	/** @id: pipeline id */
+	int id;
+	/** @avail_comps: available components mask of pipeline */
+	u32 avail_comps;
+	/**
+	 * @standalone_disabled_comps:
+	 *
+	 * When disable the pipeline, some components can not be disabled
+	 * together with others, but need a sparated and standalone disable.
+	 * The standalone_disabled_comps are the components which need to be
+	 * disabled standalone, and this concept also introduce concept of
+	 * two phase.
+	 * phase 1: for disabling the common components.
+	 * phase 2: for disabling the standalong_disabled_comps.
+	 */
+	u32 standalone_disabled_comps;
+	/** @n_layers: the number of layer on @layers */
+	int n_layers;
+	/** @layers: the pipeline layers */
+	struct komeda_layer *layers[KOMEDA_PIPELINE_MAX_LAYERS];
+	/** @n_scalers: the number of scaler on @scalers */
+	int n_scalers;
+	/** @scalers: the pipeline scalers */
+	struct komeda_scaler *scalers[KOMEDA_PIPELINE_MAX_SCALERS];
+	/** @compiz: compositor */
+	struct komeda_compiz *compiz;
+	/** @splitter: for split the compiz output to two half data flows */
+	struct komeda_splitter *splitter;
+	/** @merger: merger */
+	struct komeda_merger *merger;
+	/** @wb_layer: writeback layer */
+	struct komeda_layer  *wb_layer;
+	/** @improc: post image processor */
+	struct komeda_improc *improc;
+	/** @ctrlr: timing controller */
+	struct komeda_timing_ctrlr *ctrlr;
+	/** @funcs: chip private pipeline functions */
+	const struct komeda_pipeline_funcs *funcs;
+
+	/** @of_node: pipeline dt node */
+	struct device_node *of_node;
+	/** @of_output_port: pipeline output port */
+	struct device_node *of_output_port;
+	/** @of_output_links: output connector device nodes */
+	struct device_node *of_output_links[2];
+	/** @dual_link: true if of_output_links[0] and [1] are both valid */
+	bool dual_link;
+};
+
+/**
+ * struct komeda_pipeline_state
+ *
+ * NOTE:
+ * Unlike the pipeline, pipeline_state doesn’t gather any component_state
+ * into it. It because all component will be managed by drm_atomic_state.
+ */
+struct komeda_pipeline_state {
+	/** @obj: tracking pipeline_state by drm_atomic_state */
+	struct drm_private_state obj;
+	/** @pipe: backpointer to the pipeline */
+	struct komeda_pipeline *pipe;
+	/** @crtc: currently bound crtc */
+	struct drm_crtc *crtc;
+	/**
+	 * @active_comps:
+	 *
+	 * bitmask - BIT(component->id) of active components
+	 */
+	u32 active_comps;
+};
+
+#define to_layer(c)	container_of(c, struct komeda_layer, base)
+#define to_compiz(c)	container_of(c, struct komeda_compiz, base)
+#define to_scaler(c)	container_of(c, struct komeda_scaler, base)
+#define to_splitter(c)	container_of(c, struct komeda_splitter, base)
+#define to_merger(c)	container_of(c, struct komeda_merger, base)
+#define to_improc(c)	container_of(c, struct komeda_improc, base)
+#define to_ctrlr(c)	container_of(c, struct komeda_timing_ctrlr, base)
+
+#define to_layer_st(c)	container_of(c, struct komeda_layer_state, base)
+#define to_compiz_st(c)	container_of(c, struct komeda_compiz_state, base)
+#define to_scaler_st(c)	container_of(c, struct komeda_scaler_state, base)
+#define to_splitter_st(c) container_of(c, struct komeda_splitter_state, base)
+#define to_merger_st(c)	container_of(c, struct komeda_merger_state, base)
+#define to_improc_st(c)	container_of(c, struct komeda_improc_state, base)
+#define to_ctrlr_st(c)	container_of(c, struct komeda_timing_ctrlr_state, base)
+
+#define priv_to_comp_st(o) container_of(o, struct komeda_component_state, obj)
+#define priv_to_pipe_st(o) container_of(o, struct komeda_pipeline_state, obj)
+
+/* pipeline APIs */
+struct komeda_pipeline *
+komeda_pipeline_add(struct komeda_dev *mdev, size_t size,
+		    const struct komeda_pipeline_funcs *funcs);
+void komeda_pipeline_destroy(struct komeda_dev *mdev,
+			     struct komeda_pipeline *pipe);
+struct komeda_pipeline *
+komeda_pipeline_get_slave(struct komeda_pipeline *master);
+int komeda_assemble_pipelines(struct komeda_dev *mdev);
+struct komeda_component *
+komeda_pipeline_get_component(struct komeda_pipeline *pipe, int id);
+struct komeda_component *
+komeda_pipeline_get_first_component(struct komeda_pipeline *pipe,
+				    u32 comp_mask);
+
+void komeda_pipeline_dump_register(struct komeda_pipeline *pipe,
+				   struct seq_file *sf);
+
+/* component APIs */
+extern __printf(10, 11)
+struct komeda_component *
+komeda_component_add(struct komeda_pipeline *pipe,
+		     size_t comp_sz, u32 id, u32 hw_id,
+		     const struct komeda_component_funcs *funcs,
+		     u8 max_active_inputs, u32 supported_inputs,
+		     u8 max_active_outputs, u32 __iomem *reg,
+		     const char *name_fmt, ...);
+
+void komeda_component_destroy(struct komeda_dev *mdev,
+			      struct komeda_component *c);
+
+static inline struct komeda_component *
+komeda_component_pickup_output(struct komeda_component *c, u32 avail_comps)
+{
+	u32 avail_inputs = c->supported_outputs & (avail_comps);
+
+	return komeda_pipeline_get_first_component(c->pipeline, avail_inputs);
+}
+
+struct komeda_plane_state;
+struct komeda_crtc_state;
+struct komeda_crtc;
+
+void pipeline_composition_size(struct komeda_crtc_state *kcrtc_st,
+			       u16 *hsize, u16 *vsize);
+
+int komeda_build_layer_data_flow(struct komeda_layer *layer,
+				 struct komeda_plane_state *kplane_st,
+				 struct komeda_crtc_state *kcrtc_st,
+				 struct komeda_data_flow_cfg *dflow);
+int komeda_build_wb_data_flow(struct komeda_layer *wb_layer,
+			      struct drm_connector_state *conn_st,
+			      struct komeda_crtc_state *kcrtc_st,
+			      struct komeda_data_flow_cfg *dflow);
+int komeda_build_display_data_flow(struct komeda_crtc *kcrtc,
+				   struct komeda_crtc_state *kcrtc_st);
+
+int komeda_build_layer_split_data_flow(struct komeda_layer *left,
+				       struct komeda_plane_state *kplane_st,
+				       struct komeda_crtc_state *kcrtc_st,
+				       struct komeda_data_flow_cfg *dflow);
+int komeda_build_wb_split_data_flow(struct komeda_layer *wb_layer,
+				    struct drm_connector_state *conn_st,
+				    struct komeda_crtc_state *kcrtc_st,
+				    struct komeda_data_flow_cfg *dflow);
+
+int komeda_release_unclaimed_resources(struct komeda_pipeline *pipe,
+				       struct komeda_crtc_state *kcrtc_st);
+
+struct komeda_pipeline_state *
+komeda_pipeline_get_old_state(struct komeda_pipeline *pipe,
+			      struct drm_atomic_state *state);
+bool komeda_pipeline_disable(struct komeda_pipeline *pipe,
+			     struct drm_atomic_state *old_state);
+void komeda_pipeline_update(struct komeda_pipeline *pipe,
+			    struct drm_atomic_state *old_state);
+
+void komeda_complete_data_flow_cfg(struct komeda_layer *layer,
+				   struct komeda_data_flow_cfg *dflow,
+				   struct drm_framebuffer *fb);
+
+#endif /* _KOMEDA_PIPELINE_H_*/
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_pipeline_state.c b/drivers/gpu/drm/arm/display/komeda/komeda_pipeline_state.c
new file mode 100644
index 000000000..1e922703e
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_pipeline_state.c
@@ -0,0 +1,1356 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+
+#include <drm/drm_print.h>
+#include <linux/clk.h>
+#include "komeda_dev.h"
+#include "komeda_kms.h"
+#include "komeda_pipeline.h"
+#include "komeda_framebuffer.h"
+
+static inline bool is_switching_user(void *old, void *new)
+{
+	if (!old || !new)
+		return false;
+
+	return old != new;
+}
+
+static struct komeda_pipeline_state *
+komeda_pipeline_get_state(struct komeda_pipeline *pipe,
+			  struct drm_atomic_state *state)
+{
+	struct drm_private_state *priv_st;
+
+	priv_st = drm_atomic_get_private_obj_state(state, &pipe->obj);
+	if (IS_ERR(priv_st))
+		return ERR_CAST(priv_st);
+
+	return priv_to_pipe_st(priv_st);
+}
+
+struct komeda_pipeline_state *
+komeda_pipeline_get_old_state(struct komeda_pipeline *pipe,
+			      struct drm_atomic_state *state)
+{
+	struct drm_private_state *priv_st;
+
+	priv_st = drm_atomic_get_old_private_obj_state(state, &pipe->obj);
+	if (priv_st)
+		return priv_to_pipe_st(priv_st);
+	return NULL;
+}
+
+static struct komeda_pipeline_state *
+komeda_pipeline_get_new_state(struct komeda_pipeline *pipe,
+			      struct drm_atomic_state *state)
+{
+	struct drm_private_state *priv_st;
+
+	priv_st = drm_atomic_get_new_private_obj_state(state, &pipe->obj);
+	if (priv_st)
+		return priv_to_pipe_st(priv_st);
+	return NULL;
+}
+
+/* Assign pipeline for crtc */
+static struct komeda_pipeline_state *
+komeda_pipeline_get_state_and_set_crtc(struct komeda_pipeline *pipe,
+				       struct drm_atomic_state *state,
+				       struct drm_crtc *crtc)
+{
+	struct komeda_pipeline_state *st;
+
+	st = komeda_pipeline_get_state(pipe, state);
+	if (IS_ERR(st))
+		return st;
+
+	if (is_switching_user(crtc, st->crtc)) {
+		DRM_DEBUG_ATOMIC("CRTC%d required pipeline%d is busy.\n",
+				 drm_crtc_index(crtc), pipe->id);
+		return ERR_PTR(-EBUSY);
+	}
+
+	/* pipeline only can be disabled when the it is free or unused */
+	if (!crtc && st->active_comps) {
+		DRM_DEBUG_ATOMIC("Disabling a busy pipeline:%d.\n", pipe->id);
+		return ERR_PTR(-EBUSY);
+	}
+
+	st->crtc = crtc;
+
+	if (crtc) {
+		struct komeda_crtc_state *kcrtc_st;
+
+		kcrtc_st = to_kcrtc_st(drm_atomic_get_new_crtc_state(state,
+								     crtc));
+
+		kcrtc_st->active_pipes |= BIT(pipe->id);
+		kcrtc_st->affected_pipes |= BIT(pipe->id);
+	}
+	return st;
+}
+
+static struct komeda_component_state *
+komeda_component_get_state(struct komeda_component *c,
+			   struct drm_atomic_state *state)
+{
+	struct drm_private_state *priv_st;
+
+	WARN_ON(!drm_modeset_is_locked(&c->pipeline->obj.lock));
+
+	priv_st = drm_atomic_get_private_obj_state(state, &c->obj);
+	if (IS_ERR(priv_st))
+		return ERR_CAST(priv_st);
+
+	return priv_to_comp_st(priv_st);
+}
+
+static struct komeda_component_state *
+komeda_component_get_old_state(struct komeda_component *c,
+			       struct drm_atomic_state *state)
+{
+	struct drm_private_state *priv_st;
+
+	priv_st = drm_atomic_get_old_private_obj_state(state, &c->obj);
+	if (priv_st)
+		return priv_to_comp_st(priv_st);
+	return NULL;
+}
+
+/**
+ * komeda_component_get_state_and_set_user()
+ *
+ * @c: component to get state and set user
+ * @state: global atomic state
+ * @user: direct user, the binding user
+ * @crtc: the CRTC user, the big boss :)
+ *
+ * This function accepts two users:
+ * -   The direct user: can be plane/crtc/wb_connector depends on component
+ * -   The big boss (CRTC)
+ * CRTC is the big boss (the final user), because all component resources
+ * eventually will be assigned to CRTC, like the layer will be binding to
+ * kms_plane, but kms plane will be binding to a CRTC eventually.
+ *
+ * The big boss (CRTC) is for pipeline assignment, since &komeda_component isn't
+ * independent and can be assigned to CRTC freely, but belongs to a specific
+ * pipeline, only pipeline can be shared between crtc, and pipeline as a whole
+ * (include all the internal components) assigned to a specific CRTC.
+ *
+ * So when set a user to komeda_component, need first to check the status of
+ * component->pipeline to see if the pipeline is available on this specific
+ * CRTC. if the pipeline is busy (assigned to another CRTC), even the required
+ * component is free, the component still cannot be assigned to the direct user.
+ */
+static struct komeda_component_state *
+komeda_component_get_state_and_set_user(struct komeda_component *c,
+					struct drm_atomic_state *state,
+					void *user,
+					struct drm_crtc *crtc)
+{
+	struct komeda_pipeline_state *pipe_st;
+	struct komeda_component_state *st;
+
+	/* First check if the pipeline is available */
+	pipe_st = komeda_pipeline_get_state_and_set_crtc(c->pipeline,
+							 state, crtc);
+	if (IS_ERR(pipe_st))
+		return ERR_CAST(pipe_st);
+
+	st = komeda_component_get_state(c, state);
+	if (IS_ERR(st))
+		return st;
+
+	/* check if the component has been occupied */
+	if (is_switching_user(user, st->binding_user)) {
+		DRM_DEBUG_ATOMIC("required %s is busy.\n", c->name);
+		return ERR_PTR(-EBUSY);
+	}
+
+	st->binding_user = user;
+	/* mark the component as active if user is valid */
+	if (st->binding_user)
+		pipe_st->active_comps |= BIT(c->id);
+
+	return st;
+}
+
+static void
+komeda_component_add_input(struct komeda_component_state *state,
+			   struct komeda_component_output *input,
+			   int idx)
+{
+	struct komeda_component *c = state->component;
+
+	WARN_ON((idx < 0 || idx >= c->max_active_inputs));
+
+	/* since the inputs[i] is only valid when it is active. So if a input[i]
+	 * is a newly enabled input which switches from disable to enable, then
+	 * the old inputs[i] is undefined (NOT zeroed), we can not rely on
+	 * memcmp, but directly mark it changed
+	 */
+	if (!has_bit(idx, state->affected_inputs) ||
+	    memcmp(&state->inputs[idx], input, sizeof(*input))) {
+		memcpy(&state->inputs[idx], input, sizeof(*input));
+		state->changed_active_inputs |= BIT(idx);
+	}
+	state->active_inputs |= BIT(idx);
+	state->affected_inputs |= BIT(idx);
+}
+
+static int
+komeda_component_check_input(struct komeda_component_state *state,
+			     struct komeda_component_output *input,
+			     int idx)
+{
+	struct komeda_component *c = state->component;
+
+	if ((idx < 0) || (idx >= c->max_active_inputs)) {
+		DRM_DEBUG_ATOMIC("%s required an invalid %s-input[%d].\n",
+				 input->component->name, c->name, idx);
+		return -EINVAL;
+	}
+
+	if (has_bit(idx, state->active_inputs)) {
+		DRM_DEBUG_ATOMIC("%s required %s-input[%d] has been occupied already.\n",
+				 input->component->name, c->name, idx);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static void
+komeda_component_set_output(struct komeda_component_output *output,
+			    struct komeda_component *comp,
+			    u8 output_port)
+{
+	output->component = comp;
+	output->output_port = output_port;
+}
+
+static int
+komeda_component_validate_private(struct komeda_component *c,
+				  struct komeda_component_state *st)
+{
+	int err;
+
+	if (!c->funcs->validate)
+		return 0;
+
+	err = c->funcs->validate(c, st);
+	if (err)
+		DRM_DEBUG_ATOMIC("%s validate private failed.\n", c->name);
+
+	return err;
+}
+
+/* Get current available scaler from the component->supported_outputs */
+static struct komeda_scaler *
+komeda_component_get_avail_scaler(struct komeda_component *c,
+				  struct drm_atomic_state *state)
+{
+	struct komeda_pipeline_state *pipe_st;
+	u32 avail_scalers;
+
+	pipe_st = komeda_pipeline_get_state(c->pipeline, state);
+	if (!pipe_st)
+		return NULL;
+
+	avail_scalers = (pipe_st->active_comps & KOMEDA_PIPELINE_SCALERS) ^
+			KOMEDA_PIPELINE_SCALERS;
+
+	c = komeda_component_pickup_output(c, avail_scalers);
+
+	return to_scaler(c);
+}
+
+static void
+komeda_rotate_data_flow(struct komeda_data_flow_cfg *dflow, u32 rot)
+{
+	if (drm_rotation_90_or_270(rot)) {
+		swap(dflow->in_h, dflow->in_w);
+		swap(dflow->total_in_h, dflow->total_in_w);
+	}
+}
+
+static int
+komeda_layer_check_cfg(struct komeda_layer *layer,
+		       struct komeda_fb *kfb,
+		       struct komeda_data_flow_cfg *dflow)
+{
+	u32 src_x, src_y, src_w, src_h;
+	u32 line_sz, max_line_sz;
+
+	if (!komeda_fb_is_layer_supported(kfb, layer->layer_type, dflow->rot))
+		return -EINVAL;
+
+	if (layer->base.id == KOMEDA_COMPONENT_WB_LAYER) {
+		src_x = dflow->out_x;
+		src_y = dflow->out_y;
+		src_w = dflow->out_w;
+		src_h = dflow->out_h;
+	} else {
+		src_x = dflow->in_x;
+		src_y = dflow->in_y;
+		src_w = dflow->in_w;
+		src_h = dflow->in_h;
+	}
+
+	if (komeda_fb_check_src_coords(kfb, src_x, src_y, src_w, src_h))
+		return -EINVAL;
+
+	if (!in_range(&layer->hsize_in, src_w)) {
+		DRM_DEBUG_ATOMIC("invalidate src_w %d.\n", src_w);
+		return -EINVAL;
+	}
+
+	if (!in_range(&layer->vsize_in, src_h)) {
+		DRM_DEBUG_ATOMIC("invalidate src_h %d.\n", src_h);
+		return -EINVAL;
+	}
+
+	if (drm_rotation_90_or_270(dflow->rot))
+		line_sz = dflow->in_h;
+	else
+		line_sz = dflow->in_w;
+
+	if (kfb->base.format->hsub > 1)
+		max_line_sz = layer->yuv_line_sz;
+	else
+		max_line_sz = layer->line_sz;
+
+	if (line_sz > max_line_sz) {
+		DRM_DEBUG_ATOMIC("Required line_sz: %d exceeds the max size %d\n",
+				 line_sz, max_line_sz);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int
+komeda_layer_validate(struct komeda_layer *layer,
+		      struct komeda_plane_state *kplane_st,
+		      struct komeda_data_flow_cfg *dflow)
+{
+	struct drm_plane_state *plane_st = &kplane_st->base;
+	struct drm_framebuffer *fb = plane_st->fb;
+	struct komeda_fb *kfb = to_kfb(fb);
+	struct komeda_component_state *c_st;
+	struct komeda_layer_state *st;
+	int i, err;
+
+	err = komeda_layer_check_cfg(layer, kfb, dflow);
+	if (err)
+		return err;
+
+	c_st = komeda_component_get_state_and_set_user(&layer->base,
+			plane_st->state, plane_st->plane, plane_st->crtc);
+	if (IS_ERR(c_st))
+		return PTR_ERR(c_st);
+
+	st = to_layer_st(c_st);
+
+	st->rot = dflow->rot;
+
+	if (fb->modifier) {
+		st->hsize = kfb->aligned_w;
+		st->vsize = kfb->aligned_h;
+		st->afbc_crop_l = dflow->in_x;
+		st->afbc_crop_r = kfb->aligned_w - dflow->in_x - dflow->in_w;
+		st->afbc_crop_t = dflow->in_y;
+		st->afbc_crop_b = kfb->aligned_h - dflow->in_y - dflow->in_h;
+	} else {
+		st->hsize = dflow->in_w;
+		st->vsize = dflow->in_h;
+		st->afbc_crop_l = 0;
+		st->afbc_crop_r = 0;
+		st->afbc_crop_t = 0;
+		st->afbc_crop_b = 0;
+	}
+
+	for (i = 0; i < fb->format->num_planes; i++)
+		st->addr[i] = komeda_fb_get_pixel_addr(kfb, dflow->in_x,
+						       dflow->in_y, i);
+
+	err = komeda_component_validate_private(&layer->base, c_st);
+	if (err)
+		return err;
+
+	/* update the data flow for the next stage */
+	komeda_component_set_output(&dflow->input, &layer->base, 0);
+
+	/*
+	 * The rotation has been handled by layer, so adjusted the data flow for
+	 * the next stage.
+	 */
+	komeda_rotate_data_flow(dflow, st->rot);
+
+	return 0;
+}
+
+static int
+komeda_wb_layer_validate(struct komeda_layer *wb_layer,
+			 struct drm_connector_state *conn_st,
+			 struct komeda_data_flow_cfg *dflow)
+{
+	struct komeda_fb *kfb = to_kfb(conn_st->writeback_job->fb);
+	struct komeda_component_state *c_st;
+	struct komeda_layer_state *st;
+	int i, err;
+
+	err = komeda_layer_check_cfg(wb_layer, kfb, dflow);
+	if (err)
+		return err;
+
+	c_st = komeda_component_get_state_and_set_user(&wb_layer->base,
+			conn_st->state, conn_st->connector, conn_st->crtc);
+	if (IS_ERR(c_st))
+		return PTR_ERR(c_st);
+
+	st = to_layer_st(c_st);
+
+	st->hsize = dflow->out_w;
+	st->vsize = dflow->out_h;
+
+	for (i = 0; i < kfb->base.format->num_planes; i++)
+		st->addr[i] = komeda_fb_get_pixel_addr(kfb, dflow->out_x,
+						       dflow->out_y, i);
+
+	komeda_component_add_input(&st->base, &dflow->input, 0);
+	komeda_component_set_output(&dflow->input, &wb_layer->base, 0);
+
+	return 0;
+}
+
+static bool scaling_ratio_valid(u32 size_in, u32 size_out,
+				u32 max_upscaling, u32 max_downscaling)
+{
+	if (size_out > size_in * max_upscaling)
+		return false;
+	else if (size_in > size_out * max_downscaling)
+		return false;
+	return true;
+}
+
+static int
+komeda_scaler_check_cfg(struct komeda_scaler *scaler,
+			struct komeda_crtc_state *kcrtc_st,
+			struct komeda_data_flow_cfg *dflow)
+{
+	u32 hsize_in, vsize_in, hsize_out, vsize_out;
+	u32 max_upscaling;
+
+	hsize_in = dflow->in_w;
+	vsize_in = dflow->in_h;
+	hsize_out = dflow->out_w;
+	vsize_out = dflow->out_h;
+
+	if (!in_range(&scaler->hsize, hsize_in) ||
+	    !in_range(&scaler->hsize, hsize_out)) {
+		DRM_DEBUG_ATOMIC("Invalid horizontal sizes");
+		return -EINVAL;
+	}
+
+	if (!in_range(&scaler->vsize, vsize_in) ||
+	    !in_range(&scaler->vsize, vsize_out)) {
+		DRM_DEBUG_ATOMIC("Invalid vertical sizes");
+		return -EINVAL;
+	}
+
+	/* If input comes from compiz that means the scaling is for writeback
+	 * and scaler can not do upscaling for writeback
+	 */
+	if (has_bit(dflow->input.component->id, KOMEDA_PIPELINE_COMPIZS))
+		max_upscaling = 1;
+	else
+		max_upscaling = scaler->max_upscaling;
+
+	if (!scaling_ratio_valid(hsize_in, hsize_out, max_upscaling,
+				 scaler->max_downscaling)) {
+		DRM_DEBUG_ATOMIC("Invalid horizontal scaling ratio");
+		return -EINVAL;
+	}
+
+	if (!scaling_ratio_valid(vsize_in, vsize_out, max_upscaling,
+				 scaler->max_downscaling)) {
+		DRM_DEBUG_ATOMIC("Invalid vertical scaling ratio");
+		return -EINVAL;
+	}
+
+	if (hsize_in > hsize_out || vsize_in > vsize_out) {
+		struct komeda_pipeline *pipe = scaler->base.pipeline;
+		int err;
+
+		err = pipe->funcs->downscaling_clk_check(pipe,
+					&kcrtc_st->base.adjusted_mode,
+					komeda_crtc_get_aclk(kcrtc_st), dflow);
+		if (err) {
+			DRM_DEBUG_ATOMIC("aclk can't satisfy the clock requirement of the downscaling\n");
+			return err;
+		}
+	}
+
+	return 0;
+}
+
+static int
+komeda_scaler_validate(void *user,
+		       struct komeda_crtc_state *kcrtc_st,
+		       struct komeda_data_flow_cfg *dflow)
+{
+	struct drm_atomic_state *drm_st = kcrtc_st->base.state;
+	struct komeda_component_state *c_st;
+	struct komeda_scaler_state *st;
+	struct komeda_scaler *scaler;
+	int err = 0;
+
+	if (!(dflow->en_scaling || dflow->en_img_enhancement))
+		return 0;
+
+	scaler = komeda_component_get_avail_scaler(dflow->input.component,
+						   drm_st);
+	if (!scaler) {
+		DRM_DEBUG_ATOMIC("No scaler available");
+		return -EINVAL;
+	}
+
+	err = komeda_scaler_check_cfg(scaler, kcrtc_st, dflow);
+	if (err)
+		return err;
+
+	c_st = komeda_component_get_state_and_set_user(&scaler->base,
+			drm_st, user, kcrtc_st->base.crtc);
+	if (IS_ERR(c_st))
+		return PTR_ERR(c_st);
+
+	st = to_scaler_st(c_st);
+
+	st->hsize_in = dflow->in_w;
+	st->vsize_in = dflow->in_h;
+	st->hsize_out = dflow->out_w;
+	st->vsize_out = dflow->out_h;
+	st->right_crop = dflow->right_crop;
+	st->left_crop = dflow->left_crop;
+	st->total_vsize_in = dflow->total_in_h;
+	st->total_hsize_in = dflow->total_in_w;
+	st->total_hsize_out = dflow->total_out_w;
+
+	/* Enable alpha processing if the next stage needs the pixel alpha */
+	st->en_alpha = dflow->pixel_blend_mode != DRM_MODE_BLEND_PIXEL_NONE;
+	st->en_scaling = dflow->en_scaling;
+	st->en_img_enhancement = dflow->en_img_enhancement;
+	st->en_split = dflow->en_split;
+	st->right_part = dflow->right_part;
+
+	komeda_component_add_input(&st->base, &dflow->input, 0);
+	komeda_component_set_output(&dflow->input, &scaler->base, 0);
+	return err;
+}
+
+static void komeda_split_data_flow(struct komeda_scaler *scaler,
+				   struct komeda_data_flow_cfg *dflow,
+				   struct komeda_data_flow_cfg *l_dflow,
+				   struct komeda_data_flow_cfg *r_dflow);
+
+static int
+komeda_splitter_validate(struct komeda_splitter *splitter,
+			 struct drm_connector_state *conn_st,
+			 struct komeda_data_flow_cfg *dflow,
+			 struct komeda_data_flow_cfg *l_output,
+			 struct komeda_data_flow_cfg *r_output)
+{
+	struct komeda_component_state *c_st;
+	struct komeda_splitter_state *st;
+
+	if (!splitter) {
+		DRM_DEBUG_ATOMIC("Current HW doesn't support splitter.\n");
+		return -EINVAL;
+	}
+
+	if (!in_range(&splitter->hsize, dflow->in_w)) {
+		DRM_DEBUG_ATOMIC("split in_w:%d is out of the acceptable range.\n",
+				 dflow->in_w);
+		return -EINVAL;
+	}
+
+	if (!in_range(&splitter->vsize, dflow->in_h)) {
+		DRM_DEBUG_ATOMIC("split in_h: %d exceeds the acceptable range.\n",
+				 dflow->in_h);
+		return -EINVAL;
+	}
+
+	c_st = komeda_component_get_state_and_set_user(&splitter->base,
+			conn_st->state, conn_st->connector, conn_st->crtc);
+
+	if (IS_ERR(c_st))
+		return PTR_ERR(c_st);
+
+	komeda_split_data_flow(splitter->base.pipeline->scalers[0],
+			       dflow, l_output, r_output);
+
+	st = to_splitter_st(c_st);
+	st->hsize = dflow->in_w;
+	st->vsize = dflow->in_h;
+	st->overlap = dflow->overlap;
+
+	komeda_component_add_input(&st->base, &dflow->input, 0);
+	komeda_component_set_output(&l_output->input, &splitter->base, 0);
+	komeda_component_set_output(&r_output->input, &splitter->base, 1);
+
+	return 0;
+}
+
+static int
+komeda_merger_validate(struct komeda_merger *merger,
+		       void *user,
+		       struct komeda_crtc_state *kcrtc_st,
+		       struct komeda_data_flow_cfg *left_input,
+		       struct komeda_data_flow_cfg *right_input,
+		       struct komeda_data_flow_cfg *output)
+{
+	struct komeda_component_state *c_st;
+	struct komeda_merger_state *st;
+	int err = 0;
+
+	if (!merger) {
+		DRM_DEBUG_ATOMIC("No merger is available");
+		return -EINVAL;
+	}
+
+	if (!in_range(&merger->hsize_merged, output->out_w)) {
+		DRM_DEBUG_ATOMIC("merged_w: %d is out of the accepted range.\n",
+				 output->out_w);
+		return -EINVAL;
+	}
+
+	if (!in_range(&merger->vsize_merged, output->out_h)) {
+		DRM_DEBUG_ATOMIC("merged_h: %d is out of the accepted range.\n",
+				 output->out_h);
+		return -EINVAL;
+	}
+
+	c_st = komeda_component_get_state_and_set_user(&merger->base,
+			kcrtc_st->base.state, kcrtc_st->base.crtc, kcrtc_st->base.crtc);
+
+	if (IS_ERR(c_st))
+		return PTR_ERR(c_st);
+
+	st = to_merger_st(c_st);
+	st->hsize_merged = output->out_w;
+	st->vsize_merged = output->out_h;
+
+	komeda_component_add_input(c_st, &left_input->input, 0);
+	komeda_component_add_input(c_st, &right_input->input, 1);
+	komeda_component_set_output(&output->input, &merger->base, 0);
+
+	return err;
+}
+
+void pipeline_composition_size(struct komeda_crtc_state *kcrtc_st,
+			       u16 *hsize, u16 *vsize)
+{
+	struct drm_display_mode *m = &kcrtc_st->base.adjusted_mode;
+
+	if (hsize)
+		*hsize = m->hdisplay;
+	if (vsize)
+		*vsize = m->vdisplay;
+}
+
+static int
+komeda_compiz_set_input(struct komeda_compiz *compiz,
+			struct komeda_crtc_state *kcrtc_st,
+			struct komeda_data_flow_cfg *dflow)
+{
+	struct drm_atomic_state *drm_st = kcrtc_st->base.state;
+	struct komeda_component_state *c_st, *old_st;
+	struct komeda_compiz_input_cfg *cin;
+	u16 compiz_w, compiz_h;
+	int idx = dflow->blending_zorder;
+
+	pipeline_composition_size(kcrtc_st, &compiz_w, &compiz_h);
+	/* check display rect */
+	if ((dflow->out_x + dflow->out_w > compiz_w) ||
+	    (dflow->out_y + dflow->out_h > compiz_h) ||
+	     dflow->out_w == 0 || dflow->out_h == 0) {
+		DRM_DEBUG_ATOMIC("invalid disp rect [x=%d, y=%d, w=%d, h=%d]\n",
+				 dflow->out_x, dflow->out_y,
+				 dflow->out_w, dflow->out_h);
+		return -EINVAL;
+	}
+
+	c_st = komeda_component_get_state_and_set_user(&compiz->base, drm_st,
+			kcrtc_st->base.crtc, kcrtc_st->base.crtc);
+	if (IS_ERR(c_st))
+		return PTR_ERR(c_st);
+
+	if (komeda_component_check_input(c_st, &dflow->input, idx))
+		return -EINVAL;
+
+	cin = &(to_compiz_st(c_st)->cins[idx]);
+
+	cin->hsize   = dflow->out_w;
+	cin->vsize   = dflow->out_h;
+	cin->hoffset = dflow->out_x;
+	cin->voffset = dflow->out_y;
+	cin->pixel_blend_mode = dflow->pixel_blend_mode;
+	cin->layer_alpha = dflow->layer_alpha;
+
+	old_st = komeda_component_get_old_state(&compiz->base, drm_st);
+	WARN_ON(!old_st);
+
+	/* compare with old to check if this input has been changed */
+	if (memcmp(&(to_compiz_st(old_st)->cins[idx]), cin, sizeof(*cin)))
+		c_st->changed_active_inputs |= BIT(idx);
+
+	komeda_component_add_input(c_st, &dflow->input, idx);
+	komeda_component_set_output(&dflow->input, &compiz->base, 0);
+
+	return 0;
+}
+
+static int
+komeda_compiz_validate(struct komeda_compiz *compiz,
+		       struct komeda_crtc_state *state,
+		       struct komeda_data_flow_cfg *dflow)
+{
+	struct komeda_component_state *c_st;
+	struct komeda_compiz_state *st;
+
+	c_st = komeda_component_get_state_and_set_user(&compiz->base,
+			state->base.state, state->base.crtc, state->base.crtc);
+	if (IS_ERR(c_st))
+		return PTR_ERR(c_st);
+
+	st = to_compiz_st(c_st);
+
+	pipeline_composition_size(state, &st->hsize, &st->vsize);
+
+	komeda_component_set_output(&dflow->input, &compiz->base, 0);
+
+	/* compiz output dflow will be fed to the next pipeline stage, prepare
+	 * the data flow configuration for the next stage
+	 */
+	if (dflow) {
+		dflow->in_w = st->hsize;
+		dflow->in_h = st->vsize;
+		dflow->out_w = dflow->in_w;
+		dflow->out_h = dflow->in_h;
+		/* the output data of compiz doesn't have alpha, it only can be
+		 * used as bottom layer when blend it with master layers
+		 */
+		dflow->pixel_blend_mode = DRM_MODE_BLEND_PIXEL_NONE;
+		dflow->layer_alpha = 0xFF;
+		dflow->blending_zorder = 0;
+	}
+
+	return 0;
+}
+
+static int
+komeda_improc_validate(struct komeda_improc *improc,
+		       struct komeda_crtc_state *kcrtc_st,
+		       struct komeda_data_flow_cfg *dflow)
+{
+	struct drm_crtc *crtc = kcrtc_st->base.crtc;
+	struct drm_crtc_state *crtc_st = &kcrtc_st->base;
+	struct komeda_component_state *c_st;
+	struct komeda_improc_state *st;
+
+	c_st = komeda_component_get_state_and_set_user(&improc->base,
+			kcrtc_st->base.state, crtc, crtc);
+	if (IS_ERR(c_st))
+		return PTR_ERR(c_st);
+
+	st = to_improc_st(c_st);
+
+	st->hsize = dflow->in_w;
+	st->vsize = dflow->in_h;
+
+	if (drm_atomic_crtc_needs_modeset(crtc_st)) {
+		u32 output_depths, output_formats;
+		u32 avail_depths, avail_formats;
+
+		komeda_crtc_get_color_config(crtc_st, &output_depths,
+					     &output_formats);
+
+		avail_depths = output_depths & improc->supported_color_depths;
+		if (avail_depths == 0) {
+			DRM_DEBUG_ATOMIC("No available color depths, conn depths: 0x%x & display: 0x%x\n",
+					 output_depths,
+					 improc->supported_color_depths);
+			return -EINVAL;
+		}
+
+		avail_formats = output_formats &
+				improc->supported_color_formats;
+		if (!avail_formats) {
+			DRM_DEBUG_ATOMIC("No available color_formats, conn formats 0x%x & display: 0x%x\n",
+					 output_formats,
+					 improc->supported_color_formats);
+			return -EINVAL;
+		}
+
+		st->color_depth = __fls(avail_depths);
+		st->color_format = BIT(__ffs(avail_formats));
+	}
+
+	if (kcrtc_st->base.color_mgmt_changed) {
+		drm_lut_to_fgamma_coeffs(kcrtc_st->base.gamma_lut,
+					 st->fgamma_coeffs);
+		drm_ctm_to_coeffs(kcrtc_st->base.ctm, st->ctm_coeffs);
+	}
+
+	komeda_component_add_input(&st->base, &dflow->input, 0);
+	komeda_component_set_output(&dflow->input, &improc->base, 0);
+
+	return 0;
+}
+
+static int
+komeda_timing_ctrlr_validate(struct komeda_timing_ctrlr *ctrlr,
+			     struct komeda_crtc_state *kcrtc_st,
+			     struct komeda_data_flow_cfg *dflow)
+{
+	struct drm_crtc *crtc = kcrtc_st->base.crtc;
+	struct komeda_timing_ctrlr_state *st;
+	struct komeda_component_state *c_st;
+
+	c_st = komeda_component_get_state_and_set_user(&ctrlr->base,
+			kcrtc_st->base.state, crtc, crtc);
+	if (IS_ERR(c_st))
+		return PTR_ERR(c_st);
+
+	st = to_ctrlr_st(c_st);
+
+	komeda_component_add_input(&st->base, &dflow->input, 0);
+	komeda_component_set_output(&dflow->input, &ctrlr->base, 0);
+
+	return 0;
+}
+
+void komeda_complete_data_flow_cfg(struct komeda_layer *layer,
+				   struct komeda_data_flow_cfg *dflow,
+				   struct drm_framebuffer *fb)
+{
+	struct komeda_scaler *scaler = layer->base.pipeline->scalers[0];
+	u32 w = dflow->in_w;
+	u32 h = dflow->in_h;
+
+	dflow->total_in_w = dflow->in_w;
+	dflow->total_in_h = dflow->in_h;
+	dflow->total_out_w = dflow->out_w;
+
+	/* if format doesn't have alpha, fix blend mode to PIXEL_NONE */
+	if (!fb->format->has_alpha)
+		dflow->pixel_blend_mode = DRM_MODE_BLEND_PIXEL_NONE;
+
+	if (drm_rotation_90_or_270(dflow->rot))
+		swap(w, h);
+
+	dflow->en_scaling = (w != dflow->out_w) || (h != dflow->out_h);
+	dflow->is_yuv = fb->format->is_yuv;
+
+	/* try to enable image enhancer if data flow is a 2x+ upscaling */
+	dflow->en_img_enhancement = dflow->out_w >= 2 * w ||
+				    dflow->out_h >= 2 * h;
+
+	/* try to enable split if scaling exceed the scaler's acceptable
+	 * input/output range.
+	 */
+	if (dflow->en_scaling && scaler)
+		dflow->en_split = !in_range(&scaler->hsize, dflow->in_w) ||
+				  !in_range(&scaler->hsize, dflow->out_w);
+}
+
+static bool merger_is_available(struct komeda_pipeline *pipe,
+				struct komeda_data_flow_cfg *dflow)
+{
+	u32 avail_inputs = pipe->merger ?
+			   pipe->merger->base.supported_inputs : 0;
+
+	return has_bit(dflow->input.component->id, avail_inputs);
+}
+
+int komeda_build_layer_data_flow(struct komeda_layer *layer,
+				 struct komeda_plane_state *kplane_st,
+				 struct komeda_crtc_state *kcrtc_st,
+				 struct komeda_data_flow_cfg *dflow)
+{
+	struct drm_plane *plane = kplane_st->base.plane;
+	struct komeda_pipeline *pipe = layer->base.pipeline;
+	int err;
+
+	DRM_DEBUG_ATOMIC("%s handling [PLANE:%d:%s]: src[x/y:%d/%d, w/h:%d/%d] disp[x/y:%d/%d, w/h:%d/%d]",
+			 layer->base.name, plane->base.id, plane->name,
+			 dflow->in_x, dflow->in_y, dflow->in_w, dflow->in_h,
+			 dflow->out_x, dflow->out_y, dflow->out_w, dflow->out_h);
+
+	err = komeda_layer_validate(layer, kplane_st, dflow);
+	if (err)
+		return err;
+
+	err = komeda_scaler_validate(plane, kcrtc_st, dflow);
+	if (err)
+		return err;
+
+	/* if split, check if can put the data flow into merger */
+	if (dflow->en_split && merger_is_available(pipe, dflow))
+		return 0;
+
+	err = komeda_compiz_set_input(pipe->compiz, kcrtc_st, dflow);
+
+	return err;
+}
+
+/*
+ * Split is introduced for workaround scaler's input/output size limitation.
+ * The idea is simple, if one scaler can not fit the requirement, use two.
+ * So split splits the big source image to two half parts (left/right) and do
+ * the scaling by two scaler separately and independently.
+ * But split also imports an edge problem in the middle of the image when
+ * scaling, to avoid it, split isn't a simple half-and-half, but add an extra
+ * pixels (overlap) to both side, after split the left/right will be:
+ * - left: [0, src_length/2 + overlap]
+ * - right: [src_length/2 - overlap, src_length]
+ * The extra overlap do eliminate the edge problem, but which may also generates
+ * unnecessary pixels when scaling, we need to crop them before scaler output
+ * the result to the next stage. and for the how to crop, it depends on the
+ * unneeded pixels, another words the position where overlay has been added.
+ * - left: crop the right
+ * - right: crop the left
+ *
+ * The diagram for how to do the split
+ *
+ *  <---------------------left->out_w ---------------->
+ * |--------------------------------|---right_crop-----| <- left after split
+ *  \                                \                /
+ *   \                                \<--overlap--->/
+ *   |-----------------|-------------|(Middle)------|-----------------| <- src
+ *                     /<---overlap--->\                               \
+ *                    /                 \                               \
+ * right after split->|-----left_crop---|--------------------------------|
+ *                    ^<------------------- right->out_w --------------->^
+ *
+ * NOTE: To consistent with HW the output_w always contains the crop size.
+ */
+
+static void komeda_split_data_flow(struct komeda_scaler *scaler,
+				   struct komeda_data_flow_cfg *dflow,
+				   struct komeda_data_flow_cfg *l_dflow,
+				   struct komeda_data_flow_cfg *r_dflow)
+{
+	bool r90 = drm_rotation_90_or_270(dflow->rot);
+	bool flip_h = has_flip_h(dflow->rot);
+	u32 l_out, r_out, overlap;
+
+	memcpy(l_dflow, dflow, sizeof(*dflow));
+	memcpy(r_dflow, dflow, sizeof(*dflow));
+
+	l_dflow->right_part = false;
+	r_dflow->right_part = true;
+	r_dflow->blending_zorder = dflow->blending_zorder + 1;
+
+	overlap = 0;
+	if (dflow->en_scaling && scaler)
+		overlap += scaler->scaling_split_overlap;
+
+	/* original dflow may fed into splitter, and which doesn't need
+	 * enhancement overlap
+	 */
+	dflow->overlap = overlap;
+
+	if (dflow->en_img_enhancement && scaler)
+		overlap += scaler->enh_split_overlap;
+
+	l_dflow->overlap = overlap;
+	r_dflow->overlap = overlap;
+
+	/* split the origin content */
+	/* left/right here always means the left/right part of display image,
+	 * not the source Image
+	 */
+	/* DRM rotation is anti-clockwise */
+	if (r90) {
+		if (dflow->en_scaling) {
+			l_dflow->in_h = ALIGN(dflow->in_h, 2) / 2 + l_dflow->overlap;
+			r_dflow->in_h = l_dflow->in_h;
+		} else if (dflow->en_img_enhancement) {
+			/* enhancer only */
+			l_dflow->in_h = ALIGN(dflow->in_h, 2) / 2 + l_dflow->overlap;
+			r_dflow->in_h = dflow->in_h / 2 + r_dflow->overlap;
+		} else {
+			/* split without scaler, no overlap */
+			l_dflow->in_h = ALIGN(((dflow->in_h + 1) >> 1), 2);
+			r_dflow->in_h = dflow->in_h - l_dflow->in_h;
+		}
+
+		/* Consider YUV format, after split, the split source w/h
+		 * may not aligned to 2. we have two choices for such case.
