Adding upstream version 6.1.76.upstream/6.1.76upstream

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

14 files changed, 9893 insertions, 0 deletions

diff --git a/drivers/media/platform/chips-media/Kconfig b/drivers/media/platform/chips-media/Kconfig
new file mode 100644
index 000000000..57f8f8a22
--- /dev/null
+++ b/drivers/media/platform/chips-media/Kconfig
@@ -0,0 +1,20 @@
+# SPDX-License-Identifier: GPL-2.0-only
+
+comment "Chips&Media media platform drivers"
+
+config VIDEO_CODA
+	tristate "Chips&Media Coda multi-standard codec IP"
+	depends on V4L_MEM2MEM_DRIVERS
+	depends on VIDEO_DEV && OF && (ARCH_MXC || COMPILE_TEST)
+	select SRAM
+	select VIDEOBUF2_DMA_CONTIG
+	select VIDEOBUF2_VMALLOC
+	select V4L2_JPEG_HELPER
+	select V4L2_MEM2MEM_DEV
+	select GENERIC_ALLOCATOR
+	help
+	   Coda is a range of video codec IPs that supports
+	   H.264, MPEG-4, and other video formats.
+
+config VIDEO_IMX_VDOA
+	def_tristate VIDEO_CODA if SOC_IMX6Q || COMPILE_TEST
diff --git a/drivers/media/platform/chips-media/Makefile b/drivers/media/platform/chips-media/Makefile
new file mode 100644
index 000000000..bbb16425a
--- /dev/null
+++ b/drivers/media/platform/chips-media/Makefile
@@ -0,0 +1,6 @@
+# SPDX-License-Identifier: GPL-2.0-only
+
+coda-vpu-objs := coda-common.o coda-bit.o coda-gdi.o coda-h264.o coda-mpeg2.o coda-mpeg4.o coda-jpeg.o
+
+obj-$(CONFIG_VIDEO_CODA) += coda-vpu.o
+obj-$(CONFIG_VIDEO_IMX_VDOA) += imx-vdoa.o
diff --git a/drivers/media/platform/chips-media/coda-bit.c b/drivers/media/platform/chips-media/coda-bit.c
new file mode 100644
index 000000000..ed47d5bd8
--- /dev/null
+++ b/drivers/media/platform/chips-media/coda-bit.c
@@ -0,0 +1,2666 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Coda multi-standard codec IP - BIT processor functions
+ *
+ * Copyright (C) 2012 Vista Silicon S.L.
+ *    Javier Martin, <javier.martin@vista-silicon.com>
+ *    Xavier Duret
+ * Copyright (C) 2012-2014 Philipp Zabel, Pengutronix
+ */
+
+#include <linux/clk.h>
+#include <linux/irqreturn.h>
+#include <linux/kernel.h>
+#include <linux/log2.h>
+#include <linux/platform_device.h>
+#include <linux/ratelimit.h>
+#include <linux/reset.h>
+#include <linux/slab.h>
+#include <linux/videodev2.h>
+
+#include <media/v4l2-common.h>
+#include <media/v4l2-ctrls.h>
+#include <media/v4l2-fh.h>
+#include <media/v4l2-mem2mem.h>
+#include <media/videobuf2-v4l2.h>
+#include <media/videobuf2-dma-contig.h>
+#include <media/videobuf2-vmalloc.h>
+
+#include "coda.h"
+#include "imx-vdoa.h"
+#define CREATE_TRACE_POINTS
+#include "trace.h"
+
+#define CODA_PARA_BUF_SIZE	(10 * 1024)
+#define CODA7_PS_BUF_SIZE	0x28000
+#define CODA9_PS_SAVE_SIZE	(512 * 1024)
+
+#define CODA_DEFAULT_GAMMA	4096
+#define CODA9_DEFAULT_GAMMA	24576	/* 0.75 * 32768 */
+
+static void coda_free_bitstream_buffer(struct coda_ctx *ctx);
+
+static inline int coda_is_initialized(struct coda_dev *dev)
+{
+	return coda_read(dev, CODA_REG_BIT_CUR_PC) != 0;
+}
+
+static inline unsigned long coda_isbusy(struct coda_dev *dev)
+{
+	return coda_read(dev, CODA_REG_BIT_BUSY);
+}
+
+static int coda_wait_timeout(struct coda_dev *dev)
+{
+	unsigned long timeout = jiffies + msecs_to_jiffies(1000);
+
+	while (coda_isbusy(dev)) {
+		if (time_after(jiffies, timeout))
+			return -ETIMEDOUT;
+	}
+	return 0;
+}
+
+static void coda_command_async(struct coda_ctx *ctx, int cmd)
+{
+	struct coda_dev *dev = ctx->dev;
+
+	if (dev->devtype->product == CODA_HX4 ||
+	    dev->devtype->product == CODA_7541 ||
+	    dev->devtype->product == CODA_960) {
+		/* Restore context related registers to CODA */
+		coda_write(dev, ctx->bit_stream_param,
+				CODA_REG_BIT_BIT_STREAM_PARAM);
+		coda_write(dev, ctx->frm_dis_flg,
+				CODA_REG_BIT_FRM_DIS_FLG(ctx->reg_idx));
+		coda_write(dev, ctx->frame_mem_ctrl,
+				CODA_REG_BIT_FRAME_MEM_CTRL);
+		coda_write(dev, ctx->workbuf.paddr, CODA_REG_BIT_WORK_BUF_ADDR);
+	}
+
+	if (dev->devtype->product == CODA_960) {
+		coda_write(dev, 1, CODA9_GDI_WPROT_ERR_CLR);
+		coda_write(dev, 0, CODA9_GDI_WPROT_RGN_EN);
+	}
+
+	coda_write(dev, CODA_REG_BIT_BUSY_FLAG, CODA_REG_BIT_BUSY);
+
+	coda_write(dev, ctx->idx, CODA_REG_BIT_RUN_INDEX);
+	coda_write(dev, ctx->params.codec_mode, CODA_REG_BIT_RUN_COD_STD);
+	coda_write(dev, ctx->params.codec_mode_aux, CODA7_REG_BIT_RUN_AUX_STD);
+
+	trace_coda_bit_run(ctx, cmd);
+
+	coda_write(dev, cmd, CODA_REG_BIT_RUN_COMMAND);
+}
+
+static int coda_command_sync(struct coda_ctx *ctx, int cmd)
+{
+	struct coda_dev *dev = ctx->dev;
+	int ret;
+
+	lockdep_assert_held(&dev->coda_mutex);
+
+	coda_command_async(ctx, cmd);
+	ret = coda_wait_timeout(dev);
+	trace_coda_bit_done(ctx);
+
+	return ret;
+}
+
+int coda_hw_reset(struct coda_ctx *ctx)
+{
+	struct coda_dev *dev = ctx->dev;
+	unsigned long timeout;
+	unsigned int idx;
+	int ret;
+
+	lockdep_assert_held(&dev->coda_mutex);
+
+	if (!dev->rstc)
+		return -ENOENT;
+
+	idx = coda_read(dev, CODA_REG_BIT_RUN_INDEX);
+
+	if (dev->devtype->product == CODA_960) {
+		timeout = jiffies + msecs_to_jiffies(100);
+		coda_write(dev, 0x11, CODA9_GDI_BUS_CTRL);
+		while (coda_read(dev, CODA9_GDI_BUS_STATUS) != 0x77) {
+			if (time_after(jiffies, timeout))
+				return -ETIME;
+			cpu_relax();
+		}
+	}
+
+	ret = reset_control_reset(dev->rstc);
+	if (ret < 0)
+		return ret;
+
+	if (dev->devtype->product == CODA_960)
+		coda_write(dev, 0x00, CODA9_GDI_BUS_CTRL);
+	coda_write(dev, CODA_REG_BIT_BUSY_FLAG, CODA_REG_BIT_BUSY);
+	coda_write(dev, CODA_REG_RUN_ENABLE, CODA_REG_BIT_CODE_RUN);
+	ret = coda_wait_timeout(dev);
+	coda_write(dev, idx, CODA_REG_BIT_RUN_INDEX);
+
+	return ret;
+}
+
+static void coda_kfifo_sync_from_device(struct coda_ctx *ctx)
+{
+	struct __kfifo *kfifo = &ctx->bitstream_fifo.kfifo;
+	struct coda_dev *dev = ctx->dev;
+	u32 rd_ptr;
+
+	rd_ptr = coda_read(dev, CODA_REG_BIT_RD_PTR(ctx->reg_idx));
+	kfifo->out = (kfifo->in & ~kfifo->mask) |
+		      (rd_ptr - ctx->bitstream.paddr);
+	if (kfifo->out > kfifo->in)
+		kfifo->out -= kfifo->mask + 1;
+}
+
+static void coda_kfifo_sync_to_device_full(struct coda_ctx *ctx)
+{
+	struct __kfifo *kfifo = &ctx->bitstream_fifo.kfifo;
+	struct coda_dev *dev = ctx->dev;
+	u32 rd_ptr, wr_ptr;
+
+	rd_ptr = ctx->bitstream.paddr + (kfifo->out & kfifo->mask);
+	coda_write(dev, rd_ptr, CODA_REG_BIT_RD_PTR(ctx->reg_idx));
+	wr_ptr = ctx->bitstream.paddr + (kfifo->in & kfifo->mask);
+	coda_write(dev, wr_ptr, CODA_REG_BIT_WR_PTR(ctx->reg_idx));
+}
+
+static void coda_kfifo_sync_to_device_write(struct coda_ctx *ctx)
+{
+	struct __kfifo *kfifo = &ctx->bitstream_fifo.kfifo;
+	struct coda_dev *dev = ctx->dev;
+	u32 wr_ptr;
+
+	wr_ptr = ctx->bitstream.paddr + (kfifo->in & kfifo->mask);
+	coda_write(dev, wr_ptr, CODA_REG_BIT_WR_PTR(ctx->reg_idx));
+}
+
+static int coda_h264_bitstream_pad(struct coda_ctx *ctx, u32 size)
+{
+	unsigned char *buf;
+	u32 n;
+
+	if (size < 6)
+		size = 6;
+
+	buf = kmalloc(size, GFP_KERNEL);
+	if (!buf)
+		return -ENOMEM;
+
+	coda_h264_filler_nal(size, buf);
+	n = kfifo_in(&ctx->bitstream_fifo, buf, size);
+	kfree(buf);
+
+	return (n < size) ? -ENOSPC : 0;
+}
+
+int coda_bitstream_flush(struct coda_ctx *ctx)
+{
+	int ret;
+
+	if (ctx->inst_type != CODA_INST_DECODER || !ctx->use_bit)
+		return 0;
+
+	ret = coda_command_sync(ctx, CODA_COMMAND_DEC_BUF_FLUSH);
+	if (ret < 0) {
+		v4l2_err(&ctx->dev->v4l2_dev, "failed to flush bitstream\n");
+		return ret;
+	}
+
+	kfifo_init(&ctx->bitstream_fifo, ctx->bitstream.vaddr,
+		   ctx->bitstream.size);
+	coda_kfifo_sync_to_device_full(ctx);
+
+	return 0;
+}
+
+static int coda_bitstream_queue(struct coda_ctx *ctx, const u8 *buf, u32 size)
+{
+	u32 n = kfifo_in(&ctx->bitstream_fifo, buf, size);
+
+	return (n < size) ? -ENOSPC : 0;
+}
+
+static u32 coda_buffer_parse_headers(struct coda_ctx *ctx,
+				     struct vb2_v4l2_buffer *src_buf,
+				     u32 payload)
+{
+	u8 *vaddr = vb2_plane_vaddr(&src_buf->vb2_buf, 0);
+	u32 size = 0;
+
+	switch (ctx->codec->src_fourcc) {
+	case V4L2_PIX_FMT_MPEG2:
+		size = coda_mpeg2_parse_headers(ctx, vaddr, payload);
+		break;
+	case V4L2_PIX_FMT_MPEG4:
+		size = coda_mpeg4_parse_headers(ctx, vaddr, payload);
+		break;
+	default:
+		break;
+	}
+
+	return size;
+}
+
+static bool coda_bitstream_try_queue(struct coda_ctx *ctx,
+				     struct vb2_v4l2_buffer *src_buf)
+{
+	unsigned long payload = vb2_get_plane_payload(&src_buf->vb2_buf, 0);
+	u8 *vaddr = vb2_plane_vaddr(&src_buf->vb2_buf, 0);
+	int ret;
+	int i;
+
+	if (coda_get_bitstream_payload(ctx) + payload + 512 >=
+	    ctx->bitstream.size)
+		return false;
+
+	if (!vaddr) {
+		v4l2_err(&ctx->dev->v4l2_dev, "trying to queue empty buffer\n");
+		return true;
+	}
+
+	if (ctx->qsequence == 0 && payload < 512) {
+		/*
+		 * Add padding after the first buffer, if it is too small to be
+		 * fetched by the CODA, by repeating the headers. Without
+		 * repeated headers, or the first frame already queued, decoder
+		 * sequence initialization fails with error code 0x2000 on i.MX6
+		 * or error code 0x1 on i.MX51.
+		 */
+		u32 header_size = coda_buffer_parse_headers(ctx, src_buf,
+							    payload);
+
+		if (header_size) {
+			coda_dbg(1, ctx, "pad with %u-byte header\n",
+				 header_size);
+			for (i = payload; i < 512; i += header_size) {
+				ret = coda_bitstream_queue(ctx, vaddr,
+							   header_size);
+				if (ret < 0) {
+					v4l2_err(&ctx->dev->v4l2_dev,
+						 "bitstream buffer overflow\n");
+					return false;
+				}
+				if (ctx->dev->devtype->product == CODA_960)
+					break;
+			}
+		} else {
+			coda_dbg(1, ctx,
+				 "could not parse header, sequence initialization might fail\n");
+		}
+
+		/* Add padding before the first buffer, if it is too small */
+		if (ctx->codec->src_fourcc == V4L2_PIX_FMT_H264)
+			coda_h264_bitstream_pad(ctx, 512 - payload);
+	}
+
+	ret = coda_bitstream_queue(ctx, vaddr, payload);
+	if (ret < 0) {
+		v4l2_err(&ctx->dev->v4l2_dev, "bitstream buffer overflow\n");
+		return false;
+	}
+
+	src_buf->sequence = ctx->qsequence++;
+
+	/* Sync read pointer to device */
+	if (ctx == v4l2_m2m_get_curr_priv(ctx->dev->m2m_dev))
+		coda_kfifo_sync_to_device_write(ctx);
+
+	/* Set the stream-end flag after the last buffer is queued */
+	if (src_buf->flags & V4L2_BUF_FLAG_LAST)
+		coda_bit_stream_end_flag(ctx);
+	ctx->hold = false;
+
+	return true;
+}
+
+void coda_fill_bitstream(struct coda_ctx *ctx, struct list_head *buffer_list)
+{
+	struct vb2_v4l2_buffer *src_buf;
+	struct coda_buffer_meta *meta;
+	u32 start;
+
+	lockdep_assert_held(&ctx->bitstream_mutex);
+
+	if (ctx->bit_stream_param & CODA_BIT_STREAM_END_FLAG)
+		return;
+
+	while (v4l2_m2m_num_src_bufs_ready(ctx->fh.m2m_ctx) > 0) {
+		/*
+		 * Only queue two JPEGs into the bitstream buffer to keep
+		 * latency low. We need at least one complete buffer and the
+		 * header of another buffer (for prescan) in the bitstream.
+		 */
+		if (ctx->codec->src_fourcc == V4L2_PIX_FMT_JPEG &&
+		    ctx->num_metas > 1)
+			break;
+
+		if (ctx->num_internal_frames &&
+		    ctx->num_metas >= ctx->num_internal_frames) {
+			meta = list_first_entry(&ctx->buffer_meta_list,
+						struct coda_buffer_meta, list);
+
+			/*
+			 * If we managed to fill in at least a full reorder
+			 * window of buffers (num_internal_frames is a
+			 * conservative estimate for this) and the bitstream
+			 * prefetcher has at least 2 256 bytes periods beyond
+			 * the first buffer to fetch, we can safely stop queuing
+			 * in order to limit the decoder drain latency.
+			 */
+			if (coda_bitstream_can_fetch_past(ctx, meta->end))
+				break;
+		}
+
+		src_buf = v4l2_m2m_next_src_buf(ctx->fh.m2m_ctx);
+
+		/* Drop frames that do not start/end with a SOI/EOI markers */
+		if (ctx->codec->src_fourcc == V4L2_PIX_FMT_JPEG &&
+		    !coda_jpeg_check_buffer(ctx, &src_buf->vb2_buf)) {
+			v4l2_err(&ctx->dev->v4l2_dev,
+				 "dropping invalid JPEG frame %d\n",
+				 ctx->qsequence);
+			src_buf = v4l2_m2m_src_buf_remove(ctx->fh.m2m_ctx);
+			if (buffer_list) {
+				struct v4l2_m2m_buffer *m2m_buf;
+
+				m2m_buf = container_of(src_buf,
+						       struct v4l2_m2m_buffer,
+						       vb);
+				list_add_tail(&m2m_buf->list, buffer_list);
+			} else {
+				v4l2_m2m_buf_done(src_buf, VB2_BUF_STATE_ERROR);
+			}
+			continue;
+		}
+
+		/* Dump empty buffers */
+		if (!vb2_get_plane_payload(&src_buf->vb2_buf, 0)) {
+			src_buf = v4l2_m2m_src_buf_remove(ctx->fh.m2m_ctx);
+			v4l2_m2m_buf_done(src_buf, VB2_BUF_STATE_DONE);
+			continue;
+		}
+
+		/* Buffer start position */
+		start = ctx->bitstream_fifo.kfifo.in;
+
+		if (coda_bitstream_try_queue(ctx, src_buf)) {
+			/*
+			 * Source buffer is queued in the bitstream ringbuffer;
+			 * queue the timestamp and mark source buffer as done
+			 */
+			src_buf = v4l2_m2m_src_buf_remove(ctx->fh.m2m_ctx);
+
+			meta = kmalloc(sizeof(*meta), GFP_KERNEL);
+			if (meta) {
+				meta->sequence = src_buf->sequence;
+				meta->timecode = src_buf->timecode;
+				meta->timestamp = src_buf->vb2_buf.timestamp;
+				meta->start = start;
+				meta->end = ctx->bitstream_fifo.kfifo.in;
+				meta->last = src_buf->flags & V4L2_BUF_FLAG_LAST;
+				if (meta->last)
+					coda_dbg(1, ctx, "marking last meta");
+				spin_lock(&ctx->buffer_meta_lock);
+				list_add_tail(&meta->list,
+					      &ctx->buffer_meta_list);
+				ctx->num_metas++;
+				spin_unlock(&ctx->buffer_meta_lock);
+
+				trace_coda_bit_queue(ctx, src_buf, meta);
+			}
+
+			if (buffer_list) {
+				struct v4l2_m2m_buffer *m2m_buf;
+
+				m2m_buf = container_of(src_buf,
+						       struct v4l2_m2m_buffer,
+						       vb);
+				list_add_tail(&m2m_buf->list, buffer_list);
+			} else {
+				v4l2_m2m_buf_done(src_buf, VB2_BUF_STATE_DONE);
+			}
+		} else {
+			break;
+		}
+	}
+}
+
+void coda_bit_stream_end_flag(struct coda_ctx *ctx)
+{
+	struct coda_dev *dev = ctx->dev;
+
+	ctx->bit_stream_param |= CODA_BIT_STREAM_END_FLAG;
+
+	/* If this context is currently running, update the hardware flag */
+	if ((dev->devtype->product == CODA_960) &&
+	    coda_isbusy(dev) &&
+	    (ctx->idx == coda_read(dev, CODA_REG_BIT_RUN_INDEX))) {
+		coda_write(dev, ctx->bit_stream_param,
+			   CODA_REG_BIT_BIT_STREAM_PARAM);
+	}
+}
+
+static void coda_parabuf_write(struct coda_ctx *ctx, int index, u32 value)
+{
+	struct coda_dev *dev = ctx->dev;
+	u32 *p = ctx->parabuf.vaddr;
+
+	if (dev->devtype->product == CODA_DX6)
+		p[index] = value;
+	else
+		p[index ^ 1] = value;
+}
+
+static inline int coda_alloc_context_buf(struct coda_ctx *ctx,
+					 struct coda_aux_buf *buf, size_t size,
+					 const char *name)
+{
+	return coda_alloc_aux_buf(ctx->dev, buf, size, name, ctx->debugfs_entry);
+}
+
+
+static void coda_free_framebuffers(struct coda_ctx *ctx)
+{
+	int i;
+
+	for (i = 0; i < CODA_MAX_FRAMEBUFFERS; i++)
+		coda_free_aux_buf(ctx->dev, &ctx->internal_frames[i].buf);
+}
+
+static int coda_alloc_framebuffers(struct coda_ctx *ctx,
+				   struct coda_q_data *q_data, u32 fourcc)
+{
+	struct coda_dev *dev = ctx->dev;
+	unsigned int ysize, ycbcr_size;
+	int ret;
+	int i;
+
+	if (ctx->codec->src_fourcc == V4L2_PIX_FMT_H264 ||
+	    ctx->codec->dst_fourcc == V4L2_PIX_FMT_H264 ||
+	    ctx->codec->src_fourcc == V4L2_PIX_FMT_MPEG4 ||
+	    ctx->codec->dst_fourcc == V4L2_PIX_FMT_MPEG4)
+		ysize = round_up(q_data->rect.width, 16) *
+			round_up(q_data->rect.height, 16);
+	else
+		ysize = round_up(q_data->rect.width, 8) * q_data->rect.height;
+
+	if (ctx->tiled_map_type == GDI_TILED_FRAME_MB_RASTER_MAP)
+		ycbcr_size = round_up(ysize, 4096) + ysize / 2;
+	else
+		ycbcr_size = ysize + ysize / 2;
+
+	/* Allocate frame buffers */
+	for (i = 0; i < ctx->num_internal_frames; i++) {
+		size_t size = ycbcr_size;
+		char *name;
+
+		/* Add space for mvcol buffers */
+		if (dev->devtype->product != CODA_DX6 &&
+		    (ctx->codec->src_fourcc == V4L2_PIX_FMT_H264 ||
+		     (ctx->codec->src_fourcc == V4L2_PIX_FMT_MPEG4 && i == 0)))
+			size += ysize / 4;
+		name = kasprintf(GFP_KERNEL, "fb%d", i);
+		if (!name) {
+			coda_free_framebuffers(ctx);
+			return -ENOMEM;
+		}
+		ret = coda_alloc_context_buf(ctx, &ctx->internal_frames[i].buf,
+					     size, name);
+		kfree(name);
+		if (ret < 0) {
+			coda_free_framebuffers(ctx);
+			return ret;
+		}
+	}
+
+	/* Register frame buffers in the parameter buffer */
+	for (i = 0; i < ctx->num_internal_frames; i++) {
+		u32 y, cb, cr, mvcol;
+
+		/* Start addresses of Y, Cb, Cr planes */
+		y = ctx->internal_frames[i].buf.paddr;
+		cb = y + ysize;
+		cr = y + ysize + ysize/4;
+		mvcol = y + ysize + ysize/4 + ysize/4;
+		if (ctx->tiled_map_type == GDI_TILED_FRAME_MB_RASTER_MAP) {
+			cb = round_up(cb, 4096);
+			mvcol = cb + ysize/2;
+			cr = 0;
+			/* Packed 20-bit MSB of base addresses */
+			/* YYYYYCCC, CCyyyyyc, cccc.... */
+			y = (y & 0xfffff000) | cb >> 20;
+			cb = (cb & 0x000ff000) << 12;
+		}
+		coda_parabuf_write(ctx, i * 3 + 0, y);
+		coda_parabuf_write(ctx, i * 3 + 1, cb);
+		coda_parabuf_write(ctx, i * 3 + 2, cr);
+
+		if (dev->devtype->product == CODA_DX6)
+			continue;
+
+		/* mvcol buffer for h.264 and mpeg4 */
+		if (ctx->codec->src_fourcc == V4L2_PIX_FMT_H264)
+			coda_parabuf_write(ctx, 96 + i, mvcol);
+		if (ctx->codec->src_fourcc == V4L2_PIX_FMT_MPEG4 && i == 0)
+			coda_parabuf_write(ctx, 97, mvcol);
+	}
+
+	return 0;
+}
+
+static void coda_free_context_buffers(struct coda_ctx *ctx)
+{
+	struct coda_dev *dev = ctx->dev;
+
+	coda_free_aux_buf(dev, &ctx->slicebuf);
+	coda_free_aux_buf(dev, &ctx->psbuf);
+	if (dev->devtype->product != CODA_DX6)
+		coda_free_aux_buf(dev, &ctx->workbuf);
+	coda_free_aux_buf(dev, &ctx->parabuf);
+}
+
+static int coda_alloc_context_buffers(struct coda_ctx *ctx,
+				      struct coda_q_data *q_data)
+{
+	struct coda_dev *dev = ctx->dev;
+	size_t size;
+	int ret;
+
+	if (!ctx->parabuf.vaddr) {
+		ret = coda_alloc_context_buf(ctx, &ctx->parabuf,
+					     CODA_PARA_BUF_SIZE, "parabuf");
+		if (ret < 0)
+			return ret;
+	}
+
+	if (dev->devtype->product == CODA_DX6)
+		return 0;
+
+	if (!ctx->slicebuf.vaddr && q_data->fourcc == V4L2_PIX_FMT_H264) {
+		/* worst case slice size */
+		size = (DIV_ROUND_UP(q_data->rect.width, 16) *
+			DIV_ROUND_UP(q_data->rect.height, 16)) * 3200 / 8 + 512;
+		ret = coda_alloc_context_buf(ctx, &ctx->slicebuf, size,
+					     "slicebuf");
+		if (ret < 0)
+			goto err;
+	}
+
+	if (!ctx->psbuf.vaddr && (dev->devtype->product == CODA_HX4 ||
+				  dev->devtype->product == CODA_7541)) {
+		ret = coda_alloc_context_buf(ctx, &ctx->psbuf,
+					     CODA7_PS_BUF_SIZE, "psbuf");
+		if (ret < 0)
+			goto err;
+	}
+
+	if (!ctx->workbuf.vaddr) {
+		size = dev->devtype->workbuf_size;
+		if (dev->devtype->product == CODA_960 &&
+		    q_data->fourcc == V4L2_PIX_FMT_H264)
+			size += CODA9_PS_SAVE_SIZE;
+		ret = coda_alloc_context_buf(ctx, &ctx->workbuf, size,
+					     "workbuf");
+		if (ret < 0)
+			goto err;
+	}
+
+	return 0;
+
+err:
+	coda_free_context_buffers(ctx);
+	return ret;
+}
+
+static int coda_encode_header(struct coda_ctx *ctx, struct vb2_v4l2_buffer *buf,
+			      int header_code, u8 *header, int *size)
+{
+	struct vb2_buffer *vb = &buf->vb2_buf;
+	struct coda_dev *dev = ctx->dev;
+	struct coda_q_data *q_data_src;
+	struct v4l2_rect *r;
+	size_t bufsize;
+	int ret;
+	int i;
+
+	if (dev->devtype->product == CODA_960)
+		memset(vb2_plane_vaddr(vb, 0), 0, 64);
+
+	coda_write(dev, vb2_dma_contig_plane_dma_addr(vb, 0),
+		   CODA_CMD_ENC_HEADER_BB_START);
+	bufsize = vb2_plane_size(vb, 0);
+	if (dev->devtype->product == CODA_960)
+		bufsize /= 1024;
+	coda_write(dev, bufsize, CODA_CMD_ENC_HEADER_BB_SIZE);
+	if (dev->devtype->product == CODA_960 &&
+	    ctx->codec->dst_fourcc == V4L2_PIX_FMT_H264 &&
+	    header_code == CODA_HEADER_H264_SPS) {
+		q_data_src = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+		r = &q_data_src->rect;
+
+		if (r->width % 16 || r->height % 16) {
+			u32 crop_right = round_up(r->width, 16) -  r->width;
+			u32 crop_bottom = round_up(r->height, 16) - r->height;
+
+			coda_write(dev, crop_right,
+				   CODA9_CMD_ENC_HEADER_FRAME_CROP_H);
+			coda_write(dev, crop_bottom,
+				   CODA9_CMD_ENC_HEADER_FRAME_CROP_V);
+			header_code |= CODA9_HEADER_FRAME_CROP;
+		}
+	}
+	coda_write(dev, header_code, CODA_CMD_ENC_HEADER_CODE);
+	ret = coda_command_sync(ctx, CODA_COMMAND_ENCODE_HEADER);
+	if (ret < 0) {
+		v4l2_err(&dev->v4l2_dev, "CODA_COMMAND_ENCODE_HEADER timeout\n");
+		return ret;
+	}
+
+	if (dev->devtype->product == CODA_960) {
+		for (i = 63; i > 0; i--)
+			if (((char *)vb2_plane_vaddr(vb, 0))[i] != 0)
+				break;
+		*size = i + 1;
+	} else {
+		*size = coda_read(dev, CODA_REG_BIT_WR_PTR(ctx->reg_idx)) -
+			coda_read(dev, CODA_CMD_ENC_HEADER_BB_START);
+	}
+	memcpy(header, vb2_plane_vaddr(vb, 0), *size);
+
+	return 0;
+}
+
+static u32 coda_slice_mode(struct coda_ctx *ctx)
+{
+	int size, unit;
+
+	switch (ctx->params.slice_mode) {
+	case V4L2_MPEG_VIDEO_MULTI_SLICE_MODE_SINGLE:
+	default:
+		return 0;
+	case V4L2_MPEG_VIDEO_MULTI_SLICE_MODE_MAX_MB:
+		size = ctx->params.slice_max_mb;
+		unit = 1;
+		break;
+	case V4L2_MPEG_VIDEO_MULTI_SLICE_MODE_MAX_BYTES:
+		size = ctx->params.slice_max_bits;
+		unit = 0;
+		break;
+	}
+
+	return ((size & CODA_SLICING_SIZE_MASK) << CODA_SLICING_SIZE_OFFSET) |
+	       ((unit & CODA_SLICING_UNIT_MASK) << CODA_SLICING_UNIT_OFFSET) |
+	       ((1 & CODA_SLICING_MODE_MASK) << CODA_SLICING_MODE_OFFSET);
+}
+
+static int coda_enc_param_change(struct coda_ctx *ctx)
+{
+	struct coda_dev *dev = ctx->dev;
+	u32 change_enable = 0;
+	u32 success;
+	int ret;
+
+	if (ctx->params.gop_size_changed) {
+		change_enable |= CODA_PARAM_CHANGE_RC_GOP;
+		coda_write(dev, ctx->params.gop_size,
+			   CODA_CMD_ENC_PARAM_RC_GOP);
+		ctx->gopcounter = ctx->params.gop_size - 1;
+		ctx->params.gop_size_changed = false;
+	}
+	if (ctx->params.h264_intra_qp_changed) {
+		coda_dbg(1, ctx, "parameter change: intra Qp %u\n",
+			 ctx->params.h264_intra_qp);
+
+		if (ctx->params.bitrate) {
+			change_enable |= CODA_PARAM_CHANGE_RC_INTRA_QP;
+			coda_write(dev, ctx->params.h264_intra_qp,
+				   CODA_CMD_ENC_PARAM_RC_INTRA_QP);
+		}
+		ctx->params.h264_intra_qp_changed = false;
+	}
+	if (ctx->params.bitrate_changed) {
+		coda_dbg(1, ctx, "parameter change: bitrate %u kbit/s\n",
+			 ctx->params.bitrate);
+		change_enable |= CODA_PARAM_CHANGE_RC_BITRATE;
+		coda_write(dev, ctx->params.bitrate,
+			   CODA_CMD_ENC_PARAM_RC_BITRATE);
+		ctx->params.bitrate_changed = false;
+	}
+	if (ctx->params.framerate_changed) {
+		coda_dbg(1, ctx, "parameter change: frame rate %u/%u Hz\n",
+			 ctx->params.framerate & 0xffff,
+			 (ctx->params.framerate >> 16) + 1);
+		change_enable |= CODA_PARAM_CHANGE_RC_FRAME_RATE;
+		coda_write(dev, ctx->params.framerate,
+			   CODA_CMD_ENC_PARAM_RC_FRAME_RATE);
+		ctx->params.framerate_changed = false;
+	}
+	if (ctx->params.intra_refresh_changed) {
+		coda_dbg(1, ctx, "parameter change: intra refresh MBs %u\n",
+			 ctx->params.intra_refresh);
+		change_enable |= CODA_PARAM_CHANGE_INTRA_MB_NUM;
+		coda_write(dev, ctx->params.intra_refresh,
+			   CODA_CMD_ENC_PARAM_INTRA_MB_NUM);
+		ctx->params.intra_refresh_changed = false;
+	}
+	if (ctx->params.slice_mode_changed) {
+		change_enable |= CODA_PARAM_CHANGE_SLICE_MODE;
+		coda_write(dev, coda_slice_mode(ctx),
+			   CODA_CMD_ENC_PARAM_SLICE_MODE);
+		ctx->params.slice_mode_changed = false;
+	}
+
+	if (!change_enable)
+		return 0;
+
+	coda_write(dev, change_enable, CODA_CMD_ENC_PARAM_CHANGE_ENABLE);
+
+	ret = coda_command_sync(ctx, CODA_COMMAND_RC_CHANGE_PARAMETER);
+	if (ret < 0)
+		return ret;
+
+	success = coda_read(dev, CODA_RET_ENC_PARAM_CHANGE_SUCCESS);
+	if (success != 1)
+		coda_dbg(1, ctx, "parameter change failed: %u\n", success);
+
+	return 0;
+}
+
+static phys_addr_t coda_iram_alloc(struct coda_iram_info *iram, size_t size)
+{
+	phys_addr_t ret;
+
+	size = round_up(size, 1024);
+	if (size > iram->remaining)
+		return 0;
+	iram->remaining -= size;
+
+	ret = iram->next_paddr;
+	iram->next_paddr += size;
+
+	return ret;
+}
+
+static void coda_setup_iram(struct coda_ctx *ctx)
+{
+	struct coda_iram_info *iram_info = &ctx->iram_info;
+	struct coda_dev *dev = ctx->dev;
+	int w64, w128;
+	int mb_width;
+	int dbk_bits;
+	int bit_bits;
+	int ip_bits;
+	int me_bits;
+
+	memset(iram_info, 0, sizeof(*iram_info));
+	iram_info->next_paddr = dev->iram.paddr;
+	iram_info->remaining = dev->iram.size;
+
+	if (!dev->iram.vaddr)
+		return;
+
+	switch (dev->devtype->product) {
+	case CODA_HX4:
+		dbk_bits = CODA7_USE_HOST_DBK_ENABLE;
+		bit_bits = CODA7_USE_HOST_BIT_ENABLE;
+		ip_bits = CODA7_USE_HOST_IP_ENABLE;
+		me_bits = CODA7_USE_HOST_ME_ENABLE;
+		break;
+	case CODA_7541:
+		dbk_bits = CODA7_USE_HOST_DBK_ENABLE | CODA7_USE_DBK_ENABLE;
+		bit_bits = CODA7_USE_HOST_BIT_ENABLE | CODA7_USE_BIT_ENABLE;
+		ip_bits = CODA7_USE_HOST_IP_ENABLE | CODA7_USE_IP_ENABLE;
+		me_bits = CODA7_USE_HOST_ME_ENABLE | CODA7_USE_ME_ENABLE;
+		break;
+	case CODA_960:
+		dbk_bits = CODA9_USE_HOST_DBK_ENABLE | CODA9_USE_DBK_ENABLE;
+		bit_bits = CODA9_USE_HOST_BIT_ENABLE | CODA7_USE_BIT_ENABLE;
+		ip_bits = CODA9_USE_HOST_IP_ENABLE | CODA7_USE_IP_ENABLE;
+		me_bits = 0;
+		break;
+	default: /* CODA_DX6 */
+		return;
+	}
+
+	if (ctx->inst_type == CODA_INST_ENCODER) {
+		struct coda_q_data *q_data_src;
+
+		q_data_src = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+		mb_width = DIV_ROUND_UP(q_data_src->rect.width, 16);
+		w128 = mb_width * 128;
+		w64 = mb_width * 64;
+
+		/* Prioritize in case IRAM is too small for everything */
+		if (dev->devtype->product == CODA_HX4 ||
+		    dev->devtype->product == CODA_7541) {
+			iram_info->search_ram_size = round_up(mb_width * 16 *
+							      36 + 2048, 1024);
+			iram_info->search_ram_paddr = coda_iram_alloc(iram_info,
+						iram_info->search_ram_size);
+			if (!iram_info->search_ram_paddr) {
+				pr_err("IRAM is smaller than the search ram size\n");
+				goto out;
+			}
+			iram_info->axi_sram_use |= me_bits;
+		}
+
+		/* Only H.264BP and H.263P3 are considered */
+		iram_info->buf_dbk_y_use = coda_iram_alloc(iram_info, w64);
+		iram_info->buf_dbk_c_use = coda_iram_alloc(iram_info, w64);
+		if (!iram_info->buf_dbk_y_use || !iram_info->buf_dbk_c_use)
+			goto out;
+		iram_info->axi_sram_use |= dbk_bits;
+
+		iram_info->buf_bit_use = coda_iram_alloc(iram_info, w128);
+		if (!iram_info->buf_bit_use)
+			goto out;
+		iram_info->axi_sram_use |= bit_bits;
+
+		iram_info->buf_ip_ac_dc_use = coda_iram_alloc(iram_info, w128);
+		if (!iram_info->buf_ip_ac_dc_use)
+			goto out;
+		iram_info->axi_sram_use |= ip_bits;
+
+		/* OVL and BTP disabled for encoder */
+	} else if (ctx->inst_type == CODA_INST_DECODER) {
+		struct coda_q_data *q_data_dst;
+
+		q_data_dst = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_CAPTURE);
+		mb_width = DIV_ROUND_UP(q_data_dst->width, 16);
+		w128 = mb_width * 128;
+
+		iram_info->buf_dbk_y_use = coda_iram_alloc(iram_info, w128);
+		iram_info->buf_dbk_c_use = coda_iram_alloc(iram_info, w128);
+		if (!iram_info->buf_dbk_y_use || !iram_info->buf_dbk_c_use)
+			goto out;
+		iram_info->axi_sram_use |= dbk_bits;
+
+		iram_info->buf_bit_use = coda_iram_alloc(iram_info, w128);
+		if (!iram_info->buf_bit_use)
+			goto out;
+		iram_info->axi_sram_use |= bit_bits;
+
+		iram_info->buf_ip_ac_dc_use = coda_iram_alloc(iram_info, w128);