+		 * 1. scaler is enabled (overlap != 0), we can do a alignment
+		 *    both left/right and crop the extra data by scaler.
+		 * 2. scaler is not enabled, only align the split left
+		 *    src/disp, and the rest part assign to right
+		 */
+		if ((overlap != 0) && dflow->is_yuv) {
+			l_dflow->in_h = ALIGN(l_dflow->in_h, 2);
+			r_dflow->in_h = ALIGN(r_dflow->in_h, 2);
+		}
+
+		if (flip_h)
+			l_dflow->in_y = dflow->in_y + dflow->in_h - l_dflow->in_h;
+		else
+			r_dflow->in_y = dflow->in_y + dflow->in_h - r_dflow->in_h;
+	} else {
+		if (dflow->en_scaling) {
+			l_dflow->in_w = ALIGN(dflow->in_w, 2) / 2 + l_dflow->overlap;
+			r_dflow->in_w = l_dflow->in_w;
+		} else if (dflow->en_img_enhancement) {
+			l_dflow->in_w = ALIGN(dflow->in_w, 2) / 2 + l_dflow->overlap;
+			r_dflow->in_w = dflow->in_w / 2 + r_dflow->overlap;
+		} else {
+			l_dflow->in_w = ALIGN(((dflow->in_w + 1) >> 1), 2);
+			r_dflow->in_w = dflow->in_w - l_dflow->in_w;
+		}
+
+		/* do YUV alignment when scaler enabled */
+		if ((overlap != 0) && dflow->is_yuv) {
+			l_dflow->in_w = ALIGN(l_dflow->in_w, 2);
+			r_dflow->in_w = ALIGN(r_dflow->in_w, 2);
+		}
+
+		/* on flip_h, the left display content from the right-source */
+		if (flip_h)
+			l_dflow->in_x = dflow->in_w + dflow->in_x - l_dflow->in_w;
+		else
+			r_dflow->in_x = dflow->in_w + dflow->in_x - r_dflow->in_w;
+	}
+
+	/* split the disp_rect */
+	if (dflow->en_scaling || dflow->en_img_enhancement)
+		l_dflow->out_w = ((dflow->out_w + 1) >> 1);
+	else
+		l_dflow->out_w = ALIGN(((dflow->out_w + 1) >> 1), 2);
+
+	r_dflow->out_w = dflow->out_w - l_dflow->out_w;
+
+	l_dflow->out_x = dflow->out_x;
+	r_dflow->out_x = l_dflow->out_w + l_dflow->out_x;
+
+	/* calculate the scaling crop */
+	/* left scaler output more data and do crop */
+	if (r90) {
+		l_out = (dflow->out_w * l_dflow->in_h) / dflow->in_h;
+		r_out = (dflow->out_w * r_dflow->in_h) / dflow->in_h;
+	} else {
+		l_out = (dflow->out_w * l_dflow->in_w) / dflow->in_w;
+		r_out = (dflow->out_w * r_dflow->in_w) / dflow->in_w;
+	}
+
+	l_dflow->left_crop  = 0;
+	l_dflow->right_crop = l_out - l_dflow->out_w;
+	r_dflow->left_crop  = r_out - r_dflow->out_w;
+	r_dflow->right_crop = 0;
+
+	/* out_w includes the crop length */
+	l_dflow->out_w += l_dflow->right_crop + l_dflow->left_crop;
+	r_dflow->out_w += r_dflow->right_crop + r_dflow->left_crop;
+}
+
+/* For layer split, a plane state will be split to two data flows and handled
+ * by two separated komeda layer input pipelines. komeda supports two types of
+ * layer split:
+ * - none-scaling split:
+ *             / layer-left -> \
+ * plane_state                  compiz-> ...
+ *             \ layer-right-> /
+ *
+ * - scaling split:
+ *             / layer-left -> scaler->\
+ * plane_state                          merger -> compiz-> ...
+ *             \ layer-right-> scaler->/
+ *
+ * Since merger only supports scaler as input, so for none-scaling split, two
+ * layer data flows will be output to compiz directly. for scaling_split, two
+ * data flow will be merged by merger firstly, then merger outputs one merged
+ * data flow to compiz.
+ */
+int komeda_build_layer_split_data_flow(struct komeda_layer *left,
+				       struct komeda_plane_state *kplane_st,
+				       struct komeda_crtc_state *kcrtc_st,
+				       struct komeda_data_flow_cfg *dflow)
+{
+	struct drm_plane *plane = kplane_st->base.plane;
+	struct komeda_pipeline *pipe = left->base.pipeline;
+	struct komeda_layer *right = left->right;
+	struct komeda_data_flow_cfg l_dflow, r_dflow;
+	int err;
+
+	komeda_split_data_flow(pipe->scalers[0], dflow, &l_dflow, &r_dflow);
+
+	DRM_DEBUG_ATOMIC("Assign %s + %s to [PLANE:%d:%s]: "
+			 "src[x/y:%d/%d, w/h:%d/%d] disp[x/y:%d/%d, w/h:%d/%d]",
+			 left->base.name, right->base.name,
+			 plane->base.id, plane->name,
+			 dflow->in_x, dflow->in_y, dflow->in_w, dflow->in_h,
+			 dflow->out_x, dflow->out_y, dflow->out_w, dflow->out_h);
+
+	err = komeda_build_layer_data_flow(left, kplane_st, kcrtc_st, &l_dflow);
+	if (err)
+		return err;
+
+	err = komeda_build_layer_data_flow(right, kplane_st, kcrtc_st, &r_dflow);
+	if (err)
+		return err;
+
+	/* The rotation has been handled by layer, so adjusted the data flow */
+	komeda_rotate_data_flow(dflow, dflow->rot);
+
+	/* left and right dflow has been merged to compiz already,
+	 * no need merger to merge them anymore.
+	 */
+	if (r_dflow.input.component == l_dflow.input.component)
+		return 0;
+
+	/* line merger path */
+	err = komeda_merger_validate(pipe->merger, plane, kcrtc_st,
+				     &l_dflow, &r_dflow, dflow);
+	if (err)
+		return err;
+
+	err = komeda_compiz_set_input(pipe->compiz, kcrtc_st, dflow);
+
+	return err;
+}
+
+/* writeback data path: compiz -> scaler -> wb_layer -> memory */
+int komeda_build_wb_data_flow(struct komeda_layer *wb_layer,
+			      struct drm_connector_state *conn_st,
+			      struct komeda_crtc_state *kcrtc_st,
+			      struct komeda_data_flow_cfg *dflow)
+{
+	struct drm_connector *conn = conn_st->connector;
+	int err;
+
+	err = komeda_scaler_validate(conn, kcrtc_st, dflow);
+	if (err)
+		return err;
+
+	return komeda_wb_layer_validate(wb_layer, conn_st, dflow);
+}
+
+/* writeback scaling split data path:
+ *                   /-> scaler ->\
+ * compiz -> splitter              merger -> wb_layer -> memory
+ *                   \-> scaler ->/
+ */
+int komeda_build_wb_split_data_flow(struct komeda_layer *wb_layer,
+				    struct drm_connector_state *conn_st,
+				    struct komeda_crtc_state *kcrtc_st,
+				    struct komeda_data_flow_cfg *dflow)
+{
+	struct komeda_pipeline *pipe = wb_layer->base.pipeline;
+	struct drm_connector *conn = conn_st->connector;
+	struct komeda_data_flow_cfg l_dflow, r_dflow;
+	int err;
+
+	err = komeda_splitter_validate(pipe->splitter, conn_st,
+				       dflow, &l_dflow, &r_dflow);
+	if (err)
+		return err;
+	err = komeda_scaler_validate(conn, kcrtc_st, &l_dflow);
+	if (err)
+		return err;
+
+	err = komeda_scaler_validate(conn, kcrtc_st, &r_dflow);
+	if (err)
+		return err;
+
+	err = komeda_merger_validate(pipe->merger, conn_st, kcrtc_st,
+				     &l_dflow, &r_dflow, dflow);
+	if (err)
+		return err;
+
+	return komeda_wb_layer_validate(wb_layer, conn_st, dflow);
+}
+
+/* build display output data flow, the data path is:
+ * compiz -> improc -> timing_ctrlr
+ */
+int komeda_build_display_data_flow(struct komeda_crtc *kcrtc,
+				   struct komeda_crtc_state *kcrtc_st)
+{
+	struct komeda_pipeline *master = kcrtc->master;
+	struct komeda_pipeline *slave  = kcrtc->slave;
+	struct komeda_data_flow_cfg m_dflow; /* master data flow */
+	struct komeda_data_flow_cfg s_dflow; /* slave data flow */
+	int err;
+
+	memset(&m_dflow, 0, sizeof(m_dflow));
+	memset(&s_dflow, 0, sizeof(s_dflow));
+
+	if (slave && has_bit(slave->id, kcrtc_st->active_pipes)) {
+		err = komeda_compiz_validate(slave->compiz, kcrtc_st, &s_dflow);
+		if (err)
+			return err;
+
+		/* merge the slave dflow into master pipeline */
+		err = komeda_compiz_set_input(master->compiz, kcrtc_st,
+					      &s_dflow);
+		if (err)
+			return err;
+	}
+
+	err = komeda_compiz_validate(master->compiz, kcrtc_st, &m_dflow);
+	if (err)
+		return err;
+
+	err = komeda_improc_validate(master->improc, kcrtc_st, &m_dflow);
+	if (err)
+		return err;
+
+	err = komeda_timing_ctrlr_validate(master->ctrlr, kcrtc_st, &m_dflow);
+	if (err)
+		return err;
+
+	return 0;
+}
+
+static int
+komeda_pipeline_unbound_components(struct komeda_pipeline *pipe,
+				   struct komeda_pipeline_state *new)
+{
+	struct drm_atomic_state *drm_st = new->obj.state;
+	struct komeda_pipeline_state *old = priv_to_pipe_st(pipe->obj.state);
+	struct komeda_component_state *c_st;
+	struct komeda_component *c;
+	u32 id;
+	unsigned long disabling_comps;
+
+	WARN_ON(!old);
+
+	disabling_comps = (~new->active_comps) & old->active_comps;
+
+	/* unbound all disabling component */
+	for_each_set_bit(id, &disabling_comps, 32) {
+		c = komeda_pipeline_get_component(pipe, id);
+		c_st = komeda_component_get_state_and_set_user(c,
+				drm_st, NULL, new->crtc);
+		if (PTR_ERR(c_st) == -EDEADLK)
+			return -EDEADLK;
+		WARN_ON(IS_ERR(c_st));
+	}
+
+	return 0;
+}
+
+/* release unclaimed pipeline resource */
+int komeda_release_unclaimed_resources(struct komeda_pipeline *pipe,
+				       struct komeda_crtc_state *kcrtc_st)
+{
+	struct drm_atomic_state *drm_st = kcrtc_st->base.state;
+	struct komeda_pipeline_state *st;
+
+	/* ignore the pipeline which is not affected */
+	if (!pipe || !has_bit(pipe->id, kcrtc_st->affected_pipes))
+		return 0;
+
+	if (has_bit(pipe->id, kcrtc_st->active_pipes))
+		st = komeda_pipeline_get_new_state(pipe, drm_st);
+	else
+		st = komeda_pipeline_get_state_and_set_crtc(pipe, drm_st, NULL);
+
+	if (WARN_ON(IS_ERR_OR_NULL(st)))
+		return -EINVAL;
+
+	return komeda_pipeline_unbound_components(pipe, st);
+
+}
+
+/* Since standalong disabled components must be disabled separately and in the
+ * last, So a complete disable operation may needs to call pipeline_disable
+ * twice (two phase disabling).
+ * Phase 1: disable the common components, flush it.
+ * Phase 2: disable the standalone disabled components, flush it.
+ *
+ * RETURNS:
+ * true: disable is not complete, needs a phase 2 disable.
+ * false: disable is complete.
+ */
+bool komeda_pipeline_disable(struct komeda_pipeline *pipe,
+			     struct drm_atomic_state *old_state)
+{
+	struct komeda_pipeline_state *old;
+	struct komeda_component *c;
+	struct komeda_component_state *c_st;
+	u32 id;
+	unsigned long disabling_comps;
+
+	old = komeda_pipeline_get_old_state(pipe, old_state);
+
+	disabling_comps = old->active_comps &
+			  (~pipe->standalone_disabled_comps);
+	if (!disabling_comps)
+		disabling_comps = old->active_comps &
+				  pipe->standalone_disabled_comps;
+
+	DRM_DEBUG_ATOMIC("PIPE%d: active_comps: 0x%x, disabling_comps: 0x%lx.\n",
+			 pipe->id, old->active_comps, disabling_comps);
+
+	for_each_set_bit(id, &disabling_comps, 32) {
+		c = komeda_pipeline_get_component(pipe, id);
+		c_st = priv_to_comp_st(c->obj.state);
+
+		/*
+		 * If we disabled a component then all active_inputs should be
+		 * put in the list of changed_active_inputs, so they get
+		 * re-enabled.
+		 * This usually happens during a modeset when the pipeline is
+		 * first disabled and then the actual state gets committed
+		 * again.
+		 */
+		c_st->changed_active_inputs |= c_st->active_inputs;
+
+		c->funcs->disable(c);
+	}
+
+	/* Update the pipeline state, if there are components that are still
+	 * active, return true for calling the phase 2 disable.
+	 */
+	old->active_comps &= ~disabling_comps;
+
+	return old->active_comps ? true : false;
+}
+
+void komeda_pipeline_update(struct komeda_pipeline *pipe,
+			    struct drm_atomic_state *old_state)
+{
+	struct komeda_pipeline_state *new = priv_to_pipe_st(pipe->obj.state);
+	struct komeda_pipeline_state *old;
+	struct komeda_component *c;
+	u32 id;
+	unsigned long changed_comps;
+
+	old = komeda_pipeline_get_old_state(pipe, old_state);
+
+	changed_comps = new->active_comps | old->active_comps;
+
+	DRM_DEBUG_ATOMIC("PIPE%d: active_comps: 0x%x, changed: 0x%lx.\n",
+			 pipe->id, new->active_comps, changed_comps);
+
+	for_each_set_bit(id, &changed_comps, 32) {
+		c = komeda_pipeline_get_component(pipe, id);
+
+		if (new->active_comps & BIT(c->id))
+			c->funcs->update(c, priv_to_comp_st(c->obj.state));
+		else
+			c->funcs->disable(c);
+	}
+}
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_plane.c b/drivers/gpu/drm/arm/display/komeda/komeda_plane.c
new file mode 100644
index 000000000..bc3f42e91
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_plane.c
@@ -0,0 +1,343 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_plane_helper.h>
+#include <drm/drm_print.h>
+#include "komeda_dev.h"
+#include "komeda_kms.h"
+#include "komeda_framebuffer.h"
+
+static int
+komeda_plane_init_data_flow(struct drm_plane_state *st,
+			    struct komeda_crtc_state *kcrtc_st,
+			    struct komeda_data_flow_cfg *dflow)
+{
+	struct komeda_plane *kplane = to_kplane(st->plane);
+	struct drm_framebuffer *fb = st->fb;
+	const struct komeda_format_caps *caps = to_kfb(fb)->format_caps;
+	struct komeda_pipeline *pipe = kplane->layer->base.pipeline;
+
+	memset(dflow, 0, sizeof(*dflow));
+
+	dflow->blending_zorder = st->normalized_zpos;
+	if (pipe == to_kcrtc(st->crtc)->master)
+		dflow->blending_zorder -= kcrtc_st->max_slave_zorder;
+	if (dflow->blending_zorder < 0) {
+		DRM_DEBUG_ATOMIC("%s zorder:%d < max_slave_zorder: %d.\n",
+				 st->plane->name, st->normalized_zpos,
+				 kcrtc_st->max_slave_zorder);
+		return -EINVAL;
+	}
+
+	dflow->pixel_blend_mode = st->pixel_blend_mode;
+	dflow->layer_alpha = st->alpha >> 8;
+
+	dflow->out_x = st->crtc_x;
+	dflow->out_y = st->crtc_y;
+	dflow->out_w = st->crtc_w;
+	dflow->out_h = st->crtc_h;
+
+	dflow->in_x = st->src_x >> 16;
+	dflow->in_y = st->src_y >> 16;
+	dflow->in_w = st->src_w >> 16;
+	dflow->in_h = st->src_h >> 16;
+
+	dflow->rot = drm_rotation_simplify(st->rotation, caps->supported_rots);
+	if (!has_bits(dflow->rot, caps->supported_rots)) {
+		DRM_DEBUG_ATOMIC("rotation(0x%x) isn't supported by %s.\n",
+				 dflow->rot,
+				 komeda_get_format_name(caps->fourcc,
+							fb->modifier));
+		return -EINVAL;
+	}
+
+	komeda_complete_data_flow_cfg(kplane->layer, dflow, fb);
+
+	return 0;
+}
+
+/**
+ * komeda_plane_atomic_check - build input data flow
+ * @plane: DRM plane
+ * @state: the plane state object
+ *
+ * RETURNS:
+ * Zero for success or -errno
+ */
+static int
+komeda_plane_atomic_check(struct drm_plane *plane,
+			  struct drm_plane_state *state)
+{
+	struct komeda_plane *kplane = to_kplane(plane);
+	struct komeda_plane_state *kplane_st = to_kplane_st(state);
+	struct komeda_layer *layer = kplane->layer;
+	struct drm_crtc_state *crtc_st;
+	struct komeda_crtc_state *kcrtc_st;
+	struct komeda_data_flow_cfg dflow;
+	int err;
+
+	if (!state->crtc || !state->fb)
+		return 0;
+
+	crtc_st = drm_atomic_get_crtc_state(state->state, state->crtc);
+	if (IS_ERR(crtc_st) || !crtc_st->enable) {
+		DRM_DEBUG_ATOMIC("Cannot update plane on a disabled CRTC.\n");
+		return -EINVAL;
+	}
+
+	/* crtc is inactive, skip the resource assignment */
+	if (!crtc_st->active)
+		return 0;
+
+	kcrtc_st = to_kcrtc_st(crtc_st);
+
+	err = komeda_plane_init_data_flow(state, kcrtc_st, &dflow);
+	if (err)
+		return err;
+
+	if (dflow.en_split)
+		err = komeda_build_layer_split_data_flow(layer,
+				kplane_st, kcrtc_st, &dflow);
+	else
+		err = komeda_build_layer_data_flow(layer,
+				kplane_st, kcrtc_st, &dflow);
+
+	return err;
+}
+
+/* plane doesn't represent a real HW, so there is no HW update for plane.
+ * komeda handles all the HW update in crtc->atomic_flush
+ */
+static void
+komeda_plane_atomic_update(struct drm_plane *plane,
+			   struct drm_plane_state *old_state)
+{
+}
+
+static const struct drm_plane_helper_funcs komeda_plane_helper_funcs = {
+	.atomic_check	= komeda_plane_atomic_check,
+	.atomic_update	= komeda_plane_atomic_update,
+};
+
+static void komeda_plane_destroy(struct drm_plane *plane)
+{
+	drm_plane_cleanup(plane);
+
+	kfree(to_kplane(plane));
+}
+
+static void komeda_plane_reset(struct drm_plane *plane)
+{
+	struct komeda_plane_state *state;
+	struct komeda_plane *kplane = to_kplane(plane);
+
+	if (plane->state)
+		__drm_atomic_helper_plane_destroy_state(plane->state);
+
+	kfree(plane->state);
+	plane->state = NULL;
+
+	state = kzalloc(sizeof(*state), GFP_KERNEL);
+	if (state) {
+		state->base.rotation = DRM_MODE_ROTATE_0;
+		state->base.pixel_blend_mode = DRM_MODE_BLEND_PREMULTI;
+		state->base.alpha = DRM_BLEND_ALPHA_OPAQUE;
+		state->base.zpos = kplane->layer->base.id;
+		state->base.color_encoding = DRM_COLOR_YCBCR_BT601;
+		state->base.color_range = DRM_COLOR_YCBCR_LIMITED_RANGE;
+		plane->state = &state->base;
+		plane->state->plane = plane;
+	}
+}
+
+static struct drm_plane_state *
+komeda_plane_atomic_duplicate_state(struct drm_plane *plane)
+{
+	struct komeda_plane_state *new;
+
+	if (WARN_ON(!plane->state))
+		return NULL;
+
+	new = kzalloc(sizeof(*new), GFP_KERNEL);
+	if (!new)
+		return NULL;
+
+	__drm_atomic_helper_plane_duplicate_state(plane, &new->base);
+
+	return &new->base;
+}
+
+static void
+komeda_plane_atomic_destroy_state(struct drm_plane *plane,
+				  struct drm_plane_state *state)
+{
+	__drm_atomic_helper_plane_destroy_state(state);
+	kfree(to_kplane_st(state));
+}
+
+static bool
+komeda_plane_format_mod_supported(struct drm_plane *plane,
+				  u32 format, u64 modifier)
+{
+	struct komeda_dev *mdev = plane->dev->dev_private;
+	struct komeda_plane *kplane = to_kplane(plane);
+	u32 layer_type = kplane->layer->layer_type;
+
+	return komeda_format_mod_supported(&mdev->fmt_tbl, layer_type,
+					   format, modifier, 0);
+}
+
+static const struct drm_plane_funcs komeda_plane_funcs = {
+	.update_plane		= drm_atomic_helper_update_plane,
+	.disable_plane		= drm_atomic_helper_disable_plane,
+	.destroy		= komeda_plane_destroy,
+	.reset			= komeda_plane_reset,
+	.atomic_duplicate_state	= komeda_plane_atomic_duplicate_state,
+	.atomic_destroy_state	= komeda_plane_atomic_destroy_state,
+	.format_mod_supported	= komeda_plane_format_mod_supported,
+};
+
+/* for komeda, which is pipeline can be share between crtcs */
+static u32 get_possible_crtcs(struct komeda_kms_dev *kms,
+			      struct komeda_pipeline *pipe)
+{
+	struct komeda_crtc *crtc;
+	u32 possible_crtcs = 0;
+	int i;
+
+	for (i = 0; i < kms->n_crtcs; i++) {
+		crtc = &kms->crtcs[i];
+
+		if ((pipe == crtc->master) || (pipe == crtc->slave))
+			possible_crtcs |= BIT(i);
+	}
+
+	return possible_crtcs;
+}
+
+static void
+komeda_set_crtc_plane_mask(struct komeda_kms_dev *kms,
+			   struct komeda_pipeline *pipe,
+			   struct drm_plane *plane)
+{
+	struct komeda_crtc *kcrtc;
+	int i;
+
+	for (i = 0; i < kms->n_crtcs; i++) {
+		kcrtc = &kms->crtcs[i];
+
+		if (pipe == kcrtc->slave)
+			kcrtc->slave_planes |= BIT(drm_plane_index(plane));
+	}
+}
+
+/* use Layer0 as primary */
+static u32 get_plane_type(struct komeda_kms_dev *kms,
+			  struct komeda_component *c)
+{
+	bool is_primary = (c->id == KOMEDA_COMPONENT_LAYER0);
+
+	return is_primary ? DRM_PLANE_TYPE_PRIMARY : DRM_PLANE_TYPE_OVERLAY;
+}
+
+static int komeda_plane_add(struct komeda_kms_dev *kms,
+			    struct komeda_layer *layer)
+{
+	struct komeda_dev *mdev = kms->base.dev_private;
+	struct komeda_component *c = &layer->base;
+	struct komeda_plane *kplane;
+	struct drm_plane *plane;
+	u32 *formats, n_formats = 0;
+	int err;
+
+	kplane = kzalloc(sizeof(*kplane), GFP_KERNEL);
+	if (!kplane)
+		return -ENOMEM;
+
+	plane = &kplane->base;
+	kplane->layer = layer;
+
+	formats = komeda_get_layer_fourcc_list(&mdev->fmt_tbl,
+					       layer->layer_type, &n_formats);
+	if (!formats) {
+		kfree(kplane);
+		return -ENOMEM;
+	}
+
+	err = drm_universal_plane_init(&kms->base, plane,
+			get_possible_crtcs(kms, c->pipeline),
+			&komeda_plane_funcs,
+			formats, n_formats, komeda_supported_modifiers,
+			get_plane_type(kms, c),
+			"%s", c->name);
+
+	komeda_put_fourcc_list(formats);
+
+	if (err) {
+		kfree(kplane);
+		return err;
+	}
+
+	drm_plane_helper_add(plane, &komeda_plane_helper_funcs);
+
+	err = drm_plane_create_rotation_property(plane, DRM_MODE_ROTATE_0,
+						 layer->supported_rots);
+	if (err)
+		goto cleanup;
+
+	err = drm_plane_create_alpha_property(plane);
+	if (err)
+		goto cleanup;
+
+	err = drm_plane_create_blend_mode_property(plane,
+			BIT(DRM_MODE_BLEND_PIXEL_NONE) |
+			BIT(DRM_MODE_BLEND_PREMULTI)   |
+			BIT(DRM_MODE_BLEND_COVERAGE));
+	if (err)
+		goto cleanup;
+
+	err = drm_plane_create_color_properties(plane,
+			BIT(DRM_COLOR_YCBCR_BT601) |
+			BIT(DRM_COLOR_YCBCR_BT709) |
+			BIT(DRM_COLOR_YCBCR_BT2020),
+			BIT(DRM_COLOR_YCBCR_LIMITED_RANGE) |
+			BIT(DRM_COLOR_YCBCR_FULL_RANGE),
+			DRM_COLOR_YCBCR_BT601,
+			DRM_COLOR_YCBCR_LIMITED_RANGE);
+	if (err)
+		goto cleanup;
+
+	err = drm_plane_create_zpos_property(plane, layer->base.id, 0, 8);
+	if (err)
+		goto cleanup;
+
+	komeda_set_crtc_plane_mask(kms, c->pipeline, plane);
+
+	return 0;
+cleanup:
+	komeda_plane_destroy(plane);
+	return err;
+}
+
+int komeda_kms_add_planes(struct komeda_kms_dev *kms, struct komeda_dev *mdev)
+{
+	struct komeda_pipeline *pipe;
+	int i, j, err;
+
+	for (i = 0; i < mdev->n_pipelines; i++) {
+		pipe = mdev->pipelines[i];
+
+		for (j = 0; j < pipe->n_layers; j++) {
+			err = komeda_plane_add(kms, pipe->layers[j]);
+			if (err)
+				return err;
+		}
+	}
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_private_obj.c b/drivers/gpu/drm/arm/display/komeda/komeda_private_obj.c
new file mode 100644
index 000000000..914400c4a
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_private_obj.c
@@ -0,0 +1,432 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#include "komeda_dev.h"
+#include "komeda_kms.h"
+
+static void
+komeda_component_state_reset(struct komeda_component_state *st)
+{
+	st->binding_user = NULL;
+	st->affected_inputs = st->active_inputs;
+	st->active_inputs = 0;
+	st->changed_active_inputs = 0;
+}
+
+static struct drm_private_state *
+komeda_layer_atomic_duplicate_state(struct drm_private_obj *obj)
+{
+	struct komeda_layer_state *st;
+
+	st = kmemdup(obj->state, sizeof(*st), GFP_KERNEL);
+	if (!st)
+		return NULL;
+
+	komeda_component_state_reset(&st->base);
+	__drm_atomic_helper_private_obj_duplicate_state(obj, &st->base.obj);
+
+	return &st->base.obj;
+}
+
+static void
+komeda_layer_atomic_destroy_state(struct drm_private_obj *obj,
+				  struct drm_private_state *state)
+{
+	struct komeda_layer_state *st = to_layer_st(priv_to_comp_st(state));
+
+	kfree(st);
+}
+
+static const struct drm_private_state_funcs komeda_layer_obj_funcs = {
+	.atomic_duplicate_state	= komeda_layer_atomic_duplicate_state,
+	.atomic_destroy_state	= komeda_layer_atomic_destroy_state,
+};
+
+static int komeda_layer_obj_add(struct komeda_kms_dev *kms,
+				struct komeda_layer *layer)
+{
+	struct komeda_layer_state *st;
+
+	st = kzalloc(sizeof(*st), GFP_KERNEL);
+	if (!st)
+		return -ENOMEM;
+
+	st->base.component = &layer->base;
+	drm_atomic_private_obj_init(&kms->base, &layer->base.obj, &st->base.obj,
+				    &komeda_layer_obj_funcs);
+	return 0;
+}
+
+static struct drm_private_state *
+komeda_scaler_atomic_duplicate_state(struct drm_private_obj *obj)
+{
+	struct komeda_scaler_state *st;
+
+	st = kmemdup(obj->state, sizeof(*st), GFP_KERNEL);
+	if (!st)
+		return NULL;
+
+	komeda_component_state_reset(&st->base);
+	__drm_atomic_helper_private_obj_duplicate_state(obj, &st->base.obj);
+
+	return &st->base.obj;
+}
+
+static void
+komeda_scaler_atomic_destroy_state(struct drm_private_obj *obj,
+				   struct drm_private_state *state)
+{
+	kfree(to_scaler_st(priv_to_comp_st(state)));
+}
+
+static const struct drm_private_state_funcs komeda_scaler_obj_funcs = {
+	.atomic_duplicate_state	= komeda_scaler_atomic_duplicate_state,
+	.atomic_destroy_state	= komeda_scaler_atomic_destroy_state,
+};
+
+static int komeda_scaler_obj_add(struct komeda_kms_dev *kms,
+				 struct komeda_scaler *scaler)
+{
+	struct komeda_scaler_state *st;
+
+	st = kzalloc(sizeof(*st), GFP_KERNEL);
+	if (!st)
+		return -ENOMEM;
+
+	st->base.component = &scaler->base;
+	drm_atomic_private_obj_init(&kms->base,
+				    &scaler->base.obj, &st->base.obj,
+				    &komeda_scaler_obj_funcs);
+	return 0;
+}
+
+static struct drm_private_state *
+komeda_compiz_atomic_duplicate_state(struct drm_private_obj *obj)
+{
+	struct komeda_compiz_state *st;
+
+	st = kmemdup(obj->state, sizeof(*st), GFP_KERNEL);
+	if (!st)
+		return NULL;
+
+	komeda_component_state_reset(&st->base);
+	__drm_atomic_helper_private_obj_duplicate_state(obj, &st->base.obj);
+
+	return &st->base.obj;
+}
+
+static void
+komeda_compiz_atomic_destroy_state(struct drm_private_obj *obj,
+				   struct drm_private_state *state)
+{
+	kfree(to_compiz_st(priv_to_comp_st(state)));
+}
+
+static const struct drm_private_state_funcs komeda_compiz_obj_funcs = {
+	.atomic_duplicate_state	= komeda_compiz_atomic_duplicate_state,
+	.atomic_destroy_state	= komeda_compiz_atomic_destroy_state,
+};
+
+static int komeda_compiz_obj_add(struct komeda_kms_dev *kms,
+				 struct komeda_compiz *compiz)
+{
+	struct komeda_compiz_state *st;
+
+	st = kzalloc(sizeof(*st), GFP_KERNEL);
+	if (!st)
+		return -ENOMEM;
+
+	st->base.component = &compiz->base;
+	drm_atomic_private_obj_init(&kms->base, &compiz->base.obj, &st->base.obj,
+				    &komeda_compiz_obj_funcs);
+
+	return 0;
+}
+
+static struct drm_private_state *
+komeda_splitter_atomic_duplicate_state(struct drm_private_obj *obj)
+{
+	struct komeda_splitter_state *st;
+
+	st = kmemdup(obj->state, sizeof(*st), GFP_KERNEL);
+	if (!st)
+		return NULL;
+
+	komeda_component_state_reset(&st->base);
+	__drm_atomic_helper_private_obj_duplicate_state(obj, &st->base.obj);
+
+	return &st->base.obj;
+}
+
+static void
+komeda_splitter_atomic_destroy_state(struct drm_private_obj *obj,
+				     struct drm_private_state *state)
+{
+	kfree(to_splitter_st(priv_to_comp_st(state)));
+}
+
+static const struct drm_private_state_funcs komeda_splitter_obj_funcs = {
+	.atomic_duplicate_state	= komeda_splitter_atomic_duplicate_state,
+	.atomic_destroy_state	= komeda_splitter_atomic_destroy_state,
+};
+
+static int komeda_splitter_obj_add(struct komeda_kms_dev *kms,
+				   struct komeda_splitter *splitter)
+{
+	struct komeda_splitter_state *st;
+
+	st = kzalloc(sizeof(*st), GFP_KERNEL);
+	if (!st)
+		return -ENOMEM;
+
+	st->base.component = &splitter->base;
+	drm_atomic_private_obj_init(&kms->base,
+				    &splitter->base.obj, &st->base.obj,
+				    &komeda_splitter_obj_funcs);
+
+	return 0;
+}
+
+static struct drm_private_state *
+komeda_merger_atomic_duplicate_state(struct drm_private_obj *obj)
+{
+	struct komeda_merger_state *st;
+
+	st = kmemdup(obj->state, sizeof(*st), GFP_KERNEL);
+	if (!st)
+		return NULL;
+
+	komeda_component_state_reset(&st->base);
+	__drm_atomic_helper_private_obj_duplicate_state(obj, &st->base.obj);
+
+	return &st->base.obj;
+}
+
+static void komeda_merger_atomic_destroy_state(struct drm_private_obj *obj,
+					       struct drm_private_state *state)
+{
+	kfree(to_merger_st(priv_to_comp_st(state)));
+}
+
+static const struct drm_private_state_funcs komeda_merger_obj_funcs = {
+	.atomic_duplicate_state	= komeda_merger_atomic_duplicate_state,
+	.atomic_destroy_state	= komeda_merger_atomic_destroy_state,
+};
+
+static int komeda_merger_obj_add(struct komeda_kms_dev *kms,
+				 struct komeda_merger *merger)
+{
+	struct komeda_merger_state *st;
+
+	st = kzalloc(sizeof(*st), GFP_KERNEL);
+	if (!st)
+		return -ENOMEM;
+
+	st->base.component = &merger->base;
+	drm_atomic_private_obj_init(&kms->base,
+				    &merger->base.obj, &st->base.obj,
+				    &komeda_merger_obj_funcs);
+
+	return 0;
+}
+
+static struct drm_private_state *
+komeda_improc_atomic_duplicate_state(struct drm_private_obj *obj)
+{
+	struct komeda_improc_state *st;
+
+	st = kmemdup(obj->state, sizeof(*st), GFP_KERNEL);
+	if (!st)
+		return NULL;
+
+	komeda_component_state_reset(&st->base);
+	__drm_atomic_helper_private_obj_duplicate_state(obj, &st->base.obj);
+
+	return &st->base.obj;
+}
+
+static void
+komeda_improc_atomic_destroy_state(struct drm_private_obj *obj,
+				   struct drm_private_state *state)
+{
+	kfree(to_improc_st(priv_to_comp_st(state)));
+}
+
+static const struct drm_private_state_funcs komeda_improc_obj_funcs = {
+	.atomic_duplicate_state	= komeda_improc_atomic_duplicate_state,
+	.atomic_destroy_state	= komeda_improc_atomic_destroy_state,
+};
+
+static int komeda_improc_obj_add(struct komeda_kms_dev *kms,
+				 struct komeda_improc *improc)
+{
+	struct komeda_improc_state *st;
+
+	st = kzalloc(sizeof(*st), GFP_KERNEL);
+	if (!st)
+		return -ENOMEM;
+
+	st->base.component = &improc->base;
+	drm_atomic_private_obj_init(&kms->base, &improc->base.obj, &st->base.obj,
+				    &komeda_improc_obj_funcs);
+
+	return 0;
+}
+
+static struct drm_private_state *
+komeda_timing_ctrlr_atomic_duplicate_state(struct drm_private_obj *obj)
+{
+	struct komeda_timing_ctrlr_state *st;
+
+	st = kmemdup(obj->state, sizeof(*st), GFP_KERNEL);
+	if (!st)
+		return NULL;
+
+	komeda_component_state_reset(&st->base);
+	__drm_atomic_helper_private_obj_duplicate_state(obj, &st->base.obj);
+
+	return &st->base.obj;
+}
+
+static void
+komeda_timing_ctrlr_atomic_destroy_state(struct drm_private_obj *obj,
+					 struct drm_private_state *state)
+{
+	kfree(to_ctrlr_st(priv_to_comp_st(state)));
+}
+
+static const struct drm_private_state_funcs komeda_timing_ctrlr_obj_funcs = {
+	.atomic_duplicate_state	= komeda_timing_ctrlr_atomic_duplicate_state,
+	.atomic_destroy_state	= komeda_timing_ctrlr_atomic_destroy_state,
+};
+
+static int komeda_timing_ctrlr_obj_add(struct komeda_kms_dev *kms,
+				       struct komeda_timing_ctrlr *ctrlr)
+{
+	struct komeda_compiz_state *st;
+
+	st = kzalloc(sizeof(*st), GFP_KERNEL);
+	if (!st)
+		return -ENOMEM;
+
+	st->base.component = &ctrlr->base;
+	drm_atomic_private_obj_init(&kms->base, &ctrlr->base.obj, &st->base.obj,
+				    &komeda_timing_ctrlr_obj_funcs);
+
+	return 0;
+}
+
+static struct drm_private_state *
+komeda_pipeline_atomic_duplicate_state(struct drm_private_obj *obj)
+{
+	struct komeda_pipeline_state *st;
+
+	st = kmemdup(obj->state, sizeof(*st), GFP_KERNEL);
+	if (!st)
+		return NULL;
+
+	st->active_comps = 0;
+
+	__drm_atomic_helper_private_obj_duplicate_state(obj, &st->obj);
+
+	return &st->obj;
+}
+
+static void
+komeda_pipeline_atomic_destroy_state(struct drm_private_obj *obj,
+				     struct drm_private_state *state)
+{
+	kfree(priv_to_pipe_st(state));
+}
+
+static const struct drm_private_state_funcs komeda_pipeline_obj_funcs = {
+	.atomic_duplicate_state	= komeda_pipeline_atomic_duplicate_state,
+	.atomic_destroy_state	= komeda_pipeline_atomic_destroy_state,
+};
+
+static int komeda_pipeline_obj_add(struct komeda_kms_dev *kms,
+				   struct komeda_pipeline *pipe)
+{
+	struct komeda_pipeline_state *st;
+
+	st = kzalloc(sizeof(*st), GFP_KERNEL);
+	if (!st)
+		return -ENOMEM;
+
+	st->pipe = pipe;
+	drm_atomic_private_obj_init(&kms->base, &pipe->obj, &st->obj,
+				    &komeda_pipeline_obj_funcs);
+
+	return 0;
+}
+
+int komeda_kms_add_private_objs(struct komeda_kms_dev *kms,
+				struct komeda_dev *mdev)
+{
+	struct komeda_pipeline *pipe;
+	int i, j, err;
+
+	for (i = 0; i < mdev->n_pipelines; i++) {
+		pipe = mdev->pipelines[i];
+
+		err = komeda_pipeline_obj_add(kms, pipe);
+		if (err)
+			return err;
+
+		for (j = 0; j < pipe->n_layers; j++) {
+			err = komeda_layer_obj_add(kms, pipe->layers[j]);
+			if (err)
+				return err;
+		}
+
+		if (pipe->wb_layer) {
+			err = komeda_layer_obj_add(kms, pipe->wb_layer);
+			if (err)
+				return err;
+		}
+
+		for (j = 0; j < pipe->n_scalers; j++) {
+			err = komeda_scaler_obj_add(kms, pipe->scalers[j]);
+			if (err)
+				return err;
+		}
+
+		err = komeda_compiz_obj_add(kms, pipe->compiz);
+		if (err)
+			return err;
+
+		if (pipe->splitter) {
+			err = komeda_splitter_obj_add(kms, pipe->splitter);
+			if (err)
+				return err;
+		}
+
+		if (pipe->merger) {
+			err = komeda_merger_obj_add(kms, pipe->merger);
+			if (err)
+				return err;
+		}
+
+		err = komeda_improc_obj_add(kms, pipe->improc);
+		if (err)
+			return err;
+
+		err = komeda_timing_ctrlr_obj_add(kms, pipe->ctrlr);
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
+
+void komeda_kms_cleanup_private_objs(struct komeda_kms_dev *kms)
+{
+	struct drm_mode_config *config = &kms->base.mode_config;
+	struct drm_private_obj *obj, *next;
+
+	list_for_each_entry_safe(obj, next, &config->privobj_list, head)
+		drm_atomic_private_obj_fini(obj);
+}
diff --git a/drivers/gpu/drm/arm/display/komeda/komeda_wb_connector.c b/drivers/gpu/drm/arm/display/komeda/komeda_wb_connector.c
new file mode 100644
index 000000000..e465cc487
--- /dev/null
+++ b/drivers/gpu/drm/arm/display/komeda/komeda_wb_connector.c
@@ -0,0 +1,197 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2018 ARM Limited. All rights reserved.
+ * Author: James.Qian.Wang <james.qian.wang@arm.com>
+ *
+ */
+#include "komeda_dev.h"
+#include "komeda_kms.h"
+
+static int
+komeda_wb_init_data_flow(struct komeda_layer *wb_layer,
+			 struct drm_connector_state *conn_st,
+			 struct komeda_crtc_state *kcrtc_st,
+			 struct komeda_data_flow_cfg *dflow)
+{
+	struct drm_framebuffer *fb = conn_st->writeback_job->fb;
+
+	memset(dflow, 0, sizeof(*dflow));
+
+	dflow->out_w = fb->width;
+	dflow->out_h = fb->height;
+
+	/* the write back data comes from the compiz */
+	pipeline_composition_size(kcrtc_st, &dflow->in_w, &dflow->in_h);
+	dflow->input.component = &wb_layer->base.pipeline->compiz->base;
+	/* compiz doesn't output alpha */
+	dflow->pixel_blend_mode = DRM_MODE_BLEND_PIXEL_NONE;
+	dflow->rot = DRM_MODE_ROTATE_0;
+
+	komeda_complete_data_flow_cfg(wb_layer, dflow, fb);
+
+	return 0;
+}
+
+static int
+komeda_wb_encoder_atomic_check(struct drm_encoder *encoder,
+			       struct drm_crtc_state *crtc_st,
+			       struct drm_connector_state *conn_st)
+{
+	struct komeda_crtc_state *kcrtc_st = to_kcrtc_st(crtc_st);
+	struct drm_writeback_job *writeback_job = conn_st->writeback_job;
+	struct komeda_layer *wb_layer;
+	struct komeda_data_flow_cfg dflow;
+	int err;
+
+	if (!writeback_job)
+		return 0;
+
+	if (!crtc_st->active) {
+		DRM_DEBUG_ATOMIC("Cannot write the composition result out on a inactive CRTC.\n");
+		return -EINVAL;
+	}
+
+	wb_layer = to_kconn(to_wb_conn(conn_st->connector))->wb_layer;
+
+	/*
+	 * No need for a full modested when the only connector changed is the
+	 * writeback connector.
+	 */
+	if (crtc_st->connectors_changed &&
+	    is_only_changed_connector(crtc_st, conn_st->connector))
+		crtc_st->connectors_changed = false;
+
+	err = komeda_wb_init_data_flow(wb_layer, conn_st, kcrtc_st, &dflow);
+	if (err)
+		return err;
+
+	if (dflow.en_split)
+		err = komeda_build_wb_split_data_flow(wb_layer,
+				conn_st, kcrtc_st, &dflow);
+	else
+		err = komeda_build_wb_data_flow(wb_layer,
+				conn_st, kcrtc_st, &dflow);
+
+	return err;
+}
+
+static const struct drm_encoder_helper_funcs komeda_wb_encoder_helper_funcs = {
+	.atomic_check = komeda_wb_encoder_atomic_check,
+};
+
+static int
+komeda_wb_connector_get_modes(struct drm_connector *connector)
+{
+	return 0;
+}
+
+static enum drm_mode_status
+komeda_wb_connector_mode_valid(struct drm_connector *connector,
+			       struct drm_display_mode *mode)
+{
+	struct drm_device *dev = connector->dev;
+	struct drm_mode_config *mode_config = &dev->mode_config;
+	int w = mode->hdisplay, h = mode->vdisplay;
+
+	if ((w < mode_config->min_width) || (w > mode_config->max_width))
+		return MODE_BAD_HVALUE;
+
+	if ((h < mode_config->min_height) || (h > mode_config->max_height))
+		return MODE_BAD_VVALUE;
+
+	return MODE_OK;
+}
+
+static const struct drm_connector_helper_funcs komeda_wb_conn_helper_funcs = {
+	.get_modes	= komeda_wb_connector_get_modes,
+	.mode_valid	= komeda_wb_connector_mode_valid,
+};
+
+static enum drm_connector_status
+komeda_wb_connector_detect(struct drm_connector *connector, bool force)
+{
+	return connector_status_connected;
+}
+
+static int
+komeda_wb_connector_fill_modes(struct drm_connector *connector,
+			       uint32_t maxX, uint32_t maxY)
+{
+	return 0;
+}
+
+static void komeda_wb_connector_destroy(struct drm_connector *connector)
+{
+	drm_connector_cleanup(connector);
+	kfree(to_kconn(to_wb_conn(connector)));
+}
+
+static const struct drm_connector_funcs komeda_wb_connector_funcs = {
+	.reset			= drm_atomic_helper_connector_reset,
+	.detect			= komeda_wb_connector_detect,
+	.fill_modes		= komeda_wb_connector_fill_modes,
+	.destroy		= komeda_wb_connector_destroy,
+	.atomic_duplicate_state	= drm_atomic_helper_connector_duplicate_state,
+	.atomic_destroy_state	= drm_atomic_helper_connector_destroy_state,
+};
+
+static int komeda_wb_connector_add(struct komeda_kms_dev *kms,
+				   struct komeda_crtc *kcrtc)
+{
+	struct komeda_dev *mdev = kms->base.dev_private;
+	struct komeda_wb_connector *kwb_conn;
+	struct drm_writeback_connector *wb_conn;
+	struct drm_display_info *info;
+	u32 *formats, n_formats = 0;
+	int err;
+
+	if (!kcrtc->master->wb_layer)
+		return 0;
+
+	kwb_conn = kzalloc(sizeof(*kwb_conn), GFP_KERNEL);
+	if (!kwb_conn)
+		return -ENOMEM;
+
+	kwb_conn->wb_layer = kcrtc->master->wb_layer;
+
+	wb_conn = &kwb_conn->base;
+	wb_conn->encoder.possible_crtcs = BIT(drm_crtc_index(&kcrtc->base));
+
+	formats = komeda_get_layer_fourcc_list(&mdev->fmt_tbl,
+					       kwb_conn->wb_layer->layer_type,
+					       &n_formats);
+
+	err = drm_writeback_connector_init(&kms->base, wb_conn,
+					   &komeda_wb_connector_funcs,
+					   &komeda_wb_encoder_helper_funcs,
+					   formats, n_formats);
+	komeda_put_fourcc_list(formats);
+	if (err) {
+		kfree(kwb_conn);
+		return err;
+	}
+
+	drm_connector_helper_add(&wb_conn->base, &komeda_wb_conn_helper_funcs);
+
+	info = &kwb_conn->base.base.display_info;
+	info->bpc = __fls(kcrtc->master->improc->supported_color_depths);
+	info->color_formats = kcrtc->master->improc->supported_color_formats;
+
+	kcrtc->wb_conn = kwb_conn;
+
+	return 0;
+}
+
+int komeda_kms_add_wb_connectors(struct komeda_kms_dev *kms,
+				 struct komeda_dev *mdev)
+{
+	int i, err;
+
+	for (i = 0; i < kms->n_crtcs; i++) {
+		err = komeda_wb_connector_add(kms, &kms->crtcs[i]);
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/arm/hdlcd_crtc.c b/drivers/gpu/drm/arm/hdlcd_crtc.c
new file mode 100644
index 000000000..af67fefed
--- /dev/null
+++ b/drivers/gpu/drm/arm/hdlcd_crtc.c
@@ -0,0 +1,337 @@
+/*
+ * Copyright (C) 2013-2015 ARM Limited
+ * Author: Liviu Dudau <Liviu.Dudau@arm.com>
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file COPYING in the main directory of this archive
+ * for more details.
+ *
+ *  Implementation of a CRTC class for the HDLCD driver.
+ */
+
+#include <linux/clk.h>
+#include <linux/of_graph.h>
+#include <linux/platform_data/simplefb.h>
+
+#include <video/videomode.h>
+
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_fb_cma_helper.h>
+#include <drm/drm_fb_helper.h>
+#include <drm/drm_gem_cma_helper.h>
+#include <drm/drm_of.h>
+#include <drm/drm_plane_helper.h>
+#include <drm/drm_probe_helper.h>
+#include <drm/drm_vblank.h>
+
+#include "hdlcd_drv.h"
+#include "hdlcd_regs.h"
+
+/*
+ * The HDLCD controller is a dumb RGB streamer that gets connected to
+ * a single HDMI transmitter or in the case of the ARM Models it gets
+ * emulated by the software that does the actual rendering.
+ *
+ */
+
+static void hdlcd_crtc_cleanup(struct drm_crtc *crtc)
+{
+	struct hdlcd_drm_private *hdlcd = crtc_to_hdlcd_priv(crtc);
+
+	/* stop the controller on cleanup */
+	hdlcd_write(hdlcd, HDLCD_REG_COMMAND, 0);
+	drm_crtc_cleanup(crtc);
+}
+
+static int hdlcd_crtc_enable_vblank(struct drm_crtc *crtc)
+{
+	struct hdlcd_drm_private *hdlcd = crtc_to_hdlcd_priv(crtc);
+	unsigned int mask = hdlcd_read(hdlcd, HDLCD_REG_INT_MASK);
+
+	hdlcd_write(hdlcd, HDLCD_REG_INT_MASK, mask | HDLCD_INTERRUPT_VSYNC);
+
+	return 0;
+}
+
+static void hdlcd_crtc_disable_vblank(struct drm_crtc *crtc)
+{
+	struct hdlcd_drm_private *hdlcd = crtc_to_hdlcd_priv(crtc);
+	unsigned int mask = hdlcd_read(hdlcd, HDLCD_REG_INT_MASK);
+
+	hdlcd_write(hdlcd, HDLCD_REG_INT_MASK, mask & ~HDLCD_INTERRUPT_VSYNC);
+}
+
+static const struct drm_crtc_funcs hdlcd_crtc_funcs = {
+	.destroy = hdlcd_crtc_cleanup,
+	.set_config = drm_atomic_helper_set_config,
+	.page_flip = drm_atomic_helper_page_flip,
+	.reset = drm_atomic_helper_crtc_reset,
+	.atomic_duplicate_state = drm_atomic_helper_crtc_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_crtc_destroy_state,
+	.enable_vblank = hdlcd_crtc_enable_vblank,
+	.disable_vblank = hdlcd_crtc_disable_vblank,
+};
+
+static struct simplefb_format supported_formats[] = SIMPLEFB_FORMATS;
+
+/*
+ * Setup the HDLCD registers for decoding the pixels out of the framebuffer
+ */
+static int hdlcd_set_pxl_fmt(struct drm_crtc *crtc)
+{
+	unsigned int btpp;
+	struct hdlcd_drm_private *hdlcd = crtc_to_hdlcd_priv(crtc);
+	const struct drm_framebuffer *fb = crtc->primary->state->fb;
+	uint32_t pixel_format;
+	struct simplefb_format *format = NULL;
+	int i;
+
+	pixel_format = fb->format->format;
+
+	for (i = 0; i < ARRAY_SIZE(supported_formats); i++) {
+		if (supported_formats[i].fourcc == pixel_format)
+			format = &supported_formats[i];
+	}
+
+	if (WARN_ON(!format))
+		return 0;
+
+	/* HDLCD uses 'bytes per pixel', zero means 1 byte */
+	btpp = (format->bits_per_pixel + 7) / 8;
+	hdlcd_write(hdlcd, HDLCD_REG_PIXEL_FORMAT, (btpp - 1) << 3);
+
+	/*
+	 * The format of the HDLCD_REG_<color>_SELECT register is:
+	 *   - bits[23:16] - default value for that color component
+	 *   - bits[11:8]  - number of bits to extract for each color component
+	 *   - bits[4:0]   - index of the lowest bit to extract
+	 *
+	 * The default color value is used when bits[11:8] are zero, when the
+	 * pixel is outside the visible frame area or when there is a
+	 * buffer underrun.
+	 */
+	hdlcd_write(hdlcd, HDLCD_REG_RED_SELECT, format->red.offset |
+#ifdef CONFIG_DRM_HDLCD_SHOW_UNDERRUN
+		    0x00ff0000 |	/* show underruns in red */
+#endif
+		    ((format->red.length & 0xf) << 8));
+	hdlcd_write(hdlcd, HDLCD_REG_GREEN_SELECT, format->green.offset |
+		    ((format->green.length & 0xf) << 8));
+	hdlcd_write(hdlcd, HDLCD_REG_BLUE_SELECT, format->blue.offset |
+		    ((format->blue.length & 0xf) << 8));
+
+	return 0;
+}
+
+static void hdlcd_crtc_mode_set_nofb(struct drm_crtc *crtc)
+{
+	struct hdlcd_drm_private *hdlcd = crtc_to_hdlcd_priv(crtc);
+	struct drm_display_mode *m = &crtc->state->adjusted_mode;
+	struct videomode vm;
+	unsigned int polarities, err;
+
+	vm.vfront_porch = m->crtc_vsync_start - m->crtc_vdisplay;
+	vm.vback_porch = m->crtc_vtotal - m->crtc_vsync_end;
+	vm.vsync_len = m->crtc_vsync_end - m->crtc_vsync_start;
+	vm.hfront_porch = m->crtc_hsync_start - m->crtc_hdisplay;
+	vm.hback_porch = m->crtc_htotal - m->crtc_hsync_end;
+	vm.hsync_len = m->crtc_hsync_end - m->crtc_hsync_start;
+
+	polarities = HDLCD_POLARITY_DATAEN | HDLCD_POLARITY_DATA;
+
+	if (m->flags & DRM_MODE_FLAG_PHSYNC)
+		polarities |= HDLCD_POLARITY_HSYNC;
+	if (m->flags & DRM_MODE_FLAG_PVSYNC)
+		polarities |= HDLCD_POLARITY_VSYNC;
+
+	/* Allow max number of outstanding requests and largest burst size */
+	hdlcd_write(hdlcd, HDLCD_REG_BUS_OPTIONS,
+		    HDLCD_BUS_MAX_OUTSTAND | HDLCD_BUS_BURST_16);
+
+	hdlcd_write(hdlcd, HDLCD_REG_V_DATA, m->crtc_vdisplay - 1);
+	hdlcd_write(hdlcd, HDLCD_REG_V_BACK_PORCH, vm.vback_porch - 1);
+	hdlcd_write(hdlcd, HDLCD_REG_V_FRONT_PORCH, vm.vfront_porch - 1);
+	hdlcd_write(hdlcd, HDLCD_REG_V_SYNC, vm.vsync_len - 1);
+	hdlcd_write(hdlcd, HDLCD_REG_H_DATA, m->crtc_hdisplay - 1);
+	hdlcd_write(hdlcd, HDLCD_REG_H_BACK_PORCH, vm.hback_porch - 1);
+	hdlcd_write(hdlcd, HDLCD_REG_H_FRONT_PORCH, vm.hfront_porch - 1);
+	hdlcd_write(hdlcd, HDLCD_REG_H_SYNC, vm.hsync_len - 1);
+	hdlcd_write(hdlcd, HDLCD_REG_POLARITIES, polarities);
+
+	err = hdlcd_set_pxl_fmt(crtc);
+	if (err)
+		return;
+
+	clk_set_rate(hdlcd->clk, m->crtc_clock * 1000);
+}
+
+static void hdlcd_crtc_atomic_enable(struct drm_crtc *crtc,
+				     struct drm_crtc_state *old_state)
+{
+	struct hdlcd_drm_private *hdlcd = crtc_to_hdlcd_priv(crtc);
+
+	clk_prepare_enable(hdlcd->clk);
+	hdlcd_crtc_mode_set_nofb(crtc);
+	hdlcd_write(hdlcd, HDLCD_REG_COMMAND, 1);
+	drm_crtc_vblank_on(crtc);
+}
+
+static void hdlcd_crtc_atomic_disable(struct drm_crtc *crtc,
+				      struct drm_crtc_state *old_state)
+{
+	struct hdlcd_drm_private *hdlcd = crtc_to_hdlcd_priv(crtc);
+
+	drm_crtc_vblank_off(crtc);
+	hdlcd_write(hdlcd, HDLCD_REG_COMMAND, 0);
+	clk_disable_unprepare(hdlcd->clk);
+}
+
+static enum drm_mode_status hdlcd_crtc_mode_valid(struct drm_crtc *crtc,
+		const struct drm_display_mode *mode)
+{
+	struct hdlcd_drm_private *hdlcd = crtc_to_hdlcd_priv(crtc);
+	long rate, clk_rate = mode->clock * 1000;
+
+	rate = clk_round_rate(hdlcd->clk, clk_rate);
+	/* 0.1% seems a close enough tolerance for the TDA19988 on Juno */
+	if (abs(rate - clk_rate) * 1000 > clk_rate) {
+		/* clock required by mode not supported by hardware */
+		return MODE_NOCLOCK;
+	}
+
+	return MODE_OK;
+}
+
+static void hdlcd_crtc_atomic_begin(struct drm_crtc *crtc,
+				    struct drm_crtc_state *state)
+{
+	struct drm_pending_vblank_event *event = crtc->state->event;
+
+	if (event) {
+		crtc->state->event = NULL;
+
+		spin_lock_irq(&crtc->dev->event_lock);
+		if (drm_crtc_vblank_get(crtc) == 0)
+			drm_crtc_arm_vblank_event(crtc, event);
+		else
+			drm_crtc_send_vblank_event(crtc, event);
+		spin_unlock_irq(&crtc->dev->event_lock);
+	}
+}
+
+static const struct drm_crtc_helper_funcs hdlcd_crtc_helper_funcs = {
+	.mode_valid	= hdlcd_crtc_mode_valid,
+	.atomic_begin	= hdlcd_crtc_atomic_begin,
+	.atomic_enable	= hdlcd_crtc_atomic_enable,
+	.atomic_disable	= hdlcd_crtc_atomic_disable,
+};
+
+static int hdlcd_plane_atomic_check(struct drm_plane *plane,
+				    struct drm_plane_state *state)
+{
+	int i;
+	struct drm_crtc *crtc;
+	struct drm_crtc_state *crtc_state;
+	u32 src_h = state->src_h >> 16;
+
+	/* only the HDLCD_REG_FB_LINE_COUNT register has a limit */
+	if (src_h >= HDLCD_MAX_YRES) {
+		DRM_DEBUG_KMS("Invalid source width: %d\n", src_h);
+		return -EINVAL;
+	}
+
+	for_each_new_crtc_in_state(state->state, crtc, crtc_state, i) {
+		/* we cannot disable the plane while the CRTC is active */
+		if (!state->fb && crtc_state->active)
+			return -EINVAL;
+		return drm_atomic_helper_check_plane_state(state, crtc_state,
+						DRM_PLANE_HELPER_NO_SCALING,
+						DRM_PLANE_HELPER_NO_SCALING,
+						false, true);
+	}
+
+	return 0;
+}
+
+static void hdlcd_plane_atomic_update(struct drm_plane *plane,
+				      struct drm_plane_state *state)
+{
+	struct drm_framebuffer *fb = plane->state->fb;
+	struct hdlcd_drm_private *hdlcd;
+	u32 dest_h;
+	dma_addr_t scanout_start;
+
+	if (!fb)
+		return;
+
+	dest_h = drm_rect_height(&plane->state->dst);
+	scanout_start = drm_fb_cma_get_gem_addr(fb, plane->state, 0);
+
+	hdlcd = plane->dev->dev_private;
+	hdlcd_write(hdlcd, HDLCD_REG_FB_LINE_LENGTH, fb->pitches[0]);
+	hdlcd_write(hdlcd, HDLCD_REG_FB_LINE_PITCH, fb->pitches[0]);
+	hdlcd_write(hdlcd, HDLCD_REG_FB_LINE_COUNT, dest_h - 1);
+	hdlcd_write(hdlcd, HDLCD_REG_FB_BASE, scanout_start);
+}
+
+static const struct drm_plane_helper_funcs hdlcd_plane_helper_funcs = {
+	.atomic_check = hdlcd_plane_atomic_check,
+	.atomic_update = hdlcd_plane_atomic_update,
+};
+
+static const struct drm_plane_funcs hdlcd_plane_funcs = {
+	.update_plane		= drm_atomic_helper_update_plane,
+	.disable_plane		= drm_atomic_helper_disable_plane,
+	.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,
+};
+
+static struct drm_plane *hdlcd_plane_init(struct drm_device *drm)
+{
+	struct hdlcd_drm_private *hdlcd = drm->dev_private;
+	struct drm_plane *plane = NULL;
+	u32 formats[ARRAY_SIZE(supported_formats)], i;
+	int ret;
+
+	plane = devm_kzalloc(drm->dev, sizeof(*plane), GFP_KERNEL);
+	if (!plane)
+		return ERR_PTR(-ENOMEM);
+
+	for (i = 0; i < ARRAY_SIZE(supported_formats); i++)
+		formats[i] = supported_formats[i].fourcc;
+
+	ret = drm_universal_plane_init(drm, plane, 0xff, &hdlcd_plane_funcs,
+				       formats, ARRAY_SIZE(formats),
+				       NULL,
+				       DRM_PLANE_TYPE_PRIMARY, NULL);
+	if (ret)
+		return ERR_PTR(ret);
+
+	drm_plane_helper_add(plane, &hdlcd_plane_helper_funcs);
+	hdlcd->plane = plane;
+
+	return plane;
+}
+
+int hdlcd_setup_crtc(struct drm_device *drm)
+{
+	struct hdlcd_drm_private *hdlcd = drm->dev_private;
+	struct drm_plane *primary;
+	int ret;
+
+	primary = hdlcd_plane_init(drm);
+	if (IS_ERR(primary))
+		return PTR_ERR(primary);
+
+	ret = drm_crtc_init_with_planes(drm, &hdlcd->crtc, primary, NULL,
+					&hdlcd_crtc_funcs, NULL);
+	if (ret)
+		return ret;
+
+	drm_crtc_helper_add(&hdlcd->crtc, &hdlcd_crtc_helper_funcs);
+	return 0;
+}
diff --git a/drivers/gpu/drm/arm/hdlcd_drv.c b/drivers/gpu/drm/arm/hdlcd_drv.c
new file mode 100644
index 000000000..faa8a5a75
--- /dev/null
+++ b/drivers/gpu/drm/arm/hdlcd_drv.c
@@ -0,0 +1,423 @@
+/*
+ * Copyright (C) 2013-2015 ARM Limited
+ * Author: Liviu Dudau <Liviu.Dudau@arm.com>
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file COPYING in the main directory of this archive
+ * for more details.
+ *
+ *  ARM HDLCD Driver
+ */
+
+#include <linux/module.h>
+#include <linux/spinlock.h>
+#include <linux/clk.h>
+#include <linux/component.h>
+#include <linux/console.h>
+#include <linux/dma-mapping.h>
+#include <linux/list.h>
+#include <linux/of_graph.h>
+#include <linux/of_reserved_mem.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_debugfs.h>
+#include <drm/drm_drv.h>
+#include <drm/drm_fb_cma_helper.h>
+#include <drm/drm_fb_helper.h>
+#include <drm/drm_gem_cma_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
+#include <drm/drm_irq.h>
+#include <drm/drm_modeset_helper.h>
+#include <drm/drm_of.h>
+#include <drm/drm_probe_helper.h>
+#include <drm/drm_vblank.h>
+
+#include "hdlcd_drv.h"
+#include "hdlcd_regs.h"
+
+static int hdlcd_load(struct drm_device *drm, unsigned long flags)
+{
+	struct hdlcd_drm_private *hdlcd = drm->dev_private;
+	struct platform_device *pdev = to_platform_device(drm->dev);
+	struct resource *res;
+	u32 version;
+	int ret;
+
+	hdlcd->clk = devm_clk_get(drm->dev, "pxlclk");
+	if (IS_ERR(hdlcd->clk))
+		return PTR_ERR(hdlcd->clk);
+
+#ifdef CONFIG_DEBUG_FS
+	atomic_set(&hdlcd->buffer_underrun_count, 0);
+	atomic_set(&hdlcd->bus_error_count, 0);
+	atomic_set(&hdlcd->vsync_count, 0);
+	atomic_set(&hdlcd->dma_end_count, 0);
+#endif
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	hdlcd->mmio = devm_ioremap_resource(drm->dev, res);
+	if (IS_ERR(hdlcd->mmio)) {
+		DRM_ERROR("failed to map control registers area\n");
+		ret = PTR_ERR(hdlcd->mmio);
+		hdlcd->mmio = NULL;
+		return ret;
+	}
+
+	version = hdlcd_read(hdlcd, HDLCD_REG_VERSION);
+	if ((version & HDLCD_PRODUCT_MASK) != HDLCD_PRODUCT_ID) {
+		DRM_ERROR("unknown product id: 0x%x\n", version);
+		return -EINVAL;
+	}
+	DRM_INFO("found ARM HDLCD version r%dp%d\n",
+		(version & HDLCD_VERSION_MAJOR_MASK) >> 8,
+		version & HDLCD_VERSION_MINOR_MASK);
+
+	/* Get the optional framebuffer memory resource */
+	ret = of_reserved_mem_device_init(drm->dev);
+	if (ret && ret != -ENODEV)
+		return ret;
+
+	ret = dma_set_mask_and_coherent(drm->dev, DMA_BIT_MASK(32));
+	if (ret)
+		goto setup_fail;
+
+	ret = hdlcd_setup_crtc(drm);
+	if (ret < 0) {
+		DRM_ERROR("failed to create crtc\n");
+		goto setup_fail;
+	}
+
+	ret = drm_irq_install(drm, platform_get_irq(pdev, 0));
+	if (ret < 0) {
+		DRM_ERROR("failed to install IRQ handler\n");
+		goto irq_fail;
+	}
+
+	return 0;
+
+irq_fail:
+	drm_crtc_cleanup(&hdlcd->crtc);
+setup_fail:
+	of_reserved_mem_device_release(drm->dev);
+
+	return ret;
+}
+
+static const struct drm_mode_config_funcs hdlcd_mode_config_funcs = {
+	.fb_create = drm_gem_fb_create,
+	.atomic_check = drm_atomic_helper_check,
+	.atomic_commit = drm_atomic_helper_commit,
+};
+
+static void hdlcd_setup_mode_config(struct drm_device *drm)
+{
+	drm_mode_config_init(drm);
+	drm->mode_config.min_width = 0;
+	drm->mode_config.min_height = 0;
+	drm->mode_config.max_width = HDLCD_MAX_XRES;
+	drm->mode_config.max_height = HDLCD_MAX_YRES;
+	drm->mode_config.funcs = &hdlcd_mode_config_funcs;
+}
+
+static irqreturn_t hdlcd_irq(int irq, void *arg)
+{
+	struct drm_device *drm = arg;
+	struct hdlcd_drm_private *hdlcd = drm->dev_private;
+	unsigned long irq_status;
+
+	irq_status = hdlcd_read(hdlcd, HDLCD_REG_INT_STATUS);
+
+#ifdef CONFIG_DEBUG_FS
+	if (irq_status & HDLCD_INTERRUPT_UNDERRUN)
+		atomic_inc(&hdlcd->buffer_underrun_count);
+
+	if (irq_status & HDLCD_INTERRUPT_DMA_END)
+		atomic_inc(&hdlcd->dma_end_count);
+
+	if (irq_status & HDLCD_INTERRUPT_BUS_ERROR)
+		atomic_inc(&hdlcd->bus_error_count);
+
+	if (irq_status & HDLCD_INTERRUPT_VSYNC)
+		atomic_inc(&hdlcd->vsync_count);
+
+#endif
+	if (irq_status & HDLCD_INTERRUPT_VSYNC)
+		drm_crtc_handle_vblank(&hdlcd->crtc);
+
+	/* acknowledge interrupt(s) */
+	hdlcd_write(hdlcd, HDLCD_REG_INT_CLEAR, irq_status);
+
+	return IRQ_HANDLED;
+}
+
+static void hdlcd_irq_preinstall(struct drm_device *drm)
+{
+	struct hdlcd_drm_private *hdlcd = drm->dev_private;
+	/* Ensure interrupts are disabled */
+	hdlcd_write(hdlcd, HDLCD_REG_INT_MASK, 0);
+	hdlcd_write(hdlcd, HDLCD_REG_INT_CLEAR, ~0);
+}
+
+static int hdlcd_irq_postinstall(struct drm_device *drm)
+{
+#ifdef CONFIG_DEBUG_FS
+	struct hdlcd_drm_private *hdlcd = drm->dev_private;
+	unsigned long irq_mask = hdlcd_read(hdlcd, HDLCD_REG_INT_MASK);
+
+	/* enable debug interrupts */
+	irq_mask |= HDLCD_DEBUG_INT_MASK;
+
+	hdlcd_write(hdlcd, HDLCD_REG_INT_MASK, irq_mask);
+#endif
+	return 0;
+}
+
+static void hdlcd_irq_uninstall(struct drm_device *drm)
+{
+	struct hdlcd_drm_private *hdlcd = drm->dev_private;
+	/* disable all the interrupts that we might have enabled */
+	unsigned long irq_mask = hdlcd_read(hdlcd, HDLCD_REG_INT_MASK);
+
+#ifdef CONFIG_DEBUG_FS
+	/* disable debug interrupts */
+	irq_mask &= ~HDLCD_DEBUG_INT_MASK;
+#endif
+
+	/* disable vsync interrupts */
+	irq_mask &= ~HDLCD_INTERRUPT_VSYNC;
+
+	hdlcd_write(hdlcd, HDLCD_REG_INT_MASK, irq_mask);
+}
+
+#ifdef CONFIG_DEBUG_FS
+static int hdlcd_show_underrun_count(struct seq_file *m, void *arg)
+{
+	struct drm_info_node *node = (struct drm_info_node *)m->private;
+	struct drm_device *drm = node->minor->dev;
+	struct hdlcd_drm_private *hdlcd = drm->dev_private;
+
+	seq_printf(m, "underrun : %d\n", atomic_read(&hdlcd->buffer_underrun_count));
+	seq_printf(m, "dma_end  : %d\n", atomic_read(&hdlcd->dma_end_count));
+	seq_printf(m, "bus_error: %d\n", atomic_read(&hdlcd->bus_error_count));
+	seq_printf(m, "vsync    : %d\n", atomic_read(&hdlcd->vsync_count));
+	return 0;
+}
+
+static int hdlcd_show_pxlclock(struct seq_file *m, void *arg)
+{
+	struct drm_info_node *node = (struct drm_info_node *)m->private;
+	struct drm_device *drm = node->minor->dev;
+	struct hdlcd_drm_private *hdlcd = drm->dev_private;
+	unsigned long clkrate = clk_get_rate(hdlcd->clk);
+	unsigned long mode_clock = hdlcd->crtc.mode.crtc_clock * 1000;
+
+	seq_printf(m, "hw  : %lu\n", clkrate);
+	seq_printf(m, "mode: %lu\n", mode_clock);
+	return 0;
+}
+
+static struct drm_info_list hdlcd_debugfs_list[] = {
+	{ "interrupt_count", hdlcd_show_underrun_count, 0 },
+	{ "clocks", hdlcd_show_pxlclock, 0 },
+};
+
+static void hdlcd_debugfs_init(struct drm_minor *minor)
+{
+	drm_debugfs_create_files(hdlcd_debugfs_list,
+				 ARRAY_SIZE(hdlcd_debugfs_list),
+				 minor->debugfs_root, minor);
+}
+#endif
+
+DEFINE_DRM_GEM_CMA_FOPS(fops);
+
+static struct drm_driver hdlcd_driver = {
+	.driver_features = DRIVER_GEM | DRIVER_MODESET | DRIVER_ATOMIC,
+	.irq_handler = hdlcd_irq,
+	.irq_preinstall = hdlcd_irq_preinstall,
+	.irq_postinstall = hdlcd_irq_postinstall,
+	.irq_uninstall = hdlcd_irq_uninstall,
+	DRM_GEM_CMA_DRIVER_OPS,
+#ifdef CONFIG_DEBUG_FS
+	.debugfs_init = hdlcd_debugfs_init,
+#endif
+	.fops = &fops,
+	.name = "hdlcd",
+	.desc = "ARM HDLCD Controller DRM",
+	.date = "20151021",
+	.major = 1,
+	.minor = 0,
+};
+
+static int hdlcd_drm_bind(struct device *dev)
+{
+	struct drm_device *drm;
+	struct hdlcd_drm_private *hdlcd;
+	int ret;
+
+	hdlcd = devm_kzalloc(dev, sizeof(*hdlcd), GFP_KERNEL);
+	if (!hdlcd)
+		return -ENOMEM;
+
+	drm = drm_dev_alloc(&hdlcd_driver, dev);
+	if (IS_ERR(drm))
+		return PTR_ERR(drm);
+
+	drm->dev_private = hdlcd;
+	dev_set_drvdata(dev, drm);
+
+	hdlcd_setup_mode_config(drm);
+	ret = hdlcd_load(drm, 0);
+	if (ret)
+		goto err_free;
+
+	/* Set the CRTC's port so that the encoder component can find it */
+	hdlcd->crtc.port = of_graph_get_port_by_id(dev->of_node, 0);
+
+	ret = component_bind_all(dev, drm);
+	if (ret) {
+		DRM_ERROR("Failed to bind all components\n");
+		goto err_unload;
+	}
+
+	ret = pm_runtime_set_active(dev);
+	if (ret)
+		goto err_pm_active;
+
+	pm_runtime_enable(dev);
+
+	ret = drm_vblank_init(drm, drm->mode_config.num_crtc);
+	if (ret < 0) {
+		DRM_ERROR("failed to initialise vblank\n");
+		goto err_vblank;
+	}
+
+	drm_mode_config_reset(drm);
+	drm_kms_helper_poll_init(drm);
+
+	ret = drm_dev_register(drm, 0);
+	if (ret)
+		goto err_register;
+
+	drm_fbdev_generic_setup(drm, 32);
+
+	return 0;
+
+err_register:
+	drm_kms_helper_poll_fini(drm);
+err_vblank:
+	pm_runtime_disable(drm->dev);
+err_pm_active:
+	drm_atomic_helper_shutdown(drm);
+	component_unbind_all(dev, drm);
+err_unload:
+	of_node_put(hdlcd->crtc.port);
+	hdlcd->crtc.port = NULL;
+	drm_irq_uninstall(drm);
+	of_reserved_mem_device_release(drm->dev);
+err_free:
+	drm_mode_config_cleanup(drm);
+	dev_set_drvdata(dev, NULL);
+	drm_dev_put(drm);
+
+	return ret;
+}
+
+static void hdlcd_drm_unbind(struct device *dev)
+{
+	struct drm_device *drm = dev_get_drvdata(dev);
+	struct hdlcd_drm_private *hdlcd = drm->dev_private;
+
+	drm_dev_unregister(drm);
+	drm_kms_helper_poll_fini(drm);
+	component_unbind_all(dev, drm);
+	of_node_put(hdlcd->crtc.port);
+	hdlcd->crtc.port = NULL;
+	pm_runtime_get_sync(dev);
+	drm_atomic_helper_shutdown(drm);
+	drm_irq_uninstall(drm);
+	pm_runtime_put(dev);
+	if (pm_runtime_enabled(dev))
+		pm_runtime_disable(dev);
+	of_reserved_mem_device_release(dev);
+	drm_mode_config_cleanup(drm);
+	drm->dev_private = NULL;
+	dev_set_drvdata(dev, NULL);
+	drm_dev_put(drm);
+}
+
+static const struct component_master_ops hdlcd_master_ops = {
+	.bind		= hdlcd_drm_bind,
+	.unbind		= hdlcd_drm_unbind,
+};
+
+static int compare_dev(struct device *dev, void *data)
+{
+	return dev->of_node == data;
+}
+
+static int hdlcd_probe(struct platform_device *pdev)
+{
+	struct device_node *port;
+	struct component_match *match = NULL;
+
+	/* there is only one output port inside each device, find it */
+	port = of_graph_get_remote_node(pdev->dev.of_node, 0, 0);
+	if (!port)
+		return -ENODEV;
+
+	drm_of_component_match_add(&pdev->dev, &match, compare_dev, port);
+	of_node_put(port);
+
+	return component_master_add_with_match(&pdev->dev, &hdlcd_master_ops,
+					       match);
+}
+
+static int hdlcd_remove(struct platform_device *pdev)
+{
+	component_master_del(&pdev->dev, &hdlcd_master_ops);
+	return 0;
+}
+
+static const struct of_device_id  hdlcd_of_match[] = {
+	{ .compatible	= "arm,hdlcd" },
+	{},
+};
+MODULE_DEVICE_TABLE(of, hdlcd_of_match);
+
+static int __maybe_unused hdlcd_pm_suspend(struct device *dev)
+{
+	struct drm_device *drm = dev_get_drvdata(dev);
+
+	return drm_mode_config_helper_suspend(drm);
+}
+
+static int __maybe_unused hdlcd_pm_resume(struct device *dev)
+{
+	struct drm_device *drm = dev_get_drvdata(dev);
+
+	drm_mode_config_helper_resume(drm);
+
+	return 0;
+}
+
+static SIMPLE_DEV_PM_OPS(hdlcd_pm_ops, hdlcd_pm_suspend, hdlcd_pm_resume);
+
+static struct platform_driver hdlcd_platform_driver = {
+	.probe		= hdlcd_probe,
+	.remove		= hdlcd_remove,
+	.driver	= {
+		.name = "hdlcd",
+		.pm = &hdlcd_pm_ops,
+		.of_match_table	= hdlcd_of_match,
+	},
+};
+
+module_platform_driver(hdlcd_platform_driver);
+
+MODULE_AUTHOR("Liviu Dudau");
+MODULE_DESCRIPTION("ARM HDLCD DRM driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/gpu/drm/arm/hdlcd_drv.h b/drivers/gpu/drm/arm/hdlcd_drv.h
new file mode 100644
index 000000000..fd438d177
--- /dev/null
+++ b/drivers/gpu/drm/arm/hdlcd_drv.h
@@ -0,0 +1,38 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ *  ARM HDLCD Controller register definition
+ */
+
+#ifndef __HDLCD_DRV_H__
+#define __HDLCD_DRV_H__
+
+struct hdlcd_drm_private {
+	void __iomem			*mmio;
+	struct clk			*clk;
+	struct drm_crtc			crtc;
+	struct drm_plane		*plane;
+#ifdef CONFIG_DEBUG_FS
+	atomic_t buffer_underrun_count;
+	atomic_t bus_error_count;
+	atomic_t vsync_count;
+	atomic_t dma_end_count;
+#endif
+};
+
+#define crtc_to_hdlcd_priv(x)	container_of(x, struct hdlcd_drm_private, crtc)
+
+static inline void hdlcd_write(struct hdlcd_drm_private *hdlcd,
+			       unsigned int reg, u32 value)
+{
+	writel(value, hdlcd->mmio + reg);
+}
+
+static inline u32 hdlcd_read(struct hdlcd_drm_private *hdlcd, unsigned int reg)
+{
+	return readl(hdlcd->mmio + reg);
+}
+
+int hdlcd_setup_crtc(struct drm_device *dev);
+void hdlcd_set_scanout(struct hdlcd_drm_private *hdlcd);
+
+#endif /* __HDLCD_DRV_H__ */
diff --git a/drivers/gpu/drm/arm/hdlcd_regs.h b/drivers/gpu/drm/arm/hdlcd_regs.h
new file mode 100644
index 000000000..66799ebef
--- /dev/null
+++ b/drivers/gpu/drm/arm/hdlcd_regs.h
@@ -0,0 +1,87 @@
+/*
+ * Copyright (C) 2013,2014 ARM Limited
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file COPYING in the main directory of this archive
+ * for more details.
+ *
+ *  ARM HDLCD Controller register definition
+ */
+
+#ifndef __HDLCD_REGS_H__
+#define __HDLCD_REGS_H__
+
+/* register offsets */
+#define HDLCD_REG_VERSION		0x0000	/* ro */
+#define HDLCD_REG_INT_RAWSTAT		0x0010	/* rw */
+#define HDLCD_REG_INT_CLEAR		0x0014	/* wo */
+#define HDLCD_REG_INT_MASK		0x0018	/* rw */
+#define HDLCD_REG_INT_STATUS		0x001c	/* ro */
+#define HDLCD_REG_FB_BASE		0x0100	/* rw */
+#define HDLCD_REG_FB_LINE_LENGTH	0x0104	/* rw */
+#define HDLCD_REG_FB_LINE_COUNT		0x0108	/* rw */
+#define HDLCD_REG_FB_LINE_PITCH		0x010c	/* rw */
+#define HDLCD_REG_BUS_OPTIONS		0x0110	/* rw */
+#define HDLCD_REG_V_SYNC		0x0200	/* rw */
+#define HDLCD_REG_V_BACK_PORCH		0x0204	/* rw */
+#define HDLCD_REG_V_DATA		0x0208	/* rw */
+#define HDLCD_REG_V_FRONT_PORCH		0x020c	/* rw */
+#define HDLCD_REG_H_SYNC		0x0210	/* rw */
+#define HDLCD_REG_H_BACK_PORCH		0x0214	/* rw */
+#define HDLCD_REG_H_DATA		0x0218	/* rw */
+#define HDLCD_REG_H_FRONT_PORCH		0x021c	/* rw */
+#define HDLCD_REG_POLARITIES		0x0220	/* rw */
+#define HDLCD_REG_COMMAND		0x0230	/* rw */
+#define HDLCD_REG_PIXEL_FORMAT		0x0240	/* rw */
+#define HDLCD_REG_RED_SELECT		0x0244	/* rw */
+#define HDLCD_REG_GREEN_SELECT		0x0248	/* rw */
+#define HDLCD_REG_BLUE_SELECT		0x024c	/* rw */
+
+/* version */
+#define HDLCD_PRODUCT_ID		0x1CDC0000
+#define HDLCD_PRODUCT_MASK		0xFFFF0000
+#define HDLCD_VERSION_MAJOR_MASK	0x0000FF00
+#define HDLCD_VERSION_MINOR_MASK	0x000000FF
+
+/* interrupts */
+#define HDLCD_INTERRUPT_DMA_END		(1 << 0)
+#define HDLCD_INTERRUPT_BUS_ERROR	(1 << 1)
+#define HDLCD_INTERRUPT_VSYNC		(1 << 2)
+#define HDLCD_INTERRUPT_UNDERRUN	(1 << 3)
+#define HDLCD_DEBUG_INT_MASK		(HDLCD_INTERRUPT_DMA_END |  \
+					HDLCD_INTERRUPT_BUS_ERROR | \
+					HDLCD_INTERRUPT_UNDERRUN)
+
+/* polarities */
+#define HDLCD_POLARITY_VSYNC		(1 << 0)
+#define HDLCD_POLARITY_HSYNC		(1 << 1)
+#define HDLCD_POLARITY_DATAEN		(1 << 2)
+#define HDLCD_POLARITY_DATA		(1 << 3)
+#define HDLCD_POLARITY_PIXELCLK		(1 << 4)
+
+/* commands */
+#define HDLCD_COMMAND_DISABLE		(0 << 0)
+#define HDLCD_COMMAND_ENABLE		(1 << 0)
+
+/* pixel format */
+#define HDLCD_PIXEL_FMT_LITTLE_ENDIAN	(0 << 31)
+#define HDLCD_PIXEL_FMT_BIG_ENDIAN	(1 << 31)
+#define HDLCD_BYTES_PER_PIXEL_MASK	(3 << 3)
+
+/* bus options */
+#define HDLCD_BUS_BURST_MASK		0x01f
+#define HDLCD_BUS_MAX_OUTSTAND		0xf00
+#define HDLCD_BUS_BURST_NONE		(0 << 0)
+#define HDLCD_BUS_BURST_1		(1 << 0)
+#define HDLCD_BUS_BURST_2		(1 << 1)
+#define HDLCD_BUS_BURST_4		(1 << 2)
+#define HDLCD_BUS_BURST_8		(1 << 3)
+#define HDLCD_BUS_BURST_16		(1 << 4)
+
+/* Max resolution supported is 4096x4096, 32bpp */
+#define HDLCD_MAX_XRES			4096
+#define HDLCD_MAX_YRES			4096
+
+#define NR_PALETTE			256
+
+#endif /* __HDLCD_REGS_H__ */
diff --git a/drivers/gpu/drm/arm/malidp_crtc.c b/drivers/gpu/drm/arm/malidp_crtc.c
new file mode 100644
index 000000000..af7290942
--- /dev/null
+++ b/drivers/gpu/drm/arm/malidp_crtc.c
@@ -0,0 +1,560 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * (C) COPYRIGHT 2016 ARM Limited. All rights reserved.
+ * Author: Liviu Dudau <Liviu.Dudau@arm.com>
+ *
+ * ARM Mali DP500/DP550/DP650 driver (crtc operations)
+ */
+
+#include <linux/clk.h>
+#include <linux/pm_runtime.h>
+
+#include <video/videomode.h>
+
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_print.h>
+#include <drm/drm_probe_helper.h>
+#include <drm/drm_vblank.h>
+
+#include "malidp_drv.h"
+#include "malidp_hw.h"
+
+static enum drm_mode_status malidp_crtc_mode_valid(struct drm_crtc *crtc,
+						   const struct drm_display_mode *mode)
+{
+	struct malidp_drm *malidp = crtc_to_malidp_device(crtc);
+	struct malidp_hw_device *hwdev = malidp->dev;
+
+	/*
+	 * check that the hardware can drive the required clock rate,
+	 * but skip the check if the clock is meant to be disabled (req_rate = 0)
+	 */
+	long rate, req_rate = mode->crtc_clock * 1000;
+
+	if (req_rate) {
+		rate = clk_round_rate(hwdev->pxlclk, req_rate);
+		if (rate != req_rate) {
+			DRM_DEBUG_DRIVER("pxlclk doesn't support %ld Hz\n",
+					 req_rate);
+			return MODE_NOCLOCK;
+		}
+	}
+
+	return MODE_OK;
+}
+
+static void malidp_crtc_atomic_enable(struct drm_crtc *crtc,
+				      struct drm_crtc_state *old_state)
+{
+	struct malidp_drm *malidp = crtc_to_malidp_device(crtc);
+	struct malidp_hw_device *hwdev = malidp->dev;
+	struct videomode vm;
+	int err = pm_runtime_get_sync(crtc->dev->dev);
+
+	if (err < 0) {
+		DRM_DEBUG_DRIVER("Failed to enable runtime power management: %d\n", err);
+		return;
+	}
+
+	drm_display_mode_to_videomode(&crtc->state->adjusted_mode, &vm);
+	clk_prepare_enable(hwdev->pxlclk);
+
+	/* We rely on firmware to set mclk to a sensible level. */
+	clk_set_rate(hwdev->pxlclk, crtc->state->adjusted_mode.crtc_clock * 1000);
+
+	hwdev->hw->modeset(hwdev, &vm);
+	hwdev->hw->leave_config_mode(hwdev);
+	drm_crtc_vblank_on(crtc);
+}
+
+static void malidp_crtc_atomic_disable(struct drm_crtc *crtc,
+				       struct drm_crtc_state *old_state)
+{
+	struct malidp_drm *malidp = crtc_to_malidp_device(crtc);
+	struct malidp_hw_device *hwdev = malidp->dev;
+	int err;
+
+	/* always disable planes on the CRTC that is being turned off */
+	drm_atomic_helper_disable_planes_on_crtc(old_state, false);
+
+	drm_crtc_vblank_off(crtc);
+	hwdev->hw->enter_config_mode(hwdev);
+
+	clk_disable_unprepare(hwdev->pxlclk);
+
+	err = pm_runtime_put(crtc->dev->dev);
+	if (err < 0) {
+		DRM_DEBUG_DRIVER("Failed to disable runtime power management: %d\n", err);
+	}
+}
+
+static const struct gamma_curve_segment {
+	u16 start;
+	u16 end;
+} segments[MALIDP_COEFFTAB_NUM_COEFFS] = {
+	/* sector 0 */
+	{    0,    0 }, {    1,    1 }, {    2,    2 }, {    3,    3 },
+	{    4,    4 }, {    5,    5 }, {    6,    6 }, {    7,    7 },
+	{    8,    8 }, {    9,    9 }, {   10,   10 }, {   11,   11 },
+	{   12,   12 }, {   13,   13 }, {   14,   14 }, {   15,   15 },
+	/* sector 1 */
+	{   16,   19 }, {   20,   23 }, {   24,   27 }, {   28,   31 },
+	/* sector 2 */
+	{   32,   39 }, {   40,   47 }, {   48,   55 }, {   56,   63 },
+	/* sector 3 */
+	{   64,   79 }, {   80,   95 }, {   96,  111 }, {  112,  127 },
+	/* sector 4 */
+	{  128,  159 }, {  160,  191 }, {  192,  223 }, {  224,  255 },
+	/* sector 5 */
+	{  256,  319 }, {  320,  383 }, {  384,  447 }, {  448,  511 },
+	/* sector 6 */
+	{  512,  639 }, {  640,  767 }, {  768,  895 }, {  896, 1023 },
+	{ 1024, 1151 }, { 1152, 1279 }, { 1280, 1407 }, { 1408, 1535 },
+	{ 1536, 1663 }, { 1664, 1791 }, { 1792, 1919 }, { 1920, 2047 },
+	{ 2048, 2175 }, { 2176, 2303 }, { 2304, 2431 }, { 2432, 2559 },
+	{ 2560, 2687 }, { 2688, 2815 }, { 2816, 2943 }, { 2944, 3071 },
+	{ 3072, 3199 }, { 3200, 3327 }, { 3328, 3455 }, { 3456, 3583 },
+	{ 3584, 3711 }, { 3712, 3839 }, { 3840, 3967 }, { 3968, 4095 },
+};
+
+#define DE_COEFTAB_DATA(a, b) ((((a) & 0xfff) << 16) | (((b) & 0xfff)))
+
+static void malidp_generate_gamma_table(struct drm_property_blob *lut_blob,
+					u32 coeffs[MALIDP_COEFFTAB_NUM_COEFFS])
+{
+	struct drm_color_lut *lut = (struct drm_color_lut *)lut_blob->data;
+	int i;
+
+	for (i = 0; i < MALIDP_COEFFTAB_NUM_COEFFS; ++i) {
+		u32 a, b, delta_in, out_start, out_end;
+
+		delta_in = segments[i].end - segments[i].start;
+		/* DP has 12-bit internal precision for its LUTs. */
+		out_start = drm_color_lut_extract(lut[segments[i].start].green,
+						  12);
+		out_end = drm_color_lut_extract(lut[segments[i].end].green, 12);
+		a = (delta_in == 0) ? 0 : ((out_end - out_start) * 256) / delta_in;
+		b = out_start;
+		coeffs[i] = DE_COEFTAB_DATA(a, b);
+	}
+}
+
+/*
+ * Check if there is a new gamma LUT and if it is of an acceptable size. Also,
+ * reject any LUTs that use distinct red, green, and blue curves.