+		if (!iram_info->buf_ip_ac_dc_use)
+			goto out;
+		iram_info->axi_sram_use |= ip_bits;
+
+		/* OVL and BTP unused as there is no VC1 support yet */
+	}
+
+out:
+	if (!(iram_info->axi_sram_use & CODA7_USE_HOST_IP_ENABLE))
+		coda_dbg(1, ctx, "IRAM smaller than needed\n");
+
+	if (dev->devtype->product == CODA_HX4 ||
+	    dev->devtype->product == CODA_7541) {
+		/* TODO - Enabling these causes picture errors on CODA7541 */
+		if (ctx->inst_type == CODA_INST_DECODER) {
+			/* fw 1.4.50 */
+			iram_info->axi_sram_use &= ~(CODA7_USE_HOST_IP_ENABLE |
+						     CODA7_USE_IP_ENABLE);
+		} else {
+			/* fw 13.4.29 */
+			iram_info->axi_sram_use &= ~(CODA7_USE_HOST_IP_ENABLE |
+						     CODA7_USE_HOST_DBK_ENABLE |
+						     CODA7_USE_IP_ENABLE |
+						     CODA7_USE_DBK_ENABLE);
+		}
+	}
+}
+
+static u32 coda_supported_firmwares[] = {
+	CODA_FIRMWARE_VERNUM(CODA_DX6, 2, 2, 5),
+	CODA_FIRMWARE_VERNUM(CODA_HX4, 1, 4, 50),
+	CODA_FIRMWARE_VERNUM(CODA_7541, 1, 4, 50),
+	CODA_FIRMWARE_VERNUM(CODA_960, 2, 1, 5),
+	CODA_FIRMWARE_VERNUM(CODA_960, 2, 1, 9),
+	CODA_FIRMWARE_VERNUM(CODA_960, 2, 3, 10),
+	CODA_FIRMWARE_VERNUM(CODA_960, 3, 1, 1),
+};
+
+static bool coda_firmware_supported(u32 vernum)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(coda_supported_firmwares); i++)
+		if (vernum == coda_supported_firmwares[i])
+			return true;
+	return false;
+}
+
+int coda_check_firmware(struct coda_dev *dev)
+{
+	u16 product, major, minor, release;
+	u32 data;
+	int ret;
+
+	ret = clk_prepare_enable(dev->clk_per);
+	if (ret)
+		goto err_clk_per;
+
+	ret = clk_prepare_enable(dev->clk_ahb);
+	if (ret)
+		goto err_clk_ahb;
+
+	coda_write(dev, 0, CODA_CMD_FIRMWARE_VERNUM);
+	coda_write(dev, CODA_REG_BIT_BUSY_FLAG, CODA_REG_BIT_BUSY);
+	coda_write(dev, 0, CODA_REG_BIT_RUN_INDEX);
+	coda_write(dev, 0, CODA_REG_BIT_RUN_COD_STD);
+	coda_write(dev, CODA_COMMAND_FIRMWARE_GET, CODA_REG_BIT_RUN_COMMAND);
+	if (coda_wait_timeout(dev)) {
+		v4l2_err(&dev->v4l2_dev, "firmware get command error\n");
+		ret = -EIO;
+		goto err_run_cmd;
+	}
+
+	if (dev->devtype->product == CODA_960) {
+		data = coda_read(dev, CODA9_CMD_FIRMWARE_CODE_REV);
+		v4l2_info(&dev->v4l2_dev, "Firmware code revision: %d\n",
+			  data);
+	}
+
+	/* Check we are compatible with the loaded firmware */
+	data = coda_read(dev, CODA_CMD_FIRMWARE_VERNUM);
+	product = CODA_FIRMWARE_PRODUCT(data);
+	major = CODA_FIRMWARE_MAJOR(data);
+	minor = CODA_FIRMWARE_MINOR(data);
+	release = CODA_FIRMWARE_RELEASE(data);
+
+	clk_disable_unprepare(dev->clk_per);
+	clk_disable_unprepare(dev->clk_ahb);
+
+	if (product != dev->devtype->product) {
+		v4l2_err(&dev->v4l2_dev,
+			 "Wrong firmware. Hw: %s, Fw: %s, Version: %u.%u.%u\n",
+			 coda_product_name(dev->devtype->product),
+			 coda_product_name(product), major, minor, release);
+		return -EINVAL;
+	}
+
+	v4l2_info(&dev->v4l2_dev, "Initialized %s.\n",
+		  coda_product_name(product));
+
+	if (coda_firmware_supported(data)) {
+		v4l2_info(&dev->v4l2_dev, "Firmware version: %u.%u.%u\n",
+			  major, minor, release);
+	} else {
+		v4l2_warn(&dev->v4l2_dev,
+			  "Unsupported firmware version: %u.%u.%u\n",
+			  major, minor, release);
+	}
+
+	return 0;
+
+err_run_cmd:
+	clk_disable_unprepare(dev->clk_ahb);
+err_clk_ahb:
+	clk_disable_unprepare(dev->clk_per);
+err_clk_per:
+	return ret;
+}
+
+static void coda9_set_frame_cache(struct coda_ctx *ctx, u32 fourcc)
+{
+	u32 cache_size, cache_config;
+
+	if (ctx->tiled_map_type == GDI_LINEAR_FRAME_MAP) {
+		/* Luma 2x0 page, 2x6 cache, chroma 2x0 page, 2x4 cache size */
+		cache_size = 0x20262024;
+		cache_config = 2 << CODA9_CACHE_PAGEMERGE_OFFSET;
+	} else {
+		/* Luma 0x2 page, 4x4 cache, chroma 0x2 page, 4x3 cache size */
+		cache_size = 0x02440243;
+		cache_config = 1 << CODA9_CACHE_PAGEMERGE_OFFSET;
+	}
+	coda_write(ctx->dev, cache_size, CODA9_CMD_SET_FRAME_CACHE_SIZE);
+	if (fourcc == V4L2_PIX_FMT_NV12 || fourcc == V4L2_PIX_FMT_YUYV) {
+		cache_config |= 32 << CODA9_CACHE_LUMA_BUFFER_SIZE_OFFSET |
+				16 << CODA9_CACHE_CR_BUFFER_SIZE_OFFSET |
+				0 << CODA9_CACHE_CB_BUFFER_SIZE_OFFSET;
+	} else {
+		cache_config |= 32 << CODA9_CACHE_LUMA_BUFFER_SIZE_OFFSET |
+				8 << CODA9_CACHE_CR_BUFFER_SIZE_OFFSET |
+				8 << CODA9_CACHE_CB_BUFFER_SIZE_OFFSET;
+	}
+	coda_write(ctx->dev, cache_config, CODA9_CMD_SET_FRAME_CACHE_CONFIG);
+}
+
+/*
+ * Encoder context operations
+ */
+
+static int coda_encoder_reqbufs(struct coda_ctx *ctx,
+				struct v4l2_requestbuffers *rb)
+{
+	struct coda_q_data *q_data_src;
+	int ret;
+
+	if (rb->type != V4L2_BUF_TYPE_VIDEO_OUTPUT)
+		return 0;
+
+	if (rb->count) {
+		q_data_src = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+		ret = coda_alloc_context_buffers(ctx, q_data_src);
+		if (ret < 0)
+			return ret;
+	} else {
+		coda_free_context_buffers(ctx);
+	}
+
+	return 0;
+}
+
+static int coda_start_encoding(struct coda_ctx *ctx)
+{
+	struct coda_dev *dev = ctx->dev;
+	struct v4l2_device *v4l2_dev = &dev->v4l2_dev;
+	struct coda_q_data *q_data_src, *q_data_dst;
+	u32 bitstream_buf, bitstream_size;
+	struct vb2_v4l2_buffer *buf;
+	int gamma, ret, value;
+	u32 dst_fourcc;
+	int num_fb;
+	u32 stride;
+
+	q_data_src = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+	q_data_dst = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_CAPTURE);
+	dst_fourcc = q_data_dst->fourcc;
+
+	buf = v4l2_m2m_next_dst_buf(ctx->fh.m2m_ctx);
+	bitstream_buf = vb2_dma_contig_plane_dma_addr(&buf->vb2_buf, 0);
+	bitstream_size = q_data_dst->sizeimage;
+
+	if (!coda_is_initialized(dev)) {
+		v4l2_err(v4l2_dev, "coda is not initialized.\n");
+		return -EFAULT;
+	}
+
+	if (dst_fourcc == V4L2_PIX_FMT_JPEG) {
+		if (!ctx->params.jpeg_qmat_tab[0]) {
+			ctx->params.jpeg_qmat_tab[0] = kmalloc(64, GFP_KERNEL);
+			if (!ctx->params.jpeg_qmat_tab[0])
+				return -ENOMEM;
+		}
+		if (!ctx->params.jpeg_qmat_tab[1]) {
+			ctx->params.jpeg_qmat_tab[1] = kmalloc(64, GFP_KERNEL);
+			if (!ctx->params.jpeg_qmat_tab[1])
+				return -ENOMEM;
+		}
+		coda_set_jpeg_compression_quality(ctx, ctx->params.jpeg_quality);
+	}
+
+	mutex_lock(&dev->coda_mutex);
+
+	coda_write(dev, ctx->parabuf.paddr, CODA_REG_BIT_PARA_BUF_ADDR);
+	coda_write(dev, bitstream_buf, CODA_REG_BIT_RD_PTR(ctx->reg_idx));
+	coda_write(dev, bitstream_buf, CODA_REG_BIT_WR_PTR(ctx->reg_idx));
+	switch (dev->devtype->product) {
+	case CODA_DX6:
+		coda_write(dev, CODADX6_STREAM_BUF_DYNALLOC_EN |
+			CODADX6_STREAM_BUF_PIC_RESET, CODA_REG_BIT_STREAM_CTRL);
+		break;
+	case CODA_960:
+		coda_write(dev, 0, CODA9_GDI_WPROT_RGN_EN);
+		fallthrough;
+	case CODA_HX4:
+	case CODA_7541:
+		coda_write(dev, CODA7_STREAM_BUF_DYNALLOC_EN |
+			CODA7_STREAM_BUF_PIC_RESET, CODA_REG_BIT_STREAM_CTRL);
+		break;
+	}
+
+	ctx->frame_mem_ctrl &= ~(CODA_FRAME_CHROMA_INTERLEAVE | (0x3 << 9) |
+				 CODA9_FRAME_TILED2LINEAR);
+	if (q_data_src->fourcc == V4L2_PIX_FMT_NV12)
+		ctx->frame_mem_ctrl |= CODA_FRAME_CHROMA_INTERLEAVE;
+	if (ctx->tiled_map_type == GDI_TILED_FRAME_MB_RASTER_MAP)
+		ctx->frame_mem_ctrl |= (0x3 << 9) | CODA9_FRAME_TILED2LINEAR;
+	coda_write(dev, ctx->frame_mem_ctrl, CODA_REG_BIT_FRAME_MEM_CTRL);
+
+	if (dev->devtype->product == CODA_DX6) {
+		/* Configure the coda */
+		coda_write(dev, dev->iram.paddr,
+			   CODADX6_REG_BIT_SEARCH_RAM_BASE_ADDR);
+	}
+
+	/* Could set rotation here if needed */
+	value = 0;
+	switch (dev->devtype->product) {
+	case CODA_DX6:
+		value = (q_data_src->rect.width & CODADX6_PICWIDTH_MASK)
+			<< CODADX6_PICWIDTH_OFFSET;
+		value |= (q_data_src->rect.height & CODADX6_PICHEIGHT_MASK)
+			 << CODA_PICHEIGHT_OFFSET;
+		break;
+	case CODA_HX4:
+	case CODA_7541:
+		if (dst_fourcc == V4L2_PIX_FMT_H264) {
+			value = (round_up(q_data_src->rect.width, 16) &
+				 CODA7_PICWIDTH_MASK) << CODA7_PICWIDTH_OFFSET;
+			value |= (round_up(q_data_src->rect.height, 16) &
+				 CODA7_PICHEIGHT_MASK) << CODA_PICHEIGHT_OFFSET;
+			break;
+		}
+		fallthrough;
+	case CODA_960:
+		value = (q_data_src->rect.width & CODA7_PICWIDTH_MASK)
+			<< CODA7_PICWIDTH_OFFSET;
+		value |= (q_data_src->rect.height & CODA7_PICHEIGHT_MASK)
+			 << CODA_PICHEIGHT_OFFSET;
+	}
+	coda_write(dev, value, CODA_CMD_ENC_SEQ_SRC_SIZE);
+	if (dst_fourcc == V4L2_PIX_FMT_JPEG)
+		ctx->params.framerate = 0;
+	coda_write(dev, ctx->params.framerate,
+		   CODA_CMD_ENC_SEQ_SRC_F_RATE);
+
+	ctx->params.codec_mode = ctx->codec->mode;
+	switch (dst_fourcc) {
+	case V4L2_PIX_FMT_MPEG4:
+		if (dev->devtype->product == CODA_960)
+			coda_write(dev, CODA9_STD_MPEG4,
+				   CODA_CMD_ENC_SEQ_COD_STD);
+		else
+			coda_write(dev, CODA_STD_MPEG4,
+				   CODA_CMD_ENC_SEQ_COD_STD);
+		coda_write(dev, 0, CODA_CMD_ENC_SEQ_MP4_PARA);
+		break;
+	case V4L2_PIX_FMT_H264:
+		if (dev->devtype->product == CODA_960)
+			coda_write(dev, CODA9_STD_H264,
+				   CODA_CMD_ENC_SEQ_COD_STD);
+		else
+			coda_write(dev, CODA_STD_H264,
+				   CODA_CMD_ENC_SEQ_COD_STD);
+		value = ((ctx->params.h264_disable_deblocking_filter_idc &
+			  CODA_264PARAM_DISABLEDEBLK_MASK) <<
+			 CODA_264PARAM_DISABLEDEBLK_OFFSET) |
+			((ctx->params.h264_slice_alpha_c0_offset_div2 &
+			  CODA_264PARAM_DEBLKFILTEROFFSETALPHA_MASK) <<
+			 CODA_264PARAM_DEBLKFILTEROFFSETALPHA_OFFSET) |
+			((ctx->params.h264_slice_beta_offset_div2 &
+			  CODA_264PARAM_DEBLKFILTEROFFSETBETA_MASK) <<
+			 CODA_264PARAM_DEBLKFILTEROFFSETBETA_OFFSET) |
+			(ctx->params.h264_constrained_intra_pred_flag <<
+			 CODA_264PARAM_CONSTRAINEDINTRAPREDFLAG_OFFSET) |
+			(ctx->params.h264_chroma_qp_index_offset &
+			 CODA_264PARAM_CHROMAQPOFFSET_MASK);
+		coda_write(dev, value, CODA_CMD_ENC_SEQ_264_PARA);
+		break;
+	case V4L2_PIX_FMT_JPEG:
+		coda_write(dev, 0, CODA_CMD_ENC_SEQ_JPG_PARA);
+		coda_write(dev, ctx->params.jpeg_restart_interval,
+				CODA_CMD_ENC_SEQ_JPG_RST_INTERVAL);
+		coda_write(dev, 0, CODA_CMD_ENC_SEQ_JPG_THUMB_EN);
+		coda_write(dev, 0, CODA_CMD_ENC_SEQ_JPG_THUMB_SIZE);
+		coda_write(dev, 0, CODA_CMD_ENC_SEQ_JPG_THUMB_OFFSET);
+
+		coda_jpeg_write_tables(ctx);
+		break;
+	default:
+		v4l2_err(v4l2_dev,
+			 "dst format (0x%08x) invalid.\n", dst_fourcc);
+		ret = -EINVAL;
+		goto out;
+	}
+
+	/*
+	 * slice mode and GOP size registers are used for thumb size/offset
+	 * in JPEG mode
+	 */
+	if (dst_fourcc != V4L2_PIX_FMT_JPEG) {
+		value = coda_slice_mode(ctx);
+		coda_write(dev, value, CODA_CMD_ENC_SEQ_SLICE_MODE);
+		value = ctx->params.gop_size;
+		coda_write(dev, value, CODA_CMD_ENC_SEQ_GOP_SIZE);
+	}
+
+	if (ctx->params.bitrate && (ctx->params.frame_rc_enable ||
+				    ctx->params.mb_rc_enable)) {
+		ctx->params.bitrate_changed = false;
+		ctx->params.h264_intra_qp_changed = false;
+
+		/* Rate control enabled */
+		value = (ctx->params.bitrate & CODA_RATECONTROL_BITRATE_MASK)
+			<< CODA_RATECONTROL_BITRATE_OFFSET;
+		value |=  1 & CODA_RATECONTROL_ENABLE_MASK;
+		value |= (ctx->params.vbv_delay &
+			  CODA_RATECONTROL_INITIALDELAY_MASK)
+			 << CODA_RATECONTROL_INITIALDELAY_OFFSET;
+		if (dev->devtype->product == CODA_960)
+			value |= BIT(31); /* disable autoskip */
+	} else {
+		value = 0;
+	}
+	coda_write(dev, value, CODA_CMD_ENC_SEQ_RC_PARA);
+
+	coda_write(dev, ctx->params.vbv_size, CODA_CMD_ENC_SEQ_RC_BUF_SIZE);
+	coda_write(dev, ctx->params.intra_refresh,
+		   CODA_CMD_ENC_SEQ_INTRA_REFRESH);
+
+	coda_write(dev, bitstream_buf, CODA_CMD_ENC_SEQ_BB_START);
+	coda_write(dev, bitstream_size / 1024, CODA_CMD_ENC_SEQ_BB_SIZE);
+
+
+	value = 0;
+	if (dev->devtype->product == CODA_960)
+		gamma = CODA9_DEFAULT_GAMMA;
+	else
+		gamma = CODA_DEFAULT_GAMMA;
+	if (gamma > 0) {
+		coda_write(dev, (gamma & CODA_GAMMA_MASK) << CODA_GAMMA_OFFSET,
+			   CODA_CMD_ENC_SEQ_RC_GAMMA);
+	}
+
+	if (ctx->params.h264_min_qp || ctx->params.h264_max_qp) {
+		coda_write(dev,
+			   ctx->params.h264_min_qp << CODA_QPMIN_OFFSET |
+			   ctx->params.h264_max_qp << CODA_QPMAX_OFFSET,
+			   CODA_CMD_ENC_SEQ_RC_QP_MIN_MAX);
+	}
+	if (dev->devtype->product == CODA_960) {
+		if (ctx->params.h264_max_qp)
+			value |= 1 << CODA9_OPTION_RCQPMAX_OFFSET;
+		if (CODA_DEFAULT_GAMMA > 0)
+			value |= 1 << CODA9_OPTION_GAMMA_OFFSET;
+	} else {
+		if (CODA_DEFAULT_GAMMA > 0) {
+			if (dev->devtype->product == CODA_DX6)
+				value |= 1 << CODADX6_OPTION_GAMMA_OFFSET;
+			else
+				value |= 1 << CODA7_OPTION_GAMMA_OFFSET;
+		}
+		if (ctx->params.h264_min_qp)
+			value |= 1 << CODA7_OPTION_RCQPMIN_OFFSET;
+		if (ctx->params.h264_max_qp)
+			value |= 1 << CODA7_OPTION_RCQPMAX_OFFSET;
+	}
+	coda_write(dev, value, CODA_CMD_ENC_SEQ_OPTION);
+
+	if (ctx->params.frame_rc_enable && !ctx->params.mb_rc_enable)
+		value = 1;
+	else
+		value = 0;
+	coda_write(dev, value, CODA_CMD_ENC_SEQ_RC_INTERVAL_MODE);
+
+	coda_setup_iram(ctx);
+
+	if (dst_fourcc == V4L2_PIX_FMT_H264) {
+		switch (dev->devtype->product) {
+		case CODA_DX6:
+			value = FMO_SLICE_SAVE_BUF_SIZE << 7;
+			coda_write(dev, value, CODADX6_CMD_ENC_SEQ_FMO);
+			break;
+		case CODA_HX4:
+		case CODA_7541:
+			coda_write(dev, ctx->iram_info.search_ram_paddr,
+					CODA7_CMD_ENC_SEQ_SEARCH_BASE);
+			coda_write(dev, ctx->iram_info.search_ram_size,
+					CODA7_CMD_ENC_SEQ_SEARCH_SIZE);
+			break;
+		case CODA_960:
+			coda_write(dev, 0, CODA9_CMD_ENC_SEQ_ME_OPTION);
+			coda_write(dev, 0, CODA9_CMD_ENC_SEQ_INTRA_WEIGHT);
+		}
+	}
+
+	ret = coda_command_sync(ctx, CODA_COMMAND_SEQ_INIT);
+	if (ret < 0) {
+		v4l2_err(v4l2_dev, "CODA_COMMAND_SEQ_INIT timeout\n");
+		goto out;
+	}
+
+	if (coda_read(dev, CODA_RET_ENC_SEQ_SUCCESS) == 0) {
+		v4l2_err(v4l2_dev, "CODA_COMMAND_SEQ_INIT failed\n");
+		ret = -EFAULT;
+		goto out;
+	}
+	ctx->initialized = 1;
+
+	if (dst_fourcc != V4L2_PIX_FMT_JPEG) {
+		if (dev->devtype->product == CODA_960)
+			ctx->num_internal_frames = 4;
+		else
+			ctx->num_internal_frames = 2;
+		ret = coda_alloc_framebuffers(ctx, q_data_src, dst_fourcc);
+		if (ret < 0) {
+			v4l2_err(v4l2_dev, "failed to allocate framebuffers\n");
+			goto out;
+		}
+		num_fb = 2;
+		stride = q_data_src->bytesperline;
+	} else {
+		ctx->num_internal_frames = 0;
+		num_fb = 0;
+		stride = 0;
+	}
+	coda_write(dev, num_fb, CODA_CMD_SET_FRAME_BUF_NUM);
+	coda_write(dev, stride, CODA_CMD_SET_FRAME_BUF_STRIDE);
+
+	if (dev->devtype->product == CODA_HX4 ||
+	    dev->devtype->product == CODA_7541) {
+		coda_write(dev, q_data_src->bytesperline,
+				CODA7_CMD_SET_FRAME_SOURCE_BUF_STRIDE);
+	}
+	if (dev->devtype->product != CODA_DX6) {
+		coda_write(dev, ctx->iram_info.buf_bit_use,
+				CODA7_CMD_SET_FRAME_AXI_BIT_ADDR);
+		coda_write(dev, ctx->iram_info.buf_ip_ac_dc_use,
+				CODA7_CMD_SET_FRAME_AXI_IPACDC_ADDR);
+		coda_write(dev, ctx->iram_info.buf_dbk_y_use,
+				CODA7_CMD_SET_FRAME_AXI_DBKY_ADDR);
+		coda_write(dev, ctx->iram_info.buf_dbk_c_use,
+				CODA7_CMD_SET_FRAME_AXI_DBKC_ADDR);
+		coda_write(dev, ctx->iram_info.buf_ovl_use,
+				CODA7_CMD_SET_FRAME_AXI_OVL_ADDR);
+		if (dev->devtype->product == CODA_960) {
+			coda_write(dev, ctx->iram_info.buf_btp_use,
+					CODA9_CMD_SET_FRAME_AXI_BTP_ADDR);
+
+			coda9_set_frame_cache(ctx, q_data_src->fourcc);
+
+			/* FIXME */
+			coda_write(dev, ctx->internal_frames[2].buf.paddr,
+				   CODA9_CMD_SET_FRAME_SUBSAMP_A);
+			coda_write(dev, ctx->internal_frames[3].buf.paddr,
+				   CODA9_CMD_SET_FRAME_SUBSAMP_B);
+		}
+	}
+
+	ret = coda_command_sync(ctx, CODA_COMMAND_SET_FRAME_BUF);
+	if (ret < 0) {
+		v4l2_err(v4l2_dev, "CODA_COMMAND_SET_FRAME_BUF timeout\n");
+		goto out;
+	}
+
+	coda_dbg(1, ctx, "start encoding %dx%d %4.4s->%4.4s @ %d/%d Hz\n",
+		 q_data_src->rect.width, q_data_src->rect.height,
+		 (char *)&ctx->codec->src_fourcc, (char *)&dst_fourcc,
+		 ctx->params.framerate & 0xffff,
+		 (ctx->params.framerate >> 16) + 1);
+
+	/* Save stream headers */
+	buf = v4l2_m2m_next_dst_buf(ctx->fh.m2m_ctx);
+	switch (dst_fourcc) {
+	case V4L2_PIX_FMT_H264:
+		/*
+		 * Get SPS in the first frame and copy it to an
+		 * intermediate buffer.