+ */
+static int malidp_crtc_atomic_check_gamma(struct drm_crtc *crtc,
+					  struct drm_crtc_state *state)
+{
+	struct malidp_crtc_state *mc = to_malidp_crtc_state(state);
+	struct drm_color_lut *lut;
+	size_t lut_size;
+	int i;
+
+	if (!state->color_mgmt_changed || !state->gamma_lut)
+		return 0;
+
+	if (crtc->state->gamma_lut &&
+	    (crtc->state->gamma_lut->base.id == state->gamma_lut->base.id))
+		return 0;
+
+	if (state->gamma_lut->length % sizeof(struct drm_color_lut))
+		return -EINVAL;
+
+	lut_size = state->gamma_lut->length / sizeof(struct drm_color_lut);
+	if (lut_size != MALIDP_GAMMA_LUT_SIZE)
+		return -EINVAL;
+
+	lut = (struct drm_color_lut *)state->gamma_lut->data;
+	for (i = 0; i < lut_size; ++i)
+		if (!((lut[i].red == lut[i].green) &&
+		      (lut[i].red == lut[i].blue)))
+			return -EINVAL;
+
+	if (!state->mode_changed) {
+		int ret;
+
+		state->mode_changed = true;
+		/*
+		 * Kerneldoc for drm_atomic_helper_check_modeset mandates that
+		 * it be invoked when the driver sets ->mode_changed. Since
+		 * changing the gamma LUT doesn't depend on any external
+		 * resources, it is safe to call it only once.
+		 */
+		ret = drm_atomic_helper_check_modeset(crtc->dev, state->state);
+		if (ret)
+			return ret;
+	}
+
+	malidp_generate_gamma_table(state->gamma_lut, mc->gamma_coeffs);
+	return 0;
+}
+
+/*
+ * Check if there is a new CTM and if it contains valid input. Valid here means
+ * that the number is inside the representable range for a Q3.12 number,
+ * excluding truncating the fractional part of the input data.
+ *
+ * The COLORADJ registers can be changed atomically.
+ */
+static int malidp_crtc_atomic_check_ctm(struct drm_crtc *crtc,
+					struct drm_crtc_state *state)
+{
+	struct malidp_crtc_state *mc = to_malidp_crtc_state(state);
+	struct drm_color_ctm *ctm;
+	int i;
+
+	if (!state->color_mgmt_changed)
+		return 0;
+
+	if (!state->ctm)
+		return 0;
+
+	if (crtc->state->ctm && (crtc->state->ctm->base.id ==
+				 state->ctm->base.id))
+		return 0;
+
+	/*
+	 * The size of the ctm is checked in
+	 * drm_atomic_replace_property_blob_from_id.
+	 */
+	ctm = (struct drm_color_ctm *)state->ctm->data;
+	for (i = 0; i < ARRAY_SIZE(ctm->matrix); ++i) {
+		/* Convert from S31.32 to Q3.12. */
+		s64 val = ctm->matrix[i];
+		u32 mag = ((((u64)val) & ~BIT_ULL(63)) >> 20) &
+			  GENMASK_ULL(14, 0);
+
+		/*
+		 * Convert to 2s complement and check the destination's top bit
+		 * for overflow. NB: Can't check before converting or it'd
+		 * incorrectly reject the case:
+		 * sign == 1
+		 * mag == 0x2000
+		 */
+		if (val & BIT_ULL(63))
+			mag = ~mag + 1;
+		if (!!(val & BIT_ULL(63)) != !!(mag & BIT(14)))
+			return -EINVAL;
+		mc->coloradj_coeffs[i] = mag;
+	}
+
+	return 0;
+}
+
+static int malidp_crtc_atomic_check_scaling(struct drm_crtc *crtc,
+					    struct drm_crtc_state *state)
+{
+	struct malidp_drm *malidp = crtc_to_malidp_device(crtc);
+	struct malidp_hw_device *hwdev = malidp->dev;
+	struct malidp_crtc_state *cs = to_malidp_crtc_state(state);
+	struct malidp_se_config *s = &cs->scaler_config;
+	struct drm_plane *plane;
+	struct videomode vm;
+	const struct drm_plane_state *pstate;
+	u32 h_upscale_factor = 0; /* U16.16 */
+	u32 v_upscale_factor = 0; /* U16.16 */
+	u8 scaling = cs->scaled_planes_mask;
+	int ret;
+
+	if (!scaling) {
+		s->scale_enable = false;
+		goto mclk_calc;
+	}
+
+	/* The scaling engine can only handle one plane at a time. */
+	if (scaling & (scaling - 1))
+		return -EINVAL;
+
+	drm_atomic_crtc_state_for_each_plane_state(plane, pstate, state) {
+		struct malidp_plane *mp = to_malidp_plane(plane);
+		u32 phase;
+
+		if (!(mp->layer->id & scaling))
+			continue;
+
+		/*
+		 * Convert crtc_[w|h] to U32.32, then divide by U16.16 src_[w|h]
+		 * to get the U16.16 result.
+		 */
+		h_upscale_factor = div_u64((u64)pstate->crtc_w << 32,
+					   pstate->src_w);
+		v_upscale_factor = div_u64((u64)pstate->crtc_h << 32,
+					   pstate->src_h);
+
+		s->enhancer_enable = ((h_upscale_factor >> 16) >= 2 ||
+				      (v_upscale_factor >> 16) >= 2);
+
+		if (pstate->rotation & MALIDP_ROTATED_MASK) {
+			s->input_w = pstate->src_h >> 16;
+			s->input_h = pstate->src_w >> 16;
+		} else {
+			s->input_w = pstate->src_w >> 16;
+			s->input_h = pstate->src_h >> 16;
+		}
+
+		s->output_w = pstate->crtc_w;
+		s->output_h = pstate->crtc_h;
+
+#define SE_N_PHASE 4
+#define SE_SHIFT_N_PHASE 12
+		/* Calculate initial_phase and delta_phase for horizontal. */
+		phase = s->input_w;
+		s->h_init_phase =
+				((phase << SE_N_PHASE) / s->output_w + 1) / 2;
+
+		phase = s->input_w;
+		phase <<= (SE_SHIFT_N_PHASE + SE_N_PHASE);
+		s->h_delta_phase = phase / s->output_w;
+
+		/* Same for vertical. */
+		phase = s->input_h;
+		s->v_init_phase =
+				((phase << SE_N_PHASE) / s->output_h + 1) / 2;
+
+		phase = s->input_h;
+		phase <<= (SE_SHIFT_N_PHASE + SE_N_PHASE);
+		s->v_delta_phase = phase / s->output_h;
+#undef SE_N_PHASE
+#undef SE_SHIFT_N_PHASE
+		s->plane_src_id = mp->layer->id;
+	}
+
+	s->scale_enable = true;
+	s->hcoeff = malidp_se_select_coeffs(h_upscale_factor);
+	s->vcoeff = malidp_se_select_coeffs(v_upscale_factor);
+
+mclk_calc:
+	drm_display_mode_to_videomode(&state->adjusted_mode, &vm);
+	ret = hwdev->hw->se_calc_mclk(hwdev, s, &vm);
+	if (ret < 0)
+		return -EINVAL;
+	return 0;
+}
+
+static int malidp_crtc_atomic_check(struct drm_crtc *crtc,
+				    struct drm_crtc_state *state)
+{
+	struct malidp_drm *malidp = crtc_to_malidp_device(crtc);
+	struct malidp_hw_device *hwdev = malidp->dev;
+	struct drm_plane *plane;
+	const struct drm_plane_state *pstate;
+	u32 rot_mem_free, rot_mem_usable;
+	int rotated_planes = 0;
+	int ret;
+
+	/*
+	 * check if there is enough rotation memory available for planes
+	 * that need 90° and 270° rotion or planes that are compressed.
+	 * Each plane has set its required memory size in the ->plane_check()
+	 * callback, here we only make sure that the sums are less that the
+	 * total usable memory.
+	 *
+	 * The rotation memory allocation algorithm (for each plane):
+	 *  a. If no more rotated or compressed planes exist, all remaining
+	 *     rotate memory in the bank is available for use by the plane.
+	 *  b. If other rotated or compressed planes exist, and plane's
+	 *     layer ID is DE_VIDEO1, it can use all the memory from first bank
+	 *     if secondary rotation memory bank is available, otherwise it can
+	 *     use up to half the bank's memory.
+	 *  c. If other rotated or compressed planes exist, and plane's layer ID
+	 *     is not DE_VIDEO1, it can use half of the available memory.
+	 *
+	 * Note: this algorithm assumes that the order in which the planes are
+	 * checked always has DE_VIDEO1 plane first in the list if it is
+	 * rotated. Because that is how we create the planes in the first
+	 * place, under current DRM version things work, but if ever the order
+	 * in which drm_atomic_crtc_state_for_each_plane() iterates over planes
+	 * changes, we need to pre-sort the planes before validation.
+	 */
+
+	/* first count the number of rotated planes */
+	drm_atomic_crtc_state_for_each_plane_state(plane, pstate, state) {
+		struct drm_framebuffer *fb = pstate->fb;
+
+		if ((pstate->rotation & MALIDP_ROTATED_MASK) || fb->modifier)
+			rotated_planes++;
+	}
+
+	rot_mem_free = hwdev->rotation_memory[0];
+	/*
+	 * if we have more than 1 plane using rotation memory, use the second
+	 * block of rotation memory as well
+	 */
+	if (rotated_planes > 1)
+		rot_mem_free += hwdev->rotation_memory[1];
+
+	/* now validate the rotation memory requirements */
+	drm_atomic_crtc_state_for_each_plane_state(plane, pstate, state) {
+		struct malidp_plane *mp = to_malidp_plane(plane);
+		struct malidp_plane_state *ms = to_malidp_plane_state(pstate);
+		struct drm_framebuffer *fb = pstate->fb;
+
+		if ((pstate->rotation & MALIDP_ROTATED_MASK) || fb->modifier) {
+			/* process current plane */
+			rotated_planes--;
+
+			if (!rotated_planes) {
+				/* no more rotated planes, we can use what's left */
+				rot_mem_usable = rot_mem_free;
+			} else {
+				if ((mp->layer->id != DE_VIDEO1) ||
+				    (hwdev->rotation_memory[1] == 0))
+					rot_mem_usable = rot_mem_free / 2;
+				else
+					rot_mem_usable = hwdev->rotation_memory[0];
+			}
+
+			rot_mem_free -= rot_mem_usable;
+
+			if (ms->rotmem_size > rot_mem_usable)
+				return -EINVAL;
+		}
+	}
+
+	/* If only the writeback routing has changed, we don't need a modeset */
+	if (state->connectors_changed) {
+		u32 old_mask = crtc->state->connector_mask;
+		u32 new_mask = state->connector_mask;
+
+		if ((old_mask ^ new_mask) ==
+		    (1 << drm_connector_index(&malidp->mw_connector.base)))
+			state->connectors_changed = false;
+	}
+
+	ret = malidp_crtc_atomic_check_gamma(crtc, state);
+	ret = ret ? ret : malidp_crtc_atomic_check_ctm(crtc, state);
+	ret = ret ? ret : malidp_crtc_atomic_check_scaling(crtc, state);
+
+	return ret;
+}
+
+static const struct drm_crtc_helper_funcs malidp_crtc_helper_funcs = {
+	.mode_valid = malidp_crtc_mode_valid,
+	.atomic_check = malidp_crtc_atomic_check,
+	.atomic_enable = malidp_crtc_atomic_enable,
+	.atomic_disable = malidp_crtc_atomic_disable,
+};
+
+static struct drm_crtc_state *malidp_crtc_duplicate_state(struct drm_crtc *crtc)
+{
+	struct malidp_crtc_state *state, *old_state;
+
+	if (WARN_ON(!crtc->state))
+		return NULL;
+
+	old_state = to_malidp_crtc_state(crtc->state);
+	state = kmalloc(sizeof(*state), GFP_KERNEL);
+	if (!state)
+		return NULL;
+
+	__drm_atomic_helper_crtc_duplicate_state(crtc, &state->base);
+	memcpy(state->gamma_coeffs, old_state->gamma_coeffs,
+	       sizeof(state->gamma_coeffs));
+	memcpy(state->coloradj_coeffs, old_state->coloradj_coeffs,
+	       sizeof(state->coloradj_coeffs));
+	memcpy(&state->scaler_config, &old_state->scaler_config,
+	       sizeof(state->scaler_config));
+	state->scaled_planes_mask = 0;
+
+	return &state->base;
+}
+
+static void malidp_crtc_destroy_state(struct drm_crtc *crtc,
+				      struct drm_crtc_state *state)
+{
+	struct malidp_crtc_state *mali_state = NULL;
+
+	if (state) {
+		mali_state = to_malidp_crtc_state(state);
+		__drm_atomic_helper_crtc_destroy_state(state);
+	}
+
+	kfree(mali_state);
+}
+
+static void malidp_crtc_reset(struct drm_crtc *crtc)
+{
+	struct malidp_crtc_state *state =
+		kzalloc(sizeof(*state), GFP_KERNEL);
+
+	if (crtc->state)
+		malidp_crtc_destroy_state(crtc, crtc->state);
+
+	if (state)
+		__drm_atomic_helper_crtc_reset(crtc, &state->base);
+	else
+		__drm_atomic_helper_crtc_reset(crtc, NULL);
+}
+
+static int malidp_crtc_enable_vblank(struct drm_crtc *crtc)
+{
+	struct malidp_drm *malidp = crtc_to_malidp_device(crtc);
+	struct malidp_hw_device *hwdev = malidp->dev;
+
+	malidp_hw_enable_irq(hwdev, MALIDP_DE_BLOCK,
+			     hwdev->hw->map.de_irq_map.vsync_irq);
+	return 0;
+}
+
+static void malidp_crtc_disable_vblank(struct drm_crtc *crtc)
+{
+	struct malidp_drm *malidp = crtc_to_malidp_device(crtc);
+	struct malidp_hw_device *hwdev = malidp->dev;
+
+	malidp_hw_disable_irq(hwdev, MALIDP_DE_BLOCK,
+			      hwdev->hw->map.de_irq_map.vsync_irq);
+}
+
+static const struct drm_crtc_funcs malidp_crtc_funcs = {
+	.gamma_set = drm_atomic_helper_legacy_gamma_set,
+	.destroy = drm_crtc_cleanup,
+	.set_config = drm_atomic_helper_set_config,
+	.page_flip = drm_atomic_helper_page_flip,
+	.reset = malidp_crtc_reset,
+	.atomic_duplicate_state = malidp_crtc_duplicate_state,
+	.atomic_destroy_state = malidp_crtc_destroy_state,
+	.enable_vblank = malidp_crtc_enable_vblank,
+	.disable_vblank = malidp_crtc_disable_vblank,
+};
+
+int malidp_crtc_init(struct drm_device *drm)
+{
+	struct malidp_drm *malidp = drm->dev_private;
+	struct drm_plane *primary = NULL, *plane;
+	int ret;
+
+	ret = malidp_de_planes_init(drm);
+	if (ret < 0) {
+		DRM_ERROR("Failed to initialise planes\n");
+		return ret;
+	}
+
+	drm_for_each_plane(plane, drm) {
+		if (plane->type == DRM_PLANE_TYPE_PRIMARY) {
+			primary = plane;
+			break;
+		}
+	}
+
+	if (!primary) {
+		DRM_ERROR("no primary plane found\n");
+		return -EINVAL;
+	}
+
+	ret = drm_crtc_init_with_planes(drm, &malidp->crtc, primary, NULL,
+					&malidp_crtc_funcs, NULL);
+	if (ret)
+		return ret;
+
+	drm_crtc_helper_add(&malidp->crtc, &malidp_crtc_helper_funcs);
+	drm_mode_crtc_set_gamma_size(&malidp->crtc, MALIDP_GAMMA_LUT_SIZE);
+	/* No inverse-gamma: it is per-plane. */
+	drm_crtc_enable_color_mgmt(&malidp->crtc, 0, true, MALIDP_GAMMA_LUT_SIZE);
+
+	malidp_se_set_enh_coeffs(malidp->dev);
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/arm/malidp_drv.c b/drivers/gpu/drm/arm/malidp_drv.c
new file mode 100644
index 000000000..69fee05c2
--- /dev/null
+++ b/drivers/gpu/drm/arm/malidp_drv.c
@@ -0,0 +1,1017 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * (C) COPYRIGHT 2016 ARM Limited. All rights reserved.
+ * Author: Liviu Dudau <Liviu.Dudau@arm.com>
+ *
+ * ARM Mali DP500/DP550/DP650 KMS/DRM driver
+ */
+
+#include <linux/module.h>
+#include <linux/clk.h>
+#include <linux/component.h>
+#include <linux/of_device.h>
+#include <linux/of_graph.h>
+#include <linux/of_reserved_mem.h>
+#include <linux/pm_runtime.h>
+#include <linux/debugfs.h>
+
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_drv.h>
+#include <drm/drm_fb_cma_helper.h>
+#include <drm/drm_fb_helper.h>
+#include <drm/drm_fourcc.h>
+#include <drm/drm_gem_cma_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
+#include <drm/drm_modeset_helper.h>
+#include <drm/drm_of.h>
+#include <drm/drm_probe_helper.h>
+#include <drm/drm_vblank.h>
+
+#include "malidp_drv.h"
+#include "malidp_mw.h"
+#include "malidp_regs.h"
+#include "malidp_hw.h"
+
+#define MALIDP_CONF_VALID_TIMEOUT	250
+#define AFBC_HEADER_SIZE		16
+#define AFBC_SUPERBLK_ALIGNMENT		128
+
+static void malidp_write_gamma_table(struct malidp_hw_device *hwdev,
+				     u32 data[MALIDP_COEFFTAB_NUM_COEFFS])
+{
+	int i;
+	/* Update all channels with a single gamma curve. */
+	const u32 gamma_write_mask = GENMASK(18, 16);
+	/*
+	 * Always write an entire table, so the address field in
+	 * DE_COEFFTAB_ADDR is 0 and we can use the gamma_write_mask bitmask
+	 * directly.
+	 */
+	malidp_hw_write(hwdev, gamma_write_mask,
+			hwdev->hw->map.coeffs_base + MALIDP_COEF_TABLE_ADDR);
+	for (i = 0; i < MALIDP_COEFFTAB_NUM_COEFFS; ++i)
+		malidp_hw_write(hwdev, data[i],
+				hwdev->hw->map.coeffs_base +
+				MALIDP_COEF_TABLE_DATA);
+}
+
+static void malidp_atomic_commit_update_gamma(struct drm_crtc *crtc,
+					      struct drm_crtc_state *old_state)
+{
+	struct malidp_drm *malidp = crtc_to_malidp_device(crtc);
+	struct malidp_hw_device *hwdev = malidp->dev;
+
+	if (!crtc->state->color_mgmt_changed)
+		return;
+
+	if (!crtc->state->gamma_lut) {
+		malidp_hw_clearbits(hwdev,
+				    MALIDP_DISP_FUNC_GAMMA,
+				    MALIDP_DE_DISPLAY_FUNC);
+	} else {
+		struct malidp_crtc_state *mc =
+			to_malidp_crtc_state(crtc->state);
+
+		if (!old_state->gamma_lut || (crtc->state->gamma_lut->base.id !=
+					      old_state->gamma_lut->base.id))
+			malidp_write_gamma_table(hwdev, mc->gamma_coeffs);
+
+		malidp_hw_setbits(hwdev, MALIDP_DISP_FUNC_GAMMA,
+				  MALIDP_DE_DISPLAY_FUNC);
+	}
+}
+
+static
+void malidp_atomic_commit_update_coloradj(struct drm_crtc *crtc,
+					  struct drm_crtc_state *old_state)
+{
+	struct malidp_drm *malidp = crtc_to_malidp_device(crtc);
+	struct malidp_hw_device *hwdev = malidp->dev;
+	int i;
+
+	if (!crtc->state->color_mgmt_changed)
+		return;
+
+	if (!crtc->state->ctm) {
+		malidp_hw_clearbits(hwdev, MALIDP_DISP_FUNC_CADJ,
+				    MALIDP_DE_DISPLAY_FUNC);
+	} else {
+		struct malidp_crtc_state *mc =
+			to_malidp_crtc_state(crtc->state);
+
+		if (!old_state->ctm || (crtc->state->ctm->base.id !=
+					old_state->ctm->base.id))
+			for (i = 0; i < MALIDP_COLORADJ_NUM_COEFFS; ++i)
+				malidp_hw_write(hwdev,
+						mc->coloradj_coeffs[i],
+						hwdev->hw->map.coeffs_base +
+						MALIDP_COLOR_ADJ_COEF + 4 * i);
+
+		malidp_hw_setbits(hwdev, MALIDP_DISP_FUNC_CADJ,
+				  MALIDP_DE_DISPLAY_FUNC);
+	}
+}
+
+static void malidp_atomic_commit_se_config(struct drm_crtc *crtc,
+					   struct drm_crtc_state *old_state)
+{
+	struct malidp_crtc_state *cs = to_malidp_crtc_state(crtc->state);
+	struct malidp_crtc_state *old_cs = to_malidp_crtc_state(old_state);
+	struct malidp_drm *malidp = crtc_to_malidp_device(crtc);
+	struct malidp_hw_device *hwdev = malidp->dev;
+	struct malidp_se_config *s = &cs->scaler_config;
+	struct malidp_se_config *old_s = &old_cs->scaler_config;
+	u32 se_control = hwdev->hw->map.se_base +
+			 ((hwdev->hw->map.features & MALIDP_REGMAP_HAS_CLEARIRQ) ?
+			 0x10 : 0xC);
+	u32 layer_control = se_control + MALIDP_SE_LAYER_CONTROL;
+	u32 scr = se_control + MALIDP_SE_SCALING_CONTROL;
+	u32 val;
+
+	/* Set SE_CONTROL */
+	if (!s->scale_enable) {
+		val = malidp_hw_read(hwdev, se_control);
+		val &= ~MALIDP_SE_SCALING_EN;
+		malidp_hw_write(hwdev, val, se_control);
+		return;
+	}
+
+	hwdev->hw->se_set_scaling_coeffs(hwdev, s, old_s);
+	val = malidp_hw_read(hwdev, se_control);
+	val |= MALIDP_SE_SCALING_EN | MALIDP_SE_ALPHA_EN;
+
+	val &= ~MALIDP_SE_ENH(MALIDP_SE_ENH_MASK);
+	val |= s->enhancer_enable ? MALIDP_SE_ENH(3) : 0;
+
+	val |= MALIDP_SE_RGBO_IF_EN;
+	malidp_hw_write(hwdev, val, se_control);
+
+	/* Set IN_SIZE & OUT_SIZE. */
+	val = MALIDP_SE_SET_V_SIZE(s->input_h) |
+	      MALIDP_SE_SET_H_SIZE(s->input_w);
+	malidp_hw_write(hwdev, val, layer_control + MALIDP_SE_L0_IN_SIZE);
+	val = MALIDP_SE_SET_V_SIZE(s->output_h) |
+	      MALIDP_SE_SET_H_SIZE(s->output_w);
+	malidp_hw_write(hwdev, val, layer_control + MALIDP_SE_L0_OUT_SIZE);
+
+	/* Set phase regs. */
+	malidp_hw_write(hwdev, s->h_init_phase, scr + MALIDP_SE_H_INIT_PH);
+	malidp_hw_write(hwdev, s->h_delta_phase, scr + MALIDP_SE_H_DELTA_PH);
+	malidp_hw_write(hwdev, s->v_init_phase, scr + MALIDP_SE_V_INIT_PH);
+	malidp_hw_write(hwdev, s->v_delta_phase, scr + MALIDP_SE_V_DELTA_PH);
+}
+
+/*
+ * set the "config valid" bit and wait until the hardware acts on it
+ */
+static int malidp_set_and_wait_config_valid(struct drm_device *drm)
+{
+	struct malidp_drm *malidp = drm->dev_private;
+	struct malidp_hw_device *hwdev = malidp->dev;
+	int ret;
+
+	hwdev->hw->set_config_valid(hwdev, 1);
+	/* don't wait for config_valid flag if we are in config mode */
+	if (hwdev->hw->in_config_mode(hwdev)) {
+		atomic_set(&malidp->config_valid, MALIDP_CONFIG_VALID_DONE);
+		return 0;
+	}
+
+	ret = wait_event_interruptible_timeout(malidp->wq,
+			atomic_read(&malidp->config_valid) == MALIDP_CONFIG_VALID_DONE,
+			msecs_to_jiffies(MALIDP_CONF_VALID_TIMEOUT));
+
+	return (ret > 0) ? 0 : -ETIMEDOUT;
+}
+
+static void malidp_atomic_commit_hw_done(struct drm_atomic_state *state)
+{
+	struct drm_device *drm = state->dev;
+	struct malidp_drm *malidp = drm->dev_private;
+	int loop = 5;
+
+	malidp->event = malidp->crtc.state->event;
+	malidp->crtc.state->event = NULL;
+
+	if (malidp->crtc.state->active) {
+		/*
+		 * if we have an event to deliver to userspace, make sure
+		 * the vblank is enabled as we are sending it from the IRQ
+		 * handler.
+		 */
+		if (malidp->event)
+			drm_crtc_vblank_get(&malidp->crtc);
+
+		/* only set config_valid if the CRTC is enabled */
+		if (malidp_set_and_wait_config_valid(drm) < 0) {
+			/*
+			 * make a loop around the second CVAL setting and
+			 * try 5 times before giving up.
+			 */
+			while (loop--) {
+				if (!malidp_set_and_wait_config_valid(drm))
+					break;
+			}
+			DRM_DEBUG_DRIVER("timed out waiting for updated configuration\n");
+		}
+
+	} else if (malidp->event) {
+		/* CRTC inactive means vblank IRQ is disabled, send event directly */
+		spin_lock_irq(&drm->event_lock);
+		drm_crtc_send_vblank_event(&malidp->crtc, malidp->event);
+		malidp->event = NULL;
+		spin_unlock_irq(&drm->event_lock);
+	}
+	drm_atomic_helper_commit_hw_done(state);
+}
+
+static void malidp_atomic_commit_tail(struct drm_atomic_state *state)
+{
+	struct drm_device *drm = state->dev;
+	struct malidp_drm *malidp = drm->dev_private;
+	struct drm_crtc *crtc;
+	struct drm_crtc_state *old_crtc_state;
+	int i;
+
+	pm_runtime_get_sync(drm->dev);
+
+	/*
+	 * set config_valid to a special value to let IRQ handlers
+	 * know that we are updating registers
+	 */
+	atomic_set(&malidp->config_valid, MALIDP_CONFIG_START);
+	malidp->dev->hw->set_config_valid(malidp->dev, 0);
+
+	drm_atomic_helper_commit_modeset_disables(drm, state);
+
+	for_each_old_crtc_in_state(state, crtc, old_crtc_state, i) {
+		malidp_atomic_commit_update_gamma(crtc, old_crtc_state);
+		malidp_atomic_commit_update_coloradj(crtc, old_crtc_state);
+		malidp_atomic_commit_se_config(crtc, old_crtc_state);
+	}
+
+	drm_atomic_helper_commit_planes(drm, state, DRM_PLANE_COMMIT_ACTIVE_ONLY);
+
+	malidp_mw_atomic_commit(drm, state);
+
+	drm_atomic_helper_commit_modeset_enables(drm, state);
+
+	malidp_atomic_commit_hw_done(state);
+
+	pm_runtime_put(drm->dev);
+
+	drm_atomic_helper_cleanup_planes(drm, state);
+}
+
+static const struct drm_mode_config_helper_funcs malidp_mode_config_helpers = {
+	.atomic_commit_tail = malidp_atomic_commit_tail,
+};
+
+static bool
+malidp_verify_afbc_framebuffer_caps(struct drm_device *dev,
+				    const struct drm_mode_fb_cmd2 *mode_cmd)
+{
+	if (malidp_format_mod_supported(dev, mode_cmd->pixel_format,
+					mode_cmd->modifier[0]) == false)
+		return false;
+
+	if (mode_cmd->offsets[0] != 0) {
+		DRM_DEBUG_KMS("AFBC buffers' plane offset should be 0\n");
+		return false;
+	}
+
+	switch (mode_cmd->modifier[0] & AFBC_SIZE_MASK) {
+	case AFBC_SIZE_16X16:
+		if ((mode_cmd->width % 16) || (mode_cmd->height % 16)) {
+			DRM_DEBUG_KMS("AFBC buffers must be aligned to 16 pixels\n");
+			return false;
+		}
+		break;
+	default:
+		DRM_DEBUG_KMS("Unsupported AFBC block size\n");
+		return false;
+	}
+
+	return true;
+}
+
+static bool
+malidp_verify_afbc_framebuffer_size(struct drm_device *dev,
+				    struct drm_file *file,
+				    const struct drm_mode_fb_cmd2 *mode_cmd)
+{
+	int n_superblocks = 0;
+	const struct drm_format_info *info;
+	struct drm_gem_object *objs = NULL;
+	u32 afbc_superblock_size = 0, afbc_superblock_height = 0;
+	u32 afbc_superblock_width = 0, afbc_size = 0;
+	int bpp = 0;
+
+	switch (mode_cmd->modifier[0] & AFBC_SIZE_MASK) {
+	case AFBC_SIZE_16X16:
+		afbc_superblock_height = 16;
+		afbc_superblock_width = 16;
+		break;
+	default:
+		DRM_DEBUG_KMS("AFBC superblock size is not supported\n");
+		return false;
+	}
+
+	info = drm_get_format_info(dev, mode_cmd);
+
+	n_superblocks = (mode_cmd->width / afbc_superblock_width) *
+		(mode_cmd->height / afbc_superblock_height);
+
+	bpp = malidp_format_get_bpp(info->format);
+
+	afbc_superblock_size = (bpp * afbc_superblock_width * afbc_superblock_height)
+				/ BITS_PER_BYTE;
+
+	afbc_size = ALIGN(n_superblocks * AFBC_HEADER_SIZE, AFBC_SUPERBLK_ALIGNMENT);
+	afbc_size += n_superblocks * ALIGN(afbc_superblock_size, AFBC_SUPERBLK_ALIGNMENT);
+
+	if ((mode_cmd->width * bpp) != (mode_cmd->pitches[0] * BITS_PER_BYTE)) {
+		DRM_DEBUG_KMS("Invalid value of (pitch * BITS_PER_BYTE) (=%u) "
+			      "should be same as width (=%u) * bpp (=%u)\n",
+			      (mode_cmd->pitches[0] * BITS_PER_BYTE),
+			      mode_cmd->width, bpp);
+		return false;
+	}
+
+	objs = drm_gem_object_lookup(file, mode_cmd->handles[0]);
+	if (!objs) {
+		DRM_DEBUG_KMS("Failed to lookup GEM object\n");
+		return false;
+	}
+
+	if (objs->size < afbc_size) {
+		DRM_DEBUG_KMS("buffer size (%zu) too small for AFBC buffer size = %u\n",
+			      objs->size, afbc_size);
+		drm_gem_object_put(objs);
+		return false;
+	}
+
+	drm_gem_object_put(objs);
+
+	return true;
+}
+
+static bool
+malidp_verify_afbc_framebuffer(struct drm_device *dev, struct drm_file *file,
+			       const struct drm_mode_fb_cmd2 *mode_cmd)
+{
+	if (malidp_verify_afbc_framebuffer_caps(dev, mode_cmd))
+		return malidp_verify_afbc_framebuffer_size(dev, file, mode_cmd);
+
+	return false;
+}
+
+static struct drm_framebuffer *
+malidp_fb_create(struct drm_device *dev, struct drm_file *file,
+		 const struct drm_mode_fb_cmd2 *mode_cmd)
+{
+	if (mode_cmd->modifier[0]) {
+		if (!malidp_verify_afbc_framebuffer(dev, file, mode_cmd))
+			return ERR_PTR(-EINVAL);
+	}
+
+	return drm_gem_fb_create(dev, file, mode_cmd);
+}
+
+static const struct drm_mode_config_funcs malidp_mode_config_funcs = {
+	.fb_create = malidp_fb_create,
+	.atomic_check = drm_atomic_helper_check,
+	.atomic_commit = drm_atomic_helper_commit,
+};
+
+static int malidp_init(struct drm_device *drm)
+{
+	int ret;
+	struct malidp_drm *malidp = drm->dev_private;
+	struct malidp_hw_device *hwdev = malidp->dev;
+
+	drm_mode_config_init(drm);
+
+	drm->mode_config.min_width = hwdev->min_line_size;
+	drm->mode_config.min_height = hwdev->min_line_size;
+	drm->mode_config.max_width = hwdev->max_line_size;
+	drm->mode_config.max_height = hwdev->max_line_size;
+	drm->mode_config.funcs = &malidp_mode_config_funcs;
+	drm->mode_config.helper_private = &malidp_mode_config_helpers;
+	drm->mode_config.allow_fb_modifiers = true;
+
+	ret = malidp_crtc_init(drm);
+	if (ret)
+		goto crtc_fail;
+
+	ret = malidp_mw_connector_init(drm);
+	if (ret)
+		goto crtc_fail;
+
+	return 0;
+
+crtc_fail:
+	drm_mode_config_cleanup(drm);
+	return ret;
+}
+
+static void malidp_fini(struct drm_device *drm)
+{
+	drm_mode_config_cleanup(drm);
+}
+
+static int malidp_irq_init(struct platform_device *pdev)
+{
+	int irq_de, irq_se, ret = 0;
+	struct drm_device *drm = dev_get_drvdata(&pdev->dev);
+	struct malidp_drm *malidp = drm->dev_private;
+	struct malidp_hw_device *hwdev = malidp->dev;
+
+	/* fetch the interrupts from DT */
+	irq_de = platform_get_irq_byname(pdev, "DE");
+	if (irq_de < 0) {
+		DRM_ERROR("no 'DE' IRQ specified!\n");
+		return irq_de;
+	}
+	irq_se = platform_get_irq_byname(pdev, "SE");
+	if (irq_se < 0) {
+		DRM_ERROR("no 'SE' IRQ specified!\n");
+		return irq_se;
+	}
+
+	ret = malidp_de_irq_init(drm, irq_de);
+	if (ret)
+		return ret;
+
+	ret = malidp_se_irq_init(drm, irq_se);
+	if (ret) {
+		malidp_de_irq_fini(hwdev);
+		return ret;
+	}
+
+	return 0;
+}
+
+DEFINE_DRM_GEM_CMA_FOPS(fops);
+
+static int malidp_dumb_create(struct drm_file *file_priv,
+			      struct drm_device *drm,
+			      struct drm_mode_create_dumb *args)
+{
+	struct malidp_drm *malidp = drm->dev_private;
+	/* allocate for the worst case scenario, i.e. rotated buffers */
+	u8 alignment = malidp_hw_get_pitch_align(malidp->dev, 1);
+
+	args->pitch = ALIGN(DIV_ROUND_UP(args->width * args->bpp, 8), alignment);
+
+	return drm_gem_cma_dumb_create_internal(file_priv, drm, args);
+}
+
+#ifdef CONFIG_DEBUG_FS
+
+static void malidp_error_stats_init(struct malidp_error_stats *error_stats)
+{
+	error_stats->num_errors = 0;
+	error_stats->last_error_status = 0;
+	error_stats->last_error_vblank = -1;
+}
+
+void malidp_error(struct malidp_drm *malidp,
+		  struct malidp_error_stats *error_stats, u32 status,
+		  u64 vblank)
+{
+	unsigned long irqflags;
+
+	spin_lock_irqsave(&malidp->errors_lock, irqflags);
+	error_stats->last_error_status = status;
+	error_stats->last_error_vblank = vblank;
+	error_stats->num_errors++;
+	spin_unlock_irqrestore(&malidp->errors_lock, irqflags);
+}
+
+static void malidp_error_stats_dump(const char *prefix,
+				    struct malidp_error_stats error_stats,
+				    struct seq_file *m)
+{
+	seq_printf(m, "[%s] num_errors : %d\n", prefix,
+		   error_stats.num_errors);
+	seq_printf(m, "[%s] last_error_status  : 0x%08x\n", prefix,
+		   error_stats.last_error_status);
+	seq_printf(m, "[%s] last_error_vblank : %lld\n", prefix,
+		   error_stats.last_error_vblank);
+}
+
+static int malidp_show_stats(struct seq_file *m, void *arg)
+{
+	struct drm_device *drm = m->private;
+	struct malidp_drm *malidp = drm->dev_private;
+	unsigned long irqflags;
+	struct malidp_error_stats de_errors, se_errors;
+
+	spin_lock_irqsave(&malidp->errors_lock, irqflags);
+	de_errors = malidp->de_errors;
+	se_errors = malidp->se_errors;
+	spin_unlock_irqrestore(&malidp->errors_lock, irqflags);
+	malidp_error_stats_dump("DE", de_errors, m);
+	malidp_error_stats_dump("SE", se_errors, m);
+	return 0;
+}
+
+static int malidp_debugfs_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, malidp_show_stats, inode->i_private);
+}
+
+static ssize_t malidp_debugfs_write(struct file *file, const char __user *ubuf,
+				    size_t len, loff_t *offp)
+{
+	struct seq_file *m = file->private_data;
+	struct drm_device *drm = m->private;
+	struct malidp_drm *malidp = drm->dev_private;
+	unsigned long irqflags;
+
+	spin_lock_irqsave(&malidp->errors_lock, irqflags);
+	malidp_error_stats_init(&malidp->de_errors);
+	malidp_error_stats_init(&malidp->se_errors);
+	spin_unlock_irqrestore(&malidp->errors_lock, irqflags);
+	return len;
+}
+
+static const struct file_operations malidp_debugfs_fops = {
+	.owner = THIS_MODULE,
+	.open = malidp_debugfs_open,
+	.read = seq_read,
+	.write = malidp_debugfs_write,
+	.llseek = seq_lseek,
+	.release = single_release,
+};
+
+static void malidp_debugfs_init(struct drm_minor *minor)
+{
+	struct malidp_drm *malidp = minor->dev->dev_private;
+
+	malidp_error_stats_init(&malidp->de_errors);
+	malidp_error_stats_init(&malidp->se_errors);
+	spin_lock_init(&malidp->errors_lock);
+	debugfs_create_file("debug", S_IRUGO | S_IWUSR, minor->debugfs_root,
+			    minor->dev, &malidp_debugfs_fops);
+}
+
+#endif //CONFIG_DEBUG_FS
+
+static struct drm_driver malidp_driver = {
+	.driver_features = DRIVER_GEM | DRIVER_MODESET | DRIVER_ATOMIC,
+	DRM_GEM_CMA_DRIVER_OPS_WITH_DUMB_CREATE(malidp_dumb_create),
+#ifdef CONFIG_DEBUG_FS
+	.debugfs_init = malidp_debugfs_init,
+#endif
+	.fops = &fops,
+	.name = "mali-dp",
+	.desc = "ARM Mali Display Processor driver",
+	.date = "20160106",
+	.major = 1,
+	.minor = 0,
+};
+
+static const struct of_device_id  malidp_drm_of_match[] = {
+	{
+		.compatible = "arm,mali-dp500",
+		.data = &malidp_device[MALIDP_500]
+	},
+	{
+		.compatible = "arm,mali-dp550",
+		.data = &malidp_device[MALIDP_550]
+	},
+	{
+		.compatible = "arm,mali-dp650",
+		.data = &malidp_device[MALIDP_650]
+	},
+	{},
+};
+MODULE_DEVICE_TABLE(of, malidp_drm_of_match);
+
+static bool malidp_is_compatible_hw_id(struct malidp_hw_device *hwdev,
+				       const struct of_device_id *dev_id)
+{
+	u32 core_id;
+	const char *compatstr_dp500 = "arm,mali-dp500";
+	bool is_dp500;
+	bool dt_is_dp500;
+
+	/*
+	 * The DP500 CORE_ID register is in a different location, so check it
+	 * first. If the product id field matches, then this is DP500, otherwise
+	 * check the DP550/650 CORE_ID register.