+		 */
+		ret = coda_encode_header(ctx, buf, CODA_HEADER_H264_SPS,
+					 &ctx->vpu_header[0][0],
+					 &ctx->vpu_header_size[0]);
+		if (ret < 0)
+			goto out;
+
+		/*
+		 * If visible width or height are not aligned to macroblock
+		 * size, the crop_right and crop_bottom SPS fields must be set
+		 * to the difference between visible and coded size.  This is
+		 * only supported by CODA960 firmware. All others do not allow
+		 * writing frame cropping parameters, so we have to manually
+		 * fix up the SPS RBSP (Sequence Parameter Set Raw Byte
+		 * Sequence Payload) ourselves.
+		 */
+		if (ctx->dev->devtype->product != CODA_960 &&
+		    ((q_data_src->rect.width % 16) ||
+		     (q_data_src->rect.height % 16))) {
+			ret = coda_h264_sps_fixup(ctx, q_data_src->rect.width,
+						  q_data_src->rect.height,
+						  &ctx->vpu_header[0][0],
+						  &ctx->vpu_header_size[0],
+						  sizeof(ctx->vpu_header[0]));
+			if (ret < 0)
+				goto out;
+		}
+
+		/*
+		 * Get PPS in the first frame and copy it to an
+		 * intermediate buffer.
+		 */
+		ret = coda_encode_header(ctx, buf, CODA_HEADER_H264_PPS,
+					 &ctx->vpu_header[1][0],
+					 &ctx->vpu_header_size[1]);
+		if (ret < 0)
+			goto out;
+
+		/*
+		 * Length of H.264 headers is variable and thus it might not be
+		 * aligned for the coda to append the encoded frame. In that is
+		 * the case a filler NAL must be added to header 2.
+		 */
+		ctx->vpu_header_size[2] = coda_h264_padding(
+					(ctx->vpu_header_size[0] +
+					 ctx->vpu_header_size[1]),
+					 ctx->vpu_header[2]);
+		break;
+	case V4L2_PIX_FMT_MPEG4:
+		/*
+		 * Get VOS in the first frame and copy it to an
+		 * intermediate buffer
+		 */
+		ret = coda_encode_header(ctx, buf, CODA_HEADER_MP4V_VOS,
+					 &ctx->vpu_header[0][0],
+					 &ctx->vpu_header_size[0]);
+		if (ret < 0)
+			goto out;
+
+		ret = coda_encode_header(ctx, buf, CODA_HEADER_MP4V_VIS,
+					 &ctx->vpu_header[1][0],
+					 &ctx->vpu_header_size[1]);
+		if (ret < 0)
+			goto out;
+
+		ret = coda_encode_header(ctx, buf, CODA_HEADER_MP4V_VOL,
+					 &ctx->vpu_header[2][0],
+					 &ctx->vpu_header_size[2]);
+		if (ret < 0)
+			goto out;
+		break;
+	default:
+		/* No more formats need to save headers at the moment */
+		break;
+	}
+
+out:
+	mutex_unlock(&dev->coda_mutex);
+	return ret;
+}
+
+static int coda_prepare_encode(struct coda_ctx *ctx)
+{
+	struct coda_q_data *q_data_src, *q_data_dst;
+	struct vb2_v4l2_buffer *src_buf, *dst_buf;
+	struct coda_dev *dev = ctx->dev;
+	int force_ipicture;
+	int quant_param = 0;
+	u32 pic_stream_buffer_addr, pic_stream_buffer_size;
+	u32 rot_mode = 0;
+	u32 dst_fourcc;
+	u32 reg;
+	int ret;
+
+	ret = coda_enc_param_change(ctx);
+	if (ret < 0) {
+		v4l2_warn(&ctx->dev->v4l2_dev, "parameter change failed: %d\n",
+			  ret);
+	}
+
+	src_buf = v4l2_m2m_next_src_buf(ctx->fh.m2m_ctx);
+	dst_buf = v4l2_m2m_next_dst_buf(ctx->fh.m2m_ctx);
+	q_data_src = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+	q_data_dst = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_CAPTURE);
+	dst_fourcc = q_data_dst->fourcc;
+
+	src_buf->sequence = ctx->osequence;
+	dst_buf->sequence = ctx->osequence;
+	ctx->osequence++;
+
+	force_ipicture = ctx->params.force_ipicture;
+	if (force_ipicture)
+		ctx->params.force_ipicture = false;
+	else if (ctx->params.gop_size != 0 &&
+		 (src_buf->sequence % ctx->params.gop_size) == 0)
+		force_ipicture = 1;
+
+	/*
+	 * Workaround coda firmware BUG that only marks the first
+	 * frame as IDR. This is a problem for some decoders that can't
+	 * recover when a frame is lost.
+	 */
+	if (!force_ipicture) {
+		src_buf->flags |= V4L2_BUF_FLAG_PFRAME;
+		src_buf->flags &= ~V4L2_BUF_FLAG_KEYFRAME;
+	} else {
+		src_buf->flags |= V4L2_BUF_FLAG_KEYFRAME;
+		src_buf->flags &= ~V4L2_BUF_FLAG_PFRAME;
+	}
+
+	if (dev->devtype->product == CODA_960)
+		coda_set_gdi_regs(ctx);
+
+	/*
+	 * Copy headers in front of the first frame and forced I frames for
+	 * H.264 only. In MPEG4 they are already copied by the CODA.
+	 */
+	if (src_buf->sequence == 0 || force_ipicture) {
+		pic_stream_buffer_addr =
+			vb2_dma_contig_plane_dma_addr(&dst_buf->vb2_buf, 0) +
+			ctx->vpu_header_size[0] +
+			ctx->vpu_header_size[1] +
+			ctx->vpu_header_size[2];
+		pic_stream_buffer_size = q_data_dst->sizeimage -
+			ctx->vpu_header_size[0] -
+			ctx->vpu_header_size[1] -
+			ctx->vpu_header_size[2];
+		memcpy(vb2_plane_vaddr(&dst_buf->vb2_buf, 0),
+		       &ctx->vpu_header[0][0], ctx->vpu_header_size[0]);
+		memcpy(vb2_plane_vaddr(&dst_buf->vb2_buf, 0)
+			+ ctx->vpu_header_size[0], &ctx->vpu_header[1][0],
+			ctx->vpu_header_size[1]);
+		memcpy(vb2_plane_vaddr(&dst_buf->vb2_buf, 0)
+			+ ctx->vpu_header_size[0] + ctx->vpu_header_size[1],
+			&ctx->vpu_header[2][0], ctx->vpu_header_size[2]);
+	} else {
+		pic_stream_buffer_addr =
+			vb2_dma_contig_plane_dma_addr(&dst_buf->vb2_buf, 0);
+		pic_stream_buffer_size = q_data_dst->sizeimage;
+	}
+
+	if (force_ipicture) {
+		switch (dst_fourcc) {
+		case V4L2_PIX_FMT_H264:
+			quant_param = ctx->params.h264_intra_qp;
+			break;
+		case V4L2_PIX_FMT_MPEG4:
+			quant_param = ctx->params.mpeg4_intra_qp;
+			break;
+		case V4L2_PIX_FMT_JPEG:
+			quant_param = 30;
+			break;
+		default:
+			v4l2_warn(&ctx->dev->v4l2_dev,
+				"cannot set intra qp, fmt not supported\n");
+			break;
+		}
+	} else {
+		switch (dst_fourcc) {
+		case V4L2_PIX_FMT_H264:
+			quant_param = ctx->params.h264_inter_qp;
+			break;
+		case V4L2_PIX_FMT_MPEG4:
+			quant_param = ctx->params.mpeg4_inter_qp;
+			break;
+		default:
+			v4l2_warn(&ctx->dev->v4l2_dev,
+				"cannot set inter qp, fmt not supported\n");
+			break;
+		}
+	}
+
+	/* submit */
+	if (ctx->params.rot_mode)
+		rot_mode = CODA_ROT_MIR_ENABLE | ctx->params.rot_mode;
+	coda_write(dev, rot_mode, CODA_CMD_ENC_PIC_ROT_MODE);
+	coda_write(dev, quant_param, CODA_CMD_ENC_PIC_QS);
+
+	if (dev->devtype->product == CODA_960) {
+		coda_write(dev, 4/*FIXME: 0*/, CODA9_CMD_ENC_PIC_SRC_INDEX);
+		coda_write(dev, q_data_src->bytesperline,
+			   CODA9_CMD_ENC_PIC_SRC_STRIDE);
+		coda_write(dev, 0, CODA9_CMD_ENC_PIC_SUB_FRAME_SYNC);
+
+		reg = CODA9_CMD_ENC_PIC_SRC_ADDR_Y;
+	} else {
+		reg = CODA_CMD_ENC_PIC_SRC_ADDR_Y;
+	}
+	coda_write_base(ctx, q_data_src, src_buf, reg);
+
+	coda_write(dev, force_ipicture << 1 & 0x2,
+		   CODA_CMD_ENC_PIC_OPTION);
+
+	coda_write(dev, pic_stream_buffer_addr, CODA_CMD_ENC_PIC_BB_START);
+	coda_write(dev, pic_stream_buffer_size / 1024,
+		   CODA_CMD_ENC_PIC_BB_SIZE);
+
+	if (!ctx->streamon_out) {
+		/* After streamoff on the output side, set stream end flag */
+		ctx->bit_stream_param |= CODA_BIT_STREAM_END_FLAG;
+		coda_write(dev, ctx->bit_stream_param,
+			   CODA_REG_BIT_BIT_STREAM_PARAM);
+	}
+
+	if (dev->devtype->product != CODA_DX6)
+		coda_write(dev, ctx->iram_info.axi_sram_use,
+				CODA7_REG_BIT_AXI_SRAM_USE);
+
+	trace_coda_enc_pic_run(ctx, src_buf);
+
+	coda_command_async(ctx, CODA_COMMAND_PIC_RUN);
+
+	return 0;
+}
+
+static char coda_frame_type_char(u32 flags)
+{
+	return (flags & V4L2_BUF_FLAG_KEYFRAME) ? 'I' :
+	       (flags & V4L2_BUF_FLAG_PFRAME) ? 'P' :
+	       (flags & V4L2_BUF_FLAG_BFRAME) ? 'B' : '?';
+}
+
+static void coda_finish_encode(struct coda_ctx *ctx)
+{
+	struct vb2_v4l2_buffer *src_buf, *dst_buf;
+	struct coda_dev *dev = ctx->dev;
+	u32 wr_ptr, start_ptr;
+
+	if (ctx->aborting)
+		return;
+
+	/*
+	 * Lock to make sure that an encoder stop command running in parallel
+	 * will either already have marked src_buf as last, or it will wake up
+	 * the capture queue after the buffers are returned.
+	 */
+	mutex_lock(&ctx->wakeup_mutex);
+	src_buf = v4l2_m2m_src_buf_remove(ctx->fh.m2m_ctx);
+	dst_buf = v4l2_m2m_next_dst_buf(ctx->fh.m2m_ctx);
+
+	trace_coda_enc_pic_done(ctx, dst_buf);
+
+	/* Get results from the coda */
+	start_ptr = coda_read(dev, CODA_CMD_ENC_PIC_BB_START);
+	wr_ptr = coda_read(dev, CODA_REG_BIT_WR_PTR(ctx->reg_idx));
+
+	/* Calculate bytesused field */
+	if (dst_buf->sequence == 0 ||
+	    src_buf->flags & V4L2_BUF_FLAG_KEYFRAME) {
+		vb2_set_plane_payload(&dst_buf->vb2_buf, 0, wr_ptr - start_ptr +
+					ctx->vpu_header_size[0] +
+					ctx->vpu_header_size[1] +
+					ctx->vpu_header_size[2]);
+	} else {
+		vb2_set_plane_payload(&dst_buf->vb2_buf, 0, wr_ptr - start_ptr);
+	}
+
+	coda_dbg(1, ctx, "frame size = %u\n", wr_ptr - start_ptr);
+
+	coda_read(dev, CODA_RET_ENC_PIC_SLICE_NUM);
+	coda_read(dev, CODA_RET_ENC_PIC_FLAG);
+
+	dst_buf->flags &= ~(V4L2_BUF_FLAG_KEYFRAME |
+			    V4L2_BUF_FLAG_PFRAME |
+			    V4L2_BUF_FLAG_LAST);
+	if (coda_read(dev, CODA_RET_ENC_PIC_TYPE) == 0)
+		dst_buf->flags |= V4L2_BUF_FLAG_KEYFRAME;
+	else
+		dst_buf->flags |= V4L2_BUF_FLAG_PFRAME;
+	dst_buf->flags |= src_buf->flags & V4L2_BUF_FLAG_LAST;
+
+	v4l2_m2m_buf_copy_metadata(src_buf, dst_buf, false);
+
+	v4l2_m2m_buf_done(src_buf, VB2_BUF_STATE_DONE);
+
+	dst_buf = v4l2_m2m_dst_buf_remove(ctx->fh.m2m_ctx);
+	coda_m2m_buf_done(ctx, dst_buf, VB2_BUF_STATE_DONE);
+	mutex_unlock(&ctx->wakeup_mutex);
+
+	ctx->gopcounter--;
+	if (ctx->gopcounter < 0)
+		ctx->gopcounter = ctx->params.gop_size - 1;
+
+	coda_dbg(1, ctx, "job finished: encoded %c frame (%d)%s\n",
+		 coda_frame_type_char(dst_buf->flags), dst_buf->sequence,
+		 (dst_buf->flags & V4L2_BUF_FLAG_LAST) ? " (last)" : "");
+}
+
+static void coda_seq_end_work(struct work_struct *work)
+{
+	struct coda_ctx *ctx = container_of(work, struct coda_ctx, seq_end_work);
+	struct coda_dev *dev = ctx->dev;
+
+	mutex_lock(&ctx->buffer_mutex);
+	mutex_lock(&dev->coda_mutex);
+
+	if (ctx->initialized == 0)
+		goto out;
+
+	coda_dbg(1, ctx, "%s: sent command 'SEQ_END' to coda\n", __func__);
+	if (coda_command_sync(ctx, CODA_COMMAND_SEQ_END)) {
+		v4l2_err(&dev->v4l2_dev,
+			 "CODA_COMMAND_SEQ_END failed\n");
+	}
+
+	/*
+	 * FIXME: Sometimes h.264 encoding fails with 8-byte sequences missing
+	 * from the output stream after the h.264 decoder has run. Resetting the
+	 * hardware after the decoder has finished seems to help.
+	 */
+	if (dev->devtype->product == CODA_960)
+		coda_hw_reset(ctx);
+
+	kfifo_init(&ctx->bitstream_fifo,
+		ctx->bitstream.vaddr, ctx->bitstream.size);
+
+	coda_free_framebuffers(ctx);
+
+	ctx->initialized = 0;
+
+out:
+	mutex_unlock(&dev->coda_mutex);
+	mutex_unlock(&ctx->buffer_mutex);
+}
+
+static void coda_bit_release(struct coda_ctx *ctx)
+{
+	mutex_lock(&ctx->buffer_mutex);
+	coda_free_framebuffers(ctx);
+	coda_free_context_buffers(ctx);
+	coda_free_bitstream_buffer(ctx);
+	mutex_unlock(&ctx->buffer_mutex);
+}
+
+const struct coda_context_ops coda_bit_encode_ops = {
+	.queue_init = coda_encoder_queue_init,
+	.reqbufs = coda_encoder_reqbufs,
+	.start_streaming = coda_start_encoding,
+	.prepare_run = coda_prepare_encode,
+	.finish_run = coda_finish_encode,
+	.seq_end_work = coda_seq_end_work,
+	.release = coda_bit_release,
+};
+
+/*
+ * Decoder context operations
+ */
+
+static int coda_alloc_bitstream_buffer(struct coda_ctx *ctx,
+				       struct coda_q_data *q_data)
+{
+	if (ctx->bitstream.vaddr)
+		return 0;
+
+	ctx->bitstream.size = roundup_pow_of_two(q_data->sizeimage * 2);
+	ctx->bitstream.vaddr = dma_alloc_wc(ctx->dev->dev, ctx->bitstream.size,
+					    &ctx->bitstream.paddr, GFP_KERNEL);
+	if (!ctx->bitstream.vaddr) {
+		v4l2_err(&ctx->dev->v4l2_dev,
+			 "failed to allocate bitstream ringbuffer");
+		return -ENOMEM;
+	}
+	kfifo_init(&ctx->bitstream_fifo,
+		   ctx->bitstream.vaddr, ctx->bitstream.size);
+
+	return 0;
+}
+
+static void coda_free_bitstream_buffer(struct coda_ctx *ctx)
+{
+	if (ctx->bitstream.vaddr == NULL)
+		return;
+
+	dma_free_wc(ctx->dev->dev, ctx->bitstream.size, ctx->bitstream.vaddr,
+		    ctx->bitstream.paddr);
+	ctx->bitstream.vaddr = NULL;
+	kfifo_init(&ctx->bitstream_fifo, NULL, 0);
+}
+
+static int coda_decoder_reqbufs(struct coda_ctx *ctx,
+				struct v4l2_requestbuffers *rb)
+{
+	struct coda_q_data *q_data_src;
+	int ret;
+
+	if (rb->type != V4L2_BUF_TYPE_VIDEO_OUTPUT)
+		return 0;
+
+	if (rb->count) {
+		q_data_src = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+		ret = coda_alloc_context_buffers(ctx, q_data_src);
+		if (ret < 0)
+			return ret;
+		ret = coda_alloc_bitstream_buffer(ctx, q_data_src);
+		if (ret < 0) {
+			coda_free_context_buffers(ctx);
+			return ret;
+		}
+	} else {
+		coda_free_bitstream_buffer(ctx);
+		coda_free_context_buffers(ctx);
+	}
+
+	return 0;
+}
+
+static bool coda_reorder_enable(struct coda_ctx *ctx)
+{
+	struct coda_dev *dev = ctx->dev;
+	int profile;
+
+	if (dev->devtype->product != CODA_HX4 &&
+	    dev->devtype->product != CODA_7541 &&
+	    dev->devtype->product != CODA_960)
+		return false;
+
+	if (ctx->codec->src_fourcc == V4L2_PIX_FMT_JPEG)
+		return false;
+
+	if (ctx->codec->src_fourcc != V4L2_PIX_FMT_H264)
+		return true;
+
+	profile = coda_h264_profile(ctx->params.h264_profile_idc);
+	if (profile < 0)
+		v4l2_warn(&dev->v4l2_dev, "Unknown H264 Profile: %u\n",
+			  ctx->params.h264_profile_idc);
+
+	/* Baseline profile does not support reordering */
+	return profile > V4L2_MPEG_VIDEO_H264_PROFILE_BASELINE;
+}
+
+static void coda_decoder_drop_used_metas(struct coda_ctx *ctx)
+{
+	struct coda_buffer_meta *meta, *tmp;
+
+	/*
+	 * All metas that end at or before the RD pointer (fifo out),
+	 * are now consumed by the VPU and should be released.