+	 */
+	core_id = malidp_hw_read(hwdev, MALIDP500_DC_BASE + MALIDP_DE_CORE_ID);
+	/* Offset 0x18 will never read 0x500 on products other than DP500. */
+	is_dp500 = (MALIDP_PRODUCT_ID(core_id) == 0x500);
+	dt_is_dp500 = strnstr(dev_id->compatible, compatstr_dp500,
+			      sizeof(dev_id->compatible)) != NULL;
+	if (is_dp500 != dt_is_dp500) {
+		DRM_ERROR("Device-tree expects %s, but hardware %s DP500.\n",
+			  dev_id->compatible, is_dp500 ? "is" : "is not");
+		return false;
+	} else if (!dt_is_dp500) {
+		u16 product_id;
+		char buf[32];
+
+		core_id = malidp_hw_read(hwdev,
+					 MALIDP550_DC_BASE + MALIDP_DE_CORE_ID);
+		product_id = MALIDP_PRODUCT_ID(core_id);
+		snprintf(buf, sizeof(buf), "arm,mali-dp%X", product_id);
+		if (!strnstr(dev_id->compatible, buf,
+			     sizeof(dev_id->compatible))) {
+			DRM_ERROR("Device-tree expects %s, but hardware is DP%03X.\n",
+				  dev_id->compatible, product_id);
+			return false;
+		}
+	}
+	return true;
+}
+
+static bool malidp_has_sufficient_address_space(const struct resource *res,
+						const struct of_device_id *dev_id)
+{
+	resource_size_t res_size = resource_size(res);
+	const char *compatstr_dp500 = "arm,mali-dp500";
+
+	if (!strnstr(dev_id->compatible, compatstr_dp500,
+		     sizeof(dev_id->compatible)))
+		return res_size >= MALIDP550_ADDR_SPACE_SIZE;
+	else if (res_size < MALIDP500_ADDR_SPACE_SIZE)
+		return false;
+	return true;
+}
+
+static ssize_t core_id_show(struct device *dev, struct device_attribute *attr,
+			    char *buf)
+{
+	struct drm_device *drm = dev_get_drvdata(dev);
+	struct malidp_drm *malidp = drm->dev_private;
+
+	return snprintf(buf, PAGE_SIZE, "%08x\n", malidp->core_id);
+}
+
+static DEVICE_ATTR_RO(core_id);
+
+static struct attribute *mali_dp_attrs[] = {
+	&dev_attr_core_id.attr,
+	NULL,
+};
+ATTRIBUTE_GROUPS(mali_dp);
+
+#define MAX_OUTPUT_CHANNELS	3
+
+static int malidp_runtime_pm_suspend(struct device *dev)
+{
+	struct drm_device *drm = dev_get_drvdata(dev);
+	struct malidp_drm *malidp = drm->dev_private;
+	struct malidp_hw_device *hwdev = malidp->dev;
+
+	/* we can only suspend if the hardware is in config mode */
+	WARN_ON(!hwdev->hw->in_config_mode(hwdev));
+
+	malidp_se_irq_fini(hwdev);
+	malidp_de_irq_fini(hwdev);
+	hwdev->pm_suspended = true;
+	clk_disable_unprepare(hwdev->mclk);
+	clk_disable_unprepare(hwdev->aclk);
+	clk_disable_unprepare(hwdev->pclk);
+
+	return 0;
+}
+
+static int malidp_runtime_pm_resume(struct device *dev)
+{
+	struct drm_device *drm = dev_get_drvdata(dev);
+	struct malidp_drm *malidp = drm->dev_private;
+	struct malidp_hw_device *hwdev = malidp->dev;
+
+	clk_prepare_enable(hwdev->pclk);
+	clk_prepare_enable(hwdev->aclk);
+	clk_prepare_enable(hwdev->mclk);
+	hwdev->pm_suspended = false;
+	malidp_de_irq_hw_init(hwdev);
+	malidp_se_irq_hw_init(hwdev);
+
+	return 0;
+}
+
+static int malidp_bind(struct device *dev)
+{
+	struct resource *res;
+	struct drm_device *drm;
+	struct malidp_drm *malidp;
+	struct malidp_hw_device *hwdev;
+	struct platform_device *pdev = to_platform_device(dev);
+	struct of_device_id const *dev_id;
+	struct drm_encoder *encoder;
+	/* number of lines for the R, G and B output */
+	u8 output_width[MAX_OUTPUT_CHANNELS];
+	int ret = 0, i;
+	u32 version, out_depth = 0;
+
+	malidp = devm_kzalloc(dev, sizeof(*malidp), GFP_KERNEL);
+	if (!malidp)
+		return -ENOMEM;
+
+	hwdev = devm_kzalloc(dev, sizeof(*hwdev), GFP_KERNEL);
+	if (!hwdev)
+		return -ENOMEM;
+
+	hwdev->hw = (struct malidp_hw *)of_device_get_match_data(dev);
+	malidp->dev = hwdev;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	hwdev->regs = devm_ioremap_resource(dev, res);
+	if (IS_ERR(hwdev->regs))
+		return PTR_ERR(hwdev->regs);
+
+	hwdev->pclk = devm_clk_get(dev, "pclk");
+	if (IS_ERR(hwdev->pclk))
+		return PTR_ERR(hwdev->pclk);
+
+	hwdev->aclk = devm_clk_get(dev, "aclk");
+	if (IS_ERR(hwdev->aclk))
+		return PTR_ERR(hwdev->aclk);
+
+	hwdev->mclk = devm_clk_get(dev, "mclk");
+	if (IS_ERR(hwdev->mclk))
+		return PTR_ERR(hwdev->mclk);
+
+	hwdev->pxlclk = devm_clk_get(dev, "pxlclk");
+	if (IS_ERR(hwdev->pxlclk))
+		return PTR_ERR(hwdev->pxlclk);
+
+	/* Get the optional framebuffer memory resource */
+	ret = of_reserved_mem_device_init(dev);
+	if (ret && ret != -ENODEV)
+		return ret;
+
+	drm = drm_dev_alloc(&malidp_driver, dev);
+	if (IS_ERR(drm)) {
+		ret = PTR_ERR(drm);
+		goto alloc_fail;
+	}
+
+	drm->dev_private = malidp;
+	dev_set_drvdata(dev, drm);
+
+	/* Enable power management */
+	pm_runtime_enable(dev);
+
+	/* Resume device to enable the clocks */
+	if (pm_runtime_enabled(dev))
+		pm_runtime_get_sync(dev);
+	else
+		malidp_runtime_pm_resume(dev);
+
+	dev_id = of_match_device(malidp_drm_of_match, dev);
+	if (!dev_id) {
+		ret = -EINVAL;
+		goto query_hw_fail;
+	}
+
+	if (!malidp_has_sufficient_address_space(res, dev_id)) {
+		DRM_ERROR("Insufficient address space in device-tree.\n");
+		ret = -EINVAL;
+		goto query_hw_fail;
+	}
+
+	if (!malidp_is_compatible_hw_id(hwdev, dev_id)) {
+		ret = -EINVAL;
+		goto query_hw_fail;
+	}
+
+	ret = hwdev->hw->query_hw(hwdev);
+	if (ret) {
+		DRM_ERROR("Invalid HW configuration\n");
+		goto query_hw_fail;
+	}
+
+	version = malidp_hw_read(hwdev, hwdev->hw->map.dc_base + MALIDP_DE_CORE_ID);
+	DRM_INFO("found ARM Mali-DP%3x version r%dp%d\n", version >> 16,
+		 (version >> 12) & 0xf, (version >> 8) & 0xf);
+
+	malidp->core_id = version;
+
+	ret = of_property_read_u32(dev->of_node,
+					"arm,malidp-arqos-value",
+					&hwdev->arqos_value);
+	if (ret)
+		hwdev->arqos_value = 0x0;
+
+	/* set the number of lines used for output of RGB data */
+	ret = of_property_read_u8_array(dev->of_node,
+					"arm,malidp-output-port-lines",
+					output_width, MAX_OUTPUT_CHANNELS);
+	if (ret)
+		goto query_hw_fail;
+
+	for (i = 0; i < MAX_OUTPUT_CHANNELS; i++)
+		out_depth = (out_depth << 8) | (output_width[i] & 0xf);
+	malidp_hw_write(hwdev, out_depth, hwdev->hw->map.out_depth_base);
+	hwdev->output_color_depth = out_depth;
+
+	atomic_set(&malidp->config_valid, MALIDP_CONFIG_VALID_INIT);
+	init_waitqueue_head(&malidp->wq);
+
+	ret = malidp_init(drm);
+	if (ret < 0)
+		goto query_hw_fail;
+
+	/* Set the CRTC's port so that the encoder component can find it */
+	malidp->crtc.port = of_graph_get_port_by_id(dev->of_node, 0);
+
+	ret = component_bind_all(dev, drm);
+	if (ret) {
+		DRM_ERROR("Failed to bind all components\n");
+		goto bind_fail;
+	}
+
+	/* We expect to have a maximum of two encoders one for the actual
+	 * display and a virtual one for the writeback connector
+	 */
+	WARN_ON(drm->mode_config.num_encoder > 2);
+	list_for_each_entry(encoder, &drm->mode_config.encoder_list, head) {
+		encoder->possible_clones =
+				(1 << drm->mode_config.num_encoder) -  1;
+	}
+
+	ret = malidp_irq_init(pdev);
+	if (ret < 0)
+		goto irq_init_fail;
+
+	drm->irq_enabled = true;
+
+	ret = drm_vblank_init(drm, drm->mode_config.num_crtc);
+	if (ret < 0) {
+		DRM_ERROR("failed to initialise vblank\n");
+		goto vblank_fail;
+	}
+	pm_runtime_put(dev);
+
+	drm_mode_config_reset(drm);
+
+	drm_kms_helper_poll_init(drm);
+
+	ret = drm_dev_register(drm, 0);
+	if (ret)
+		goto register_fail;
+
+	drm_fbdev_generic_setup(drm, 32);
+
+	return 0;
+
+register_fail:
+	drm_kms_helper_poll_fini(drm);
+	pm_runtime_get_sync(dev);
+vblank_fail:
+	malidp_se_irq_fini(hwdev);
+	malidp_de_irq_fini(hwdev);
+	drm->irq_enabled = false;
+irq_init_fail:
+	drm_atomic_helper_shutdown(drm);
+	component_unbind_all(dev, drm);
+bind_fail:
+	of_node_put(malidp->crtc.port);
+	malidp->crtc.port = NULL;
+	malidp_fini(drm);
+query_hw_fail:
+	pm_runtime_put(dev);
+	if (pm_runtime_enabled(dev))
+		pm_runtime_disable(dev);
+	else
+		malidp_runtime_pm_suspend(dev);
+	drm->dev_private = NULL;
+	dev_set_drvdata(dev, NULL);
+	drm_dev_put(drm);
+alloc_fail:
+	of_reserved_mem_device_release(dev);
+
+	return ret;
+}
+
+static void malidp_unbind(struct device *dev)
+{
+	struct drm_device *drm = dev_get_drvdata(dev);
+	struct malidp_drm *malidp = drm->dev_private;
+	struct malidp_hw_device *hwdev = malidp->dev;
+
+	drm_dev_unregister(drm);
+	drm_kms_helper_poll_fini(drm);
+	pm_runtime_get_sync(dev);
+	drm_atomic_helper_shutdown(drm);
+	malidp_se_irq_fini(hwdev);
+	malidp_de_irq_fini(hwdev);
+	drm->irq_enabled = false;
+	component_unbind_all(dev, drm);
+	of_node_put(malidp->crtc.port);
+	malidp->crtc.port = NULL;
+	malidp_fini(drm);
+	pm_runtime_put(dev);
+	if (pm_runtime_enabled(dev))
+		pm_runtime_disable(dev);
+	else
+		malidp_runtime_pm_suspend(dev);
+	drm->dev_private = NULL;
+	dev_set_drvdata(dev, NULL);
+	drm_dev_put(drm);
+	of_reserved_mem_device_release(dev);
+}
+
+static const struct component_master_ops malidp_master_ops = {
+	.bind = malidp_bind,
+	.unbind = malidp_unbind,
+};
+
+static int malidp_compare_dev(struct device *dev, void *data)
+{
+	struct device_node *np = data;
+
+	return dev->of_node == np;
+}
+
+static int malidp_platform_probe(struct platform_device *pdev)
+{
+	struct device_node *port;
+	struct component_match *match = NULL;
+
+	if (!pdev->dev.of_node)
+		return -ENODEV;
+
+	/* there is only one output port inside each device, find it */
+	port = of_graph_get_remote_node(pdev->dev.of_node, 0, 0);
+	if (!port)
+		return -ENODEV;
+
+	drm_of_component_match_add(&pdev->dev, &match, malidp_compare_dev,
+				   port);
+	of_node_put(port);
+	return component_master_add_with_match(&pdev->dev, &malidp_master_ops,
+					       match);
+}
+
+static int malidp_platform_remove(struct platform_device *pdev)
+{
+	component_master_del(&pdev->dev, &malidp_master_ops);
+	return 0;
+}
+
+static int __maybe_unused malidp_pm_suspend(struct device *dev)
+{
+	struct drm_device *drm = dev_get_drvdata(dev);
+
+	return drm_mode_config_helper_suspend(drm);
+}
+
+static int __maybe_unused malidp_pm_resume(struct device *dev)
+{
+	struct drm_device *drm = dev_get_drvdata(dev);
+
+	drm_mode_config_helper_resume(drm);
+
+	return 0;
+}
+
+static int __maybe_unused malidp_pm_suspend_late(struct device *dev)
+{
+	if (!pm_runtime_status_suspended(dev)) {
+		malidp_runtime_pm_suspend(dev);
+		pm_runtime_set_suspended(dev);
+	}
+	return 0;
+}
+
+static int __maybe_unused malidp_pm_resume_early(struct device *dev)
+{
+	malidp_runtime_pm_resume(dev);
+	pm_runtime_set_active(dev);
+	return 0;
+}
+
+static const struct dev_pm_ops malidp_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(malidp_pm_suspend, malidp_pm_resume) \
+	SET_LATE_SYSTEM_SLEEP_PM_OPS(malidp_pm_suspend_late, malidp_pm_resume_early) \
+	SET_RUNTIME_PM_OPS(malidp_runtime_pm_suspend, malidp_runtime_pm_resume, NULL)
+};
+
+static struct platform_driver malidp_platform_driver = {
+	.probe		= malidp_platform_probe,
+	.remove		= malidp_platform_remove,
+	.driver	= {
+		.name = "mali-dp",
+		.pm = &malidp_pm_ops,
+		.of_match_table	= malidp_drm_of_match,
+		.dev_groups = mali_dp_groups,
+	},
+};
+
+module_platform_driver(malidp_platform_driver);
+
+MODULE_AUTHOR("Liviu Dudau <Liviu.Dudau@arm.com>");
+MODULE_DESCRIPTION("ARM Mali DP DRM driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/gpu/drm/arm/malidp_drv.h b/drivers/gpu/drm/arm/malidp_drv.h
new file mode 100644
index 000000000..cdfddfabf
--- /dev/null
+++ b/drivers/gpu/drm/arm/malidp_drv.h
@@ -0,0 +1,105 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * (C) COPYRIGHT 2016 ARM Limited. All rights reserved.
+ * Author: Liviu Dudau <Liviu.Dudau@arm.com>
+ *
+ * ARM Mali DP500/DP550/DP650 KMS/DRM driver structures
+ */
+
+#ifndef __MALIDP_DRV_H__
+#define __MALIDP_DRV_H__
+
+#include <linux/mutex.h>
+#include <linux/wait.h>
+#include <linux/spinlock.h>
+
+#include <drm/drm_writeback.h>
+#include <drm/drm_encoder.h>
+
+#include "malidp_hw.h"
+
+#define MALIDP_CONFIG_VALID_INIT	0
+#define MALIDP_CONFIG_VALID_DONE	1
+#define MALIDP_CONFIG_START		0xd0
+
+struct malidp_error_stats {
+	s32 num_errors;
+	u32 last_error_status;
+	s64 last_error_vblank;
+};
+
+struct malidp_drm {
+	struct malidp_hw_device *dev;
+	struct drm_crtc crtc;
+	struct drm_writeback_connector mw_connector;
+	wait_queue_head_t wq;
+	struct drm_pending_vblank_event *event;
+	atomic_t config_valid;
+	u32 core_id;
+#ifdef CONFIG_DEBUG_FS
+	struct malidp_error_stats de_errors;
+	struct malidp_error_stats se_errors;
+	/* Protects errors stats */
+	spinlock_t errors_lock;
+#endif
+};
+
+#define crtc_to_malidp_device(x) container_of(x, struct malidp_drm, crtc)
+
+struct malidp_plane {
+	struct drm_plane base;
+	struct malidp_hw_device *hwdev;
+	const struct malidp_layer *layer;
+};
+
+enum mmu_prefetch_mode {
+	MALIDP_PREFETCH_MODE_NONE,
+	MALIDP_PREFETCH_MODE_PARTIAL,
+	MALIDP_PREFETCH_MODE_FULL,
+};
+
+struct malidp_plane_state {
+	struct drm_plane_state base;
+
+	/* size of the required rotation memory if plane is rotated */
+	u32 rotmem_size;
+	/* internal format ID */
+	u8 format;
+	u8 n_planes;
+	enum mmu_prefetch_mode mmu_prefetch_mode;
+	u32 mmu_prefetch_pgsize;
+};
+
+#define to_malidp_plane(x) container_of(x, struct malidp_plane, base)
+#define to_malidp_plane_state(x) container_of(x, struct malidp_plane_state, base)
+
+struct malidp_crtc_state {
+	struct drm_crtc_state base;
+	u32 gamma_coeffs[MALIDP_COEFFTAB_NUM_COEFFS];
+	u32 coloradj_coeffs[MALIDP_COLORADJ_NUM_COEFFS];
+	struct malidp_se_config scaler_config;
+	/* Bitfield of all the planes that have requested a scaled output. */
+	u8 scaled_planes_mask;
+};
+
+#define to_malidp_crtc_state(x) container_of(x, struct malidp_crtc_state, base)
+
+int malidp_de_planes_init(struct drm_device *drm);
+int malidp_crtc_init(struct drm_device *drm);
+
+bool malidp_hw_format_is_linear_only(u32 format);
+bool malidp_hw_format_is_afbc_only(u32 format);
+
+bool malidp_format_mod_supported(struct drm_device *drm,
+				 u32 format, u64 modifier);
+
+#ifdef CONFIG_DEBUG_FS
+void malidp_error(struct malidp_drm *malidp,
+		  struct malidp_error_stats *error_stats, u32 status,
+		  u64 vblank);
+#endif
+
+/* often used combination of rotational bits */
+#define MALIDP_ROTATED_MASK	(DRM_MODE_ROTATE_90 | DRM_MODE_ROTATE_270)
+
+#endif  /* __MALIDP_DRV_H__ */
diff --git a/drivers/gpu/drm/arm/malidp_hw.c b/drivers/gpu/drm/arm/malidp_hw.c
new file mode 100644
index 000000000..e9de542f9
--- /dev/null
+++ b/drivers/gpu/drm/arm/malidp_hw.c
@@ -0,0 +1,1392 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * (C) COPYRIGHT 2016 ARM Limited. All rights reserved.
+ * Author: Liviu Dudau <Liviu.Dudau@arm.com>
+ *
+ * ARM Mali DP500/DP550/DP650 hardware manipulation routines. This is where
+ * the difference between various versions of the hardware is being dealt with
+ * in an attempt to provide to the rest of the driver code a unified view
+ */
+
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/types.h>
+#include <linux/io.h>
+
+#include <video/videomode.h>
+#include <video/display_timing.h>
+
+#include <drm/drm_fourcc.h>
+#include <drm/drm_vblank.h>
+#include <drm/drm_print.h>
+
+#include "malidp_drv.h"
+#include "malidp_hw.h"
+#include "malidp_mw.h"
+
+enum {
+	MW_NOT_ENABLED = 0,	/* SE writeback not enabled */
+	MW_ONESHOT,		/* SE in one-shot mode for writeback */
+	MW_START,		/* SE started writeback */
+	MW_RESTART,		/* SE will start another writeback after this one */
+	MW_STOP,		/* SE needs to stop after this writeback */
+};
+
+static const struct malidp_format_id malidp500_de_formats[] = {
+	/*    fourcc,   layers supporting the format,     internal id  */
+	{ DRM_FORMAT_ARGB2101010, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2 | SE_MEMWRITE,  0 },
+	{ DRM_FORMAT_ABGR2101010, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2 | SE_MEMWRITE,  1 },
+	{ DRM_FORMAT_ARGB8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2,  2 },
+	{ DRM_FORMAT_ABGR8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2,  3 },
+	{ DRM_FORMAT_XRGB8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2 | SE_MEMWRITE,  4 },
+	{ DRM_FORMAT_XBGR8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2 | SE_MEMWRITE,  5 },
+	{ DRM_FORMAT_RGB888, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2,  6 },
+	{ DRM_FORMAT_BGR888, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2,  7 },
+	{ DRM_FORMAT_RGBA5551, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2,  8 },
+	{ DRM_FORMAT_ABGR1555, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2,  9 },
+	{ DRM_FORMAT_RGB565, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2, 10 },
+	{ DRM_FORMAT_BGR565, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2, 11 },
+	{ DRM_FORMAT_UYVY, DE_VIDEO1, 12 },
+	{ DRM_FORMAT_YUYV, DE_VIDEO1, 13 },
+	{ DRM_FORMAT_NV12, DE_VIDEO1 | SE_MEMWRITE, 14 },
+	{ DRM_FORMAT_YUV420, DE_VIDEO1, 15 },
+	{ DRM_FORMAT_XYUV8888, DE_VIDEO1, 16 },
+	/* These are supported with AFBC only */
+	{ DRM_FORMAT_YUV420_8BIT, DE_VIDEO1, 14 },
+	{ DRM_FORMAT_VUY888, DE_VIDEO1, 16 },
+	{ DRM_FORMAT_VUY101010, DE_VIDEO1, 17 },
+	{ DRM_FORMAT_YUV420_10BIT, DE_VIDEO1, 18 }
+};
+
+#define MALIDP_ID(__group, __format) \
+	((((__group) & 0x7) << 3) | ((__format) & 0x7))
+
+#define AFBC_YUV_422_FORMAT_ID	MALIDP_ID(5, 1)
+
+#define MALIDP_COMMON_FORMATS \
+	/*    fourcc,   layers supporting the format,      internal id   */ \
+	{ DRM_FORMAT_ARGB2101010, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | SE_MEMWRITE, MALIDP_ID(0, 0) }, \
+	{ DRM_FORMAT_ABGR2101010, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | SE_MEMWRITE, MALIDP_ID(0, 1) }, \
+	{ DRM_FORMAT_RGBA1010102, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | SE_MEMWRITE, MALIDP_ID(0, 2) }, \
+	{ DRM_FORMAT_BGRA1010102, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | SE_MEMWRITE, MALIDP_ID(0, 3) }, \
+	{ DRM_FORMAT_ARGB8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART, MALIDP_ID(1, 0) }, \
+	{ DRM_FORMAT_ABGR8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART, MALIDP_ID(1, 1) }, \
+	{ DRM_FORMAT_RGBA8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART, MALIDP_ID(1, 2) }, \
+	{ DRM_FORMAT_BGRA8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART, MALIDP_ID(1, 3) }, \
+	{ DRM_FORMAT_XRGB8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART | SE_MEMWRITE, MALIDP_ID(2, 0) }, \
+	{ DRM_FORMAT_XBGR8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART | SE_MEMWRITE, MALIDP_ID(2, 1) }, \
+	{ DRM_FORMAT_RGBX8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART | SE_MEMWRITE, MALIDP_ID(2, 2) }, \
+	{ DRM_FORMAT_BGRX8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART | SE_MEMWRITE, MALIDP_ID(2, 3) }, \
+	{ DRM_FORMAT_RGB888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | SE_MEMWRITE, MALIDP_ID(3, 0) }, \
+	{ DRM_FORMAT_BGR888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | SE_MEMWRITE, MALIDP_ID(3, 1) }, \
+	{ DRM_FORMAT_RGBA5551, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2, MALIDP_ID(4, 0) }, \
+	{ DRM_FORMAT_ABGR1555, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2, MALIDP_ID(4, 1) }, \
+	{ DRM_FORMAT_RGB565, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2, MALIDP_ID(4, 2) }, \
+	{ DRM_FORMAT_BGR565, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2, MALIDP_ID(4, 3) }, \
+	/* This is only supported with linear modifier */	\
+	{ DRM_FORMAT_XYUV8888, DE_VIDEO1 | DE_VIDEO2, MALIDP_ID(5, 0) },\
+	/* This is only supported with AFBC modifier */		\
+	{ DRM_FORMAT_VUY888, DE_VIDEO1 | DE_VIDEO2, MALIDP_ID(5, 0) }, \
+	{ DRM_FORMAT_YUYV, DE_VIDEO1 | DE_VIDEO2, MALIDP_ID(5, 2) },	\
+	/* This is only supported with linear modifier */ \
+	{ DRM_FORMAT_UYVY, DE_VIDEO1 | DE_VIDEO2, MALIDP_ID(5, 3) },	\
+	{ DRM_FORMAT_NV12, DE_VIDEO1 | DE_VIDEO2 | SE_MEMWRITE, MALIDP_ID(5, 6) },	\
+	/* This is only supported with AFBC modifier */ \
+	{ DRM_FORMAT_YUV420_8BIT, DE_VIDEO1 | DE_VIDEO2, MALIDP_ID(5, 6) }, \
+	{ DRM_FORMAT_YUV420, DE_VIDEO1 | DE_VIDEO2, MALIDP_ID(5, 7) }, \
+	/* This is only supported with linear modifier */ \
+	{ DRM_FORMAT_XVYU2101010, DE_VIDEO1 | DE_VIDEO2, MALIDP_ID(6, 0)}, \
+	/* This is only supported with AFBC modifier */ \
+	{ DRM_FORMAT_VUY101010, DE_VIDEO1 | DE_VIDEO2, MALIDP_ID(6, 0)}, \
+	{ DRM_FORMAT_X0L2, DE_VIDEO1 | DE_VIDEO2, MALIDP_ID(6, 6)}, \
+	/* This is only supported with AFBC modifier */ \
+	{ DRM_FORMAT_YUV420_10BIT, DE_VIDEO1 | DE_VIDEO2, MALIDP_ID(6, 7)}, \
+	{ DRM_FORMAT_P010, DE_VIDEO1 | DE_VIDEO2, MALIDP_ID(6, 7)}
+
+static const struct malidp_format_id malidp550_de_formats[] = {
+	MALIDP_COMMON_FORMATS,
+};
+
+static const struct malidp_format_id malidp650_de_formats[] = {
+	MALIDP_COMMON_FORMATS,
+	{ DRM_FORMAT_X0L0, DE_VIDEO1 | DE_VIDEO2, MALIDP_ID(5, 4)},
+};
+
+static const struct malidp_layer malidp500_layers[] = {
+	/* id, base address, fb pointer address base, stride offset,
+	 *	yuv2rgb matrix offset, mmu control register offset, rotation_features
+	 */
+	{ DE_VIDEO1, MALIDP500_DE_LV_BASE, MALIDP500_DE_LV_PTR_BASE,
+		MALIDP_DE_LV_STRIDE0, MALIDP500_LV_YUV2RGB, 0, ROTATE_ANY,
+		MALIDP500_DE_LV_AD_CTRL },
+	{ DE_GRAPHICS1, MALIDP500_DE_LG1_BASE, MALIDP500_DE_LG1_PTR_BASE,
+		MALIDP_DE_LG_STRIDE, 0, 0, ROTATE_ANY,
+		MALIDP500_DE_LG1_AD_CTRL },
+	{ DE_GRAPHICS2, MALIDP500_DE_LG2_BASE, MALIDP500_DE_LG2_PTR_BASE,
+		MALIDP_DE_LG_STRIDE, 0, 0, ROTATE_ANY,
+		MALIDP500_DE_LG2_AD_CTRL },
+};
+
+static const struct malidp_layer malidp550_layers[] = {
+	/* id, base address, fb pointer address base, stride offset,
+	 *	yuv2rgb matrix offset, mmu control register offset, rotation_features
+	 */
+	{ DE_VIDEO1, MALIDP550_DE_LV1_BASE, MALIDP550_DE_LV1_PTR_BASE,
+		MALIDP_DE_LV_STRIDE0, MALIDP550_LV_YUV2RGB, 0, ROTATE_ANY,
+		MALIDP550_DE_LV1_AD_CTRL },
+	{ DE_GRAPHICS1, MALIDP550_DE_LG_BASE, MALIDP550_DE_LG_PTR_BASE,
+		MALIDP_DE_LG_STRIDE, 0, 0, ROTATE_ANY,
+		MALIDP550_DE_LG_AD_CTRL },
+	{ DE_VIDEO2, MALIDP550_DE_LV2_BASE, MALIDP550_DE_LV2_PTR_BASE,
+		MALIDP_DE_LV_STRIDE0, MALIDP550_LV_YUV2RGB, 0, ROTATE_ANY,
+		MALIDP550_DE_LV2_AD_CTRL },
+	{ DE_SMART, MALIDP550_DE_LS_BASE, MALIDP550_DE_LS_PTR_BASE,
+		MALIDP550_DE_LS_R1_STRIDE, 0, 0, ROTATE_NONE, 0 },
+};
+
+static const struct malidp_layer malidp650_layers[] = {
+	/* id, base address, fb pointer address base, stride offset,
+	 *	yuv2rgb matrix offset, mmu control register offset,
+	 *	rotation_features
+	 */
+	{ DE_VIDEO1, MALIDP550_DE_LV1_BASE, MALIDP550_DE_LV1_PTR_BASE,
+		MALIDP_DE_LV_STRIDE0, MALIDP550_LV_YUV2RGB,
+		MALIDP650_DE_LV_MMU_CTRL, ROTATE_ANY,
+		MALIDP550_DE_LV1_AD_CTRL },
+	{ DE_GRAPHICS1, MALIDP550_DE_LG_BASE, MALIDP550_DE_LG_PTR_BASE,
+		MALIDP_DE_LG_STRIDE, 0, MALIDP650_DE_LG_MMU_CTRL,
+		ROTATE_COMPRESSED, MALIDP550_DE_LG_AD_CTRL },
+	{ DE_VIDEO2, MALIDP550_DE_LV2_BASE, MALIDP550_DE_LV2_PTR_BASE,
+		MALIDP_DE_LV_STRIDE0, MALIDP550_LV_YUV2RGB,
+		MALIDP650_DE_LV_MMU_CTRL, ROTATE_ANY,
+		MALIDP550_DE_LV2_AD_CTRL },
+	{ DE_SMART, MALIDP550_DE_LS_BASE, MALIDP550_DE_LS_PTR_BASE,
+		MALIDP550_DE_LS_R1_STRIDE, 0, MALIDP650_DE_LS_MMU_CTRL,
+		ROTATE_NONE, 0 },
+};
+
+const u64 malidp_format_modifiers[] = {
+	/* All RGB formats (except XRGB, RGBX, XBGR, BGRX) */
+	DRM_FORMAT_MOD_ARM_AFBC(AFBC_SIZE_16X16 | AFBC_YTR | AFBC_SPARSE),
+	DRM_FORMAT_MOD_ARM_AFBC(AFBC_SIZE_16X16 | AFBC_YTR),
+
+	/* All RGB formats > 16bpp (except XRGB, RGBX, XBGR, BGRX) */
+	DRM_FORMAT_MOD_ARM_AFBC(AFBC_SIZE_16X16 | AFBC_YTR | AFBC_SPARSE | AFBC_SPLIT),
+
+	/* All 8 or 10 bit YUV 444 formats. */
+	/* In DP550, 10 bit YUV 420 format also supported */
+	DRM_FORMAT_MOD_ARM_AFBC(AFBC_SIZE_16X16 | AFBC_SPARSE | AFBC_SPLIT),
+
+	/* YUV 420, 422 P1 8 bit and YUV 444 8 bit/10 bit formats */
+	DRM_FORMAT_MOD_ARM_AFBC(AFBC_SIZE_16X16 | AFBC_SPARSE),
+	DRM_FORMAT_MOD_ARM_AFBC(AFBC_SIZE_16X16),
+
+	/* YUV 420, 422 P1 8, 10 bit formats */
+	DRM_FORMAT_MOD_ARM_AFBC(AFBC_SIZE_16X16 | AFBC_CBR | AFBC_SPARSE),
+	DRM_FORMAT_MOD_ARM_AFBC(AFBC_SIZE_16X16 | AFBC_CBR),
+
+	/* All formats */
+	DRM_FORMAT_MOD_LINEAR,
+
+	DRM_FORMAT_MOD_INVALID
+};
+
+#define SE_N_SCALING_COEFFS	96
+static const u16 dp500_se_scaling_coeffs[][SE_N_SCALING_COEFFS] = {
+	[MALIDP_UPSCALING_COEFFS - 1] = {
+		0x0000, 0x0001, 0x0007, 0x0011, 0x001e, 0x002e, 0x003f, 0x0052,
+		0x0064, 0x0073, 0x007d, 0x0080, 0x007a, 0x006c, 0x0053, 0x002f,
+		0x0000, 0x3fc6, 0x3f83, 0x3f39, 0x3eea, 0x3e9b, 0x3e4f, 0x3e0a,
+		0x3dd4, 0x3db0, 0x3da2, 0x3db1, 0x3dde, 0x3e2f, 0x3ea5, 0x3f40,
+		0x0000, 0x00e5, 0x01ee, 0x0315, 0x0456, 0x05aa, 0x0709, 0x086c,
+		0x09c9, 0x0b15, 0x0c4a, 0x0d5d, 0x0e4a, 0x0f06, 0x0f91, 0x0fe5,
+		0x1000, 0x0fe5, 0x0f91, 0x0f06, 0x0e4a, 0x0d5d, 0x0c4a, 0x0b15,
+		0x09c9, 0x086c, 0x0709, 0x05aa, 0x0456, 0x0315, 0x01ee, 0x00e5,
+		0x0000, 0x3f40, 0x3ea5, 0x3e2f, 0x3dde, 0x3db1, 0x3da2, 0x3db0,
+		0x3dd4, 0x3e0a, 0x3e4f, 0x3e9b, 0x3eea, 0x3f39, 0x3f83, 0x3fc6,
+		0x0000, 0x002f, 0x0053, 0x006c, 0x007a, 0x0080, 0x007d, 0x0073,
+		0x0064, 0x0052, 0x003f, 0x002e, 0x001e, 0x0011, 0x0007, 0x0001
+	},
+	[MALIDP_DOWNSCALING_1_5_COEFFS - 1] = {
+		0x0059, 0x004f, 0x0041, 0x002e, 0x0016, 0x3ffb, 0x3fd9, 0x3fb4,
+		0x3f8c, 0x3f62, 0x3f36, 0x3f09, 0x3edd, 0x3eb3, 0x3e8d, 0x3e6c,
+		0x3e52, 0x3e3f, 0x3e35, 0x3e37, 0x3e46, 0x3e61, 0x3e8c, 0x3ec5,
+		0x3f0f, 0x3f68, 0x3fd1, 0x004a, 0x00d3, 0x0169, 0x020b, 0x02b8,
+		0x036e, 0x042d, 0x04f2, 0x05b9, 0x0681, 0x0745, 0x0803, 0x08ba,
+		0x0965, 0x0a03, 0x0a91, 0x0b0d, 0x0b75, 0x0bc6, 0x0c00, 0x0c20,
+		0x0c28, 0x0c20, 0x0c00, 0x0bc6, 0x0b75, 0x0b0d, 0x0a91, 0x0a03,
+		0x0965, 0x08ba, 0x0803, 0x0745, 0x0681, 0x05b9, 0x04f2, 0x042d,
+		0x036e, 0x02b8, 0x020b, 0x0169, 0x00d3, 0x004a, 0x3fd1, 0x3f68,
+		0x3f0f, 0x3ec5, 0x3e8c, 0x3e61, 0x3e46, 0x3e37, 0x3e35, 0x3e3f,
+		0x3e52, 0x3e6c, 0x3e8d, 0x3eb3, 0x3edd, 0x3f09, 0x3f36, 0x3f62,
+		0x3f8c, 0x3fb4, 0x3fd9, 0x3ffb, 0x0016, 0x002e, 0x0041, 0x004f
+	},
+	[MALIDP_DOWNSCALING_2_COEFFS - 1] = {
+		0x3f19, 0x3f03, 0x3ef0, 0x3edf, 0x3ed0, 0x3ec5, 0x3ebd, 0x3eb9,
+		0x3eb9, 0x3ebf, 0x3eca, 0x3ed9, 0x3eef, 0x3f0a, 0x3f2c, 0x3f52,
+		0x3f7f, 0x3fb0, 0x3fe8, 0x0026, 0x006a, 0x00b4, 0x0103, 0x0158,
+		0x01b1, 0x020d, 0x026c, 0x02cd, 0x032f, 0x0392, 0x03f4, 0x0455,
+		0x04b4, 0x051e, 0x0585, 0x05eb, 0x064c, 0x06a8, 0x06fe, 0x074e,
+		0x0796, 0x07d5, 0x080c, 0x0839, 0x085c, 0x0875, 0x0882, 0x0887,
+		0x0881, 0x0887, 0x0882, 0x0875, 0x085c, 0x0839, 0x080c, 0x07d5,
+		0x0796, 0x074e, 0x06fe, 0x06a8, 0x064c, 0x05eb, 0x0585, 0x051e,
+		0x04b4, 0x0455, 0x03f4, 0x0392, 0x032f, 0x02cd, 0x026c, 0x020d,
+		0x01b1, 0x0158, 0x0103, 0x00b4, 0x006a, 0x0026, 0x3fe8, 0x3fb0,
+		0x3f7f, 0x3f52, 0x3f2c, 0x3f0a, 0x3eef, 0x3ed9, 0x3eca, 0x3ebf,
+		0x3eb9, 0x3eb9, 0x3ebd, 0x3ec5, 0x3ed0, 0x3edf, 0x3ef0, 0x3f03
+	},
+	[MALIDP_DOWNSCALING_2_75_COEFFS - 1] = {
+		0x3f51, 0x3f60, 0x3f71, 0x3f84, 0x3f98, 0x3faf, 0x3fc8, 0x3fe3,
+		0x0000, 0x001f, 0x0040, 0x0064, 0x008a, 0x00b1, 0x00da, 0x0106,
+		0x0133, 0x0160, 0x018e, 0x01bd, 0x01ec, 0x021d, 0x024e, 0x0280,
+		0x02b2, 0x02e4, 0x0317, 0x0349, 0x037c, 0x03ad, 0x03df, 0x0410,
+		0x0440, 0x0468, 0x048f, 0x04b3, 0x04d6, 0x04f8, 0x0516, 0x0533,
+		0x054e, 0x0566, 0x057c, 0x0590, 0x05a0, 0x05ae, 0x05ba, 0x05c3,
+		0x05c9, 0x05c3, 0x05ba, 0x05ae, 0x05a0, 0x0590, 0x057c, 0x0566,
+		0x054e, 0x0533, 0x0516, 0x04f8, 0x04d6, 0x04b3, 0x048f, 0x0468,
+		0x0440, 0x0410, 0x03df, 0x03ad, 0x037c, 0x0349, 0x0317, 0x02e4,
+		0x02b2, 0x0280, 0x024e, 0x021d, 0x01ec, 0x01bd, 0x018e, 0x0160,
+		0x0133, 0x0106, 0x00da, 0x00b1, 0x008a, 0x0064, 0x0040, 0x001f,
+		0x0000, 0x3fe3, 0x3fc8, 0x3faf, 0x3f98, 0x3f84, 0x3f71, 0x3f60
+	},
+	[MALIDP_DOWNSCALING_4_COEFFS - 1] = {
+		0x0094, 0x00a9, 0x00be, 0x00d4, 0x00ea, 0x0101, 0x0118, 0x012f,
+		0x0148, 0x0160, 0x017a, 0x0193, 0x01ae, 0x01c8, 0x01e4, 0x01ff,
+		0x021c, 0x0233, 0x024a, 0x0261, 0x0278, 0x028f, 0x02a6, 0x02bd,
+		0x02d4, 0x02eb, 0x0302, 0x0319, 0x032f, 0x0346, 0x035d, 0x0374,
+		0x038a, 0x0397, 0x03a3, 0x03af, 0x03bb, 0x03c6, 0x03d1, 0x03db,
+		0x03e4, 0x03ed, 0x03f6, 0x03fe, 0x0406, 0x040d, 0x0414, 0x041a,
+		0x0420, 0x041a, 0x0414, 0x040d, 0x0406, 0x03fe, 0x03f6, 0x03ed,
+		0x03e4, 0x03db, 0x03d1, 0x03c6, 0x03bb, 0x03af, 0x03a3, 0x0397,
+		0x038a, 0x0374, 0x035d, 0x0346, 0x032f, 0x0319, 0x0302, 0x02eb,
+		0x02d4, 0x02bd, 0x02a6, 0x028f, 0x0278, 0x0261, 0x024a, 0x0233,
+		0x021c, 0x01ff, 0x01e4, 0x01c8, 0x01ae, 0x0193, 0x017a, 0x0160,
+		0x0148, 0x012f, 0x0118, 0x0101, 0x00ea, 0x00d4, 0x00be, 0x00a9
+	},
+};
+
+#define MALIDP_DE_DEFAULT_PREFETCH_START	5
+
+static int malidp500_query_hw(struct malidp_hw_device *hwdev)
+{
+	u32 conf = malidp_hw_read(hwdev, MALIDP500_CONFIG_ID);
+	/* bit 4 of the CONFIG_ID register holds the line size multiplier */
+	u8 ln_size_mult = conf & 0x10 ? 2 : 1;
+
+	hwdev->min_line_size = 2;
+	hwdev->max_line_size = SZ_2K * ln_size_mult;
+	hwdev->rotation_memory[0] = SZ_1K * 64 * ln_size_mult;
+	hwdev->rotation_memory[1] = 0; /* no second rotation memory bank */
+
+	return 0;
+}
+
+static void malidp500_enter_config_mode(struct malidp_hw_device *hwdev)
+{
+	u32 status, count = 100;
+
+	malidp_hw_setbits(hwdev, MALIDP500_DC_CONFIG_REQ, MALIDP500_DC_CONTROL);
+	while (count) {
+		status = malidp_hw_read(hwdev, hwdev->hw->map.dc_base + MALIDP_REG_STATUS);
+		if ((status & MALIDP500_DC_CONFIG_REQ) == MALIDP500_DC_CONFIG_REQ)
+			break;
+		/*
+		 * entering config mode can take as long as the rendering
+		 * of a full frame, hence the long sleep here
+		 */
+		usleep_range(1000, 10000);
+		count--;
+	}
+	WARN(count == 0, "timeout while entering config mode");
+}
+
+static void malidp500_leave_config_mode(struct malidp_hw_device *hwdev)
+{
+	u32 status, count = 100;
+
+	malidp_hw_clearbits(hwdev, MALIDP_CFG_VALID, MALIDP500_CONFIG_VALID);
+	malidp_hw_clearbits(hwdev, MALIDP500_DC_CONFIG_REQ, MALIDP500_DC_CONTROL);
+	while (count) {
+		status = malidp_hw_read(hwdev, hwdev->hw->map.dc_base + MALIDP_REG_STATUS);
+		if ((status & MALIDP500_DC_CONFIG_REQ) == 0)
+			break;
+		usleep_range(100, 1000);
+		count--;
+	}
+	WARN(count == 0, "timeout while leaving config mode");
+}
+
+static bool malidp500_in_config_mode(struct malidp_hw_device *hwdev)
+{
+	u32 status;
+
+	status = malidp_hw_read(hwdev, hwdev->hw->map.dc_base + MALIDP_REG_STATUS);
+	if ((status & MALIDP500_DC_CONFIG_REQ) == MALIDP500_DC_CONFIG_REQ)
+		return true;
+
+	return false;
+}
+
+static void malidp500_set_config_valid(struct malidp_hw_device *hwdev, u8 value)
+{
+	if (value)
+		malidp_hw_setbits(hwdev, MALIDP_CFG_VALID, MALIDP500_CONFIG_VALID);
+	else
+		malidp_hw_clearbits(hwdev, MALIDP_CFG_VALID, MALIDP500_CONFIG_VALID);
+}
+
+static void malidp500_modeset(struct malidp_hw_device *hwdev, struct videomode *mode)
+{
+	u32 val = 0;
+
+	malidp_hw_write(hwdev, hwdev->output_color_depth,
+		hwdev->hw->map.out_depth_base);
+	malidp_hw_clearbits(hwdev, MALIDP500_DC_CLEAR_MASK, MALIDP500_DC_CONTROL);
+	if (mode->flags & DISPLAY_FLAGS_HSYNC_HIGH)
+		val |= MALIDP500_HSYNCPOL;
+	if (mode->flags & DISPLAY_FLAGS_VSYNC_HIGH)
+		val |= MALIDP500_VSYNCPOL;
+	val |= MALIDP_DE_DEFAULT_PREFETCH_START;
+	malidp_hw_setbits(hwdev, val, MALIDP500_DC_CONTROL);
+
+	/*
+	 * Mali-DP500 encodes the background color like this:
+	 *    - red   @ MALIDP500_BGND_COLOR[12:0]
+	 *    - green @ MALIDP500_BGND_COLOR[27:16]
+	 *    - blue  @ (MALIDP500_BGND_COLOR + 4)[12:0]
+	 */
+	val = ((MALIDP_BGND_COLOR_G & 0xfff) << 16) |
+	      (MALIDP_BGND_COLOR_R & 0xfff);
+	malidp_hw_write(hwdev, val, MALIDP500_BGND_COLOR);
+	malidp_hw_write(hwdev, MALIDP_BGND_COLOR_B, MALIDP500_BGND_COLOR + 4);
+
+	val = MALIDP_DE_H_FRONTPORCH(mode->hfront_porch) |
+		MALIDP_DE_H_BACKPORCH(mode->hback_porch);
+	malidp_hw_write(hwdev, val, MALIDP500_TIMINGS_BASE + MALIDP_DE_H_TIMINGS);
+
+	val = MALIDP500_DE_V_FRONTPORCH(mode->vfront_porch) |
+		MALIDP_DE_V_BACKPORCH(mode->vback_porch);
+	malidp_hw_write(hwdev, val, MALIDP500_TIMINGS_BASE + MALIDP_DE_V_TIMINGS);
+
+	val = MALIDP_DE_H_SYNCWIDTH(mode->hsync_len) |
+		MALIDP_DE_V_SYNCWIDTH(mode->vsync_len);
+	malidp_hw_write(hwdev, val, MALIDP500_TIMINGS_BASE + MALIDP_DE_SYNC_WIDTH);
+
+	val = MALIDP_DE_H_ACTIVE(mode->hactive) | MALIDP_DE_V_ACTIVE(mode->vactive);
+	malidp_hw_write(hwdev, val, MALIDP500_TIMINGS_BASE + MALIDP_DE_HV_ACTIVE);
+
+	if (mode->flags & DISPLAY_FLAGS_INTERLACED)
+		malidp_hw_setbits(hwdev, MALIDP_DISP_FUNC_ILACED, MALIDP_DE_DISPLAY_FUNC);
+	else
+		malidp_hw_clearbits(hwdev, MALIDP_DISP_FUNC_ILACED, MALIDP_DE_DISPLAY_FUNC);
+
+	/*
+	 * Program the RQoS register to avoid high resolutions flicker
+	 * issue on the LS1028A.