+	 */
+	spin_lock(&ctx->buffer_meta_lock);
+	list_for_each_entry_safe(meta, tmp, &ctx->buffer_meta_list, list) {
+		if (ctx->bitstream_fifo.kfifo.out >= meta->end) {
+			coda_dbg(2, ctx, "releasing meta: seq=%d start=%d end=%d\n",
+				 meta->sequence, meta->start, meta->end);
+
+			list_del(&meta->list);
+			ctx->num_metas--;
+			ctx->first_frame_sequence++;
+			kfree(meta);
+		}
+	}
+	spin_unlock(&ctx->buffer_meta_lock);
+}
+
+static int __coda_decoder_seq_init(struct coda_ctx *ctx)
+{
+	struct coda_q_data *q_data_src, *q_data_dst;
+	u32 bitstream_buf, bitstream_size;
+	struct coda_dev *dev = ctx->dev;
+	int width, height;
+	u32 src_fourcc, dst_fourcc;
+	u32 val;
+	int ret;
+
+	lockdep_assert_held(&dev->coda_mutex);
+
+	coda_dbg(1, ctx, "Video Data Order Adapter: %s\n",
+		 ctx->use_vdoa ? "Enabled" : "Disabled");
+
+	/* Start decoding */
+	q_data_src = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+	q_data_dst = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_CAPTURE);
+	bitstream_buf = ctx->bitstream.paddr;
+	bitstream_size = ctx->bitstream.size;
+	src_fourcc = q_data_src->fourcc;
+	dst_fourcc = q_data_dst->fourcc;
+
+	/* Update coda bitstream read and write pointers from kfifo */
+	coda_kfifo_sync_to_device_full(ctx);
+
+	ctx->frame_mem_ctrl &= ~(CODA_FRAME_CHROMA_INTERLEAVE | (0x3 << 9) |
+				 CODA9_FRAME_TILED2LINEAR);
+	if (dst_fourcc == V4L2_PIX_FMT_NV12 || dst_fourcc == V4L2_PIX_FMT_YUYV)
+		ctx->frame_mem_ctrl |= CODA_FRAME_CHROMA_INTERLEAVE;
+	if (ctx->tiled_map_type == GDI_TILED_FRAME_MB_RASTER_MAP)
+		ctx->frame_mem_ctrl |= (0x3 << 9) |
+			((ctx->use_vdoa) ? 0 : CODA9_FRAME_TILED2LINEAR);
+	coda_write(dev, ctx->frame_mem_ctrl, CODA_REG_BIT_FRAME_MEM_CTRL);
+
+	ctx->display_idx = -1;
+	ctx->frm_dis_flg = 0;
+	coda_write(dev, 0, CODA_REG_BIT_FRM_DIS_FLG(ctx->reg_idx));
+
+	coda_write(dev, bitstream_buf, CODA_CMD_DEC_SEQ_BB_START);
+	coda_write(dev, bitstream_size / 1024, CODA_CMD_DEC_SEQ_BB_SIZE);
+	val = 0;
+	if (coda_reorder_enable(ctx))
+		val |= CODA_REORDER_ENABLE;
+	if (ctx->codec->src_fourcc == V4L2_PIX_FMT_JPEG)
+		val |= CODA_NO_INT_ENABLE;
+	coda_write(dev, val, CODA_CMD_DEC_SEQ_OPTION);
+
+	ctx->params.codec_mode = ctx->codec->mode;
+	if (dev->devtype->product == CODA_960 &&
+	    src_fourcc == V4L2_PIX_FMT_MPEG4)
+		ctx->params.codec_mode_aux = CODA_MP4_AUX_MPEG4;
+	else
+		ctx->params.codec_mode_aux = 0;
+	if (src_fourcc == V4L2_PIX_FMT_MPEG4) {
+		coda_write(dev, CODA_MP4_CLASS_MPEG4,
+			   CODA_CMD_DEC_SEQ_MP4_ASP_CLASS);
+	}
+	if (src_fourcc == V4L2_PIX_FMT_H264) {
+		if (dev->devtype->product == CODA_HX4 ||
+		    dev->devtype->product == CODA_7541) {
+			coda_write(dev, ctx->psbuf.paddr,
+					CODA_CMD_DEC_SEQ_PS_BB_START);
+			coda_write(dev, (CODA7_PS_BUF_SIZE / 1024),
+					CODA_CMD_DEC_SEQ_PS_BB_SIZE);
+		}
+		if (dev->devtype->product == CODA_960) {
+			coda_write(dev, 0, CODA_CMD_DEC_SEQ_X264_MV_EN);
+			coda_write(dev, 512, CODA_CMD_DEC_SEQ_SPP_CHUNK_SIZE);
+		}
+	}
+	if (src_fourcc == V4L2_PIX_FMT_JPEG)
+		coda_write(dev, 0, CODA_CMD_DEC_SEQ_JPG_THUMB_EN);
+	if (dev->devtype->product != CODA_960)
+		coda_write(dev, 0, CODA_CMD_DEC_SEQ_SRC_SIZE);
+
+	ctx->bit_stream_param = CODA_BIT_DEC_SEQ_INIT_ESCAPE;
+	ret = coda_command_sync(ctx, CODA_COMMAND_SEQ_INIT);
+	ctx->bit_stream_param = 0;
+	if (ret) {
+		v4l2_err(&dev->v4l2_dev, "CODA_COMMAND_SEQ_INIT timeout\n");
+		return ret;
+	}
+	ctx->sequence_offset = ~0U;
+	ctx->initialized = 1;
+	ctx->first_frame_sequence = 0;
+
+	/* Update kfifo out pointer from coda bitstream read pointer */
+	coda_kfifo_sync_from_device(ctx);
+
+	/*
+	 * After updating the read pointer, we need to check if
+	 * any metas are consumed and should be released.
+	 */
+	coda_decoder_drop_used_metas(ctx);
+
+	if (coda_read(dev, CODA_RET_DEC_SEQ_SUCCESS) == 0) {
+		v4l2_err(&dev->v4l2_dev,
+			"CODA_COMMAND_SEQ_INIT failed, error code = 0x%x\n",
+			coda_read(dev, CODA_RET_DEC_SEQ_ERR_REASON));
+		return -EAGAIN;
+	}
+
+	val = coda_read(dev, CODA_RET_DEC_SEQ_SRC_SIZE);
+	if (dev->devtype->product == CODA_DX6) {
+		width = (val >> CODADX6_PICWIDTH_OFFSET) & CODADX6_PICWIDTH_MASK;
+		height = val & CODADX6_PICHEIGHT_MASK;
+	} else {
+		width = (val >> CODA7_PICWIDTH_OFFSET) & CODA7_PICWIDTH_MASK;
+		height = val & CODA7_PICHEIGHT_MASK;
+	}
+
+	if (width > q_data_dst->bytesperline || height > q_data_dst->height) {
+		v4l2_err(&dev->v4l2_dev, "stream is %dx%d, not %dx%d\n",
+			 width, height, q_data_dst->bytesperline,
+			 q_data_dst->height);
+		return -EINVAL;
+	}
+
+	width = round_up(width, 16);
+	height = round_up(height, 16);
+
+	coda_dbg(1, ctx, "start decoding: %dx%d\n", width, height);
+
+	ctx->num_internal_frames = coda_read(dev, CODA_RET_DEC_SEQ_FRAME_NEED);
+	/*
+	 * If the VDOA is used, the decoder needs one additional frame,
+	 * because the frames are freed when the next frame is decoded.
+	 * Otherwise there are visible errors in the decoded frames (green
+	 * regions in displayed frames) and a broken order of frames (earlier
+	 * frames are sporadically displayed after later frames).
+	 */
+	if (ctx->use_vdoa)
+		ctx->num_internal_frames += 1;
+	if (ctx->num_internal_frames > CODA_MAX_FRAMEBUFFERS) {
+		v4l2_err(&dev->v4l2_dev,
+			 "not enough framebuffers to decode (%d < %d)\n",
+			 CODA_MAX_FRAMEBUFFERS, ctx->num_internal_frames);
+		return -EINVAL;
+	}
+
+	if (src_fourcc == V4L2_PIX_FMT_H264) {
+		u32 left_right;
+		u32 top_bottom;
+
+		left_right = coda_read(dev, CODA_RET_DEC_SEQ_CROP_LEFT_RIGHT);
+		top_bottom = coda_read(dev, CODA_RET_DEC_SEQ_CROP_TOP_BOTTOM);
+
+		q_data_dst->rect.left = (left_right >> 10) & 0x3ff;
+		q_data_dst->rect.top = (top_bottom >> 10) & 0x3ff;
+		q_data_dst->rect.width = width - q_data_dst->rect.left -
+					 (left_right & 0x3ff);
+		q_data_dst->rect.height = height - q_data_dst->rect.top -
+					  (top_bottom & 0x3ff);
+	}
+
+	if (dev->devtype->product != CODA_DX6) {
+		u8 profile, level;
+
+		val = coda_read(dev, CODA7_RET_DEC_SEQ_HEADER_REPORT);
+		profile = val & 0xff;
+		level = (val >> 8) & 0x7f;
+
+		if (profile || level)
+			coda_update_profile_level_ctrls(ctx, profile, level);
+	}
+
+	return 0;
+}
+
+static void coda_dec_seq_init_work(struct work_struct *work)
+{
+	struct coda_ctx *ctx = container_of(work,
+					    struct coda_ctx, seq_init_work);
+	struct coda_dev *dev = ctx->dev;
+
+	mutex_lock(&ctx->buffer_mutex);
+	mutex_lock(&dev->coda_mutex);
+
+	if (!ctx->initialized)
+		__coda_decoder_seq_init(ctx);
+
+	mutex_unlock(&dev->coda_mutex);
+	mutex_unlock(&ctx->buffer_mutex);
+}
+
+static int __coda_start_decoding(struct coda_ctx *ctx)
+{
+	struct coda_q_data *q_data_src, *q_data_dst;
+	struct coda_dev *dev = ctx->dev;
+	u32 src_fourcc, dst_fourcc;
+	int ret;
+
+	q_data_src = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+	q_data_dst = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_CAPTURE);
+	src_fourcc = q_data_src->fourcc;
+	dst_fourcc = q_data_dst->fourcc;
+
+	if (!ctx->initialized) {
+		ret = __coda_decoder_seq_init(ctx);
+		if (ret < 0)
+			return ret;
+	} else {
+		ctx->frame_mem_ctrl &= ~(CODA_FRAME_CHROMA_INTERLEAVE | (0x3 << 9) |
+					 CODA9_FRAME_TILED2LINEAR);
+		if (dst_fourcc == V4L2_PIX_FMT_NV12 || dst_fourcc == V4L2_PIX_FMT_YUYV)
+			ctx->frame_mem_ctrl |= CODA_FRAME_CHROMA_INTERLEAVE;
+		if (ctx->tiled_map_type == GDI_TILED_FRAME_MB_RASTER_MAP)
+			ctx->frame_mem_ctrl |= (0x3 << 9) |
+				((ctx->use_vdoa) ? 0 : CODA9_FRAME_TILED2LINEAR);
+	}
+
+	coda_write(dev, ctx->parabuf.paddr, CODA_REG_BIT_PARA_BUF_ADDR);
+
+	ret = coda_alloc_framebuffers(ctx, q_data_dst, src_fourcc);
+	if (ret < 0) {
+		v4l2_err(&dev->v4l2_dev, "failed to allocate framebuffers\n");
+		return ret;
+	}
+
+	/* Tell the decoder how many frame buffers we allocated. */
+	coda_write(dev, ctx->num_internal_frames, CODA_CMD_SET_FRAME_BUF_NUM);
+	coda_write(dev, round_up(q_data_dst->rect.width, 16),
+		   CODA_CMD_SET_FRAME_BUF_STRIDE);
+
+	if (dev->devtype->product != CODA_DX6) {
+		/* Set secondary AXI IRAM */
+		coda_setup_iram(ctx);
+
+		coda_write(dev, ctx->iram_info.buf_bit_use,
+				CODA7_CMD_SET_FRAME_AXI_BIT_ADDR);
+		coda_write(dev, ctx->iram_info.buf_ip_ac_dc_use,
+				CODA7_CMD_SET_FRAME_AXI_IPACDC_ADDR);
+		coda_write(dev, ctx->iram_info.buf_dbk_y_use,
+				CODA7_CMD_SET_FRAME_AXI_DBKY_ADDR);
+		coda_write(dev, ctx->iram_info.buf_dbk_c_use,
+				CODA7_CMD_SET_FRAME_AXI_DBKC_ADDR);
+		coda_write(dev, ctx->iram_info.buf_ovl_use,
+				CODA7_CMD_SET_FRAME_AXI_OVL_ADDR);
+		if (dev->devtype->product == CODA_960) {
+			coda_write(dev, ctx->iram_info.buf_btp_use,
+					CODA9_CMD_SET_FRAME_AXI_BTP_ADDR);
+
+			coda_write(dev, -1, CODA9_CMD_SET_FRAME_DELAY);
+			coda9_set_frame_cache(ctx, dst_fourcc);
+		}
+	}
+
+	if (src_fourcc == V4L2_PIX_FMT_H264) {
+		coda_write(dev, ctx->slicebuf.paddr,
+				CODA_CMD_SET_FRAME_SLICE_BB_START);
+		coda_write(dev, ctx->slicebuf.size / 1024,
+				CODA_CMD_SET_FRAME_SLICE_BB_SIZE);
+	}
+
+	if (dev->devtype->product == CODA_HX4 ||
+	    dev->devtype->product == CODA_7541) {
+		int max_mb_x = 1920 / 16;
+		int max_mb_y = 1088 / 16;
+		int max_mb_num = max_mb_x * max_mb_y;
+
+		coda_write(dev, max_mb_num << 16 | max_mb_x << 8 | max_mb_y,
+				CODA7_CMD_SET_FRAME_MAX_DEC_SIZE);
+	} else if (dev->devtype->product == CODA_960) {
+		int max_mb_x = 1920 / 16;
+		int max_mb_y = 1088 / 16;
+		int max_mb_num = max_mb_x * max_mb_y;
+
+		coda_write(dev, max_mb_num << 16 | max_mb_x << 8 | max_mb_y,
+				CODA9_CMD_SET_FRAME_MAX_DEC_SIZE);
+	}
+
+	if (coda_command_sync(ctx, CODA_COMMAND_SET_FRAME_BUF)) {
+		v4l2_err(&ctx->dev->v4l2_dev,
+			 "CODA_COMMAND_SET_FRAME_BUF timeout\n");
+		return -ETIMEDOUT;
+	}
+
+	return 0;
+}
+
+static int coda_start_decoding(struct coda_ctx *ctx)
+{
+	struct coda_dev *dev = ctx->dev;
+	int ret;
+
+	mutex_lock(&dev->coda_mutex);
+	ret = __coda_start_decoding(ctx);
+	mutex_unlock(&dev->coda_mutex);
+
+	return ret;
+}
+
+static int coda_prepare_decode(struct coda_ctx *ctx)
+{
+	struct vb2_v4l2_buffer *dst_buf;
+	struct coda_dev *dev = ctx->dev;
+	struct coda_q_data *q_data_dst;
+	struct coda_buffer_meta *meta;
+	u32 rot_mode = 0;
+	u32 reg_addr, reg_stride;
+
+	dst_buf = v4l2_m2m_next_dst_buf(ctx->fh.m2m_ctx);
+	q_data_dst = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_CAPTURE);
+
+	/* Try to copy source buffer contents into the bitstream ringbuffer */
+	mutex_lock(&ctx->bitstream_mutex);
+	coda_fill_bitstream(ctx, NULL);
+	mutex_unlock(&ctx->bitstream_mutex);
+
+	if (coda_get_bitstream_payload(ctx) < 512 &&
+	    (!(ctx->bit_stream_param & CODA_BIT_STREAM_END_FLAG))) {
+		coda_dbg(1, ctx, "bitstream payload: %d, skipping\n",
+			 coda_get_bitstream_payload(ctx));
+		return -EAGAIN;
+	}
+
+	/* Run coda_start_decoding (again) if not yet initialized */
+	if (!ctx->initialized) {
+		int ret = __coda_start_decoding(ctx);
+
+		if (ret < 0) {
+			v4l2_err(&dev->v4l2_dev, "failed to start decoding\n");
+			return -EAGAIN;
+		} else {
+			ctx->initialized = 1;
+		}
+	}
+
+	if (dev->devtype->product == CODA_960)
+		coda_set_gdi_regs(ctx);
+
+	if (ctx->use_vdoa &&
+	    ctx->display_idx >= 0 &&
+	    ctx->display_idx < ctx->num_internal_frames) {
+		vdoa_device_run(ctx->vdoa,
+				vb2_dma_contig_plane_dma_addr(&dst_buf->vb2_buf, 0),
+				ctx->internal_frames[ctx->display_idx].buf.paddr);
+	} else {
+		if (dev->devtype->product == CODA_960) {
+			/*
+			 * It was previously assumed that the CODA960 has an
+			 * internal list of 64 buffer entries that contains
+			 * both the registered internal frame buffers as well
+			 * as the rotator buffer output, and that the ROT_INDEX
+			 * register must be set to a value between the last
+			 * internal frame buffers' index and 64.
+			 * At least on firmware version 3.1.1 it turns out that
+			 * setting ROT_INDEX to any value >= 32 causes CODA
+			 * hangups that it can not recover from with the SRC VPU
+			 * reset.
+			 * It does appear to work however, to just set it to a
+			 * fixed value in the [ctx->num_internal_frames, 31]
+			 * range, for example CODA_MAX_FRAMEBUFFERS.
+			 */
+			coda_write(dev, CODA_MAX_FRAMEBUFFERS,
+				   CODA9_CMD_DEC_PIC_ROT_INDEX);
+
+			reg_addr = CODA9_CMD_DEC_PIC_ROT_ADDR_Y;
+			reg_stride = CODA9_CMD_DEC_PIC_ROT_STRIDE;
+		} else {
+			reg_addr = CODA_CMD_DEC_PIC_ROT_ADDR_Y;
+			reg_stride = CODA_CMD_DEC_PIC_ROT_STRIDE;
+		}
+		coda_write_base(ctx, q_data_dst, dst_buf, reg_addr);
+		coda_write(dev, q_data_dst->bytesperline, reg_stride);
+
+		rot_mode = CODA_ROT_MIR_ENABLE | ctx->params.rot_mode;
+	}
+
+	coda_write(dev, rot_mode, CODA_CMD_DEC_PIC_ROT_MODE);
+
+	switch (dev->devtype->product) {
+	case CODA_DX6:
+		/* TBD */
+	case CODA_HX4:
+	case CODA_7541:
+		coda_write(dev, CODA_PRE_SCAN_EN, CODA_CMD_DEC_PIC_OPTION);
+		break;
+	case CODA_960:
+		/* 'hardcode to use interrupt disable mode'? */
+		coda_write(dev, (1 << 10), CODA_CMD_DEC_PIC_OPTION);
+		break;
+	}
+
+	coda_write(dev, 0, CODA_CMD_DEC_PIC_SKIP_NUM);
+
+	coda_write(dev, 0, CODA_CMD_DEC_PIC_BB_START);
+	coda_write(dev, 0, CODA_CMD_DEC_PIC_START_BYTE);
+
+	if (dev->devtype->product != CODA_DX6)
+		coda_write(dev, ctx->iram_info.axi_sram_use,
+				CODA7_REG_BIT_AXI_SRAM_USE);
+
+	spin_lock(&ctx->buffer_meta_lock);
+	meta = list_first_entry_or_null(&ctx->buffer_meta_list,
+					struct coda_buffer_meta, list);
+
+	if (meta && ctx->codec->src_fourcc == V4L2_PIX_FMT_JPEG) {
+
+		/* If this is the last buffer in the bitstream, add padding */
+		if (meta->end == ctx->bitstream_fifo.kfifo.in) {
+			static unsigned char buf[512];
+			unsigned int pad;
+
+			/* Pad to multiple of 256 and then add 256 more */
+			pad = ((0 - meta->end) & 0xff) + 256;
+
+			memset(buf, 0xff, sizeof(buf));
+
+			kfifo_in(&ctx->bitstream_fifo, buf, pad);
+		}
+	}
+	spin_unlock(&ctx->buffer_meta_lock);
+
+	coda_kfifo_sync_to_device_full(ctx);
+
+	/* Clear decode success flag */
+	coda_write(dev, 0, CODA_RET_DEC_PIC_SUCCESS);
+
+	/* Clear error return value */
+	coda_write(dev, 0, CODA_RET_DEC_PIC_ERR_MB);
+
+	trace_coda_dec_pic_run(ctx, meta);
+
+	coda_command_async(ctx, CODA_COMMAND_PIC_RUN);
+
+	return 0;
+}
+
+static void coda_finish_decode(struct coda_ctx *ctx)
+{
+	struct coda_dev *dev = ctx->dev;
+	struct coda_q_data *q_data_src;
+	struct coda_q_data *q_data_dst;
+	struct vb2_v4l2_buffer *dst_buf;
+	struct coda_buffer_meta *meta;
+	int width, height;
+	int decoded_idx;
+	int display_idx;
+	struct coda_internal_frame *decoded_frame = NULL;
+	u32 src_fourcc;
+	int success;
+	u32 err_mb;
+	int err_vdoa = 0;
+	u32 val;
+
+	if (ctx->aborting)
+		return;
+
+	/* Update kfifo out pointer from coda bitstream read pointer */
+	coda_kfifo_sync_from_device(ctx);
+
+	/*
+	 * in stream-end mode, the read pointer can overshoot the write pointer
+	 * by up to 512 bytes
+	 */
+	if (ctx->bit_stream_param & CODA_BIT_STREAM_END_FLAG) {
+		if (coda_get_bitstream_payload(ctx) >= ctx->bitstream.size - 512)
+			kfifo_init(&ctx->bitstream_fifo,
+				ctx->bitstream.vaddr, ctx->bitstream.size);
+	}
+
+	q_data_src = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+	src_fourcc = q_data_src->fourcc;
+
+	val = coda_read(dev, CODA_RET_DEC_PIC_SUCCESS);
+	if (val != 1)
+		pr_err("DEC_PIC_SUCCESS = %d\n", val);
+
+	success = val & 0x1;
+	if (!success)
+		v4l2_err(&dev->v4l2_dev, "decode failed\n");
+
+	if (src_fourcc == V4L2_PIX_FMT_H264) {
+		if (val & (1 << 3))
+			v4l2_err(&dev->v4l2_dev,
+				 "insufficient PS buffer space (%d bytes)\n",
+				 ctx->psbuf.size);
+		if (val & (1 << 2))
+			v4l2_err(&dev->v4l2_dev,
+				 "insufficient slice buffer space (%d bytes)\n",
+				 ctx->slicebuf.size);
+	}
+
+	val = coda_read(dev, CODA_RET_DEC_PIC_SIZE);
+	width = (val >> 16) & 0xffff;
+	height = val & 0xffff;
+
+	q_data_dst = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_CAPTURE);
+
+	/* frame crop information */
+	if (src_fourcc == V4L2_PIX_FMT_H264) {
+		u32 left_right;
+		u32 top_bottom;
+
+		left_right = coda_read(dev, CODA_RET_DEC_PIC_CROP_LEFT_RIGHT);
+		top_bottom = coda_read(dev, CODA_RET_DEC_PIC_CROP_TOP_BOTTOM);
+
+		if (left_right == 0xffffffff && top_bottom == 0xffffffff) {
+			/* Keep current crop information */
+		} else {
+			struct v4l2_rect *rect = &q_data_dst->rect;
+
+			rect->left = left_right >> 16 & 0xffff;
+			rect->top = top_bottom >> 16 & 0xffff;
+			rect->width = width - rect->left -
+				      (left_right & 0xffff);
+			rect->height = height - rect->top -
+				       (top_bottom & 0xffff);
+		}
+	} else {
+		/* no cropping */
+	}
+
+	err_mb = coda_read(dev, CODA_RET_DEC_PIC_ERR_MB);
+	if (err_mb > 0) {
+		if (__ratelimit(&dev->mb_err_rs))
+			coda_dbg(1, ctx, "errors in %d macroblocks\n", err_mb);
+		v4l2_ctrl_s_ctrl(ctx->mb_err_cnt_ctrl,
+				 v4l2_ctrl_g_ctrl(ctx->mb_err_cnt_ctrl) + err_mb);
+	}
+
+	if (dev->devtype->product == CODA_HX4 ||
+	    dev->devtype->product == CODA_7541) {
+		val = coda_read(dev, CODA_RET_DEC_PIC_OPTION);
+		if (val == 0) {
+			/* not enough bitstream data */
+			coda_dbg(1, ctx, "prescan failed: %d\n", val);
+			ctx->hold = true;
+			return;
+		}
+	}
+
+	/* Wait until the VDOA finished writing the previous display frame */
+	if (ctx->use_vdoa &&
+	    ctx->display_idx >= 0 &&
+	    ctx->display_idx < ctx->num_internal_frames) {
+		err_vdoa = vdoa_wait_for_completion(ctx->vdoa);
+	}
+
+	ctx->frm_dis_flg = coda_read(dev,
+				     CODA_REG_BIT_FRM_DIS_FLG(ctx->reg_idx));
+
+	/* The previous display frame was copied out and can be overwritten */
+	if (ctx->display_idx >= 0 &&
+	    ctx->display_idx < ctx->num_internal_frames) {
+		ctx->frm_dis_flg &= ~(1 << ctx->display_idx);
+		coda_write(dev, ctx->frm_dis_flg,
+				CODA_REG_BIT_FRM_DIS_FLG(ctx->reg_idx));
+	}
+
+	/*
+	 * The index of the last decoded frame, not necessarily in
+	 * display order, and the index of the next display frame.
+	 * The latter could have been decoded in a previous run.
+	 */
+	decoded_idx = coda_read(dev, CODA_RET_DEC_PIC_CUR_IDX);
+	display_idx = coda_read(dev, CODA_RET_DEC_PIC_FRAME_IDX);
+
+	if (decoded_idx == -1) {
+		/* no frame was decoded, but we might have a display frame */
+		if (display_idx >= 0 && display_idx < ctx->num_internal_frames)
+			ctx->sequence_offset++;
+		else if (ctx->display_idx < 0)
+			ctx->hold = true;
+	} else if (decoded_idx == -2) {
+		if (ctx->display_idx >= 0 &&
+		    ctx->display_idx < ctx->num_internal_frames)
+			ctx->sequence_offset++;
+		/* no frame was decoded, we still return remaining buffers */
+	} else if (decoded_idx < 0 || decoded_idx >= ctx->num_internal_frames) {
+		v4l2_err(&dev->v4l2_dev,
+			 "decoded frame index out of range: %d\n", decoded_idx);
+	} else {
+		int sequence;
+
+		decoded_frame = &ctx->internal_frames[decoded_idx];
+
+		val = coda_read(dev, CODA_RET_DEC_PIC_FRAME_NUM);
+		if (ctx->sequence_offset == -1)
+			ctx->sequence_offset = val;
+
+		sequence = val + ctx->first_frame_sequence
+			       - ctx->sequence_offset;
+		spin_lock(&ctx->buffer_meta_lock);
+		if (!list_empty(&ctx->buffer_meta_list)) {
+			meta = list_first_entry(&ctx->buffer_meta_list,
+					      struct coda_buffer_meta, list);
+			list_del(&meta->list);
+			ctx->num_metas--;
+			spin_unlock(&ctx->buffer_meta_lock);
+			/*
+			 * Clamp counters to 16 bits for comparison, as the HW
+			 * counter rolls over at this point for h.264. This
+			 * may be different for other formats, but using 16 bits
+			 * should be enough to detect most errors and saves us
+			 * from doing different things based on the format.
+			 */
+			if ((sequence & 0xffff) != (meta->sequence & 0xffff)) {
+				v4l2_err(&dev->v4l2_dev,
+					 "sequence number mismatch (%d(%d) != %d)\n",
+					 sequence, ctx->sequence_offset,
+					 meta->sequence);
+			}
+			decoded_frame->meta = *meta;
+			kfree(meta);
+		} else {
+			spin_unlock(&ctx->buffer_meta_lock);
+			v4l2_err(&dev->v4l2_dev, "empty timestamp list!\n");
+			memset(&decoded_frame->meta, 0,
+			       sizeof(struct coda_buffer_meta));
+			decoded_frame->meta.sequence = sequence;
+			decoded_frame->meta.last = false;
+			ctx->sequence_offset++;
+		}
+
+		trace_coda_dec_pic_done(ctx, &decoded_frame->meta);
+
+		val = coda_read(dev, CODA_RET_DEC_PIC_TYPE) & 0x7;
+		decoded_frame->type = (val == 0) ? V4L2_BUF_FLAG_KEYFRAME :
+				      (val == 1) ? V4L2_BUF_FLAG_PFRAME :
+						   V4L2_BUF_FLAG_BFRAME;
+
+		decoded_frame->error = err_mb;
+	}
+
+	if (display_idx == -1) {
+		/*
+		 * no more frames to be decoded, but there could still
+		 * be rotator output to dequeue
+		 */
+		ctx->hold = true;
+	} else if (display_idx == -3) {
+		/* possibly prescan failure */
+	} else if (display_idx < 0 || display_idx >= ctx->num_internal_frames) {
+		v4l2_err(&dev->v4l2_dev,
+			 "presentation frame index out of range: %d\n",
+			 display_idx);
+	}
+
+	/* If a frame was copied out, return it */
+	if (ctx->display_idx >= 0 &&
+	    ctx->display_idx < ctx->num_internal_frames) {
+		struct coda_internal_frame *ready_frame;
+
+		ready_frame = &ctx->internal_frames[ctx->display_idx];
+
+		dst_buf = v4l2_m2m_dst_buf_remove(ctx->fh.m2m_ctx);
+		dst_buf->sequence = ctx->osequence++;
+
+		dst_buf->field = V4L2_FIELD_NONE;
+		dst_buf->flags &= ~(V4L2_BUF_FLAG_KEYFRAME |
+					     V4L2_BUF_FLAG_PFRAME |
+					     V4L2_BUF_FLAG_BFRAME);
+		dst_buf->flags |= ready_frame->type;
+		meta = &ready_frame->meta;
+		if (meta->last && !coda_reorder_enable(ctx)) {
+			/*
+			 * If this was the last decoded frame, and reordering
+			 * is disabled, this will be the last display frame.
+			 */
+			coda_dbg(1, ctx, "last meta, marking as last frame\n");
+			dst_buf->flags |= V4L2_BUF_FLAG_LAST;
+		} else if (ctx->bit_stream_param & CODA_BIT_STREAM_END_FLAG &&
+			   display_idx == -1) {
+			/*
+			 * If there is no designated presentation frame anymore,
+			 * this frame has to be the last one.
+			 */
+			coda_dbg(1, ctx,
+				 "no more frames to return, marking as last frame\n");
+			dst_buf->flags |= V4L2_BUF_FLAG_LAST;
+		}
+		dst_buf->timecode = meta->timecode;
+		dst_buf->vb2_buf.timestamp = meta->timestamp;
+
+		trace_coda_dec_rot_done(ctx, dst_buf, meta);
+
+		vb2_set_plane_payload(&dst_buf->vb2_buf, 0,
+				      q_data_dst->sizeimage);
+
+		if (ready_frame->error || err_vdoa)
+			coda_m2m_buf_done(ctx, dst_buf, VB2_BUF_STATE_ERROR);
+		else
+			coda_m2m_buf_done(ctx, dst_buf, VB2_BUF_STATE_DONE);
+
+		if (decoded_frame) {
+			coda_dbg(1, ctx, "job finished: decoded %c frame %u, returned %c frame %u (%u/%u)%s\n",
+				 coda_frame_type_char(decoded_frame->type),
+				 decoded_frame->meta.sequence,
+				 coda_frame_type_char(dst_buf->flags),
+				 ready_frame->meta.sequence,
+				 dst_buf->sequence, ctx->qsequence,
+				 (dst_buf->flags & V4L2_BUF_FLAG_LAST) ?
+				 " (last)" : "");
+		} else {
+			coda_dbg(1, ctx, "job finished: no frame decoded (%d), returned %c frame %u (%u/%u)%s\n",
+				 decoded_idx,
+				 coda_frame_type_char(dst_buf->flags),
+				 ready_frame->meta.sequence,
+				 dst_buf->sequence, ctx->qsequence,
+				 (dst_buf->flags & V4L2_BUF_FLAG_LAST) ?
+				 " (last)" : "");
+		}
+	} else {
+		if (decoded_frame) {
+			coda_dbg(1, ctx, "job finished: decoded %c frame %u, no frame returned (%d)\n",
+				 coda_frame_type_char(decoded_frame->type),
+				 decoded_frame->meta.sequence,
+				 ctx->display_idx);
+		} else {
+			coda_dbg(1, ctx, "job finished: no frame decoded (%d) or returned (%d)\n",
+				 decoded_idx, ctx->display_idx);
+		}
+	}
+
+	/* The rotator will copy the current display frame next time */
+	ctx->display_idx = display_idx;
+
+	/*
+	 * The current decode run might have brought the bitstream fill level
+	 * below the size where we can start the next decode run. As userspace
+	 * might have filled the output queue completely and might thus be
+	 * blocked, we can't rely on the next qbuf to trigger the bitstream
+	 * refill. Check if we have data to refill the bitstream now.
+	 */
+	mutex_lock(&ctx->bitstream_mutex);
+	coda_fill_bitstream(ctx, NULL);
+	mutex_unlock(&ctx->bitstream_mutex);
+}
+
+static void coda_decode_timeout(struct coda_ctx *ctx)
+{
+	struct vb2_v4l2_buffer *dst_buf;
+
+	/*
+	 * For now this only handles the case where we would deadlock with
+	 * userspace, i.e. userspace issued DEC_CMD_STOP and waits for EOS,
+	 * but after a failed decode run we would hold the context and wait for
+	 * userspace to queue more buffers.
+	 */
+	if (!(ctx->bit_stream_param & CODA_BIT_STREAM_END_FLAG))
+		return;
+
+	dst_buf = v4l2_m2m_dst_buf_remove(ctx->fh.m2m_ctx);
+	dst_buf->sequence = ctx->qsequence - 1;
+
+	coda_m2m_buf_done(ctx, dst_buf, VB2_BUF_STATE_ERROR);
+}
+
+const struct coda_context_ops coda_bit_decode_ops = {
+	.queue_init = coda_decoder_queue_init,
+	.reqbufs = coda_decoder_reqbufs,
+	.start_streaming = coda_start_decoding,
+	.prepare_run = coda_prepare_decode,
+	.finish_run = coda_finish_decode,
+	.run_timeout = coda_decode_timeout,
+	.seq_init_work = coda_dec_seq_init_work,
+	.seq_end_work = coda_seq_end_work,
+	.release = coda_bit_release,
+};
+
+irqreturn_t coda_irq_handler(int irq, void *data)
+{
+	struct coda_dev *dev = data;
+	struct coda_ctx *ctx;
+
+	/* read status register to attend the IRQ */
+	coda_read(dev, CODA_REG_BIT_INT_STATUS);
+	coda_write(dev, 0, CODA_REG_BIT_INT_REASON);
+	coda_write(dev, CODA_REG_BIT_INT_CLEAR_SET,
+		      CODA_REG_BIT_INT_CLEAR);
+
+	ctx = v4l2_m2m_get_curr_priv(dev->m2m_dev);
+	if (ctx == NULL) {
+		v4l2_err(&dev->v4l2_dev,
+			 "Instance released before the end of transaction\n");
+		return IRQ_HANDLED;
+	}
+
+	trace_coda_bit_done(ctx);
+
+	if (ctx->aborting) {
+		coda_dbg(1, ctx, "task has been aborted\n");
+	}
+
+	if (coda_isbusy(ctx->dev)) {
+		coda_dbg(1, ctx, "coda is still busy!!!!\n");
+		return IRQ_NONE;
+	}
+
+	complete(&ctx->completion);
+
+	return IRQ_HANDLED;
+}
diff --git a/drivers/media/platform/chips-media/coda-common.c b/drivers/media/platform/chips-media/coda-common.c
new file mode 100644
index 000000000..af71eea04
--- /dev/null
+++ b/drivers/media/platform/chips-media/coda-common.c
@@ -0,0 +1,3362 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Coda multi-standard codec IP
+ *
+ * Copyright (C) 2012 Vista Silicon S.L.