+	 */
+	if (hwdev->arqos_value) {
+		val = hwdev->arqos_value;
+		malidp_hw_setbits(hwdev, val, MALIDP500_RQOS_QUALITY);
+	}
+}
+
+int malidp_format_get_bpp(u32 fmt)
+{
+	const struct drm_format_info *info = drm_format_info(fmt);
+	int bpp = info->cpp[0] * 8;
+
+	if (bpp == 0) {
+		switch (fmt) {
+		case DRM_FORMAT_VUY101010:
+			bpp = 30;
+			break;
+		case DRM_FORMAT_YUV420_10BIT:
+			bpp = 15;
+			break;
+		case DRM_FORMAT_YUV420_8BIT:
+			bpp = 12;
+			break;
+		default:
+			bpp = 0;
+		}
+	}
+
+	return bpp;
+}
+
+static int malidp500_rotmem_required(struct malidp_hw_device *hwdev, u16 w,
+				     u16 h, u32 fmt, bool has_modifier)
+{
+	/*
+	 * Each layer needs enough rotation memory to fit 8 lines
+	 * worth of pixel data. Required size is then:
+	 *    size = rotated_width * (bpp / 8) * 8;
+	 */
+	int bpp = malidp_format_get_bpp(fmt);
+
+	return w * bpp;
+}
+
+static void malidp500_se_write_pp_coefftab(struct malidp_hw_device *hwdev,
+					   u32 direction,
+					   u16 addr,
+					   u8 coeffs_id)
+{
+	int i;
+	u16 scaling_control = MALIDP500_SE_CONTROL + MALIDP_SE_SCALING_CONTROL;
+
+	malidp_hw_write(hwdev,
+			direction | (addr & MALIDP_SE_COEFFTAB_ADDR_MASK),
+			scaling_control + MALIDP_SE_COEFFTAB_ADDR);
+	for (i = 0; i < ARRAY_SIZE(dp500_se_scaling_coeffs); ++i)
+		malidp_hw_write(hwdev, MALIDP_SE_SET_COEFFTAB_DATA(
+				dp500_se_scaling_coeffs[coeffs_id][i]),
+				scaling_control + MALIDP_SE_COEFFTAB_DATA);
+}
+
+static int malidp500_se_set_scaling_coeffs(struct malidp_hw_device *hwdev,
+					   struct malidp_se_config *se_config,
+					   struct malidp_se_config *old_config)
+{
+	/* Get array indices into dp500_se_scaling_coeffs. */
+	u8 h = (u8)se_config->hcoeff - 1;
+	u8 v = (u8)se_config->vcoeff - 1;
+
+	if (WARN_ON(h >= ARRAY_SIZE(dp500_se_scaling_coeffs) ||
+		    v >= ARRAY_SIZE(dp500_se_scaling_coeffs)))
+		return -EINVAL;
+
+	if ((h == v) && (se_config->hcoeff != old_config->hcoeff ||
+			 se_config->vcoeff != old_config->vcoeff)) {
+		malidp500_se_write_pp_coefftab(hwdev,
+					       (MALIDP_SE_V_COEFFTAB |
+						MALIDP_SE_H_COEFFTAB),
+					       0, v);
+	} else {
+		if (se_config->vcoeff != old_config->vcoeff)
+			malidp500_se_write_pp_coefftab(hwdev,
+						       MALIDP_SE_V_COEFFTAB,
+						       0, v);
+		if (se_config->hcoeff != old_config->hcoeff)
+			malidp500_se_write_pp_coefftab(hwdev,
+						       MALIDP_SE_H_COEFFTAB,
+						       0, h);
+	}
+
+	return 0;
+}
+
+static long malidp500_se_calc_mclk(struct malidp_hw_device *hwdev,
+				   struct malidp_se_config *se_config,
+				   struct videomode *vm)
+{
+	unsigned long mclk;
+	unsigned long pxlclk = vm->pixelclock; /* Hz */
+	unsigned long htotal = vm->hactive + vm->hfront_porch +
+			       vm->hback_porch + vm->hsync_len;
+	unsigned long input_size = se_config->input_w * se_config->input_h;
+	unsigned long a = 10;
+	long ret;
+
+	/*
+	 * mclk = max(a, 1.5) * pxlclk
+	 *
+	 * To avoid float calculaiton, using 15 instead of 1.5 and div by
+	 * 10 to get mclk.
+	 */
+	if (se_config->scale_enable) {
+		a = 15 * input_size / (htotal * se_config->output_h);
+		if (a < 15)
+			a = 15;
+	}
+	mclk = a * pxlclk / 10;
+	ret = clk_get_rate(hwdev->mclk);
+	if (ret < mclk) {
+		DRM_DEBUG_DRIVER("mclk requirement of %lu kHz can't be met.\n",
+				 mclk / 1000);
+		return -EINVAL;
+	}
+	return ret;
+}
+
+static int malidp500_enable_memwrite(struct malidp_hw_device *hwdev,
+				     dma_addr_t *addrs, s32 *pitches,
+				     int num_planes, u16 w, u16 h, u32 fmt_id,
+				     const s16 *rgb2yuv_coeffs)
+{
+	u32 base = MALIDP500_SE_MEMWRITE_BASE;
+	u32 de_base = malidp_get_block_base(hwdev, MALIDP_DE_BLOCK);
+
+	/* enable the scaling engine block */
+	malidp_hw_setbits(hwdev, MALIDP_SCALE_ENGINE_EN, de_base + MALIDP_DE_DISPLAY_FUNC);
+
+	/* restart the writeback if already enabled */
+	if (hwdev->mw_state != MW_NOT_ENABLED)
+		hwdev->mw_state = MW_RESTART;
+	else
+		hwdev->mw_state = MW_START;
+
+	malidp_hw_write(hwdev, fmt_id, base + MALIDP_MW_FORMAT);
+	switch (num_planes) {
+	case 2:
+		malidp_hw_write(hwdev, lower_32_bits(addrs[1]), base + MALIDP_MW_P2_PTR_LOW);
+		malidp_hw_write(hwdev, upper_32_bits(addrs[1]), base + MALIDP_MW_P2_PTR_HIGH);
+		malidp_hw_write(hwdev, pitches[1], base + MALIDP_MW_P2_STRIDE);
+		fallthrough;
+	case 1:
+		malidp_hw_write(hwdev, lower_32_bits(addrs[0]), base + MALIDP_MW_P1_PTR_LOW);
+		malidp_hw_write(hwdev, upper_32_bits(addrs[0]), base + MALIDP_MW_P1_PTR_HIGH);
+		malidp_hw_write(hwdev, pitches[0], base + MALIDP_MW_P1_STRIDE);
+		break;
+	default:
+		WARN(1, "Invalid number of planes");
+	}
+
+	malidp_hw_write(hwdev, MALIDP_DE_H_ACTIVE(w) | MALIDP_DE_V_ACTIVE(h),
+			MALIDP500_SE_MEMWRITE_OUT_SIZE);
+
+	if (rgb2yuv_coeffs) {
+		int i;
+
+		for (i = 0; i < MALIDP_COLORADJ_NUM_COEFFS; i++) {
+			malidp_hw_write(hwdev, rgb2yuv_coeffs[i],
+					MALIDP500_SE_RGB_YUV_COEFFS + i * 4);
+		}
+	}
+
+	malidp_hw_setbits(hwdev, MALIDP_SE_MEMWRITE_EN, MALIDP500_SE_CONTROL);
+
+	return 0;
+}
+
+static void malidp500_disable_memwrite(struct malidp_hw_device *hwdev)
+{
+	u32 base = malidp_get_block_base(hwdev, MALIDP_DE_BLOCK);
+
+	if (hwdev->mw_state == MW_START || hwdev->mw_state == MW_RESTART)
+		hwdev->mw_state = MW_STOP;
+	malidp_hw_clearbits(hwdev, MALIDP_SE_MEMWRITE_EN, MALIDP500_SE_CONTROL);
+	malidp_hw_clearbits(hwdev, MALIDP_SCALE_ENGINE_EN, base + MALIDP_DE_DISPLAY_FUNC);
+}
+
+static int malidp550_query_hw(struct malidp_hw_device *hwdev)
+{
+	u32 conf = malidp_hw_read(hwdev, MALIDP550_CONFIG_ID);
+	u8 ln_size = (conf >> 4) & 0x3, rsize;
+
+	hwdev->min_line_size = 2;
+
+	switch (ln_size) {
+	case 0:
+		hwdev->max_line_size = SZ_2K;
+		/* two banks of 64KB for rotation memory */
+		rsize = 64;
+		break;
+	case 1:
+		hwdev->max_line_size = SZ_4K;
+		/* two banks of 128KB for rotation memory */
+		rsize = 128;
+		break;
+	case 2:
+		hwdev->max_line_size = 1280;
+		/* two banks of 40KB for rotation memory */
+		rsize = 40;
+		break;
+	case 3:
+		/* reserved value */
+		hwdev->max_line_size = 0;
+		return -EINVAL;
+	}
+
+	hwdev->rotation_memory[0] = hwdev->rotation_memory[1] = rsize * SZ_1K;
+	return 0;
+}
+
+static void malidp550_enter_config_mode(struct malidp_hw_device *hwdev)
+{
+	u32 status, count = 100;
+
+	malidp_hw_setbits(hwdev, MALIDP550_DC_CONFIG_REQ, MALIDP550_DC_CONTROL);
+	while (count) {
+		status = malidp_hw_read(hwdev, hwdev->hw->map.dc_base + MALIDP_REG_STATUS);
+		if ((status & MALIDP550_DC_CONFIG_REQ) == MALIDP550_DC_CONFIG_REQ)
+			break;
+		/*
+		 * entering config mode can take as long as the rendering
+		 * of a full frame, hence the long sleep here
+		 */
+		usleep_range(1000, 10000);
+		count--;
+	}
+	WARN(count == 0, "timeout while entering config mode");
+}
+
+static void malidp550_leave_config_mode(struct malidp_hw_device *hwdev)
+{
+	u32 status, count = 100;
+
+	malidp_hw_clearbits(hwdev, MALIDP_CFG_VALID, MALIDP550_CONFIG_VALID);
+	malidp_hw_clearbits(hwdev, MALIDP550_DC_CONFIG_REQ, MALIDP550_DC_CONTROL);
+	while (count) {
+		status = malidp_hw_read(hwdev, hwdev->hw->map.dc_base + MALIDP_REG_STATUS);
+		if ((status & MALIDP550_DC_CONFIG_REQ) == 0)
+			break;
+		usleep_range(100, 1000);
+		count--;
+	}
+	WARN(count == 0, "timeout while leaving config mode");
+}
+
+static bool malidp550_in_config_mode(struct malidp_hw_device *hwdev)
+{
+	u32 status;
+
+	status = malidp_hw_read(hwdev, hwdev->hw->map.dc_base + MALIDP_REG_STATUS);
+	if ((status & MALIDP550_DC_CONFIG_REQ) == MALIDP550_DC_CONFIG_REQ)
+		return true;
+
+	return false;
+}
+
+static void malidp550_set_config_valid(struct malidp_hw_device *hwdev, u8 value)
+{
+	if (value)
+		malidp_hw_setbits(hwdev, MALIDP_CFG_VALID, MALIDP550_CONFIG_VALID);
+	else
+		malidp_hw_clearbits(hwdev, MALIDP_CFG_VALID, MALIDP550_CONFIG_VALID);
+}
+
+static void malidp550_modeset(struct malidp_hw_device *hwdev, struct videomode *mode)
+{
+	u32 val = MALIDP_DE_DEFAULT_PREFETCH_START;
+
+	malidp_hw_write(hwdev, hwdev->output_color_depth,
+		hwdev->hw->map.out_depth_base);
+	malidp_hw_write(hwdev, val, MALIDP550_DE_CONTROL);
+	/*
+	 * Mali-DP550 and Mali-DP650 encode the background color like this:
+	 *   - red   @ MALIDP550_DE_BGND_COLOR[23:16]
+	 *   - green @ MALIDP550_DE_BGND_COLOR[15:8]
+	 *   - blue  @ MALIDP550_DE_BGND_COLOR[7:0]
+	 *
+	 * We need to truncate the least significant 4 bits from the default
+	 * MALIDP_BGND_COLOR_x values
+	 */
+	val = (((MALIDP_BGND_COLOR_R >> 4) & 0xff) << 16) |
+	      (((MALIDP_BGND_COLOR_G >> 4) & 0xff) << 8) |
+	      ((MALIDP_BGND_COLOR_B >> 4) & 0xff);
+	malidp_hw_write(hwdev, val, MALIDP550_DE_BGND_COLOR);
+
+	val = MALIDP_DE_H_FRONTPORCH(mode->hfront_porch) |
+		MALIDP_DE_H_BACKPORCH(mode->hback_porch);
+	malidp_hw_write(hwdev, val, MALIDP550_TIMINGS_BASE + MALIDP_DE_H_TIMINGS);
+
+	val = MALIDP550_DE_V_FRONTPORCH(mode->vfront_porch) |
+		MALIDP_DE_V_BACKPORCH(mode->vback_porch);
+	malidp_hw_write(hwdev, val, MALIDP550_TIMINGS_BASE + MALIDP_DE_V_TIMINGS);
+
+	val = MALIDP_DE_H_SYNCWIDTH(mode->hsync_len) |
+		MALIDP_DE_V_SYNCWIDTH(mode->vsync_len);
+	if (mode->flags & DISPLAY_FLAGS_HSYNC_HIGH)
+		val |= MALIDP550_HSYNCPOL;
+	if (mode->flags & DISPLAY_FLAGS_VSYNC_HIGH)
+		val |= MALIDP550_VSYNCPOL;
+	malidp_hw_write(hwdev, val, MALIDP550_TIMINGS_BASE + MALIDP_DE_SYNC_WIDTH);
+
+	val = MALIDP_DE_H_ACTIVE(mode->hactive) | MALIDP_DE_V_ACTIVE(mode->vactive);
+	malidp_hw_write(hwdev, val, MALIDP550_TIMINGS_BASE + MALIDP_DE_HV_ACTIVE);
+
+	if (mode->flags & DISPLAY_FLAGS_INTERLACED)
+		malidp_hw_setbits(hwdev, MALIDP_DISP_FUNC_ILACED, MALIDP_DE_DISPLAY_FUNC);
+	else
+		malidp_hw_clearbits(hwdev, MALIDP_DISP_FUNC_ILACED, MALIDP_DE_DISPLAY_FUNC);
+}
+
+static int malidpx50_get_bytes_per_column(u32 fmt)
+{
+	u32 bytes_per_column;
+
+	switch (fmt) {
+	/* 8 lines at 4 bytes per pixel */
+	case DRM_FORMAT_ARGB2101010:
+	case DRM_FORMAT_ABGR2101010:
+	case DRM_FORMAT_RGBA1010102:
+	case DRM_FORMAT_BGRA1010102:
+	case DRM_FORMAT_ARGB8888:
+	case DRM_FORMAT_ABGR8888:
+	case DRM_FORMAT_RGBA8888:
+	case DRM_FORMAT_BGRA8888:
+	case DRM_FORMAT_XRGB8888:
+	case DRM_FORMAT_XBGR8888:
+	case DRM_FORMAT_RGBX8888:
+	case DRM_FORMAT_BGRX8888:
+	case DRM_FORMAT_RGB888:
+	case DRM_FORMAT_BGR888:
+	/* 16 lines at 2 bytes per pixel */
+	case DRM_FORMAT_RGBA5551:
+	case DRM_FORMAT_ABGR1555:
+	case DRM_FORMAT_RGB565:
+	case DRM_FORMAT_BGR565:
+	case DRM_FORMAT_UYVY:
+	case DRM_FORMAT_YUYV:
+	case DRM_FORMAT_X0L0:
+		bytes_per_column = 32;
+		break;
+	/* 16 lines at 1.5 bytes per pixel */
+	case DRM_FORMAT_NV12:
+	case DRM_FORMAT_YUV420:
+	/* 8 lines at 3 bytes per pixel */
+	case DRM_FORMAT_VUY888:
+	/* 16 lines at 12 bits per pixel */
+	case DRM_FORMAT_YUV420_8BIT:
+	/* 8 lines at 3 bytes per pixel */
+	case DRM_FORMAT_P010:
+		bytes_per_column = 24;
+		break;
+	/* 8 lines at 30 bits per pixel */
+	case DRM_FORMAT_VUY101010:
+	/* 16 lines at 15 bits per pixel */
+	case DRM_FORMAT_YUV420_10BIT:
+		bytes_per_column = 30;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return bytes_per_column;
+}
+
+static int malidp550_rotmem_required(struct malidp_hw_device *hwdev, u16 w,
+				     u16 h, u32 fmt, bool has_modifier)
+{
+	int bytes_per_column = 0;
+
+	switch (fmt) {
+	/* 8 lines at 15 bits per pixel */
+	case DRM_FORMAT_YUV420_10BIT:
+		bytes_per_column = 15;
+		break;
+	/* Uncompressed YUV 420 10 bit single plane cannot be rotated */
+	case DRM_FORMAT_X0L2:
+		if (has_modifier)
+			bytes_per_column = 8;
+		else
+			return -EINVAL;
+		break;
+	default:
+		bytes_per_column = malidpx50_get_bytes_per_column(fmt);
+	}
+
+	if (bytes_per_column == -EINVAL)
+		return bytes_per_column;
+
+	return w * bytes_per_column;
+}
+
+static int malidp650_rotmem_required(struct malidp_hw_device *hwdev, u16 w,
+				     u16 h, u32 fmt, bool has_modifier)
+{
+	int bytes_per_column = 0;
+
+	switch (fmt) {
+	/* 16 lines at 2 bytes per pixel */
+	case DRM_FORMAT_X0L2:
+		bytes_per_column = 32;
+		break;
+	default:
+		bytes_per_column = malidpx50_get_bytes_per_column(fmt);
+	}
+
+	if (bytes_per_column == -EINVAL)
+		return bytes_per_column;
+
+	return w * bytes_per_column;
+}
+
+static int malidp550_se_set_scaling_coeffs(struct malidp_hw_device *hwdev,
+					   struct malidp_se_config *se_config,
+					   struct malidp_se_config *old_config)
+{
+	u32 mask = MALIDP550_SE_CTL_VCSEL(MALIDP550_SE_CTL_SEL_MASK) |
+		   MALIDP550_SE_CTL_HCSEL(MALIDP550_SE_CTL_SEL_MASK);
+	u32 new_value = MALIDP550_SE_CTL_VCSEL(se_config->vcoeff) |
+			MALIDP550_SE_CTL_HCSEL(se_config->hcoeff);
+
+	malidp_hw_clearbits(hwdev, mask, MALIDP550_SE_CONTROL);
+	malidp_hw_setbits(hwdev, new_value, MALIDP550_SE_CONTROL);
+	return 0;
+}
+
+static long malidp550_se_calc_mclk(struct malidp_hw_device *hwdev,
+				   struct malidp_se_config *se_config,
+				   struct videomode *vm)
+{
+	unsigned long mclk;
+	unsigned long pxlclk = vm->pixelclock;
+	unsigned long htotal = vm->hactive + vm->hfront_porch +
+			       vm->hback_porch + vm->hsync_len;
+	unsigned long numerator = 1, denominator = 1;
+	long ret;
+
+	if (se_config->scale_enable) {
+		numerator = max(se_config->input_w, se_config->output_w) *
+			    se_config->input_h;
+		numerator += se_config->output_w *
+			     (se_config->output_h -
+			      min(se_config->input_h, se_config->output_h));
+		denominator = (htotal - 2) * se_config->output_h;
+	}
+
+	/* mclk can't be slower than pxlclk. */
+	if (numerator < denominator)
+		numerator = denominator = 1;
+	mclk = (pxlclk * numerator) / denominator;
+	ret = clk_get_rate(hwdev->mclk);
+	if (ret < mclk) {
+		DRM_DEBUG_DRIVER("mclk requirement of %lu kHz can't be met.\n",
+				 mclk / 1000);
+		return -EINVAL;
+	}
+	return ret;
+}
+
+static int malidp550_enable_memwrite(struct malidp_hw_device *hwdev,
+				     dma_addr_t *addrs, s32 *pitches,
+				     int num_planes, u16 w, u16 h, u32 fmt_id,
+				     const s16 *rgb2yuv_coeffs)
+{
+	u32 base = MALIDP550_SE_MEMWRITE_BASE;
+	u32 de_base = malidp_get_block_base(hwdev, MALIDP_DE_BLOCK);
+
+	/* enable the scaling engine block */
+	malidp_hw_setbits(hwdev, MALIDP_SCALE_ENGINE_EN, de_base + MALIDP_DE_DISPLAY_FUNC);
+
+	hwdev->mw_state = MW_ONESHOT;
+
+	malidp_hw_write(hwdev, fmt_id, base + MALIDP_MW_FORMAT);
+	switch (num_planes) {
+	case 2:
+		malidp_hw_write(hwdev, lower_32_bits(addrs[1]), base + MALIDP_MW_P2_PTR_LOW);
+		malidp_hw_write(hwdev, upper_32_bits(addrs[1]), base + MALIDP_MW_P2_PTR_HIGH);
+		malidp_hw_write(hwdev, pitches[1], base + MALIDP_MW_P2_STRIDE);
+		fallthrough;
+	case 1:
+		malidp_hw_write(hwdev, lower_32_bits(addrs[0]), base + MALIDP_MW_P1_PTR_LOW);
+		malidp_hw_write(hwdev, upper_32_bits(addrs[0]), base + MALIDP_MW_P1_PTR_HIGH);
+		malidp_hw_write(hwdev, pitches[0], base + MALIDP_MW_P1_STRIDE);
+		break;
+	default:
+		WARN(1, "Invalid number of planes");
+	}
+
+	malidp_hw_write(hwdev, MALIDP_DE_H_ACTIVE(w) | MALIDP_DE_V_ACTIVE(h),
+			MALIDP550_SE_MEMWRITE_OUT_SIZE);
+	malidp_hw_setbits(hwdev, MALIDP550_SE_MEMWRITE_ONESHOT | MALIDP_SE_MEMWRITE_EN,
+			  MALIDP550_SE_CONTROL);
+
+	if (rgb2yuv_coeffs) {
+		int i;
+
+		for (i = 0; i < MALIDP_COLORADJ_NUM_COEFFS; i++) {
+			malidp_hw_write(hwdev, rgb2yuv_coeffs[i],
+					MALIDP550_SE_RGB_YUV_COEFFS + i * 4);
+		}
+	}
+
+	return 0;
+}
+
+static void malidp550_disable_memwrite(struct malidp_hw_device *hwdev)
+{
+	u32 base = malidp_get_block_base(hwdev, MALIDP_DE_BLOCK);
+
+	malidp_hw_clearbits(hwdev, MALIDP550_SE_MEMWRITE_ONESHOT | MALIDP_SE_MEMWRITE_EN,
+			    MALIDP550_SE_CONTROL);
+	malidp_hw_clearbits(hwdev, MALIDP_SCALE_ENGINE_EN, base + MALIDP_DE_DISPLAY_FUNC);
+}
+
+static int malidp650_query_hw(struct malidp_hw_device *hwdev)
+{
+	u32 conf = malidp_hw_read(hwdev, MALIDP550_CONFIG_ID);
+	u8 ln_size = (conf >> 4) & 0x3, rsize;
+
+	hwdev->min_line_size = 4;
+
+	switch (ln_size) {
+	case 0:
+	case 2:
+		/* reserved values */
+		hwdev->max_line_size = 0;
+		return -EINVAL;
+	case 1:
+		hwdev->max_line_size = SZ_4K;
+		/* two banks of 128KB for rotation memory */
+		rsize = 128;
+		break;
+	case 3:
+		hwdev->max_line_size = 2560;
+		/* two banks of 80KB for rotation memory */
+		rsize = 80;
+	}
+
+	hwdev->rotation_memory[0] = hwdev->rotation_memory[1] = rsize * SZ_1K;
+	return 0;
+}
+
+const struct malidp_hw malidp_device[MALIDP_MAX_DEVICES] = {
+	[MALIDP_500] = {
+		.map = {
+			.coeffs_base = MALIDP500_COEFFS_BASE,
+			.se_base = MALIDP500_SE_BASE,
+			.dc_base = MALIDP500_DC_BASE,
+			.out_depth_base = MALIDP500_OUTPUT_DEPTH,
+			.features = 0,	/* no CLEARIRQ register */
+			.n_layers = ARRAY_SIZE(malidp500_layers),
+			.layers = malidp500_layers,
+			.de_irq_map = {
+				.irq_mask = MALIDP_DE_IRQ_UNDERRUN |
+					    MALIDP500_DE_IRQ_AXI_ERR |
+					    MALIDP500_DE_IRQ_VSYNC |
+					    MALIDP500_DE_IRQ_GLOBAL,
+				.vsync_irq = MALIDP500_DE_IRQ_VSYNC,
+				.err_mask = MALIDP_DE_IRQ_UNDERRUN |
+					    MALIDP500_DE_IRQ_AXI_ERR |
+					    MALIDP500_DE_IRQ_SATURATION,
+			},
+			.se_irq_map = {
+				.irq_mask = MALIDP500_SE_IRQ_CONF_MODE |
+					    MALIDP500_SE_IRQ_CONF_VALID |
+					    MALIDP500_SE_IRQ_GLOBAL,
+				.vsync_irq = MALIDP500_SE_IRQ_CONF_VALID,
+				.err_mask = MALIDP500_SE_IRQ_INIT_BUSY |
+					    MALIDP500_SE_IRQ_AXI_ERROR |
+					    MALIDP500_SE_IRQ_OVERRUN,
+			},
+			.dc_irq_map = {
+				.irq_mask = MALIDP500_DE_IRQ_CONF_VALID,
+				.vsync_irq = MALIDP500_DE_IRQ_CONF_VALID,
+			},
+			.pixel_formats = malidp500_de_formats,
+			.n_pixel_formats = ARRAY_SIZE(malidp500_de_formats),
+			.bus_align_bytes = 8,
+		},
+		.query_hw = malidp500_query_hw,
+		.enter_config_mode = malidp500_enter_config_mode,
+		.leave_config_mode = malidp500_leave_config_mode,
+		.in_config_mode = malidp500_in_config_mode,
+		.set_config_valid = malidp500_set_config_valid,
+		.modeset = malidp500_modeset,
+		.rotmem_required = malidp500_rotmem_required,
+		.se_set_scaling_coeffs = malidp500_se_set_scaling_coeffs,
+		.se_calc_mclk = malidp500_se_calc_mclk,
+		.enable_memwrite = malidp500_enable_memwrite,
+		.disable_memwrite = malidp500_disable_memwrite,
+		.features = MALIDP_DEVICE_LV_HAS_3_STRIDES,
+	},
+	[MALIDP_550] = {
+		.map = {
+			.coeffs_base = MALIDP550_COEFFS_BASE,
+			.se_base = MALIDP550_SE_BASE,
+			.dc_base = MALIDP550_DC_BASE,
+			.out_depth_base = MALIDP550_DE_OUTPUT_DEPTH,
+			.features = MALIDP_REGMAP_HAS_CLEARIRQ |
+				    MALIDP_DEVICE_AFBC_SUPPORT_SPLIT |
+				    MALIDP_DEVICE_AFBC_YUV_420_10_SUPPORT_SPLIT |
+				    MALIDP_DEVICE_AFBC_YUYV_USE_422_P2,
+			.n_layers = ARRAY_SIZE(malidp550_layers),
+			.layers = malidp550_layers,
+			.de_irq_map = {
+				.irq_mask = MALIDP_DE_IRQ_UNDERRUN |
+					    MALIDP550_DE_IRQ_VSYNC,
+				.vsync_irq = MALIDP550_DE_IRQ_VSYNC,
+				.err_mask = MALIDP_DE_IRQ_UNDERRUN |
+					    MALIDP550_DE_IRQ_SATURATION |
+					    MALIDP550_DE_IRQ_AXI_ERR,
+			},
+			.se_irq_map = {
+				.irq_mask = MALIDP550_SE_IRQ_EOW,
+				.vsync_irq = MALIDP550_SE_IRQ_EOW,
+				.err_mask  = MALIDP550_SE_IRQ_AXI_ERR |
+					     MALIDP550_SE_IRQ_OVR |
+					     MALIDP550_SE_IRQ_IBSY,
+			},
+			.dc_irq_map = {
+				.irq_mask = MALIDP550_DC_IRQ_CONF_VALID |
+					    MALIDP550_DC_IRQ_SE,
+				.vsync_irq = MALIDP550_DC_IRQ_CONF_VALID,
+			},
+			.pixel_formats = malidp550_de_formats,
+			.n_pixel_formats = ARRAY_SIZE(malidp550_de_formats),
+			.bus_align_bytes = 8,
+		},
+		.query_hw = malidp550_query_hw,
+		.enter_config_mode = malidp550_enter_config_mode,
+		.leave_config_mode = malidp550_leave_config_mode,
+		.in_config_mode = malidp550_in_config_mode,
+		.set_config_valid = malidp550_set_config_valid,
+		.modeset = malidp550_modeset,
+		.rotmem_required = malidp550_rotmem_required,
+		.se_set_scaling_coeffs = malidp550_se_set_scaling_coeffs,
+		.se_calc_mclk = malidp550_se_calc_mclk,
+		.enable_memwrite = malidp550_enable_memwrite,
+		.disable_memwrite = malidp550_disable_memwrite,
+		.features = 0,
+	},
+	[MALIDP_650] = {
+		.map = {
+			.coeffs_base = MALIDP550_COEFFS_BASE,
+			.se_base = MALIDP550_SE_BASE,
+			.dc_base = MALIDP550_DC_BASE,
+			.out_depth_base = MALIDP550_DE_OUTPUT_DEPTH,
+			.features = MALIDP_REGMAP_HAS_CLEARIRQ |
+				    MALIDP_DEVICE_AFBC_SUPPORT_SPLIT |
+				    MALIDP_DEVICE_AFBC_YUYV_USE_422_P2,
+			.n_layers = ARRAY_SIZE(malidp650_layers),
+			.layers = malidp650_layers,
+			.de_irq_map = {
+				.irq_mask = MALIDP_DE_IRQ_UNDERRUN |
+					    MALIDP650_DE_IRQ_DRIFT |
+					    MALIDP550_DE_IRQ_VSYNC,
+				.vsync_irq = MALIDP550_DE_IRQ_VSYNC,
+				.err_mask = MALIDP_DE_IRQ_UNDERRUN |
+					    MALIDP650_DE_IRQ_DRIFT |
+					    MALIDP550_DE_IRQ_SATURATION |
+					    MALIDP550_DE_IRQ_AXI_ERR |
+					    MALIDP650_DE_IRQ_ACEV1 |
+					    MALIDP650_DE_IRQ_ACEV2 |
+					    MALIDP650_DE_IRQ_ACEG |
+					    MALIDP650_DE_IRQ_AXIEP,
+			},
+			.se_irq_map = {
+				.irq_mask = MALIDP550_SE_IRQ_EOW,
+				.vsync_irq = MALIDP550_SE_IRQ_EOW,
+				.err_mask = MALIDP550_SE_IRQ_AXI_ERR |
+					    MALIDP550_SE_IRQ_OVR |
+					    MALIDP550_SE_IRQ_IBSY,
+			},
+			.dc_irq_map = {
+				.irq_mask = MALIDP550_DC_IRQ_CONF_VALID |
+					    MALIDP550_DC_IRQ_SE,
+				.vsync_irq = MALIDP550_DC_IRQ_CONF_VALID,
+			},
+			.pixel_formats = malidp650_de_formats,
+			.n_pixel_formats = ARRAY_SIZE(malidp650_de_formats),
+			.bus_align_bytes = 16,
+		},
+		.query_hw = malidp650_query_hw,
+		.enter_config_mode = malidp550_enter_config_mode,
+		.leave_config_mode = malidp550_leave_config_mode,
+		.in_config_mode = malidp550_in_config_mode,
+		.set_config_valid = malidp550_set_config_valid,
+		.modeset = malidp550_modeset,
+		.rotmem_required = malidp650_rotmem_required,
+		.se_set_scaling_coeffs = malidp550_se_set_scaling_coeffs,
+		.se_calc_mclk = malidp550_se_calc_mclk,
+		.enable_memwrite = malidp550_enable_memwrite,
+		.disable_memwrite = malidp550_disable_memwrite,
+		.features = 0,
+	},
+};
+
+u8 malidp_hw_get_format_id(const struct malidp_hw_regmap *map,
+			   u8 layer_id, u32 format, bool has_modifier)
+{
+	unsigned int i;
+
+	for (i = 0; i < map->n_pixel_formats; i++) {
+		if (((map->pixel_formats[i].layer & layer_id) == layer_id) &&
+		    (map->pixel_formats[i].format == format)) {
+			/*
+			 * In some DP550 and DP650, DRM_FORMAT_YUYV + AFBC modifier
+			 * is supported by a different h/w format id than
+			 * DRM_FORMAT_YUYV (only).
+			 */
+			if (format == DRM_FORMAT_YUYV &&
+			    (has_modifier) &&
+			    (map->features & MALIDP_DEVICE_AFBC_YUYV_USE_422_P2))
+				return AFBC_YUV_422_FORMAT_ID;
+			else
+				return map->pixel_formats[i].id;
+		}
+	}
+
+	return MALIDP_INVALID_FORMAT_ID;
+}
+
+bool malidp_hw_format_is_linear_only(u32 format)
+{
+	switch (format) {
+	case DRM_FORMAT_ARGB2101010:
+	case DRM_FORMAT_RGBA1010102:
+	case DRM_FORMAT_BGRA1010102:
+	case DRM_FORMAT_ARGB8888:
+	case DRM_FORMAT_RGBA8888:
+	case DRM_FORMAT_BGRA8888:
+	case DRM_FORMAT_XBGR8888:
+	case DRM_FORMAT_XRGB8888:
+	case DRM_FORMAT_RGBX8888:
+	case DRM_FORMAT_BGRX8888:
+	case DRM_FORMAT_RGB888:
+	case DRM_FORMAT_RGB565:
+	case DRM_FORMAT_ARGB1555:
+	case DRM_FORMAT_RGBA5551:
+	case DRM_FORMAT_BGRA5551:
+	case DRM_FORMAT_UYVY:
+	case DRM_FORMAT_XYUV8888:
+	case DRM_FORMAT_XVYU2101010:
+	case DRM_FORMAT_X0L2:
+	case DRM_FORMAT_X0L0:
+		return true;
+	default:
+		return false;
+	}
+}
+
+bool malidp_hw_format_is_afbc_only(u32 format)
+{
+	switch (format) {
+	case DRM_FORMAT_VUY888:
+	case DRM_FORMAT_VUY101010:
+	case DRM_FORMAT_YUV420_8BIT:
+	case DRM_FORMAT_YUV420_10BIT:
+		return true;
+	default:
+		return false;
+	}
+}
+
+static void malidp_hw_clear_irq(struct malidp_hw_device *hwdev, u8 block, u32 irq)
+{
+	u32 base = malidp_get_block_base(hwdev, block);
+
+	if (hwdev->hw->map.features & MALIDP_REGMAP_HAS_CLEARIRQ)
+		malidp_hw_write(hwdev, irq, base + MALIDP_REG_CLEARIRQ);
+	else
+		malidp_hw_write(hwdev, irq, base + MALIDP_REG_STATUS);
+}
+
+static irqreturn_t malidp_de_irq(int irq, void *arg)
+{
+	struct drm_device *drm = arg;
+	struct malidp_drm *malidp = drm->dev_private;
+	struct malidp_hw_device *hwdev;
+	struct malidp_hw *hw;
+	const struct malidp_irq_map *de;
+	u32 status, mask, dc_status;
+	irqreturn_t ret = IRQ_NONE;
+
+	hwdev = malidp->dev;
+	hw = hwdev->hw;
+	de = &hw->map.de_irq_map;
+
+	/*
+	 * if we are suspended it is likely that we were invoked because
+	 * we share an interrupt line with some other driver, don't try
+	 * to read the hardware registers
+	 */
+	if (hwdev->pm_suspended)
+		return IRQ_NONE;
+
+	/* first handle the config valid IRQ */
+	dc_status = malidp_hw_read(hwdev, hw->map.dc_base + MALIDP_REG_STATUS);
+	if (dc_status & hw->map.dc_irq_map.vsync_irq) {
+		malidp_hw_clear_irq(hwdev, MALIDP_DC_BLOCK, dc_status);
+		/* do we have a page flip event? */
+		if (malidp->event != NULL) {
+			spin_lock(&drm->event_lock);
+			drm_crtc_send_vblank_event(&malidp->crtc, malidp->event);
+			malidp->event = NULL;
+			spin_unlock(&drm->event_lock);
+		}
+		atomic_set(&malidp->config_valid, MALIDP_CONFIG_VALID_DONE);
+		ret = IRQ_WAKE_THREAD;
+	}
+
+	status = malidp_hw_read(hwdev, MALIDP_REG_STATUS);
+	if (!(status & de->irq_mask))
+		return ret;
+
+	mask = malidp_hw_read(hwdev, MALIDP_REG_MASKIRQ);
+	/* keep the status of the enabled interrupts, plus the error bits */
+	status &= (mask | de->err_mask);
+	if ((status & de->vsync_irq) && malidp->crtc.enabled)
+		drm_crtc_handle_vblank(&malidp->crtc);
+
+#ifdef CONFIG_DEBUG_FS
+	if (status & de->err_mask) {
+		malidp_error(malidp, &malidp->de_errors, status,
+			     drm_crtc_vblank_count(&malidp->crtc));
+	}
+#endif
+	malidp_hw_clear_irq(hwdev, MALIDP_DE_BLOCK, status);
+
+	return (ret == IRQ_NONE) ? IRQ_HANDLED : ret;
+}
+
+static irqreturn_t malidp_de_irq_thread_handler(int irq, void *arg)
+{
+	struct drm_device *drm = arg;
+	struct malidp_drm *malidp = drm->dev_private;
+
+	wake_up(&malidp->wq);
+
+	return IRQ_HANDLED;
+}
+
+void malidp_de_irq_hw_init(struct malidp_hw_device *hwdev)
+{
+	/* ensure interrupts are disabled */
+	malidp_hw_disable_irq(hwdev, MALIDP_DE_BLOCK, 0xffffffff);
+	malidp_hw_clear_irq(hwdev, MALIDP_DE_BLOCK, 0xffffffff);
+	malidp_hw_disable_irq(hwdev, MALIDP_DC_BLOCK, 0xffffffff);
+	malidp_hw_clear_irq(hwdev, MALIDP_DC_BLOCK, 0xffffffff);
+
+	/* first enable the DC block IRQs */
+	malidp_hw_enable_irq(hwdev, MALIDP_DC_BLOCK,
+			     hwdev->hw->map.dc_irq_map.irq_mask);
+
+	/* now enable the DE block IRQs */
+	malidp_hw_enable_irq(hwdev, MALIDP_DE_BLOCK,
+			     hwdev->hw->map.de_irq_map.irq_mask);
+}
+
+int malidp_de_irq_init(struct drm_device *drm, int irq)
+{
+	struct malidp_drm *malidp = drm->dev_private;
+	struct malidp_hw_device *hwdev = malidp->dev;
+	int ret;
+
+	/* ensure interrupts are disabled */
+	malidp_hw_disable_irq(hwdev, MALIDP_DE_BLOCK, 0xffffffff);
+	malidp_hw_clear_irq(hwdev, MALIDP_DE_BLOCK, 0xffffffff);
+	malidp_hw_disable_irq(hwdev, MALIDP_DC_BLOCK, 0xffffffff);
+	malidp_hw_clear_irq(hwdev, MALIDP_DC_BLOCK, 0xffffffff);
+
+	ret = devm_request_threaded_irq(drm->dev, irq, malidp_de_irq,
+					malidp_de_irq_thread_handler,
+					IRQF_SHARED, "malidp-de", drm);
+	if (ret < 0) {
+		DRM_ERROR("failed to install DE IRQ handler\n");
+		return ret;
+	}
+
+	malidp_de_irq_hw_init(hwdev);
+
+	return 0;
+}
+
+void malidp_de_irq_fini(struct malidp_hw_device *hwdev)
+{
+	malidp_hw_disable_irq(hwdev, MALIDP_DE_BLOCK,
+			      hwdev->hw->map.de_irq_map.irq_mask);
+	malidp_hw_disable_irq(hwdev, MALIDP_DC_BLOCK,
+			      hwdev->hw->map.dc_irq_map.irq_mask);
+}
+
+static irqreturn_t malidp_se_irq(int irq, void *arg)
+{
+	struct drm_device *drm = arg;
+	struct malidp_drm *malidp = drm->dev_private;
+	struct malidp_hw_device *hwdev = malidp->dev;
+	struct malidp_hw *hw = hwdev->hw;
+	const struct malidp_irq_map *se = &hw->map.se_irq_map;
+	u32 status, mask;
+
+	/*
+	 * if we are suspended it is likely that we were invoked because
+	 * we share an interrupt line with some other driver, don't try
+	 * to read the hardware registers
+	 */
+	if (hwdev->pm_suspended)
+		return IRQ_NONE;
+
+	status = malidp_hw_read(hwdev, hw->map.se_base + MALIDP_REG_STATUS);
+	if (!(status & (se->irq_mask | se->err_mask)))
+		return IRQ_NONE;
+
+#ifdef CONFIG_DEBUG_FS
+	if (status & se->err_mask)
+		malidp_error(malidp, &malidp->se_errors, status,
+			     drm_crtc_vblank_count(&malidp->crtc));
+#endif
+	mask = malidp_hw_read(hwdev, hw->map.se_base + MALIDP_REG_MASKIRQ);
+	status &= mask;
+
+	if (status & se->vsync_irq) {
+		switch (hwdev->mw_state) {
+		case MW_ONESHOT:
+			drm_writeback_signal_completion(&malidp->mw_connector, 0);
+			break;
+		case MW_STOP:
+			drm_writeback_signal_completion(&malidp->mw_connector, 0);
+			/* disable writeback after stop */
+			hwdev->mw_state = MW_NOT_ENABLED;
+			break;
+		case MW_RESTART:
+			drm_writeback_signal_completion(&malidp->mw_connector, 0);
+			fallthrough;	/* to a new start */
+		case MW_START:
+			/* writeback started, need to emulate one-shot mode */
+			hw->disable_memwrite(hwdev);
+			/*
+			 * only set config_valid HW bit if there is no other update
+			 * in progress or if we raced ahead of the DE IRQ handler
+			 * and config_valid flag will not be update until later
+			 */
+			status = malidp_hw_read(hwdev, hw->map.dc_base + MALIDP_REG_STATUS);
+			if ((atomic_read(&malidp->config_valid) != MALIDP_CONFIG_START) ||
+			    (status & hw->map.dc_irq_map.vsync_irq))
+				hw->set_config_valid(hwdev, 1);
+			break;
+		}
+	}
+
+	malidp_hw_clear_irq(hwdev, MALIDP_SE_BLOCK, status);
+
+	return IRQ_HANDLED;
+}
+
+void malidp_se_irq_hw_init(struct malidp_hw_device *hwdev)
+{
+	/* ensure interrupts are disabled */
+	malidp_hw_disable_irq(hwdev, MALIDP_SE_BLOCK, 0xffffffff);
+	malidp_hw_clear_irq(hwdev, MALIDP_SE_BLOCK, 0xffffffff);
+
+	malidp_hw_enable_irq(hwdev, MALIDP_SE_BLOCK,
+			     hwdev->hw->map.se_irq_map.irq_mask);
+}
+
+static irqreturn_t malidp_se_irq_thread_handler(int irq, void *arg)
+{
+	return IRQ_HANDLED;
+}
+
+int malidp_se_irq_init(struct drm_device *drm, int irq)
+{
+	struct malidp_drm *malidp = drm->dev_private;
+	struct malidp_hw_device *hwdev = malidp->dev;
+	int ret;
+
+	/* ensure interrupts are disabled */
+	malidp_hw_disable_irq(hwdev, MALIDP_SE_BLOCK, 0xffffffff);
+	malidp_hw_clear_irq(hwdev, MALIDP_SE_BLOCK, 0xffffffff);
+
+	ret = devm_request_threaded_irq(drm->dev, irq, malidp_se_irq,
+					malidp_se_irq_thread_handler,
+					IRQF_SHARED, "malidp-se", drm);
+	if (ret < 0) {
+		DRM_ERROR("failed to install SE IRQ handler\n");
+		return ret;
+	}
+
+	hwdev->mw_state = MW_NOT_ENABLED;
+	malidp_se_irq_hw_init(hwdev);
+
+	return 0;
+}
+
+void malidp_se_irq_fini(struct malidp_hw_device *hwdev)
+{
+	malidp_hw_disable_irq(hwdev, MALIDP_SE_BLOCK,
+			      hwdev->hw->map.se_irq_map.irq_mask);
+}
diff --git a/drivers/gpu/drm/arm/malidp_hw.h b/drivers/gpu/drm/arm/malidp_hw.h
new file mode 100644
index 000000000..e4c36bc90
--- /dev/null
+++ b/drivers/gpu/drm/arm/malidp_hw.h
@@ -0,0 +1,412 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ *
+ * (C) COPYRIGHT 2013-2016 ARM Limited. All rights reserved.