+ *    Javier Martin, <javier.martin@vista-silicon.com>
+ *    Xavier Duret
+ */
+
+#include <linux/clk.h>
+#include <linux/debugfs.h>
+#include <linux/delay.h>
+#include <linux/firmware.h>
+#include <linux/gcd.h>
+#include <linux/genalloc.h>
+#include <linux/idr.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/irq.h>
+#include <linux/kfifo.h>
+#include <linux/module.h>
+#include <linux/of_device.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/slab.h>
+#include <linux/videodev2.h>
+#include <linux/of.h>
+#include <linux/ratelimit.h>
+#include <linux/reset.h>
+
+#include <media/v4l2-ctrls.h>
+#include <media/v4l2-device.h>
+#include <media/v4l2-event.h>
+#include <media/v4l2-ioctl.h>
+#include <media/v4l2-mem2mem.h>
+#include <media/videobuf2-v4l2.h>
+#include <media/videobuf2-dma-contig.h>
+#include <media/videobuf2-vmalloc.h>
+
+#include "coda.h"
+#include "imx-vdoa.h"
+
+#define CODA_NAME		"coda"
+
+#define CODADX6_MAX_INSTANCES	4
+#define CODA_MAX_FORMATS	5
+
+#define CODA_ISRAM_SIZE	(2048 * 2)
+
+#define MIN_W 48
+#define MIN_H 16
+
+#define S_ALIGN		1 /* multiple of 2 */
+#define W_ALIGN		1 /* multiple of 2 */
+#define H_ALIGN		1 /* multiple of 2 */
+
+#define fh_to_ctx(__fh)	container_of(__fh, struct coda_ctx, fh)
+
+int coda_debug;
+module_param(coda_debug, int, 0644);
+MODULE_PARM_DESC(coda_debug, "Debug level (0-2)");
+
+static int disable_tiling;
+module_param(disable_tiling, int, 0644);
+MODULE_PARM_DESC(disable_tiling, "Disable tiled frame buffers");
+
+static int disable_vdoa;
+module_param(disable_vdoa, int, 0644);
+MODULE_PARM_DESC(disable_vdoa, "Disable Video Data Order Adapter tiled to raster-scan conversion");
+
+static int enable_bwb = 0;
+module_param(enable_bwb, int, 0644);
+MODULE_PARM_DESC(enable_bwb, "Enable BWB unit for decoding, may crash on certain streams");
+
+void coda_write(struct coda_dev *dev, u32 data, u32 reg)
+{
+	v4l2_dbg(3, coda_debug, &dev->v4l2_dev,
+		 "%s: data=0x%x, reg=0x%x\n", __func__, data, reg);
+	writel(data, dev->regs_base + reg);
+}
+
+unsigned int coda_read(struct coda_dev *dev, u32 reg)
+{
+	u32 data;
+
+	data = readl(dev->regs_base + reg);
+	v4l2_dbg(3, coda_debug, &dev->v4l2_dev,
+		 "%s: data=0x%x, reg=0x%x\n", __func__, data, reg);
+	return data;
+}
+
+void coda_write_base(struct coda_ctx *ctx, struct coda_q_data *q_data,
+		     struct vb2_v4l2_buffer *buf, unsigned int reg_y)
+{
+	u32 base_y = vb2_dma_contig_plane_dma_addr(&buf->vb2_buf, 0);
+	u32 base_cb, base_cr;
+
+	switch (q_data->fourcc) {
+	case V4L2_PIX_FMT_YUYV:
+		/* Fallthrough: IN -H264-> CODA -NV12 MB-> VDOA -YUYV-> OUT */
+	case V4L2_PIX_FMT_NV12:
+	case V4L2_PIX_FMT_YUV420:
+	default:
+		base_cb = base_y + q_data->bytesperline * q_data->height;
+		base_cr = base_cb + q_data->bytesperline * q_data->height / 4;
+		break;
+	case V4L2_PIX_FMT_YVU420:
+		/* Switch Cb and Cr for YVU420 format */
+		base_cr = base_y + q_data->bytesperline * q_data->height;
+		base_cb = base_cr + q_data->bytesperline * q_data->height / 4;
+		break;
+	case V4L2_PIX_FMT_YUV422P:
+		base_cb = base_y + q_data->bytesperline * q_data->height;
+		base_cr = base_cb + q_data->bytesperline * q_data->height / 2;
+	}
+
+	coda_write(ctx->dev, base_y, reg_y);
+	coda_write(ctx->dev, base_cb, reg_y + 4);
+	coda_write(ctx->dev, base_cr, reg_y + 8);
+}
+
+#define CODA_CODEC(mode, src_fourcc, dst_fourcc, max_w, max_h) \
+	{ mode, src_fourcc, dst_fourcc, max_w, max_h }
+
+/*
+ * Arrays of codecs supported by each given version of Coda:
+ *  i.MX27 -> codadx6
+ *  i.MX51 -> codahx4
+ *  i.MX53 -> coda7
+ *  i.MX6  -> coda960
+ * Use V4L2_PIX_FMT_YUV420 as placeholder for all supported YUV 4:2:0 variants
+ */
+static const struct coda_codec codadx6_codecs[] = {
+	CODA_CODEC(CODADX6_MODE_ENCODE_H264, V4L2_PIX_FMT_YUV420, V4L2_PIX_FMT_H264,  720, 576),
+	CODA_CODEC(CODADX6_MODE_ENCODE_MP4,  V4L2_PIX_FMT_YUV420, V4L2_PIX_FMT_MPEG4, 720, 576),
+};
+
+static const struct coda_codec codahx4_codecs[] = {
+	CODA_CODEC(CODA7_MODE_ENCODE_H264, V4L2_PIX_FMT_YUV420, V4L2_PIX_FMT_H264,   720, 576),
+	CODA_CODEC(CODA7_MODE_DECODE_H264, V4L2_PIX_FMT_H264,   V4L2_PIX_FMT_YUV420, 1920, 1088),
+	CODA_CODEC(CODA7_MODE_DECODE_MP2,  V4L2_PIX_FMT_MPEG2,  V4L2_PIX_FMT_YUV420, 1920, 1088),
+	CODA_CODEC(CODA7_MODE_DECODE_MP4,  V4L2_PIX_FMT_MPEG4,  V4L2_PIX_FMT_YUV420, 1280, 720),
+};
+
+static const struct coda_codec coda7_codecs[] = {
+	CODA_CODEC(CODA7_MODE_ENCODE_H264, V4L2_PIX_FMT_YUV420, V4L2_PIX_FMT_H264,   1280, 720),
+	CODA_CODEC(CODA7_MODE_ENCODE_MP4,  V4L2_PIX_FMT_YUV420, V4L2_PIX_FMT_MPEG4,  1280, 720),
+	CODA_CODEC(CODA7_MODE_ENCODE_MJPG, V4L2_PIX_FMT_YUV420, V4L2_PIX_FMT_JPEG,   8192, 8192),
+	CODA_CODEC(CODA7_MODE_DECODE_H264, V4L2_PIX_FMT_H264,   V4L2_PIX_FMT_YUV420, 1920, 1088),
+	CODA_CODEC(CODA7_MODE_DECODE_MP2,  V4L2_PIX_FMT_MPEG2,  V4L2_PIX_FMT_YUV420, 1920, 1088),
+	CODA_CODEC(CODA7_MODE_DECODE_MP4,  V4L2_PIX_FMT_MPEG4,  V4L2_PIX_FMT_YUV420, 1920, 1088),
+	CODA_CODEC(CODA7_MODE_DECODE_MJPG, V4L2_PIX_FMT_JPEG,   V4L2_PIX_FMT_YUV420, 8192, 8192),
+};
+
+static const struct coda_codec coda9_codecs[] = {
+	CODA_CODEC(CODA9_MODE_ENCODE_H264, V4L2_PIX_FMT_YUV420, V4L2_PIX_FMT_H264,   1920, 1088),
+	CODA_CODEC(CODA9_MODE_ENCODE_MP4,  V4L2_PIX_FMT_YUV420, V4L2_PIX_FMT_MPEG4,  1920, 1088),
+	CODA_CODEC(CODA9_MODE_ENCODE_MJPG, V4L2_PIX_FMT_YUV420, V4L2_PIX_FMT_JPEG,   8192, 8192),
+	CODA_CODEC(CODA9_MODE_DECODE_H264, V4L2_PIX_FMT_H264,   V4L2_PIX_FMT_YUV420, 1920, 1088),
+	CODA_CODEC(CODA9_MODE_DECODE_MP2,  V4L2_PIX_FMT_MPEG2,  V4L2_PIX_FMT_YUV420, 1920, 1088),
+	CODA_CODEC(CODA9_MODE_DECODE_MP4,  V4L2_PIX_FMT_MPEG4,  V4L2_PIX_FMT_YUV420, 1920, 1088),
+	CODA_CODEC(CODA9_MODE_DECODE_MJPG, V4L2_PIX_FMT_JPEG,	V4L2_PIX_FMT_YUV420, 8192, 8192),
+};
+
+struct coda_video_device {
+	const char *name;
+	enum coda_inst_type type;
+	const struct coda_context_ops *ops;
+	bool direct;
+	u32 src_formats[CODA_MAX_FORMATS];
+	u32 dst_formats[CODA_MAX_FORMATS];
+};
+
+static const struct coda_video_device coda_bit_encoder = {
+	.name = "coda-video-encoder",
+	.type = CODA_INST_ENCODER,
+	.ops = &coda_bit_encode_ops,
+	.src_formats = {
+		V4L2_PIX_FMT_NV12,
+		V4L2_PIX_FMT_YUV420,
+		V4L2_PIX_FMT_YVU420,
+	},
+	.dst_formats = {
+		V4L2_PIX_FMT_H264,
+		V4L2_PIX_FMT_MPEG4,
+	},
+};
+
+static const struct coda_video_device coda_bit_jpeg_encoder = {
+	.name = "coda-jpeg-encoder",
+	.type = CODA_INST_ENCODER,
+	.ops = &coda_bit_encode_ops,
+	.src_formats = {
+		V4L2_PIX_FMT_NV12,
+		V4L2_PIX_FMT_YUV420,
+		V4L2_PIX_FMT_YVU420,
+		V4L2_PIX_FMT_YUV422P,
+	},
+	.dst_formats = {
+		V4L2_PIX_FMT_JPEG,
+	},
+};
+
+static const struct coda_video_device coda_bit_decoder = {
+	.name = "coda-video-decoder",
+	.type = CODA_INST_DECODER,
+	.ops = &coda_bit_decode_ops,
+	.src_formats = {
+		V4L2_PIX_FMT_H264,
+		V4L2_PIX_FMT_MPEG2,
+		V4L2_PIX_FMT_MPEG4,
+	},
+	.dst_formats = {
+		V4L2_PIX_FMT_NV12,
+		V4L2_PIX_FMT_YUV420,
+		V4L2_PIX_FMT_YVU420,
+		/*
+		 * If V4L2_PIX_FMT_YUYV should be default,
+		 * set_default_params() must be adjusted.
+		 */
+		V4L2_PIX_FMT_YUYV,
+	},
+};
+
+static const struct coda_video_device coda_bit_jpeg_decoder = {
+	.name = "coda-jpeg-decoder",
+	.type = CODA_INST_DECODER,
+	.ops = &coda_bit_decode_ops,
+	.src_formats = {
+		V4L2_PIX_FMT_JPEG,
+	},
+	.dst_formats = {
+		V4L2_PIX_FMT_NV12,
+		V4L2_PIX_FMT_YUV420,
+		V4L2_PIX_FMT_YVU420,
+		V4L2_PIX_FMT_YUV422P,
+	},
+};
+
+static const struct coda_video_device coda9_jpeg_encoder = {
+	.name = "coda-jpeg-encoder",
+	.type = CODA_INST_ENCODER,
+	.ops = &coda9_jpeg_encode_ops,
+	.direct = true,
+	.src_formats = {
+		V4L2_PIX_FMT_NV12,
+		V4L2_PIX_FMT_YUV420,
+		V4L2_PIX_FMT_YVU420,
+		V4L2_PIX_FMT_YUV422P,
+		V4L2_PIX_FMT_GREY,
+	},
+	.dst_formats = {
+		V4L2_PIX_FMT_JPEG,
+	},
+};
+
+static const struct coda_video_device coda9_jpeg_decoder = {
+	.name = "coda-jpeg-decoder",
+	.type = CODA_INST_DECODER,
+	.ops = &coda9_jpeg_decode_ops,
+	.direct = true,
+	.src_formats = {
+		V4L2_PIX_FMT_JPEG,
+	},
+	.dst_formats = {
+		V4L2_PIX_FMT_NV12,
+		V4L2_PIX_FMT_YUV420,
+		V4L2_PIX_FMT_YVU420,
+		V4L2_PIX_FMT_YUV422P,
+	},
+};
+
+static const struct coda_video_device *codadx6_video_devices[] = {
+	&coda_bit_encoder,
+};
+
+static const struct coda_video_device *codahx4_video_devices[] = {
+	&coda_bit_encoder,
+	&coda_bit_decoder,
+};
+
+static const struct coda_video_device *coda7_video_devices[] = {
+	&coda_bit_jpeg_encoder,
+	&coda_bit_jpeg_decoder,
+	&coda_bit_encoder,
+	&coda_bit_decoder,
+};
+
+static const struct coda_video_device *coda9_video_devices[] = {
+	&coda9_jpeg_encoder,
+	&coda9_jpeg_decoder,
+	&coda_bit_encoder,
+	&coda_bit_decoder,
+};
+
+/*
+ * Normalize all supported YUV 4:2:0 formats to the value used in the codec
+ * tables.
+ */
+static u32 coda_format_normalize_yuv(u32 fourcc)
+{
+	switch (fourcc) {
+	case V4L2_PIX_FMT_NV12:
+	case V4L2_PIX_FMT_YUV420:
+	case V4L2_PIX_FMT_YVU420:
+	case V4L2_PIX_FMT_YUV422P:
+	case V4L2_PIX_FMT_YUYV:
+		return V4L2_PIX_FMT_YUV420;
+	default:
+		return fourcc;
+	}
+}
+
+static const struct coda_codec *coda_find_codec(struct coda_dev *dev,
+						int src_fourcc, int dst_fourcc)
+{
+	const struct coda_codec *codecs = dev->devtype->codecs;
+	int num_codecs = dev->devtype->num_codecs;
+	int k;
+
+	src_fourcc = coda_format_normalize_yuv(src_fourcc);
+	dst_fourcc = coda_format_normalize_yuv(dst_fourcc);
+	if (src_fourcc == dst_fourcc)
+		return NULL;
+
+	for (k = 0; k < num_codecs; k++) {
+		if (codecs[k].src_fourcc == src_fourcc &&
+		    codecs[k].dst_fourcc == dst_fourcc)
+			break;
+	}
+
+	if (k == num_codecs)
+		return NULL;
+
+	return &codecs[k];
+}
+
+static void coda_get_max_dimensions(struct coda_dev *dev,
+				    const struct coda_codec *codec,
+				    int *max_w, int *max_h)
+{
+	const struct coda_codec *codecs = dev->devtype->codecs;
+	int num_codecs = dev->devtype->num_codecs;
+	unsigned int w, h;
+	int k;
+
+	if (codec) {
+		w = codec->max_w;
+		h = codec->max_h;
+	} else {
+		for (k = 0, w = 0, h = 0; k < num_codecs; k++) {
+			w = max(w, codecs[k].max_w);
+			h = max(h, codecs[k].max_h);
+		}
+	}
+
+	if (max_w)
+		*max_w = w;
+	if (max_h)
+		*max_h = h;
+}
+
+static const struct coda_video_device *to_coda_video_device(struct video_device
+							    *vdev)
+{
+	struct coda_dev *dev = video_get_drvdata(vdev);
+	unsigned int i = vdev - dev->vfd;
+
+	if (i >= dev->devtype->num_vdevs)
+		return NULL;
+
+	return dev->devtype->vdevs[i];
+}
+
+const char *coda_product_name(int product)
+{
+	static char buf[9];
+
+	switch (product) {
+	case CODA_DX6:
+		return "CodaDx6";
+	case CODA_HX4:
+		return "CodaHx4";
+	case CODA_7541:
+		return "CODA7541";
+	case CODA_960:
+		return "CODA960";
+	default:
+		snprintf(buf, sizeof(buf), "(0x%04x)", product);
+		return buf;
+	}
+}
+
+static struct vdoa_data *coda_get_vdoa_data(void)
+{
+	struct device_node *vdoa_node;
+	struct platform_device *vdoa_pdev;
+	struct vdoa_data *vdoa_data = NULL;
+
+	vdoa_node = of_find_compatible_node(NULL, NULL, "fsl,imx6q-vdoa");
+	if (!vdoa_node)
+		return NULL;
+
+	vdoa_pdev = of_find_device_by_node(vdoa_node);
+	if (!vdoa_pdev)
+		goto out;
+
+	vdoa_data = platform_get_drvdata(vdoa_pdev);
+	if (!vdoa_data)
+		vdoa_data = ERR_PTR(-EPROBE_DEFER);
+
+	put_device(&vdoa_pdev->dev);
+out:
+	of_node_put(vdoa_node);
+
+	return vdoa_data;
+}
+
+/*
+ * V4L2 ioctl() operations.
+ */
+static int coda_querycap(struct file *file, void *priv,
+			 struct v4l2_capability *cap)
+{
+	struct coda_ctx *ctx = fh_to_ctx(priv);
+
+	strscpy(cap->driver, CODA_NAME, sizeof(cap->driver));
+	strscpy(cap->card, coda_product_name(ctx->dev->devtype->product),
+		sizeof(cap->card));
+	strscpy(cap->bus_info, "platform:" CODA_NAME, sizeof(cap->bus_info));
+	return 0;
+}
+
+static const u32 coda_formats_420[CODA_MAX_FORMATS] = {
+		V4L2_PIX_FMT_NV12,
+		V4L2_PIX_FMT_YUV420,
+		V4L2_PIX_FMT_YVU420,
+};
+
+static int coda_enum_fmt(struct file *file, void *priv,
+			 struct v4l2_fmtdesc *f)
+{
+	struct video_device *vdev = video_devdata(file);
+	const struct coda_video_device *cvd = to_coda_video_device(vdev);
+	struct coda_ctx *ctx = fh_to_ctx(priv);
+	const u32 *formats;
+
+	if (f->type == V4L2_BUF_TYPE_VIDEO_OUTPUT)
+		formats = cvd->src_formats;
+	else if (f->type == V4L2_BUF_TYPE_VIDEO_CAPTURE) {
+		struct coda_q_data *q_data_src;
+		struct vb2_queue *src_vq;
+
+		formats = cvd->dst_formats;
+
+		/*
+		 * If the source format is already fixed, only allow the same
+		 * chroma subsampling.
+		 */
+		q_data_src = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+		src_vq = v4l2_m2m_get_vq(ctx->fh.m2m_ctx,
+					 V4L2_BUF_TYPE_VIDEO_OUTPUT);
+		if (q_data_src->fourcc == V4L2_PIX_FMT_JPEG &&
+		    vb2_is_streaming(src_vq)) {
+			if (ctx->params.jpeg_chroma_subsampling ==
+			    V4L2_JPEG_CHROMA_SUBSAMPLING_420) {
+				formats = coda_formats_420;
+			} else if (ctx->params.jpeg_chroma_subsampling ==
+				   V4L2_JPEG_CHROMA_SUBSAMPLING_422) {
+				f->pixelformat = V4L2_PIX_FMT_YUV422P;
+				return f->index ? -EINVAL : 0;
+			}
+		}
+	} else {
+		return -EINVAL;
+	}
+
+	if (f->index >= CODA_MAX_FORMATS || formats[f->index] == 0)
+		return -EINVAL;
+
+	/* Skip YUYV if the vdoa is not available */
+	if (!ctx->vdoa && f->type == V4L2_BUF_TYPE_VIDEO_CAPTURE &&
+	    formats[f->index] == V4L2_PIX_FMT_YUYV)
+		return -EINVAL;
+
+	f->pixelformat = formats[f->index];
+
+	return 0;
+}
+
+static int coda_g_fmt(struct file *file, void *priv,
+		      struct v4l2_format *f)
+{
+	struct coda_q_data *q_data;
+	struct coda_ctx *ctx = fh_to_ctx(priv);
+
+	q_data = get_q_data(ctx, f->type);
+	if (!q_data)
+		return -EINVAL;
+
+	f->fmt.pix.field	= V4L2_FIELD_NONE;
+	f->fmt.pix.pixelformat	= q_data->fourcc;
+	f->fmt.pix.width	= q_data->width;
+	f->fmt.pix.height	= q_data->height;
+	f->fmt.pix.bytesperline = q_data->bytesperline;
+
+	f->fmt.pix.sizeimage	= q_data->sizeimage;
+	f->fmt.pix.colorspace	= ctx->colorspace;
+	f->fmt.pix.xfer_func	= ctx->xfer_func;
+	f->fmt.pix.ycbcr_enc	= ctx->ycbcr_enc;
+	f->fmt.pix.quantization	= ctx->quantization;
+
+	return 0;
+}
+
+static int coda_try_pixelformat(struct coda_ctx *ctx, struct v4l2_format *f)
+{
+	struct coda_q_data *q_data;
+	const u32 *formats;
+	int i;
+
+	if (f->type == V4L2_BUF_TYPE_VIDEO_OUTPUT)
+		formats = ctx->cvd->src_formats;
+	else if (f->type == V4L2_BUF_TYPE_VIDEO_CAPTURE)
+		formats = ctx->cvd->dst_formats;
+	else
+		return -EINVAL;
+
+	for (i = 0; i < CODA_MAX_FORMATS; i++) {
+		/* Skip YUYV if the vdoa is not available */
+		if (!ctx->vdoa && f->type == V4L2_BUF_TYPE_VIDEO_CAPTURE &&
+		    formats[i] == V4L2_PIX_FMT_YUYV)
+			continue;
+
+		if (formats[i] == f->fmt.pix.pixelformat) {
+			f->fmt.pix.pixelformat = formats[i];
+			return 0;
+		}
+	}
+
+	/* Fall back to currently set pixelformat */
+	q_data = get_q_data(ctx, f->type);
+	f->fmt.pix.pixelformat = q_data->fourcc;
+
+	return 0;
+}
+
+static int coda_try_fmt_vdoa(struct coda_ctx *ctx, struct v4l2_format *f,
+			     bool *use_vdoa)
+{
+	int err;
+
+	if (f->type != V4L2_BUF_TYPE_VIDEO_CAPTURE)
+		return -EINVAL;
+
+	if (!use_vdoa)
+		return -EINVAL;
+
+	if (!ctx->vdoa) {
+		*use_vdoa = false;
+		return 0;
+	}
+
+	err = vdoa_context_configure(NULL, round_up(f->fmt.pix.width, 16),
+				     f->fmt.pix.height, f->fmt.pix.pixelformat);
+	if (err) {
+		*use_vdoa = false;
+		return 0;
+	}
+
+	*use_vdoa = true;
+	return 0;
+}
+
+static unsigned int coda_estimate_sizeimage(struct coda_ctx *ctx, u32 sizeimage,
+					    u32 width, u32 height)
+{
+	/*
+	 * This is a rough estimate for sensible compressed buffer
+	 * sizes (between 1 and 16 bits per pixel). This could be
+	 * improved by better format specific worst case estimates.