+ *
+ * ARM Mali DP hardware manipulation routines.
+ */
+
+#ifndef __MALIDP_HW_H__
+#define __MALIDP_HW_H__
+
+#include <linux/bitops.h>
+#include "malidp_regs.h"
+
+struct videomode;
+struct clk;
+
+/* Mali DP IP blocks */
+enum {
+	MALIDP_DE_BLOCK = 0,
+	MALIDP_SE_BLOCK,
+	MALIDP_DC_BLOCK
+};
+
+/* Mali DP layer IDs */
+enum {
+	DE_VIDEO1 = BIT(0),
+	DE_GRAPHICS1 = BIT(1),
+	DE_GRAPHICS2 = BIT(2), /* used only in DP500 */
+	DE_VIDEO2 = BIT(3),
+	DE_SMART = BIT(4),
+	SE_MEMWRITE = BIT(5),
+};
+
+enum rotation_features {
+	ROTATE_NONE,		/* does not support rotation at all */
+	ROTATE_ANY,		/* supports rotation on any buffers */
+	ROTATE_COMPRESSED,	/* supports rotation only on compressed buffers */
+};
+
+struct malidp_format_id {
+	u32 format;		/* DRM fourcc */
+	u8 layer;		/* bitmask of layers supporting it */
+	u8 id;			/* used internally */
+};
+
+#define MALIDP_INVALID_FORMAT_ID	0xff
+
+/*
+ * hide the differences between register maps
+ * by using a common structure to hold the
+ * base register offsets
+ */
+
+struct malidp_irq_map {
+	u32 irq_mask;		/* mask of IRQs that can be enabled in the block */
+	u32 vsync_irq;		/* IRQ bit used for signaling during VSYNC */
+	u32 err_mask;		/* mask of bits that represent errors */
+};
+
+struct malidp_layer {
+	u16 id;			/* layer ID */
+	u16 base;		/* address offset for the register bank */
+	u16 ptr;		/* address offset for the pointer register */
+	u16 stride_offset;	/* offset to the first stride register. */
+	s16 yuv2rgb_offset;	/* offset to the YUV->RGB matrix entries */
+	u16 mmu_ctrl_offset;    /* offset to the MMU control register */
+	enum rotation_features rot;	/* type of rotation supported */
+	/* address offset for the AFBC decoder registers */
+	u16 afbc_decoder_offset;
+};
+
+enum malidp_scaling_coeff_set {
+	MALIDP_UPSCALING_COEFFS = 1,
+	MALIDP_DOWNSCALING_1_5_COEFFS = 2,
+	MALIDP_DOWNSCALING_2_COEFFS = 3,
+	MALIDP_DOWNSCALING_2_75_COEFFS = 4,
+	MALIDP_DOWNSCALING_4_COEFFS = 5,
+};
+
+struct malidp_se_config {
+	u8 scale_enable : 1;
+	u8 enhancer_enable : 1;
+	u8 hcoeff : 3;
+	u8 vcoeff : 3;
+	u8 plane_src_id;
+	u16 input_w, input_h;
+	u16 output_w, output_h;
+	u32 h_init_phase, h_delta_phase;
+	u32 v_init_phase, v_delta_phase;
+};
+
+/* regmap features */
+#define MALIDP_REGMAP_HAS_CLEARIRQ				BIT(0)
+#define MALIDP_DEVICE_AFBC_SUPPORT_SPLIT			BIT(1)
+#define MALIDP_DEVICE_AFBC_YUV_420_10_SUPPORT_SPLIT		BIT(2)
+#define MALIDP_DEVICE_AFBC_YUYV_USE_422_P2			BIT(3)
+
+struct malidp_hw_regmap {
+	/* address offset of the DE register bank */
+	/* is always 0x0000 */
+	/* address offset of the DE coefficients registers */
+	const u16 coeffs_base;
+	/* address offset of the SE registers bank */
+	const u16 se_base;
+	/* address offset of the DC registers bank */
+	const u16 dc_base;
+
+	/* address offset for the output depth register */
+	const u16 out_depth_base;
+
+	/* bitmap with register map features */
+	const u8 features;
+
+	/* list of supported layers */
+	const u8 n_layers;
+	const struct malidp_layer *layers;
+
+	const struct malidp_irq_map de_irq_map;
+	const struct malidp_irq_map se_irq_map;
+	const struct malidp_irq_map dc_irq_map;
+
+	/* list of supported pixel formats for each layer */
+	const struct malidp_format_id *pixel_formats;
+	const u8 n_pixel_formats;
+
+	/* pitch alignment requirement in bytes */
+	const u8 bus_align_bytes;
+};
+
+/* device features */
+/* Unlike DP550/650, DP500 has 3 stride registers in its video layer. */
+#define MALIDP_DEVICE_LV_HAS_3_STRIDES	BIT(0)
+
+struct malidp_hw_device;
+
+/*
+ * Static structure containing hardware specific data and pointers to
+ * functions that behave differently between various versions of the IP.
+ */
+struct malidp_hw {
+	const struct malidp_hw_regmap map;
+
+	/*
+	 * Validate the driver instance against the hardware bits
+	 */
+	int (*query_hw)(struct malidp_hw_device *hwdev);
+
+	/*
+	 * Set the hardware into config mode, ready to accept mode changes
+	 */
+	void (*enter_config_mode)(struct malidp_hw_device *hwdev);
+
+	/*
+	 * Tell hardware to exit configuration mode
+	 */
+	void (*leave_config_mode)(struct malidp_hw_device *hwdev);
+
+	/*
+	 * Query if hardware is in configuration mode
+	 */
+	bool (*in_config_mode)(struct malidp_hw_device *hwdev);
+
+	/*
+	 * Set/clear configuration valid flag for hardware parameters that can
+	 * be changed outside the configuration mode to the given value.
+	 * Hardware will use the new settings when config valid is set,
+	 * after the end of the current buffer scanout, and will ignore
+	 * any new values for those parameters if config valid flag is cleared
+	 */
+	void (*set_config_valid)(struct malidp_hw_device *hwdev, u8 value);
+
+	/*
+	 * Set a new mode in hardware. Requires the hardware to be in
+	 * configuration mode before this function is called.
+	 */
+	void (*modeset)(struct malidp_hw_device *hwdev, struct videomode *m);
+
+	/*
+	 * Calculate the required rotation memory given the active area
+	 * and the buffer format.
+	 */
+	int (*rotmem_required)(struct malidp_hw_device *hwdev, u16 w, u16 h,
+			       u32 fmt, bool has_modifier);
+
+	int (*se_set_scaling_coeffs)(struct malidp_hw_device *hwdev,
+				     struct malidp_se_config *se_config,
+				     struct malidp_se_config *old_config);
+
+	long (*se_calc_mclk)(struct malidp_hw_device *hwdev,
+			     struct malidp_se_config *se_config,
+			     struct videomode *vm);
+	/*
+	 * Enable writing to memory the content of the next frame
+	 * @param hwdev - malidp_hw_device structure containing the HW description
+	 * @param addrs - array of addresses for each plane
+	 * @param pitches - array of pitches for each plane
+	 * @param num_planes - number of planes to be written
+	 * @param w - width of the output frame
+	 * @param h - height of the output frame
+	 * @param fmt_id - internal format ID of output buffer
+	 */
+	int (*enable_memwrite)(struct malidp_hw_device *hwdev, dma_addr_t *addrs,
+			       s32 *pitches, int num_planes, u16 w, u16 h, u32 fmt_id,
+			       const s16 *rgb2yuv_coeffs);
+
+	/*
+	 * Disable the writing to memory of the next frame's content.
+	 */
+	void (*disable_memwrite)(struct malidp_hw_device *hwdev);
+
+	u8 features;
+};
+
+/* Supported variants of the hardware */
+enum {
+	MALIDP_500 = 0,
+	MALIDP_550,
+	MALIDP_650,
+	/* keep the next entry last */
+	MALIDP_MAX_DEVICES
+};
+
+extern const struct malidp_hw malidp_device[MALIDP_MAX_DEVICES];
+
+/*
+ * Structure used by the driver during runtime operation.
+ */
+struct malidp_hw_device {
+	struct malidp_hw *hw;
+	void __iomem *regs;
+
+	/* APB clock */
+	struct clk *pclk;
+	/* AXI clock */
+	struct clk *aclk;
+	/* main clock for display core */
+	struct clk *mclk;
+	/* pixel clock for display core */
+	struct clk *pxlclk;
+
+	u8 min_line_size;
+	u16 max_line_size;
+	u32 output_color_depth;
+
+	/* track the device PM state */
+	bool pm_suspended;
+
+	/* track the SE memory writeback state */
+	u8 mw_state;
+
+	/* size of memory used for rotating layers, up to two banks available */
+	u32 rotation_memory[2];
+
+	/* priority level of RQOS register used for driven the ARQOS signal */
+	u32 arqos_value;
+};
+
+static inline u32 malidp_hw_read(struct malidp_hw_device *hwdev, u32 reg)
+{
+	WARN_ON(hwdev->pm_suspended);
+	return readl(hwdev->regs + reg);
+}
+
+static inline void malidp_hw_write(struct malidp_hw_device *hwdev,
+				   u32 value, u32 reg)
+{
+	WARN_ON(hwdev->pm_suspended);
+	writel(value, hwdev->regs + reg);
+}
+
+static inline void malidp_hw_setbits(struct malidp_hw_device *hwdev,
+				     u32 mask, u32 reg)
+{
+	u32 data = malidp_hw_read(hwdev, reg);
+
+	data |= mask;
+	malidp_hw_write(hwdev, data, reg);
+}
+
+static inline void malidp_hw_clearbits(struct malidp_hw_device *hwdev,
+				       u32 mask, u32 reg)
+{
+	u32 data = malidp_hw_read(hwdev, reg);
+
+	data &= ~mask;
+	malidp_hw_write(hwdev, data, reg);
+}
+
+static inline u32 malidp_get_block_base(struct malidp_hw_device *hwdev,
+					u8 block)
+{
+	switch (block) {
+	case MALIDP_SE_BLOCK:
+		return hwdev->hw->map.se_base;
+	case MALIDP_DC_BLOCK:
+		return hwdev->hw->map.dc_base;
+	}
+
+	return 0;
+}
+
+static inline void malidp_hw_disable_irq(struct malidp_hw_device *hwdev,
+					 u8 block, u32 irq)
+{
+	u32 base = malidp_get_block_base(hwdev, block);
+
+	malidp_hw_clearbits(hwdev, irq, base + MALIDP_REG_MASKIRQ);
+}
+
+static inline void malidp_hw_enable_irq(struct malidp_hw_device *hwdev,
+					u8 block, u32 irq)
+{
+	u32 base = malidp_get_block_base(hwdev, block);
+
+	malidp_hw_setbits(hwdev, irq, base + MALIDP_REG_MASKIRQ);
+}
+
+int malidp_de_irq_init(struct drm_device *drm, int irq);
+void malidp_se_irq_hw_init(struct malidp_hw_device *hwdev);
+void malidp_de_irq_hw_init(struct malidp_hw_device *hwdev);
+void malidp_de_irq_fini(struct malidp_hw_device *hwdev);
+int malidp_se_irq_init(struct drm_device *drm, int irq);
+void malidp_se_irq_fini(struct malidp_hw_device *hwdev);
+
+u8 malidp_hw_get_format_id(const struct malidp_hw_regmap *map,
+			   u8 layer_id, u32 format, bool has_modifier);
+
+int malidp_format_get_bpp(u32 fmt);
+
+static inline u8 malidp_hw_get_pitch_align(struct malidp_hw_device *hwdev, bool rotated)
+{
+	/*
+	 * only hardware that cannot do 8 bytes bus alignments have further
+	 * constraints on rotated planes
+	 */
+	if (hwdev->hw->map.bus_align_bytes == 8)
+		return 8;
+	else
+		return hwdev->hw->map.bus_align_bytes << (rotated ? 2 : 0);
+}
+
+/* U16.16 */
+#define FP_1_00000	0x00010000	/* 1.0 */
+#define FP_0_66667	0x0000AAAA	/* 0.6667 = 1/1.5 */
+#define FP_0_50000	0x00008000	/* 0.5 = 1/2 */
+#define FP_0_36363	0x00005D17	/* 0.36363 = 1/2.75 */
+#define FP_0_25000	0x00004000	/* 0.25 = 1/4 */
+
+static inline enum malidp_scaling_coeff_set
+malidp_se_select_coeffs(u32 upscale_factor)
+{
+	return (upscale_factor >= FP_1_00000) ? MALIDP_UPSCALING_COEFFS :
+	       (upscale_factor >= FP_0_66667) ? MALIDP_DOWNSCALING_1_5_COEFFS :
+	       (upscale_factor >= FP_0_50000) ? MALIDP_DOWNSCALING_2_COEFFS :
+	       (upscale_factor >= FP_0_36363) ? MALIDP_DOWNSCALING_2_75_COEFFS :
+	       MALIDP_DOWNSCALING_4_COEFFS;
+}
+
+#undef FP_0_25000
+#undef FP_0_36363
+#undef FP_0_50000
+#undef FP_0_66667
+#undef FP_1_00000
+
+static inline void malidp_se_set_enh_coeffs(struct malidp_hw_device *hwdev)
+{
+	static const s32 enhancer_coeffs[] = {
+		-8, -8, -8, -8, 128, -8, -8, -8, -8
+	};
+	u32 val = MALIDP_SE_SET_ENH_LIMIT_LOW(MALIDP_SE_ENH_LOW_LEVEL) |
+		  MALIDP_SE_SET_ENH_LIMIT_HIGH(MALIDP_SE_ENH_HIGH_LEVEL);
+	u32 image_enh = hwdev->hw->map.se_base +
+			((hwdev->hw->map.features & MALIDP_REGMAP_HAS_CLEARIRQ) ?
+			 0x10 : 0xC) + MALIDP_SE_IMAGE_ENH;
+	u32 enh_coeffs = image_enh + MALIDP_SE_ENH_COEFF0;
+	int i;
+
+	malidp_hw_write(hwdev, val, image_enh);
+	for (i = 0; i < ARRAY_SIZE(enhancer_coeffs); ++i)
+		malidp_hw_write(hwdev, enhancer_coeffs[i], enh_coeffs + i * 4);
+}
+
+/*
+ * background color components are defined as 12bits values,
+ * they will be shifted right when stored on hardware that
+ * supports only 8bits per channel
+ */
+#define MALIDP_BGND_COLOR_R		0x000
+#define MALIDP_BGND_COLOR_G		0x000
+#define MALIDP_BGND_COLOR_B		0x000
+
+#define MALIDP_COLORADJ_NUM_COEFFS	12
+#define MALIDP_COEFFTAB_NUM_COEFFS	64
+
+#define MALIDP_GAMMA_LUT_SIZE		4096
+
+#define AFBC_SIZE_MASK		AFBC_FORMAT_MOD_BLOCK_SIZE_MASK
+#define AFBC_SIZE_16X16		AFBC_FORMAT_MOD_BLOCK_SIZE_16x16
+#define AFBC_YTR		AFBC_FORMAT_MOD_YTR
+#define AFBC_SPARSE		AFBC_FORMAT_MOD_SPARSE
+#define AFBC_CBR		AFBC_FORMAT_MOD_CBR
+#define AFBC_SPLIT		AFBC_FORMAT_MOD_SPLIT
+#define AFBC_TILED		AFBC_FORMAT_MOD_TILED
+#define AFBC_SC			AFBC_FORMAT_MOD_SC
+
+#define AFBC_MOD_VALID_BITS	(AFBC_SIZE_MASK | AFBC_YTR | AFBC_SPLIT | \
+				 AFBC_SPARSE | AFBC_CBR | AFBC_TILED | AFBC_SC)
+
+extern const u64 malidp_format_modifiers[];
+
+#endif  /* __MALIDP_HW_H__ */
diff --git a/drivers/gpu/drm/arm/malidp_mw.c b/drivers/gpu/drm/arm/malidp_mw.c
new file mode 100644
index 000000000..7d0e7b031
--- /dev/null
+++ b/drivers/gpu/drm/arm/malidp_mw.c
@@ -0,0 +1,272 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2016 ARM Limited. All rights reserved.
+ * Author: Brian Starkey <brian.starkey@arm.com>
+ *
+ * ARM Mali DP Writeback connector implementation
+ */
+
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_fb_cma_helper.h>
+#include <drm/drm_fourcc.h>
+#include <drm/drm_gem_cma_helper.h>
+#include <drm/drm_probe_helper.h>
+#include <drm/drm_writeback.h>
+
+#include "malidp_drv.h"
+#include "malidp_hw.h"
+#include "malidp_mw.h"
+
+#define to_mw_state(_state) (struct malidp_mw_connector_state *)(_state)
+
+struct malidp_mw_connector_state {
+	struct drm_connector_state base;
+	dma_addr_t addrs[2];
+	s32 pitches[2];
+	u8 format;
+	u8 n_planes;
+	bool rgb2yuv_initialized;
+	const s16 *rgb2yuv_coeffs;
+};
+
+static int malidp_mw_connector_get_modes(struct drm_connector *connector)
+{
+	struct drm_device *dev = connector->dev;
+
+	return drm_add_modes_noedid(connector, dev->mode_config.max_width,
+				    dev->mode_config.max_height);
+}
+
+static enum drm_mode_status
+malidp_mw_connector_mode_valid(struct drm_connector *connector,
+			       struct drm_display_mode *mode)
+{
+	struct drm_device *dev = connector->dev;
+	struct drm_mode_config *mode_config = &dev->mode_config;
+	int w = mode->hdisplay, h = mode->vdisplay;
+
+	if ((w < mode_config->min_width) || (w > mode_config->max_width))
+		return MODE_BAD_HVALUE;
+
+	if ((h < mode_config->min_height) || (h > mode_config->max_height))
+		return MODE_BAD_VVALUE;
+
+	return MODE_OK;
+}
+
+static const struct drm_connector_helper_funcs malidp_mw_connector_helper_funcs = {
+	.get_modes = malidp_mw_connector_get_modes,
+	.mode_valid = malidp_mw_connector_mode_valid,
+};
+
+static void malidp_mw_connector_reset(struct drm_connector *connector)
+{
+	struct malidp_mw_connector_state *mw_state =
+		kzalloc(sizeof(*mw_state), GFP_KERNEL);
+
+	if (connector->state)
+		__drm_atomic_helper_connector_destroy_state(connector->state);
+
+	kfree(connector->state);
+	__drm_atomic_helper_connector_reset(connector, &mw_state->base);
+}
+
+static enum drm_connector_status
+malidp_mw_connector_detect(struct drm_connector *connector, bool force)
+{
+	return connector_status_connected;
+}
+
+static void malidp_mw_connector_destroy(struct drm_connector *connector)
+{
+	drm_connector_cleanup(connector);
+}
+
+static struct drm_connector_state *
+malidp_mw_connector_duplicate_state(struct drm_connector *connector)
+{
+	struct malidp_mw_connector_state *mw_state, *mw_current_state;
+
+	if (WARN_ON(!connector->state))
+		return NULL;
+
+	mw_state = kzalloc(sizeof(*mw_state), GFP_KERNEL);
+	if (!mw_state)
+		return NULL;
+
+	mw_current_state = to_mw_state(connector->state);
+	mw_state->rgb2yuv_coeffs = mw_current_state->rgb2yuv_coeffs;
+	mw_state->rgb2yuv_initialized = mw_current_state->rgb2yuv_initialized;
+
+	__drm_atomic_helper_connector_duplicate_state(connector, &mw_state->base);
+
+	return &mw_state->base;
+}
+
+static const struct drm_connector_funcs malidp_mw_connector_funcs = {
+	.reset = malidp_mw_connector_reset,
+	.detect = malidp_mw_connector_detect,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.destroy = malidp_mw_connector_destroy,
+	.atomic_duplicate_state = malidp_mw_connector_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
+};
+
+static const s16 rgb2yuv_coeffs_bt709_limited[MALIDP_COLORADJ_NUM_COEFFS] = {
+	47,  157,   16,
+	-26,  -87,  112,
+	112, -102,  -10,
+	16,  128,  128
+};
+
+static int
+malidp_mw_encoder_atomic_check(struct drm_encoder *encoder,
+			       struct drm_crtc_state *crtc_state,
+			       struct drm_connector_state *conn_state)
+{
+	struct malidp_mw_connector_state *mw_state = to_mw_state(conn_state);
+	struct malidp_drm *malidp = encoder->dev->dev_private;
+	struct drm_framebuffer *fb;
+	int i, n_planes;
+
+	if (!conn_state->writeback_job)
+		return 0;
+
+	fb = conn_state->writeback_job->fb;
+	if ((fb->width != crtc_state->mode.hdisplay) ||
+	    (fb->height != crtc_state->mode.vdisplay)) {
+		DRM_DEBUG_KMS("Invalid framebuffer size %ux%u\n",
+				fb->width, fb->height);
+		return -EINVAL;
+	}
+
+	if (fb->modifier) {
+		DRM_DEBUG_KMS("Writeback framebuffer does not support modifiers\n");
+		return -EINVAL;
+	}
+
+	mw_state->format =
+		malidp_hw_get_format_id(&malidp->dev->hw->map, SE_MEMWRITE,
+					fb->format->format, !!fb->modifier);
+	if (mw_state->format == MALIDP_INVALID_FORMAT_ID) {
+		struct drm_format_name_buf format_name;
+
+		DRM_DEBUG_KMS("Invalid pixel format %s\n",
+			      drm_get_format_name(fb->format->format,
+						  &format_name));
+		return -EINVAL;
+	}
+
+	n_planes = fb->format->num_planes;
+	for (i = 0; i < n_planes; i++) {
+		struct drm_gem_cma_object *obj = drm_fb_cma_get_gem_obj(fb, i);
+		/* memory write buffers are never rotated */
+		u8 alignment = malidp_hw_get_pitch_align(malidp->dev, 0);
+
+		if (fb->pitches[i] & (alignment - 1)) {
+			DRM_DEBUG_KMS("Invalid pitch %u for plane %d\n",
+				      fb->pitches[i], i);
+			return -EINVAL;
+		}
+		mw_state->pitches[i] = fb->pitches[i];
+		mw_state->addrs[i] = obj->paddr + fb->offsets[i];
+	}
+	mw_state->n_planes = n_planes;
+
+	if (fb->format->is_yuv)
+		mw_state->rgb2yuv_coeffs = rgb2yuv_coeffs_bt709_limited;
+
+	return 0;
+}
+
+static const struct drm_encoder_helper_funcs malidp_mw_encoder_helper_funcs = {
+	.atomic_check = malidp_mw_encoder_atomic_check,
+};
+
+static u32 *get_writeback_formats(struct malidp_drm *malidp, int *n_formats)
+{
+	const struct malidp_hw_regmap *map = &malidp->dev->hw->map;
+	u32 *formats;
+	int n, i;
+
+	formats = kcalloc(map->n_pixel_formats, sizeof(*formats),
+			  GFP_KERNEL);
+	if (!formats)
+		return NULL;
+
+	for (n = 0, i = 0;  i < map->n_pixel_formats; i++) {
+		if (map->pixel_formats[i].layer & SE_MEMWRITE)
+			formats[n++] = map->pixel_formats[i].format;
+	}
+
+	*n_formats = n;
+
+	return formats;
+}
+
+int malidp_mw_connector_init(struct drm_device *drm)
+{
+	struct malidp_drm *malidp = drm->dev_private;
+	u32 *formats;
+	int ret, n_formats;
+
+	if (!malidp->dev->hw->enable_memwrite)
+		return 0;
+
+	malidp->mw_connector.encoder.possible_crtcs = 1 << drm_crtc_index(&malidp->crtc);
+	drm_connector_helper_add(&malidp->mw_connector.base,
+				 &malidp_mw_connector_helper_funcs);
+
+	formats = get_writeback_formats(malidp, &n_formats);
+	if (!formats)
+		return -ENOMEM;
+
+	ret = drm_writeback_connector_init(drm, &malidp->mw_connector,
+					   &malidp_mw_connector_funcs,
+					   &malidp_mw_encoder_helper_funcs,
+					   formats, n_formats);
+	kfree(formats);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+void malidp_mw_atomic_commit(struct drm_device *drm,
+			     struct drm_atomic_state *old_state)
+{
+	struct malidp_drm *malidp = drm->dev_private;
+	struct drm_writeback_connector *mw_conn = &malidp->mw_connector;
+	struct drm_connector_state *conn_state = mw_conn->base.state;
+	struct malidp_hw_device *hwdev = malidp->dev;
+	struct malidp_mw_connector_state *mw_state;
+
+	if (!conn_state)
+		return;
+
+	mw_state = to_mw_state(conn_state);
+
+	if (conn_state->writeback_job) {
+		struct drm_framebuffer *fb = conn_state->writeback_job->fb;
+
+		DRM_DEV_DEBUG_DRIVER(drm->dev,
+				     "Enable memwrite %ux%u:%d %pad fmt: %u\n",
+				     fb->width, fb->height,
+				     mw_state->pitches[0],
+				     &mw_state->addrs[0],
+				     mw_state->format);
+
+		drm_writeback_queue_job(mw_conn, conn_state);
+		hwdev->hw->enable_memwrite(hwdev, mw_state->addrs,
+					   mw_state->pitches, mw_state->n_planes,
+					   fb->width, fb->height, mw_state->format,
+					   !mw_state->rgb2yuv_initialized ?
+					   mw_state->rgb2yuv_coeffs : NULL);
+		mw_state->rgb2yuv_initialized = !!mw_state->rgb2yuv_coeffs;
+	} else {
+		DRM_DEV_DEBUG_DRIVER(drm->dev, "Disable memwrite\n");
+		hwdev->hw->disable_memwrite(hwdev);
+	}
+}
diff --git a/drivers/gpu/drm/arm/malidp_mw.h b/drivers/gpu/drm/arm/malidp_mw.h
new file mode 100644
index 000000000..19a007676
--- /dev/null
+++ b/drivers/gpu/drm/arm/malidp_mw.h
@@ -0,0 +1,14 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * (C) COPYRIGHT 2016 ARM Limited. All rights reserved.
+ * Author: Brian Starkey <brian.starkey@arm.com>
+ *
+ */
+
+#ifndef __MALIDP_MW_H__
+#define __MALIDP_MW_H__
+
+int malidp_mw_connector_init(struct drm_device *drm);
+void malidp_mw_atomic_commit(struct drm_device *drm,
+			     struct drm_atomic_state *old_state);
+#endif
diff --git a/drivers/gpu/drm/arm/malidp_planes.c b/drivers/gpu/drm/arm/malidp_planes.c
new file mode 100644
index 000000000..24604b410
--- /dev/null
+++ b/drivers/gpu/drm/arm/malidp_planes.c
@@ -0,0 +1,1045 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * (C) COPYRIGHT 2016 ARM Limited. All rights reserved.
+ * Author: Liviu Dudau <Liviu.Dudau@arm.com>
+ *
+ * ARM Mali DP plane manipulation routines.
+ */
+
+#include <linux/iommu.h>
+#include <linux/platform_device.h>
+
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_drv.h>
+#include <drm/drm_fb_cma_helper.h>
+#include <drm/drm_fourcc.h>
+#include <drm/drm_gem_cma_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
+#include <drm/drm_plane_helper.h>
+#include <drm/drm_print.h>
+
+#include "malidp_hw.h"
+#include "malidp_drv.h"
+
+/* Layer specific register offsets */
+#define MALIDP_LAYER_FORMAT		0x000
+#define   LAYER_FORMAT_MASK		0x3f
+#define MALIDP_LAYER_CONTROL		0x004
+#define   LAYER_ENABLE			(1 << 0)
+#define   LAYER_FLOWCFG_MASK		7
+#define   LAYER_FLOWCFG(x)		(((x) & LAYER_FLOWCFG_MASK) << 1)
+#define     LAYER_FLOWCFG_SCALE_SE	3
+#define   LAYER_ROT_OFFSET		8
+#define   LAYER_H_FLIP			(1 << 10)
+#define   LAYER_V_FLIP			(1 << 11)
+#define   LAYER_ROT_MASK		(0xf << 8)
+#define   LAYER_COMP_MASK		(0x3 << 12)
+#define   LAYER_COMP_PIXEL		(0x3 << 12)
+#define   LAYER_COMP_PLANE		(0x2 << 12)
+#define   LAYER_PMUL_ENABLE		(0x1 << 14)
+#define   LAYER_ALPHA_OFFSET		(16)
+#define   LAYER_ALPHA_MASK		(0xff)
+#define   LAYER_ALPHA(x)		(((x) & LAYER_ALPHA_MASK) << LAYER_ALPHA_OFFSET)
+#define MALIDP_LAYER_COMPOSE		0x008
+#define MALIDP_LAYER_SIZE		0x00c
+#define   LAYER_H_VAL(x)		(((x) & 0x1fff) << 0)
+#define   LAYER_V_VAL(x)		(((x) & 0x1fff) << 16)
+#define MALIDP_LAYER_COMP_SIZE		0x010
+#define MALIDP_LAYER_OFFSET		0x014
+#define MALIDP550_LS_ENABLE		0x01c
+#define MALIDP550_LS_R1_IN_SIZE		0x020
+
+#define MODIFIERS_COUNT_MAX		15
+
+/*
+ * This 4-entry look-up-table is used to determine the full 8-bit alpha value
+ * for formats with 1- or 2-bit alpha channels.
+ * We set it to give 100%/0% opacity for 1-bit formats and 100%/66%/33%/0%
+ * opacity for 2-bit formats.
+ */
+#define MALIDP_ALPHA_LUT 0xffaa5500
+
+/* page sizes the MMU prefetcher can support */
+#define MALIDP_MMU_PREFETCH_PARTIAL_PGSIZES	(SZ_4K | SZ_64K)
+#define MALIDP_MMU_PREFETCH_FULL_PGSIZES	(SZ_1M | SZ_2M)
+
+/* readahead for partial-frame prefetch */
+#define MALIDP_MMU_PREFETCH_READAHEAD		8
+
+static void malidp_de_plane_destroy(struct drm_plane *plane)
+{
+	struct malidp_plane *mp = to_malidp_plane(plane);
+
+	drm_plane_cleanup(plane);
+	kfree(mp);
+}
+
+/*
+ * Replicate what the default ->reset hook does: free the state pointer and
+ * allocate a new empty object. We just need enough space to store
+ * a malidp_plane_state instead of a drm_plane_state.
+ */
+static void malidp_plane_reset(struct drm_plane *plane)
+{
+	struct malidp_plane_state *state = to_malidp_plane_state(plane->state);
+
+	if (state)
+		__drm_atomic_helper_plane_destroy_state(&state->base);
+	kfree(state);
+	plane->state = NULL;
+	state = kzalloc(sizeof(*state), GFP_KERNEL);
+	if (state)
+		__drm_atomic_helper_plane_reset(plane, &state->base);
+}
+
+static struct
+drm_plane_state *malidp_duplicate_plane_state(struct drm_plane *plane)
+{
+	struct malidp_plane_state *state, *m_state;
+
+	if (!plane->state)
+		return NULL;
+
+	state = kmalloc(sizeof(*state), GFP_KERNEL);
+	if (!state)
+		return NULL;
+
+	m_state = to_malidp_plane_state(plane->state);
+	__drm_atomic_helper_plane_duplicate_state(plane, &state->base);
+	state->rotmem_size = m_state->rotmem_size;
+	state->format = m_state->format;
+	state->n_planes = m_state->n_planes;
+
+	state->mmu_prefetch_mode = m_state->mmu_prefetch_mode;
+	state->mmu_prefetch_pgsize = m_state->mmu_prefetch_pgsize;
+
+	return &state->base;
+}
+
+static void malidp_destroy_plane_state(struct drm_plane *plane,
+				       struct drm_plane_state *state)
+{
+	struct malidp_plane_state *m_state = to_malidp_plane_state(state);
+
+	__drm_atomic_helper_plane_destroy_state(state);
+	kfree(m_state);
+}
+
+static const char * const prefetch_mode_names[] = {
+	[MALIDP_PREFETCH_MODE_NONE] = "MMU_PREFETCH_NONE",
+	[MALIDP_PREFETCH_MODE_PARTIAL] = "MMU_PREFETCH_PARTIAL",
+	[MALIDP_PREFETCH_MODE_FULL] = "MMU_PREFETCH_FULL",
+};
+
+static void malidp_plane_atomic_print_state(struct drm_printer *p,
+					    const struct drm_plane_state *state)
+{
+	struct malidp_plane_state *ms = to_malidp_plane_state(state);
+
+	drm_printf(p, "\trotmem_size=%u\n", ms->rotmem_size);
+	drm_printf(p, "\tformat_id=%u\n", ms->format);
+	drm_printf(p, "\tn_planes=%u\n", ms->n_planes);
+	drm_printf(p, "\tmmu_prefetch_mode=%s\n",
+		   prefetch_mode_names[ms->mmu_prefetch_mode]);
+	drm_printf(p, "\tmmu_prefetch_pgsize=%d\n", ms->mmu_prefetch_pgsize);
+}
+
+bool malidp_format_mod_supported(struct drm_device *drm,
+				 u32 format, u64 modifier)
+{
+	const struct drm_format_info *info;
+	const u64 *modifiers;
+	struct malidp_drm *malidp = drm->dev_private;
+	const struct malidp_hw_regmap *map = &malidp->dev->hw->map;
+
+	if (WARN_ON(modifier == DRM_FORMAT_MOD_INVALID))
+		return false;
+
+	/* Some pixel formats are supported without any modifier */
+	if (modifier == DRM_FORMAT_MOD_LINEAR) {
+		/*
+		 * However these pixel formats need to be supported with
+		 * modifiers only
+		 */
+		return !malidp_hw_format_is_afbc_only(format);
+	}
+
+	if ((modifier >> 56) != DRM_FORMAT_MOD_VENDOR_ARM) {
+		DRM_ERROR("Unknown modifier (not Arm)\n");
+		return false;
+	}
+
+	if (modifier &
+	    ~DRM_FORMAT_MOD_ARM_AFBC(AFBC_MOD_VALID_BITS)) {
+		DRM_DEBUG_KMS("Unsupported modifiers\n");
+		return false;
+	}
+
+	modifiers = malidp_format_modifiers;
+
+	/* SPLIT buffers must use SPARSE layout */
+	if (WARN_ON_ONCE((modifier & AFBC_SPLIT) && !(modifier & AFBC_SPARSE)))
+		return false;
+
+	/* CBR only applies to YUV formats, where YTR should be always 0 */
+	if (WARN_ON_ONCE((modifier & AFBC_CBR) && (modifier & AFBC_YTR)))
+		return false;
+
+	while (*modifiers != DRM_FORMAT_MOD_INVALID) {
+		if (*modifiers == modifier)
+			break;
+
+		modifiers++;
+	}
+
+	/* return false, if the modifier was not found */
+	if (*modifiers == DRM_FORMAT_MOD_INVALID) {
+		DRM_DEBUG_KMS("Unsupported modifier\n");
+		return false;
+	}
+
+	info = drm_format_info(format);
+
+	if (info->num_planes != 1) {
+		DRM_DEBUG_KMS("AFBC buffers expect one plane\n");
+		return false;
+	}
+
+	if (malidp_hw_format_is_linear_only(format) == true) {
+		DRM_DEBUG_KMS("Given format (0x%x) is supported is linear mode only\n",
+			      format);
+		return false;
+	}
+
+	/*
+	 * RGB formats need to provide YTR modifier and YUV formats should not
+	 * provide YTR modifier.
+	 */
+	if (!(info->is_yuv) != !!(modifier & AFBC_FORMAT_MOD_YTR)) {
+		DRM_DEBUG_KMS("AFBC_FORMAT_MOD_YTR is %s for %s formats\n",
+			      info->is_yuv ? "disallowed" : "mandatory",
+			      info->is_yuv ? "YUV" : "RGB");
+		return false;
+	}
+
+	if (modifier & AFBC_SPLIT) {
+		if (!info->is_yuv) {
+			if (info->cpp[0] <= 2) {
+				DRM_DEBUG_KMS("RGB formats <= 16bpp are not supported with SPLIT\n");
+				return false;
+			}
+		}
+
+		if ((info->hsub != 1) || (info->vsub != 1)) {
+			if (!(format == DRM_FORMAT_YUV420_10BIT &&
+			      (map->features & MALIDP_DEVICE_AFBC_YUV_420_10_SUPPORT_SPLIT))) {
+				DRM_DEBUG_KMS("Formats which are sub-sampled should never be split\n");
+				return false;
+			}
+		}
+	}
+
+	if (modifier & AFBC_CBR) {
+		if ((info->hsub == 1) || (info->vsub == 1)) {
+			DRM_DEBUG_KMS("Formats which are not sub-sampled should not have CBR set\n");
+			return false;
+		}
+	}
+
+	return true;
+}
+
+static bool malidp_format_mod_supported_per_plane(struct drm_plane *plane,
+						  u32 format, u64 modifier)
+{
+	return malidp_format_mod_supported(plane->dev, format, modifier);
+}
+
+static const struct drm_plane_funcs malidp_de_plane_funcs = {
+	.update_plane = drm_atomic_helper_update_plane,
+	.disable_plane = drm_atomic_helper_disable_plane,
+	.destroy = malidp_de_plane_destroy,
+	.reset = malidp_plane_reset,
+	.atomic_duplicate_state = malidp_duplicate_plane_state,
+	.atomic_destroy_state = malidp_destroy_plane_state,
+	.atomic_print_state = malidp_plane_atomic_print_state,
+	.format_mod_supported = malidp_format_mod_supported_per_plane,
+};
+
+static int malidp_se_check_scaling(struct malidp_plane *mp,
+				   struct drm_plane_state *state)
+{
+	struct drm_crtc_state *crtc_state =
+		drm_atomic_get_existing_crtc_state(state->state, state->crtc);
+	struct malidp_crtc_state *mc;
+	u32 src_w, src_h;
+	int ret;
+
+	if (!crtc_state)
+		return -EINVAL;
+
+	mc = to_malidp_crtc_state(crtc_state);
+
+	ret = drm_atomic_helper_check_plane_state(state, crtc_state,
+						  0, INT_MAX, true, true);
+	if (ret)
+		return ret;
+
+	if (state->rotation & MALIDP_ROTATED_MASK) {
+		src_w = state->src_h >> 16;
+		src_h = state->src_w >> 16;
+	} else {
+		src_w = state->src_w >> 16;
+		src_h = state->src_h >> 16;
+	}
+
+	if ((state->crtc_w == src_w) && (state->crtc_h == src_h)) {
+		/* Scaling not necessary for this plane. */
+		mc->scaled_planes_mask &= ~(mp->layer->id);
+		return 0;
+	}
+
+	if (mp->layer->id & (DE_SMART | DE_GRAPHICS2))
+		return -EINVAL;
+
+	mc->scaled_planes_mask |= mp->layer->id;
+	/* Defer scaling requirements calculation to the crtc check. */
+	return 0;
+}
+
+static u32 malidp_get_pgsize_bitmap(struct malidp_plane *mp)
+{
+	u32 pgsize_bitmap = 0;
+
+	if (iommu_present(&platform_bus_type)) {
+		struct iommu_domain *mmu_dom =
+			iommu_get_domain_for_dev(mp->base.dev->dev);
+
+		if (mmu_dom)
+			pgsize_bitmap = mmu_dom->pgsize_bitmap;
+	}
+
+	return pgsize_bitmap;
+}
+
+/*
+ * Check if the framebuffer is entirely made up of pages at least pgsize in
+ * size. Only a heuristic: assumes that each scatterlist entry has been aligned
+ * to the largest page size smaller than its length and that the MMU maps to
+ * the largest page size possible.
+ */
+static bool malidp_check_pages_threshold(struct malidp_plane_state *ms,
+					 u32 pgsize)
+{
+	int i;
+
+	for (i = 0; i < ms->n_planes; i++) {
+		struct drm_gem_object *obj;
+		struct drm_gem_cma_object *cma_obj;
+		struct sg_table *sgt;
+		struct scatterlist *sgl;
+
+		obj = drm_gem_fb_get_obj(ms->base.fb, i);
+		cma_obj = to_drm_gem_cma_obj(obj);
+
+		if (cma_obj->sgt)
+			sgt = cma_obj->sgt;
+		else
+			sgt = obj->funcs->get_sg_table(obj);
+
+		if (IS_ERR(sgt))
+			return false;
+
+		sgl = sgt->sgl;
+
+		while (sgl) {
+			if (sgl->length < pgsize) {
+				if (!cma_obj->sgt)
+					kfree(sgt);
+				return false;
+			}
+
+			sgl = sg_next(sgl);
+		}
+		if (!cma_obj->sgt)
+			kfree(sgt);
+	}
+
+	return true;
+}
+
+/*
+ * Check if it is possible to enable partial-frame MMU prefetch given the
+ * current format, AFBC state and rotation.
+ */
+static bool malidp_partial_prefetch_supported(u32 format, u64 modifier,
+					      unsigned int rotation)
+{
+	bool afbc, sparse;
+
+	/* rotation and horizontal flip not supported for partial prefetch */
+	if (rotation & (DRM_MODE_ROTATE_90 | DRM_MODE_ROTATE_180 |
+			DRM_MODE_ROTATE_270 | DRM_MODE_REFLECT_X))
+		return false;
+
+	afbc = modifier & DRM_FORMAT_MOD_ARM_AFBC(0);
+	sparse = modifier & AFBC_FORMAT_MOD_SPARSE;
+
+	switch (format) {
+	case DRM_FORMAT_ARGB2101010:
+	case DRM_FORMAT_RGBA1010102:
+	case DRM_FORMAT_BGRA1010102:
+	case DRM_FORMAT_ARGB8888:
+	case DRM_FORMAT_RGBA8888:
+	case DRM_FORMAT_BGRA8888:
+	case DRM_FORMAT_XRGB8888:
+	case DRM_FORMAT_XBGR8888:
+	case DRM_FORMAT_RGBX8888:
+	case DRM_FORMAT_BGRX8888:
+	case DRM_FORMAT_RGB888:
+	case DRM_FORMAT_RGBA5551:
+	case DRM_FORMAT_RGB565:
+		/* always supported */
+		return true;
+
+	case DRM_FORMAT_ABGR2101010:
+	case DRM_FORMAT_ABGR8888:
+	case DRM_FORMAT_ABGR1555:
+	case DRM_FORMAT_BGR565:
+		/* supported, but if AFBC then must be sparse mode */
+		return (!afbc) || (afbc && sparse);
+
+	case DRM_FORMAT_BGR888:
+		/* supported, but not for AFBC */
+		return !afbc;
+
+	case DRM_FORMAT_YUYV:
+	case DRM_FORMAT_UYVY:
+	case DRM_FORMAT_NV12:
+	case DRM_FORMAT_YUV420:
+		/* not supported */
+		return false;
+
+	default:
+		return false;
+	}
+}
+
+/*
+ * Select the preferred MMU prefetch mode. Full-frame prefetch is preferred as
+ * long as the framebuffer is all large pages. Otherwise partial-frame prefetch
+ * is selected as long as it is supported for the current format. The selected
+ * page size for prefetch is returned in pgsize_bitmap.
+ */
+static enum mmu_prefetch_mode malidp_mmu_prefetch_select_mode
+		(struct malidp_plane_state *ms,	u32 *pgsize_bitmap)
+{
+	u32 pgsizes;
+
+	/* get the full-frame prefetch page size(s) supported by the MMU */
+	pgsizes = *pgsize_bitmap & MALIDP_MMU_PREFETCH_FULL_PGSIZES;
+
+	while (pgsizes) {
+		u32 largest_pgsize = 1 << __fls(pgsizes);
+
+		if (malidp_check_pages_threshold(ms, largest_pgsize)) {
+			*pgsize_bitmap = largest_pgsize;
+			return MALIDP_PREFETCH_MODE_FULL;
+		}
+
+		pgsizes -= largest_pgsize;
+	}
+
+	/* get the partial-frame prefetch page size(s) supported by the MMU */
+	pgsizes = *pgsize_bitmap & MALIDP_MMU_PREFETCH_PARTIAL_PGSIZES;
+
+	if (malidp_partial_prefetch_supported(ms->base.fb->format->format,
+					      ms->base.fb->modifier,
+					      ms->base.rotation)) {
+		/* partial prefetch using the smallest page size */
+		*pgsize_bitmap = 1 << __ffs(pgsizes);
+		return MALIDP_PREFETCH_MODE_PARTIAL;
+	}
+	*pgsize_bitmap = 0;
+	return MALIDP_PREFETCH_MODE_NONE;
+}
+
+static u32 malidp_calc_mmu_control_value(enum mmu_prefetch_mode mode,
+					 u8 readahead, u8 n_planes, u32 pgsize)
+{
+	u32 mmu_ctrl = 0;
+
+	if (mode != MALIDP_PREFETCH_MODE_NONE) {
+		mmu_ctrl |= MALIDP_MMU_CTRL_EN;
+
+		if (mode == MALIDP_PREFETCH_MODE_PARTIAL) {
+			mmu_ctrl |= MALIDP_MMU_CTRL_MODE;
+			mmu_ctrl |= MALIDP_MMU_CTRL_PP_NUM_REQ(readahead);
+		}
+
+		if (pgsize == SZ_64K || pgsize == SZ_2M) {
+			int i;
+
+			for (i = 0; i < n_planes; i++)
+				mmu_ctrl |= MALIDP_MMU_CTRL_PX_PS(i);
+		}
+	}
+
+	return mmu_ctrl;
+}
+
+static void malidp_de_prefetch_settings(struct malidp_plane *mp,
+					struct malidp_plane_state *ms)
+{
+	if (!mp->layer->mmu_ctrl_offset)
+		return;
+
+	/* get the page sizes supported by the MMU */
+	ms->mmu_prefetch_pgsize = malidp_get_pgsize_bitmap(mp);
+	ms->mmu_prefetch_mode  =
+		malidp_mmu_prefetch_select_mode(ms, &ms->mmu_prefetch_pgsize);
+}
+
+static int malidp_de_plane_check(struct drm_plane *plane,
+				 struct drm_plane_state *state)
+{
+	struct malidp_plane *mp = to_malidp_plane(plane);
+	struct malidp_plane_state *ms = to_malidp_plane_state(state);
+	bool rotated = state->rotation & MALIDP_ROTATED_MASK;
+	struct drm_framebuffer *fb;
+	u16 pixel_alpha = state->pixel_blend_mode;
+	int i, ret;
+	unsigned int block_w, block_h;
+
+	if (!state->crtc || WARN_ON(!state->fb))
+		return 0;
+
+	fb = state->fb;
+
+	ms->format = malidp_hw_get_format_id(&mp->hwdev->hw->map,
+					     mp->layer->id, fb->format->format,
+					     !!fb->modifier);
+	if (ms->format == MALIDP_INVALID_FORMAT_ID)
+		return -EINVAL;
+
+	ms->n_planes = fb->format->num_planes;
+	for (i = 0; i < ms->n_planes; i++) {
+		u8 alignment = malidp_hw_get_pitch_align(mp->hwdev, rotated);
+
+		if (((fb->pitches[i] * drm_format_info_block_height(fb->format, i))
+				& (alignment - 1)) && !(fb->modifier)) {
+			DRM_DEBUG_KMS("Invalid pitch %u for plane %d\n",
+				      fb->pitches[i], i);
+			return -EINVAL;
+		}
+	}
+
+	block_w = drm_format_info_block_width(fb->format, 0);
+	block_h = drm_format_info_block_height(fb->format, 0);
+	if (fb->width % block_w || fb->height % block_h) {
+		DRM_DEBUG_KMS("Buffer width/height needs to be a multiple of tile sizes");
+		return -EINVAL;
+	}
+	if ((state->src_x >> 16) % block_w || (state->src_y >> 16) % block_h) {
+		DRM_DEBUG_KMS("Plane src_x/src_y needs to be a multiple of tile sizes");
+		return -EINVAL;
+	}
+
+	if ((state->crtc_w > mp->hwdev->max_line_size) ||
+	    (state->crtc_h > mp->hwdev->max_line_size) ||
+	    (state->crtc_w < mp->hwdev->min_line_size) ||
+	    (state->crtc_h < mp->hwdev->min_line_size))
+		return -EINVAL;
+
+	/*
+	 * DP550/650 video layers can accept 3 plane formats only if
+	 * fb->pitches[1] == fb->pitches[2] since they don't have a
+	 * third plane stride register.
+	 */
+	if (ms->n_planes == 3 &&
+	    !(mp->hwdev->hw->features & MALIDP_DEVICE_LV_HAS_3_STRIDES) &&
+	    (state->fb->pitches[1] != state->fb->pitches[2]))
+		return -EINVAL;
+
+	ret = malidp_se_check_scaling(mp, state);
+	if (ret)
+		return ret;
+
+	/* validate the rotation constraints for each layer */
+	if (state->rotation != DRM_MODE_ROTATE_0) {
+		if (mp->layer->rot == ROTATE_NONE)
+			return -EINVAL;
+		if ((mp->layer->rot == ROTATE_COMPRESSED) && !(fb->modifier))
+			return -EINVAL;
+		/*
+		 * packed RGB888 / BGR888 can't be rotated or flipped
+		 * unless they are stored in a compressed way
+		 */
+		if ((fb->format->format == DRM_FORMAT_RGB888 ||
+		     fb->format->format == DRM_FORMAT_BGR888) && !(fb->modifier))
+			return -EINVAL;
+	}
+
+	/* SMART layer does not support AFBC */
+	if (mp->layer->id == DE_SMART && fb->modifier) {
+		DRM_ERROR("AFBC framebuffer not supported in SMART layer");
+		return -EINVAL;
+	}
+
+	ms->rotmem_size = 0;
+	if (state->rotation & MALIDP_ROTATED_MASK) {
+		int val;
+
+		val = mp->hwdev->hw->rotmem_required(mp->hwdev, state->crtc_w,
+						     state->crtc_h,
+						     fb->format->format,
+						     !!(fb->modifier));
+		if (val < 0)
+			return val;
+
+		ms->rotmem_size = val;
+	}
+
+	/* HW can't support plane + pixel blending */
+	if ((state->alpha != DRM_BLEND_ALPHA_OPAQUE) &&
+	    (pixel_alpha != DRM_MODE_BLEND_PIXEL_NONE) &&
+	    fb->format->has_alpha)
+		return -EINVAL;
+
+	malidp_de_prefetch_settings(mp, ms);
+
+	return 0;
+}
+
+static void malidp_de_set_plane_pitches(struct malidp_plane *mp,
+					int num_planes, unsigned int pitches[3])
+{
+	int i;
+	int num_strides = num_planes;
+
+	if (!mp->layer->stride_offset)
+		return;
+
+	if (num_planes == 3)
+		num_strides = (mp->hwdev->hw->features &
+			       MALIDP_DEVICE_LV_HAS_3_STRIDES) ? 3 : 2;
+
+	/*
+	 * The drm convention for pitch is that it needs to cover width * cpp,
+	 * but our hardware wants the pitch/stride to cover all rows included
+	 * in a tile.
+	 */
+	for (i = 0; i < num_strides; ++i) {
+		unsigned int block_h = drm_format_info_block_height(mp->base.state->fb->format, i);
+
+		malidp_hw_write(mp->hwdev, pitches[i] * block_h,
+				mp->layer->base +
+				mp->layer->stride_offset + i * 4);
+	}
+}
+
+static const s16
+malidp_yuv2rgb_coeffs[][DRM_COLOR_RANGE_MAX][MALIDP_COLORADJ_NUM_COEFFS] = {
+	[DRM_COLOR_YCBCR_BT601][DRM_COLOR_YCBCR_LIMITED_RANGE] = {
+		1192,    0, 1634,
+		1192, -401, -832,
+		1192, 2066,    0,
+		  64,  512,  512
+	},
+	[DRM_COLOR_YCBCR_BT601][DRM_COLOR_YCBCR_FULL_RANGE] = {
+		1024,    0, 1436,
+		1024, -352, -731,
+		1024, 1815,    0,
+		   0,  512,  512
+	},
+	[DRM_COLOR_YCBCR_BT709][DRM_COLOR_YCBCR_LIMITED_RANGE] = {
+		1192,    0, 1836,
+		1192, -218, -546,
+		1192, 2163,    0,
+		  64,  512,  512
+	},
+	[DRM_COLOR_YCBCR_BT709][DRM_COLOR_YCBCR_FULL_RANGE] = {
+		1024,    0, 1613,
+		1024, -192, -479,
+		1024, 1900,    0,
+		   0,  512,  512
+	},
+	[DRM_COLOR_YCBCR_BT2020][DRM_COLOR_YCBCR_LIMITED_RANGE] = {
+		1024,    0, 1476,
+		1024, -165, -572,
+		1024, 1884,    0,
+		   0,  512,  512
+	},
+	[DRM_COLOR_YCBCR_BT2020][DRM_COLOR_YCBCR_FULL_RANGE] = {
+		1024,    0, 1510,
+		1024, -168, -585,
+		1024, 1927,    0,
+		   0,  512,  512
+	}
+};
+
+static void malidp_de_set_color_encoding(struct malidp_plane *plane,
+					 enum drm_color_encoding enc,
+					 enum drm_color_range range)
+{
+	unsigned int i;
+
+	for (i = 0; i < MALIDP_COLORADJ_NUM_COEFFS; i++) {
+		/* coefficients are signed, two's complement values */
+		malidp_hw_write(plane->hwdev, malidp_yuv2rgb_coeffs[enc][range][i],
+				plane->layer->base + plane->layer->yuv2rgb_offset +
+				i * 4);
+	}
+}
+
+static void malidp_de_set_mmu_control(struct malidp_plane *mp,
+				      struct malidp_plane_state *ms)
+{
+	u32 mmu_ctrl;
+
+	/* check hardware supports MMU prefetch */
+	if (!mp->layer->mmu_ctrl_offset)
+		return;
+
+	mmu_ctrl = malidp_calc_mmu_control_value(ms->mmu_prefetch_mode,
+						 MALIDP_MMU_PREFETCH_READAHEAD,
+						 ms->n_planes,
+						 ms->mmu_prefetch_pgsize);
+
+	malidp_hw_write(mp->hwdev, mmu_ctrl,
+			mp->layer->base + mp->layer->mmu_ctrl_offset);
+}
+
+static void malidp_set_plane_base_addr(struct drm_framebuffer *fb,
+				       struct malidp_plane *mp,
+				       int plane_index)
+{
+	dma_addr_t paddr;
+	u16 ptr;
+	struct drm_plane *plane = &mp->base;
+	bool afbc = fb->modifier ? true : false;
+
+	ptr = mp->layer->ptr + (plane_index << 4);
+
+	/*
+	 * drm_fb_cma_get_gem_addr() alters the physical base address of the
+	 * framebuffer as per the plane's src_x, src_y co-ordinates (ie to
+	 * take care of source cropping).
+	 * For AFBC, this is not needed as the cropping is handled by _AD_CROP_H
+	 * and _AD_CROP_V registers.
+	 */
+	if (!afbc) {
+		paddr = drm_fb_cma_get_gem_addr(fb, plane->state,
+						plane_index);
+	} else {
+		struct drm_gem_cma_object *obj;
+
+		obj = drm_fb_cma_get_gem_obj(fb, plane_index);
+
+		if (WARN_ON(!obj))
+			return;
+		paddr = obj->paddr;
+	}
+
+	malidp_hw_write(mp->hwdev, lower_32_bits(paddr), ptr);
+	malidp_hw_write(mp->hwdev, upper_32_bits(paddr), ptr + 4);
+}
+
+static void malidp_de_set_plane_afbc(struct drm_plane *plane)
+{
+	struct malidp_plane *mp;
+	u32 src_w, src_h, val = 0, src_x, src_y;
+	struct drm_framebuffer *fb = plane->state->fb;
+
+	mp = to_malidp_plane(plane);
+
+	/* no afbc_decoder_offset means AFBC is not supported on this plane */
+	if (!mp->layer->afbc_decoder_offset)
+		return;
+
+	if (!fb->modifier) {
+		malidp_hw_write(mp->hwdev, 0, mp->layer->afbc_decoder_offset);
+		return;
+	}
+
+	/* convert src values from Q16 fixed point to integer */
+	src_w = plane->state->src_w >> 16;
+	src_h = plane->state->src_h >> 16;
+	src_x = plane->state->src_x >> 16;
+	src_y = plane->state->src_y >> 16;
+
+	val = ((fb->width - (src_x + src_w)) << MALIDP_AD_CROP_RIGHT_OFFSET) |
+		   src_x;
+	malidp_hw_write(mp->hwdev, val,
+			mp->layer->afbc_decoder_offset + MALIDP_AD_CROP_H);
+
+	val = ((fb->height - (src_y + src_h)) << MALIDP_AD_CROP_BOTTOM_OFFSET) |
+		   src_y;
+	malidp_hw_write(mp->hwdev, val,
+			mp->layer->afbc_decoder_offset + MALIDP_AD_CROP_V);
+
+	val = MALIDP_AD_EN;
+	if (fb->modifier & AFBC_FORMAT_MOD_SPLIT)
+		val |= MALIDP_AD_BS;
+	if (fb->modifier & AFBC_FORMAT_MOD_YTR)
+		val |= MALIDP_AD_YTR;
+
+	malidp_hw_write(mp->hwdev, val, mp->layer->afbc_decoder_offset);
+}
+
+static void malidp_de_plane_update(struct drm_plane *plane,
+				   struct drm_plane_state *old_state)
+{
+	struct malidp_plane *mp;
+	struct malidp_plane_state *ms = to_malidp_plane_state(plane->state);
+	struct drm_plane_state *state = plane->state;
+	u16 pixel_alpha = state->pixel_blend_mode;
+	u8 plane_alpha = state->alpha >> 8;
+	u32 src_w, src_h, dest_w, dest_h, val;
+	int i;
+	struct drm_framebuffer *fb = plane->state->fb;
+
+	mp = to_malidp_plane(plane);
+
+	/*
+	 * For AFBC framebuffer, use the framebuffer width and height for
+	 * configuring layer input size register.
+	 */
+	if (fb->modifier) {
+		src_w = fb->width;
+		src_h = fb->height;
+	} else {
+		/* convert src values from Q16 fixed point to integer */
+		src_w = state->src_w >> 16;
+		src_h = state->src_h >> 16;
+	}
+
+	dest_w = state->crtc_w;
+	dest_h = state->crtc_h;
+
+	val = malidp_hw_read(mp->hwdev, mp->layer->base);
+	val = (val & ~LAYER_FORMAT_MASK) | ms->format;
+	malidp_hw_write(mp->hwdev, val, mp->layer->base);
+
+	for (i = 0; i < ms->n_planes; i++)
+		malidp_set_plane_base_addr(fb, mp, i);
+
+	malidp_de_set_mmu_control(mp, ms);
+
+	malidp_de_set_plane_pitches(mp, ms->n_planes,
+				    state->fb->pitches);
+
+	if ((plane->state->color_encoding != old_state->color_encoding) ||
+	    (plane->state->color_range != old_state->color_range))
+		malidp_de_set_color_encoding(mp, plane->state->color_encoding,
+					     plane->state->color_range);
+
+	malidp_hw_write(mp->hwdev, LAYER_H_VAL(src_w) | LAYER_V_VAL(src_h),
+			mp->layer->base + MALIDP_LAYER_SIZE);
+
+	malidp_hw_write(mp->hwdev, LAYER_H_VAL(dest_w) | LAYER_V_VAL(dest_h),
+			mp->layer->base + MALIDP_LAYER_COMP_SIZE);
+
+	malidp_hw_write(mp->hwdev, LAYER_H_VAL(state->crtc_x) |
+			LAYER_V_VAL(state->crtc_y),
+			mp->layer->base + MALIDP_LAYER_OFFSET);
+
+	if (mp->layer->id == DE_SMART) {
+		/*
+		 * Enable the first rectangle in the SMART layer to be
+		 * able to use it as a drm plane.
+		 */
+		malidp_hw_write(mp->hwdev, 1,
+				mp->layer->base + MALIDP550_LS_ENABLE);
+		malidp_hw_write(mp->hwdev,
+				LAYER_H_VAL(src_w) | LAYER_V_VAL(src_h),
+				mp->layer->base + MALIDP550_LS_R1_IN_SIZE);
+	}
+
+	malidp_de_set_plane_afbc(plane);
+
+	/* first clear the rotation bits */
+	val = malidp_hw_read(mp->hwdev, mp->layer->base + MALIDP_LAYER_CONTROL);
+	val &= ~LAYER_ROT_MASK;
+
+	/* setup the rotation and axis flip bits */
+	if (state->rotation & DRM_MODE_ROTATE_MASK)
+		val |= ilog2(plane->state->rotation & DRM_MODE_ROTATE_MASK) <<
+		       LAYER_ROT_OFFSET;
+	if (state->rotation & DRM_MODE_REFLECT_X)
+		val |= LAYER_H_FLIP;
+	if (state->rotation & DRM_MODE_REFLECT_Y)
+		val |= LAYER_V_FLIP;
+
+	val &= ~(LAYER_COMP_MASK | LAYER_PMUL_ENABLE | LAYER_ALPHA(0xff));
+
+	if (state->alpha != DRM_BLEND_ALPHA_OPAQUE) {
+		val |= LAYER_COMP_PLANE;
+	} else if (state->fb->format->has_alpha) {
+		/* We only care about blend mode if the format has alpha */
+		switch (pixel_alpha) {
+		case DRM_MODE_BLEND_PREMULTI:
+			val |= LAYER_COMP_PIXEL | LAYER_PMUL_ENABLE;
+			break;
+		case DRM_MODE_BLEND_COVERAGE:
+			val |= LAYER_COMP_PIXEL;
+			break;
+		}
+	}
+	val |= LAYER_ALPHA(plane_alpha);
+
+	val &= ~LAYER_FLOWCFG(LAYER_FLOWCFG_MASK);
+	if (state->crtc) {
+		struct malidp_crtc_state *m =
+			to_malidp_crtc_state(state->crtc->state);
+
+		if (m->scaler_config.scale_enable &&
+		    m->scaler_config.plane_src_id == mp->layer->id)
+			val |= LAYER_FLOWCFG(LAYER_FLOWCFG_SCALE_SE);
+	}
+
+	/* set the 'enable layer' bit */
+	val |= LAYER_ENABLE;
+
+	malidp_hw_write(mp->hwdev, val,
+			mp->layer->base + MALIDP_LAYER_CONTROL);
+}
+
+static void malidp_de_plane_disable(struct drm_plane *plane,
+				    struct drm_plane_state *state)
+{
+	struct malidp_plane *mp = to_malidp_plane(plane);
+
+	malidp_hw_clearbits(mp->hwdev,
+			    LAYER_ENABLE | LAYER_FLOWCFG(LAYER_FLOWCFG_MASK),
+			    mp->layer->base + MALIDP_LAYER_CONTROL);
+}
+
+static const struct drm_plane_helper_funcs malidp_de_plane_helper_funcs = {
+	.atomic_check = malidp_de_plane_check,
+	.atomic_update = malidp_de_plane_update,
+	.atomic_disable = malidp_de_plane_disable,
+};
+
+static const uint64_t linear_only_modifiers[] = {
+	DRM_FORMAT_MOD_LINEAR,
+	DRM_FORMAT_MOD_INVALID
+};
+
+int malidp_de_planes_init(struct drm_device *drm)
+{
+	struct malidp_drm *malidp = drm->dev_private;
+	const struct malidp_hw_regmap *map = &malidp->dev->hw->map;
+	struct malidp_plane *plane = NULL;
+	enum drm_plane_type plane_type;
+	unsigned long crtcs = BIT(drm->mode_config.num_crtc);
+	unsigned long flags = DRM_MODE_ROTATE_0 | DRM_MODE_ROTATE_90 | DRM_MODE_ROTATE_180 |
+			      DRM_MODE_ROTATE_270 | DRM_MODE_REFLECT_X | DRM_MODE_REFLECT_Y;
+	unsigned int blend_caps = BIT(DRM_MODE_BLEND_PIXEL_NONE) |
+				  BIT(DRM_MODE_BLEND_PREMULTI)   |
+				  BIT(DRM_MODE_BLEND_COVERAGE);
+	u32 *formats;
+	int ret, i = 0, j = 0, n;
+	u64 supported_modifiers[MODIFIERS_COUNT_MAX];
+	const u64 *modifiers;
+
+	modifiers = malidp_format_modifiers;
+
+	if (!(map->features & MALIDP_DEVICE_AFBC_SUPPORT_SPLIT)) {
+		/*
+		 * Since our hardware does not support SPLIT, so build the list
+		 * of supported modifiers excluding SPLIT ones.
+		 */
+		while (*modifiers != DRM_FORMAT_MOD_INVALID) {
+			if (!(*modifiers & AFBC_SPLIT))
+				supported_modifiers[j++] = *modifiers;
+
+			modifiers++;
+		}
+		supported_modifiers[j++] = DRM_FORMAT_MOD_INVALID;
+		modifiers = supported_modifiers;
+	}
+
+	formats = kcalloc(map->n_pixel_formats, sizeof(*formats), GFP_KERNEL);
+	if (!formats) {
+		ret = -ENOMEM;
+		goto cleanup;
+	}
+
+	for (i = 0; i < map->n_layers; i++) {
+		u8 id = map->layers[i].id;
+
+		plane = kzalloc(sizeof(*plane), GFP_KERNEL);
+		if (!plane) {
+			ret = -ENOMEM;
+			goto cleanup;
+		}
+
+		/* build the list of DRM supported formats based on the map */
+		for (n = 0, j = 0;  j < map->n_pixel_formats; j++) {
+			if ((map->pixel_formats[j].layer & id) == id)
+				formats[n++] = map->pixel_formats[j].format;
+		}
+
+		plane_type = (i == 0) ? DRM_PLANE_TYPE_PRIMARY :
+					DRM_PLANE_TYPE_OVERLAY;
+
+		/*
+		 * All the layers except smart layer supports AFBC modifiers.
+		 */
+		ret = drm_universal_plane_init(drm, &plane->base, crtcs,
+				&malidp_de_plane_funcs, formats, n,
+				(id == DE_SMART) ? linear_only_modifiers : modifiers,
+				plane_type, NULL);
+
+		if (ret < 0)
+			goto cleanup;
+
+		drm_plane_helper_add(&plane->base,
+				     &malidp_de_plane_helper_funcs);
+		plane->hwdev = malidp->dev;
+		plane->layer = &map->layers[i];
+
+		drm_plane_create_alpha_property(&plane->base);
+		drm_plane_create_blend_mode_property(&plane->base, blend_caps);
+
+		if (id == DE_SMART) {
+			/* Skip the features which the SMART layer doesn't have. */
+			continue;
+		}
+
+		drm_plane_create_rotation_property(&plane->base, DRM_MODE_ROTATE_0, flags);
+		malidp_hw_write(malidp->dev, MALIDP_ALPHA_LUT,
+				plane->layer->base + MALIDP_LAYER_COMPOSE);
+
+		/* Attach the YUV->RGB property only to video layers */
+		if (id & (DE_VIDEO1 | DE_VIDEO2)) {
+			/* default encoding for YUV->RGB is BT601 NARROW */
+			enum drm_color_encoding enc = DRM_COLOR_YCBCR_BT601;
+			enum drm_color_range range = DRM_COLOR_YCBCR_LIMITED_RANGE;
+
+			ret = drm_plane_create_color_properties(&plane->base,
+					BIT(DRM_COLOR_YCBCR_BT601) | \
+					BIT(DRM_COLOR_YCBCR_BT709) | \
+					BIT(DRM_COLOR_YCBCR_BT2020),
+					BIT(DRM_COLOR_YCBCR_LIMITED_RANGE) | \
+					BIT(DRM_COLOR_YCBCR_FULL_RANGE),
+					enc, range);
+			if (!ret)
+				/* program the HW registers */
+				malidp_de_set_color_encoding(plane, enc, range);
+			else
+				DRM_WARN("Failed to create video layer %d color properties\n", id);
+		}
+	}
+
+	kfree(formats);
+
+	return 0;
+
+cleanup:
+	kfree(formats);
+
+	return ret;
+}
diff --git a/drivers/gpu/drm/arm/malidp_regs.h b/drivers/gpu/drm/arm/malidp_regs.h
new file mode 100644
index 000000000..514c50dcb
--- /dev/null
+++ b/drivers/gpu/drm/arm/malidp_regs.h
@@ -0,0 +1,301 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * (C) COPYRIGHT 2016 ARM Limited. All rights reserved.
+ * Author: Liviu Dudau <Liviu.Dudau@arm.com>
+ *
+ * ARM Mali DP500/DP550/DP650 registers definition.
+ */
+
+#ifndef __MALIDP_REGS_H__
+#define __MALIDP_REGS_H__
+
+/*
+ * abbreviations used:
+ *    - DC - display core (general settings)
+ *    - DE - display engine
+ *    - SE - scaling engine
+ */
+
+/* interrupt bit masks */
+#define MALIDP_DE_IRQ_UNDERRUN			(1 << 0)
+
+#define MALIDP500_DE_IRQ_AXI_ERR		(1 << 4)
+#define MALIDP500_DE_IRQ_VSYNC			(1 << 5)
+#define MALIDP500_DE_IRQ_PROG_LINE		(1 << 6)
+#define MALIDP500_DE_IRQ_SATURATION		(1 << 7)
+#define MALIDP500_DE_IRQ_CONF_VALID		(1 << 8)
+#define MALIDP500_DE_IRQ_CONF_MODE		(1 << 11)
+#define MALIDP500_DE_IRQ_CONF_ACTIVE		(1 << 17)
+#define MALIDP500_DE_IRQ_PM_ACTIVE		(1 << 18)
+#define MALIDP500_DE_IRQ_TESTMODE_ACTIVE	(1 << 19)
+#define MALIDP500_DE_IRQ_FORCE_BLNK_ACTIVE	(1 << 24)
+#define MALIDP500_DE_IRQ_AXI_BUSY		(1 << 28)
+#define MALIDP500_DE_IRQ_GLOBAL			(1 << 31)
+#define MALIDP500_SE_IRQ_CONF_MODE		(1 << 0)
+#define MALIDP500_SE_IRQ_CONF_VALID		(1 << 4)
+#define MALIDP500_SE_IRQ_INIT_BUSY		(1 << 5)
+#define MALIDP500_SE_IRQ_AXI_ERROR		(1 << 8)
+#define MALIDP500_SE_IRQ_OVERRUN		(1 << 9)
+#define MALIDP500_SE_IRQ_PROG_LINE1		(1 << 12)
+#define MALIDP500_SE_IRQ_PROG_LINE2		(1 << 13)
+#define MALIDP500_SE_IRQ_CONF_ACTIVE		(1 << 17)
+#define MALIDP500_SE_IRQ_PM_ACTIVE		(1 << 18)
+#define MALIDP500_SE_IRQ_AXI_BUSY		(1 << 28)
+#define MALIDP500_SE_IRQ_GLOBAL			(1 << 31)
+
+#define MALIDP550_DE_IRQ_SATURATION		(1 << 8)
+#define MALIDP550_DE_IRQ_VSYNC			(1 << 12)
+#define MALIDP550_DE_IRQ_PROG_LINE		(1 << 13)
+#define MALIDP550_DE_IRQ_AXI_ERR		(1 << 16)
+#define MALIDP550_SE_IRQ_EOW			(1 << 0)
+#define MALIDP550_SE_IRQ_AXI_ERR		(1 << 16)
+#define MALIDP550_SE_IRQ_OVR			(1 << 17)
+#define MALIDP550_SE_IRQ_IBSY			(1 << 18)
+#define MALIDP550_DC_IRQ_CONF_VALID		(1 << 0)
+#define MALIDP550_DC_IRQ_CONF_MODE		(1 << 4)
+#define MALIDP550_DC_IRQ_CONF_ACTIVE		(1 << 16)
+#define MALIDP550_DC_IRQ_DE			(1 << 20)
+#define MALIDP550_DC_IRQ_SE			(1 << 24)
+
+#define MALIDP650_DE_IRQ_DRIFT			(1 << 4)
+#define MALIDP650_DE_IRQ_ACEV1			(1 << 17)
+#define MALIDP650_DE_IRQ_ACEV2			(1 << 18)
+#define MALIDP650_DE_IRQ_ACEG			(1 << 19)
+#define MALIDP650_DE_IRQ_AXIEP			(1 << 28)
+
+/* bit masks that are common between products */
+#define   MALIDP_CFG_VALID		(1 << 0)
+#define   MALIDP_DISP_FUNC_GAMMA	(1 << 0)
+#define   MALIDP_DISP_FUNC_CADJ		(1 << 4)
+#define   MALIDP_DISP_FUNC_ILACED	(1 << 8)
+#define   MALIDP_SCALE_ENGINE_EN	(1 << 16)
+#define   MALIDP_SE_MEMWRITE_EN		(2 << 5)
+
+/* register offsets for IRQ management */
+#define MALIDP_REG_STATUS		0x00000
+#define MALIDP_REG_SETIRQ		0x00004
+#define MALIDP_REG_MASKIRQ		0x00008
+#define MALIDP_REG_CLEARIRQ		0x0000c
+
+/* register offsets */
+#define MALIDP_DE_CORE_ID		0x00018
+#define MALIDP_DE_DISPLAY_FUNC		0x00020
+
+/* these offsets are relative to MALIDP5x0_TIMINGS_BASE */
+#define MALIDP_DE_H_TIMINGS		0x0
+#define MALIDP_DE_V_TIMINGS		0x4
+#define MALIDP_DE_SYNC_WIDTH		0x8
+#define MALIDP_DE_HV_ACTIVE		0xc
+
+/* Stride register offsets relative to Lx_BASE */
+#define MALIDP_DE_LG_STRIDE		0x18
+#define MALIDP_DE_LV_STRIDE0		0x18
+#define MALIDP550_DE_LS_R1_STRIDE	0x28
+
+/* macros to set values into registers */
+#define MALIDP_DE_H_FRONTPORCH(x)	(((x) & 0xfff) << 0)
+#define MALIDP_DE_H_BACKPORCH(x)	(((x) & 0x3ff) << 16)
+#define MALIDP500_DE_V_FRONTPORCH(x)	(((x) & 0xff) << 0)
+#define MALIDP550_DE_V_FRONTPORCH(x)	(((x) & 0xfff) << 0)
+#define MALIDP_DE_V_BACKPORCH(x)	(((x) & 0xff) << 16)
+#define MALIDP_DE_H_SYNCWIDTH(x)	(((x) & 0x3ff) << 0)
+#define MALIDP_DE_V_SYNCWIDTH(x)	(((x) & 0xff) << 16)
+#define MALIDP_DE_H_ACTIVE(x)		(((x) & 0x1fff) << 0)
+#define MALIDP_DE_V_ACTIVE(x)		(((x) & 0x1fff) << 16)
+
+#define MALIDP_PRODUCT_ID(__core_id) ((u32)(__core_id) >> 16)
+
+/* register offsets relative to MALIDP5x0_COEFFS_BASE */
+#define MALIDP_COLOR_ADJ_COEF		0x00000
+#define MALIDP_COEF_TABLE_ADDR		0x00030
+#define MALIDP_COEF_TABLE_DATA		0x00034
+
+/* Scaling engine registers and masks. */
+#define   MALIDP_SE_SCALING_EN			(1 << 0)
+#define   MALIDP_SE_ALPHA_EN			(1 << 1)
+#define   MALIDP_SE_ENH_MASK			3
+#define   MALIDP_SE_ENH(x)			(((x) & MALIDP_SE_ENH_MASK) << 2)
+#define   MALIDP_SE_RGBO_IF_EN			(1 << 4)
+#define   MALIDP550_SE_CTL_SEL_MASK		7
+#define   MALIDP550_SE_CTL_VCSEL(x) \
+		(((x) & MALIDP550_SE_CTL_SEL_MASK) << 20)
+#define   MALIDP550_SE_CTL_HCSEL(x) \
+		(((x) & MALIDP550_SE_CTL_SEL_MASK) << 16)
+
+/* Blocks with offsets from SE_CONTROL register. */
+#define MALIDP_SE_LAYER_CONTROL			0x14
+#define   MALIDP_SE_L0_IN_SIZE			0x00
+#define   MALIDP_SE_L0_OUT_SIZE			0x04
+#define   MALIDP_SE_SET_V_SIZE(x)		(((x) & 0x1fff) << 16)
+#define   MALIDP_SE_SET_H_SIZE(x)		(((x) & 0x1fff) << 0)
+#define MALIDP_SE_SCALING_CONTROL		0x24
+#define   MALIDP_SE_H_INIT_PH			0x00
+#define   MALIDP_SE_H_DELTA_PH			0x04
+#define   MALIDP_SE_V_INIT_PH			0x08
+#define   MALIDP_SE_V_DELTA_PH			0x0c
+#define   MALIDP_SE_COEFFTAB_ADDR		0x10
+#define     MALIDP_SE_COEFFTAB_ADDR_MASK	0x7f
+#define     MALIDP_SE_V_COEFFTAB		(1 << 8)
+#define     MALIDP_SE_H_COEFFTAB		(1 << 9)
+#define     MALIDP_SE_SET_V_COEFFTAB_ADDR(x) \
+		(MALIDP_SE_V_COEFFTAB | ((x) & MALIDP_SE_COEFFTAB_ADDR_MASK))
+#define     MALIDP_SE_SET_H_COEFFTAB_ADDR(x) \
+		(MALIDP_SE_H_COEFFTAB | ((x) & MALIDP_SE_COEFFTAB_ADDR_MASK))
+#define   MALIDP_SE_COEFFTAB_DATA		0x14
+#define     MALIDP_SE_COEFFTAB_DATA_MASK	0x3fff
+#define     MALIDP_SE_SET_COEFFTAB_DATA(x) \
+		((x) & MALIDP_SE_COEFFTAB_DATA_MASK)
+/* Enhance coeffents reigster offset */
+#define MALIDP_SE_IMAGE_ENH			0x3C
+/* ENH_LIMITS offset 0x0 */
+#define     MALIDP_SE_ENH_LOW_LEVEL		24
+#define     MALIDP_SE_ENH_HIGH_LEVEL		63
+#define     MALIDP_SE_ENH_LIMIT_MASK		0xfff
+#define     MALIDP_SE_SET_ENH_LIMIT_LOW(x) \
+		((x) & MALIDP_SE_ENH_LIMIT_MASK)
+#define     MALIDP_SE_SET_ENH_LIMIT_HIGH(x) \
+		(((x) & MALIDP_SE_ENH_LIMIT_MASK) << 16)
+#define   MALIDP_SE_ENH_COEFF0			0x04
+
+
+/* register offsets relative to MALIDP5x0_SE_MEMWRITE_BASE */
+#define MALIDP_MW_FORMAT		0x00000
+#define MALIDP_MW_P1_STRIDE		0x00004
+#define MALIDP_MW_P2_STRIDE		0x00008
+#define MALIDP_MW_P1_PTR_LOW		0x0000c
+#define MALIDP_MW_P1_PTR_HIGH		0x00010
+#define MALIDP_MW_P2_PTR_LOW		0x0002c
+#define MALIDP_MW_P2_PTR_HIGH		0x00030
+
+/* register offsets and bits specific to DP500 */
+#define MALIDP500_ADDR_SPACE_SIZE	0x01000
+#define MALIDP500_DC_BASE		0x00000
+#define MALIDP500_DC_CONTROL		0x0000c
+#define   MALIDP500_DC_CONFIG_REQ	(1 << 17)
+#define   MALIDP500_HSYNCPOL		(1 << 20)
+#define   MALIDP500_VSYNCPOL		(1 << 21)
+#define   MALIDP500_DC_CLEAR_MASK	0x300fff
+#define MALIDP500_DE_LINE_COUNTER	0x00010
+#define MALIDP500_DE_AXI_CONTROL	0x00014
+#define MALIDP500_DE_SECURE_CTRL	0x0001c
+#define MALIDP500_DE_CHROMA_KEY		0x00024
+#define MALIDP500_TIMINGS_BASE		0x00028
+
+#define MALIDP500_CONFIG_3D		0x00038
+#define MALIDP500_BGND_COLOR		0x0003c
+#define MALIDP500_OUTPUT_DEPTH		0x00044
+#define MALIDP500_COEFFS_BASE		0x00078
+
+/*
+ * The YUV2RGB coefficients on the DP500 are not in the video layer's register
+ * block. They belong in a separate block above the layer's registers, hence
+ * the negative offset.
+ */
+#define MALIDP500_LV_YUV2RGB		((s16)(-0xB8))
+#define MALIDP500_DE_LV_BASE		0x00100
+#define MALIDP500_DE_LV_PTR_BASE	0x00124
+#define MALIDP500_DE_LV_AD_CTRL		0x00400
+#define MALIDP500_DE_LG1_BASE		0x00200
+#define MALIDP500_DE_LG1_PTR_BASE	0x0021c
+#define MALIDP500_DE_LG1_AD_CTRL	0x0040c
+#define MALIDP500_DE_LG2_BASE		0x00300
+#define MALIDP500_DE_LG2_PTR_BASE	0x0031c
+#define MALIDP500_DE_LG2_AD_CTRL	0x00418
+#define MALIDP500_SE_BASE		0x00c00
+#define MALIDP500_SE_CONTROL		0x00c0c
+#define MALIDP500_SE_MEMWRITE_OUT_SIZE	0x00c2c
+#define MALIDP500_SE_RGB_YUV_COEFFS	0x00C74
+#define MALIDP500_SE_MEMWRITE_BASE	0x00e00
+#define MALIDP500_DC_IRQ_BASE		0x00f00
+#define MALIDP500_CONFIG_VALID		0x00f00
+#define MALIDP500_CONFIG_ID		0x00fd4
+
+/*
+ * The quality of service (QoS) register on the DP500. RQOS register values
+ * are driven by the ARQOS signal, using AXI transacations, dependent on the
+ * FIFO input level.
+ * The RQOS register can also set QoS levels for:
+ *    - RED_ARQOS   @ A 4-bit signal value for close to underflow conditions
+ *    - GREEN_ARQOS @ A 4-bit signal value for normal conditions
+ */
+#define MALIDP500_RQOS_QUALITY          0x00500
+
+/* register offsets and bits specific to DP550/DP650 */
+#define MALIDP550_ADDR_SPACE_SIZE	0x10000
+#define MALIDP550_DE_CONTROL		0x00010
+#define MALIDP550_DE_LINE_COUNTER	0x00014
+#define MALIDP550_DE_AXI_CONTROL	0x00018
+#define MALIDP550_DE_QOS		0x0001c
+#define MALIDP550_TIMINGS_BASE		0x00030
+#define MALIDP550_HSYNCPOL		(1 << 12)
+#define MALIDP550_VSYNCPOL		(1 << 28)
+
+#define MALIDP550_DE_DISP_SIDEBAND	0x00040
+#define MALIDP550_DE_BGND_COLOR		0x00044
+#define MALIDP550_DE_OUTPUT_DEPTH	0x0004c
+#define MALIDP550_COEFFS_BASE		0x00050
+#define MALIDP550_LV_YUV2RGB		0x00084
+#define MALIDP550_DE_LV1_BASE		0x00100
+#define MALIDP550_DE_LV1_PTR_BASE	0x00124
+#define MALIDP550_DE_LV1_AD_CTRL	0x001B8
+#define MALIDP550_DE_LV2_BASE		0x00200
+#define MALIDP550_DE_LV2_PTR_BASE	0x00224
+#define MALIDP550_DE_LV2_AD_CTRL	0x002B8
+#define MALIDP550_DE_LG_BASE		0x00300
+#define MALIDP550_DE_LG_PTR_BASE	0x0031c
+#define MALIDP550_DE_LG_AD_CTRL		0x00330
+#define MALIDP550_DE_LS_BASE		0x00400
+#define MALIDP550_DE_LS_PTR_BASE	0x0042c
+#define MALIDP550_DE_PERF_BASE		0x00500
+#define MALIDP550_SE_BASE		0x08000
+#define MALIDP550_SE_CONTROL		0x08010
+#define   MALIDP550_SE_MEMWRITE_ONESHOT	(1 << 7)
+#define MALIDP550_SE_MEMWRITE_OUT_SIZE	0x08030
+#define MALIDP550_SE_RGB_YUV_COEFFS	0x08078
+#define MALIDP550_SE_MEMWRITE_BASE	0x08100
+#define MALIDP550_DC_BASE		0x0c000
+#define MALIDP550_DC_CONTROL		0x0c010
+#define   MALIDP550_DC_CONFIG_REQ	(1 << 16)
+#define MALIDP550_CONFIG_VALID		0x0c014
+#define MALIDP550_CONFIG_ID		0x0ffd4
+
+/* register offsets specific to DP650 */
+#define MALIDP650_DE_LV_MMU_CTRL	0x000D0
+#define MALIDP650_DE_LG_MMU_CTRL	0x00048
+#define MALIDP650_DE_LS_MMU_CTRL	0x00078
+
+/* bit masks to set the MMU control register */
+#define MALIDP_MMU_CTRL_EN		(1 << 0)
+#define MALIDP_MMU_CTRL_MODE		(1 << 4)
+#define MALIDP_MMU_CTRL_PX_PS(x)	(1 << (8 + (x)))
+#define MALIDP_MMU_CTRL_PP_NUM_REQ(x)	(((x) & 0x7f) << 12)
+
+/* AFBC register offsets relative to MALIDPXXX_DE_LX_AD_CTRL */
+/* The following register offsets are common for DP500, DP550 and DP650 */
+#define MALIDP_AD_CROP_H                0x4
+#define MALIDP_AD_CROP_V                0x8
+#define MALIDP_AD_END_PTR_LOW           0xc
+#define MALIDP_AD_END_PTR_HIGH          0x10
+
+/* AFBC decoder Registers */
+#define MALIDP_AD_EN                    BIT(0)
+#define MALIDP_AD_YTR                   BIT(4)
+#define MALIDP_AD_BS                    BIT(8)
+#define MALIDP_AD_CROP_RIGHT_OFFSET     16
+#define MALIDP_AD_CROP_BOTTOM_OFFSET    16
+
+/*
+ * Starting with DP550 the register map blocks has been standardised to the
+ * following layout:
+ *
+ *   Offset            Block registers
+ *  0x00000            Display Engine
+ *  0x08000            Scaling Engine
+ *  0x0c000            Display Core
+ *  0x10000            Secure control
+ *
+ * The old DP500 IP mixes some DC with the DE registers, hence the need
+ * for a mapping structure.
+ */
+
+#endif /* __MALIDP_REGS_H__ */