+	 */
+	return round_up(clamp(sizeimage, width * height / 8,
+					 width * height * 2), PAGE_SIZE);
+}
+
+static int coda_try_fmt(struct coda_ctx *ctx, const struct coda_codec *codec,
+			struct v4l2_format *f)
+{
+	struct coda_dev *dev = ctx->dev;
+	unsigned int max_w, max_h;
+	enum v4l2_field field;
+
+	field = f->fmt.pix.field;
+	if (field == V4L2_FIELD_ANY)
+		field = V4L2_FIELD_NONE;
+	else if (V4L2_FIELD_NONE != field)
+		return -EINVAL;
+
+	/* V4L2 specification suggests the driver corrects the format struct
+	 * if any of the dimensions is unsupported */
+	f->fmt.pix.field = field;
+
+	coda_get_max_dimensions(dev, codec, &max_w, &max_h);
+	v4l_bound_align_image(&f->fmt.pix.width, MIN_W, max_w, W_ALIGN,
+			      &f->fmt.pix.height, MIN_H, max_h, H_ALIGN,
+			      S_ALIGN);
+
+	switch (f->fmt.pix.pixelformat) {
+	case V4L2_PIX_FMT_NV12:
+	case V4L2_PIX_FMT_YUV420:
+	case V4L2_PIX_FMT_YVU420:
+		/*
+		 * Frame stride must be at least multiple of 8,
+		 * but multiple of 16 for h.264 or JPEG 4:2:x
+		 */
+		f->fmt.pix.bytesperline = round_up(f->fmt.pix.width, 16);
+		f->fmt.pix.sizeimage = f->fmt.pix.bytesperline *
+					f->fmt.pix.height * 3 / 2;
+		break;
+	case V4L2_PIX_FMT_YUYV:
+		f->fmt.pix.bytesperline = round_up(f->fmt.pix.width, 16) * 2;
+		f->fmt.pix.sizeimage = f->fmt.pix.bytesperline *
+					f->fmt.pix.height;
+		break;
+	case V4L2_PIX_FMT_YUV422P:
+		f->fmt.pix.bytesperline = round_up(f->fmt.pix.width, 16);
+		f->fmt.pix.sizeimage = f->fmt.pix.bytesperline *
+					f->fmt.pix.height * 2;
+		break;
+	case V4L2_PIX_FMT_GREY:
+		/* keep 16 pixel alignment of 8-bit pixel data */
+		f->fmt.pix.bytesperline = round_up(f->fmt.pix.width, 16);
+		f->fmt.pix.sizeimage = f->fmt.pix.bytesperline * f->fmt.pix.height;
+		break;
+	case V4L2_PIX_FMT_JPEG:
+	case V4L2_PIX_FMT_H264:
+	case V4L2_PIX_FMT_MPEG4:
+	case V4L2_PIX_FMT_MPEG2:
+		f->fmt.pix.bytesperline = 0;
+		f->fmt.pix.sizeimage = coda_estimate_sizeimage(ctx,
+							f->fmt.pix.sizeimage,
+							f->fmt.pix.width,
+							f->fmt.pix.height);
+		break;
+	default:
+		BUG();
+	}
+
+	return 0;
+}
+
+static int coda_try_fmt_vid_cap(struct file *file, void *priv,
+				struct v4l2_format *f)
+{
+	struct coda_ctx *ctx = fh_to_ctx(priv);
+	const struct coda_q_data *q_data_src;
+	const struct coda_codec *codec;
+	struct vb2_queue *src_vq;
+	int hscale = 0;
+	int vscale = 0;
+	int ret;
+	bool use_vdoa;
+
+	ret = coda_try_pixelformat(ctx, f);
+	if (ret < 0)
+		return ret;
+
+	q_data_src = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+
+	/*
+	 * If the source format is already fixed, only allow the same output
+	 * resolution. When decoding JPEG images, we also have to make sure to
+	 * use the same chroma subsampling.
+	 */
+	src_vq = v4l2_m2m_get_vq(ctx->fh.m2m_ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+	if (vb2_is_streaming(src_vq)) {
+		if (q_data_src->fourcc == V4L2_PIX_FMT_JPEG &&
+		    ctx->dev->devtype->product == CODA_960) {
+			hscale = coda_jpeg_scale(q_data_src->width, f->fmt.pix.width);
+			vscale = coda_jpeg_scale(q_data_src->height, f->fmt.pix.height);
+		}
+		f->fmt.pix.width = q_data_src->width >> hscale;
+		f->fmt.pix.height = q_data_src->height >> vscale;
+
+		if (q_data_src->fourcc == V4L2_PIX_FMT_JPEG) {
+			if (ctx->params.jpeg_chroma_subsampling ==
+			    V4L2_JPEG_CHROMA_SUBSAMPLING_420 &&
+			    f->fmt.pix.pixelformat == V4L2_PIX_FMT_YUV422P)
+				f->fmt.pix.pixelformat = V4L2_PIX_FMT_NV12;
+			else if (ctx->params.jpeg_chroma_subsampling ==
+				 V4L2_JPEG_CHROMA_SUBSAMPLING_422)
+				f->fmt.pix.pixelformat = V4L2_PIX_FMT_YUV422P;
+		}
+	}
+
+	f->fmt.pix.colorspace = ctx->colorspace;
+	f->fmt.pix.xfer_func = ctx->xfer_func;
+	f->fmt.pix.ycbcr_enc = ctx->ycbcr_enc;
+	f->fmt.pix.quantization = ctx->quantization;
+
+	q_data_src = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+	codec = coda_find_codec(ctx->dev, q_data_src->fourcc,
+				f->fmt.pix.pixelformat);
+	if (!codec)
+		return -EINVAL;
+
+	ret = coda_try_fmt(ctx, codec, f);
+	if (ret < 0)
+		return ret;
+
+	/* The decoders always write complete macroblocks or MCUs */
+	if (ctx->inst_type == CODA_INST_DECODER) {
+		f->fmt.pix.bytesperline = round_up(f->fmt.pix.width, 16 >> hscale);
+		f->fmt.pix.height = round_up(f->fmt.pix.height, 16 >> vscale);
+		if (codec->src_fourcc == V4L2_PIX_FMT_JPEG &&
+		    f->fmt.pix.pixelformat == V4L2_PIX_FMT_YUV422P) {
+			f->fmt.pix.sizeimage = f->fmt.pix.bytesperline *
+					       f->fmt.pix.height * 2;
+		} else {
+			f->fmt.pix.sizeimage = f->fmt.pix.bytesperline *
+					       f->fmt.pix.height * 3 / 2;
+		}
+
+		ret = coda_try_fmt_vdoa(ctx, f, &use_vdoa);
+		if (ret < 0)
+			return ret;
+
+		if (f->fmt.pix.pixelformat == V4L2_PIX_FMT_YUYV) {
+			if (!use_vdoa)
+				return -EINVAL;
+
+			f->fmt.pix.bytesperline = round_up(f->fmt.pix.width, 16) * 2;
+			f->fmt.pix.sizeimage = f->fmt.pix.bytesperline *
+				f->fmt.pix.height;
+		}
+	}
+
+	return 0;
+}
+
+static void coda_set_default_colorspace(struct v4l2_pix_format *fmt)
+{
+	enum v4l2_colorspace colorspace;
+
+	if (fmt->pixelformat == V4L2_PIX_FMT_JPEG)
+		colorspace = V4L2_COLORSPACE_JPEG;
+	else if (fmt->width <= 720 && fmt->height <= 576)
+		colorspace = V4L2_COLORSPACE_SMPTE170M;
+	else
+		colorspace = V4L2_COLORSPACE_REC709;
+
+	fmt->colorspace = colorspace;
+	fmt->xfer_func = V4L2_XFER_FUNC_DEFAULT;
+	fmt->ycbcr_enc = V4L2_YCBCR_ENC_DEFAULT;
+	fmt->quantization = V4L2_QUANTIZATION_DEFAULT;
+}
+
+static int coda_try_fmt_vid_out(struct file *file, void *priv,
+				struct v4l2_format *f)
+{
+	struct coda_ctx *ctx = fh_to_ctx(priv);
+	struct coda_dev *dev = ctx->dev;
+	const struct coda_q_data *q_data_dst;
+	const struct coda_codec *codec;
+	int ret;
+
+	ret = coda_try_pixelformat(ctx, f);
+	if (ret < 0)
+		return ret;
+
+	if (f->fmt.pix.colorspace == V4L2_COLORSPACE_DEFAULT)
+		coda_set_default_colorspace(&f->fmt.pix);
+
+	q_data_dst = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_CAPTURE);
+	codec = coda_find_codec(dev, f->fmt.pix.pixelformat, q_data_dst->fourcc);
+
+	return coda_try_fmt(ctx, codec, f);
+}
+
+static int coda_s_fmt(struct coda_ctx *ctx, struct v4l2_format *f,
+		      struct v4l2_rect *r)
+{
+	struct coda_q_data *q_data;
+	struct vb2_queue *vq;
+
+	vq = v4l2_m2m_get_vq(ctx->fh.m2m_ctx, f->type);
+	if (!vq)
+		return -EINVAL;
+
+	q_data = get_q_data(ctx, f->type);
+	if (!q_data)
+		return -EINVAL;
+
+	if (vb2_is_busy(vq)) {
+		v4l2_err(&ctx->dev->v4l2_dev, "%s: %s queue busy: %d\n",
+			 __func__, v4l2_type_names[f->type], vq->num_buffers);
+		return -EBUSY;
+	}
+
+	q_data->fourcc = f->fmt.pix.pixelformat;
+	q_data->width = f->fmt.pix.width;
+	q_data->height = f->fmt.pix.height;
+	q_data->bytesperline = f->fmt.pix.bytesperline;
+	q_data->sizeimage = f->fmt.pix.sizeimage;
+	if (r) {
+		q_data->rect = *r;
+	} else {
+		q_data->rect.left = 0;
+		q_data->rect.top = 0;
+		q_data->rect.width = f->fmt.pix.width;
+		q_data->rect.height = f->fmt.pix.height;
+	}
+
+	switch (f->fmt.pix.pixelformat) {
+	case V4L2_PIX_FMT_YUYV:
+		ctx->tiled_map_type = GDI_TILED_FRAME_MB_RASTER_MAP;
+		break;
+	case V4L2_PIX_FMT_NV12:
+		if (!disable_tiling && ctx->use_bit &&
+		    ctx->dev->devtype->product == CODA_960) {
+			ctx->tiled_map_type = GDI_TILED_FRAME_MB_RASTER_MAP;
+			break;
+		}
+		fallthrough;
+	case V4L2_PIX_FMT_YUV420:
+	case V4L2_PIX_FMT_YVU420:
+	case V4L2_PIX_FMT_YUV422P:
+		ctx->tiled_map_type = GDI_LINEAR_FRAME_MAP;
+		break;
+	default:
+		break;
+	}
+
+	if (ctx->tiled_map_type == GDI_TILED_FRAME_MB_RASTER_MAP &&
+	    !coda_try_fmt_vdoa(ctx, f, &ctx->use_vdoa) &&
+	    ctx->use_vdoa)
+		vdoa_context_configure(ctx->vdoa,
+				       round_up(f->fmt.pix.width, 16),
+				       f->fmt.pix.height,
+				       f->fmt.pix.pixelformat);
+	else
+		ctx->use_vdoa = false;
+
+	coda_dbg(1, ctx, "Setting %s format, wxh: %dx%d, fmt: %4.4s %c\n",
+		 v4l2_type_names[f->type], q_data->width, q_data->height,
+		 (char *)&q_data->fourcc,
+		 (ctx->tiled_map_type == GDI_LINEAR_FRAME_MAP) ? 'L' : 'T');
+
+	return 0;
+}
+
+static int coda_s_fmt_vid_cap(struct file *file, void *priv,
+			      struct v4l2_format *f)
+{
+	struct coda_ctx *ctx = fh_to_ctx(priv);
+	struct coda_q_data *q_data_src;
+	const struct coda_codec *codec;
+	struct v4l2_rect r;
+	int hscale = 0;
+	int vscale = 0;
+	int ret;
+
+	q_data_src = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+
+	if (q_data_src->fourcc == V4L2_PIX_FMT_JPEG &&
+	    ctx->dev->devtype->product == CODA_960) {
+		hscale = coda_jpeg_scale(q_data_src->width, f->fmt.pix.width);
+		vscale = coda_jpeg_scale(q_data_src->height, f->fmt.pix.height);
+	}
+
+	ret = coda_try_fmt_vid_cap(file, priv, f);
+	if (ret)
+		return ret;
+
+	r.left = 0;
+	r.top = 0;
+	r.width = q_data_src->width >> hscale;
+	r.height = q_data_src->height >> vscale;
+
+	ret = coda_s_fmt(ctx, f, &r);
+	if (ret)
+		return ret;
+
+	if (ctx->inst_type != CODA_INST_ENCODER)
+		return 0;
+
+	/* Setting the coded format determines the selected codec */
+	codec = coda_find_codec(ctx->dev, q_data_src->fourcc,
+				f->fmt.pix.pixelformat);
+	if (!codec) {
+		v4l2_err(&ctx->dev->v4l2_dev, "failed to determine codec\n");
+		return -EINVAL;
+	}
+	ctx->codec = codec;
+
+	ctx->colorspace = f->fmt.pix.colorspace;
+	ctx->xfer_func = f->fmt.pix.xfer_func;
+	ctx->ycbcr_enc = f->fmt.pix.ycbcr_enc;
+	ctx->quantization = f->fmt.pix.quantization;
+
+	return 0;
+}
+
+static int coda_s_fmt_vid_out(struct file *file, void *priv,
+			      struct v4l2_format *f)
+{
+	struct coda_ctx *ctx = fh_to_ctx(priv);
+	const struct coda_codec *codec;
+	struct v4l2_format f_cap;
+	struct vb2_queue *dst_vq;
+	int ret;
+
+	ret = coda_try_fmt_vid_out(file, priv, f);
+	if (ret)
+		return ret;
+
+	ret = coda_s_fmt(ctx, f, NULL);
+	if (ret)
+		return ret;
+
+	ctx->colorspace = f->fmt.pix.colorspace;
+	ctx->xfer_func = f->fmt.pix.xfer_func;
+	ctx->ycbcr_enc = f->fmt.pix.ycbcr_enc;
+	ctx->quantization = f->fmt.pix.quantization;
+
+	if (ctx->inst_type != CODA_INST_DECODER)
+		return 0;
+
+	/* Setting the coded format determines the selected codec */
+	codec = coda_find_codec(ctx->dev, f->fmt.pix.pixelformat,
+				V4L2_PIX_FMT_YUV420);
+	if (!codec) {
+		v4l2_err(&ctx->dev->v4l2_dev, "failed to determine codec\n");
+		return -EINVAL;
+	}
+	ctx->codec = codec;
+
+	dst_vq = v4l2_m2m_get_vq(ctx->fh.m2m_ctx, V4L2_BUF_TYPE_VIDEO_CAPTURE);
+	if (!dst_vq)
+		return -EINVAL;
+
+	/*
+	 * Setting the capture queue format is not possible while the capture
+	 * queue is still busy. This is not an error, but the user will have to
+	 * make sure themselves that the capture format is set correctly before
+	 * starting the output queue again.
+	 */
+	if (vb2_is_busy(dst_vq))
+		return 0;
+
+	memset(&f_cap, 0, sizeof(f_cap));
+	f_cap.type = V4L2_BUF_TYPE_VIDEO_CAPTURE;
+	coda_g_fmt(file, priv, &f_cap);
+	f_cap.fmt.pix.width = f->fmt.pix.width;
+	f_cap.fmt.pix.height = f->fmt.pix.height;
+
+	return coda_s_fmt_vid_cap(file, priv, &f_cap);
+}
+
+static int coda_reqbufs(struct file *file, void *priv,
+			struct v4l2_requestbuffers *rb)
+{
+	struct coda_ctx *ctx = fh_to_ctx(priv);
+	int ret;
+
+	ret = v4l2_m2m_reqbufs(file, ctx->fh.m2m_ctx, rb);
+	if (ret)
+		return ret;
+
+	/*
+	 * Allow to allocate instance specific per-context buffers, such as
+	 * bitstream ringbuffer, slice buffer, work buffer, etc. if needed.
+	 */
+	if (rb->type == V4L2_BUF_TYPE_VIDEO_OUTPUT && ctx->ops->reqbufs)
+		return ctx->ops->reqbufs(ctx, rb);
+
+	return 0;
+}
+
+static int coda_qbuf(struct file *file, void *priv,
+		     struct v4l2_buffer *buf)
+{
+	struct coda_ctx *ctx = fh_to_ctx(priv);
+
+	if (ctx->inst_type == CODA_INST_DECODER &&
+	    buf->type == V4L2_BUF_TYPE_VIDEO_OUTPUT)
+		buf->flags &= ~V4L2_BUF_FLAG_LAST;
+
+	return v4l2_m2m_qbuf(file, ctx->fh.m2m_ctx, buf);
+}
+
+static int coda_dqbuf(struct file *file, void *priv, struct v4l2_buffer *buf)
+{
+	struct coda_ctx *ctx = fh_to_ctx(priv);
+	int ret;
+
+	ret = v4l2_m2m_dqbuf(file, ctx->fh.m2m_ctx, buf);
+
+	if (ctx->inst_type == CODA_INST_DECODER &&
+	    buf->type == V4L2_BUF_TYPE_VIDEO_OUTPUT)
+		buf->flags &= ~V4L2_BUF_FLAG_LAST;
+
+	return ret;
+}
+
+void coda_m2m_buf_done(struct coda_ctx *ctx, struct vb2_v4l2_buffer *buf,
+		       enum vb2_buffer_state state)
+{
+	const struct v4l2_event eos_event = {
+		.type = V4L2_EVENT_EOS
+	};
+
+	if (buf->flags & V4L2_BUF_FLAG_LAST)
+		v4l2_event_queue_fh(&ctx->fh, &eos_event);
+
+	v4l2_m2m_buf_done(buf, state);
+}
+
+static int coda_g_selection(struct file *file, void *fh,
+			    struct v4l2_selection *s)
+{
+	struct coda_ctx *ctx = fh_to_ctx(fh);
+	struct coda_q_data *q_data;
+	struct v4l2_rect r, *rsel;
+
+	q_data = get_q_data(ctx, s->type);
+	if (!q_data)
+		return -EINVAL;
+
+	r.left = 0;
+	r.top = 0;
+	r.width = q_data->width;
+	r.height = q_data->height;
+	rsel = &q_data->rect;
+
+	switch (s->target) {
+	case V4L2_SEL_TGT_CROP_DEFAULT:
+	case V4L2_SEL_TGT_CROP_BOUNDS:
+		rsel = &r;
+		fallthrough;
+	case V4L2_SEL_TGT_CROP:
+		if (s->type != V4L2_BUF_TYPE_VIDEO_OUTPUT ||
+		    ctx->inst_type == CODA_INST_DECODER)
+			return -EINVAL;
+		break;
+	case V4L2_SEL_TGT_COMPOSE_BOUNDS:
+	case V4L2_SEL_TGT_COMPOSE_PADDED:
+		rsel = &r;
+		fallthrough;
+	case V4L2_SEL_TGT_COMPOSE:
+	case V4L2_SEL_TGT_COMPOSE_DEFAULT:
+		if (s->type != V4L2_BUF_TYPE_VIDEO_CAPTURE ||
+		    ctx->inst_type == CODA_INST_ENCODER)
+			return -EINVAL;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	s->r = *rsel;
+
+	return 0;
+}
+
+static int coda_s_selection(struct file *file, void *fh,
+			    struct v4l2_selection *s)
+{
+	struct coda_ctx *ctx = fh_to_ctx(fh);
+	struct coda_q_data *q_data;
+
+	switch (s->target) {
+	case V4L2_SEL_TGT_CROP:
+		if (ctx->inst_type == CODA_INST_ENCODER &&
+		    s->type == V4L2_BUF_TYPE_VIDEO_OUTPUT) {
+			q_data = get_q_data(ctx, s->type);
+			if (!q_data)
+				return -EINVAL;
+
+			s->r.left = 0;
+			s->r.top = 0;
+			s->r.width = clamp(s->r.width, 2U, q_data->width);
+			s->r.height = clamp(s->r.height, 2U, q_data->height);
+
+			if (s->flags & V4L2_SEL_FLAG_LE) {
+				s->r.width = round_up(s->r.width, 2);
+				s->r.height = round_up(s->r.height, 2);
+			} else {
+				s->r.width = round_down(s->r.width, 2);
+				s->r.height = round_down(s->r.height, 2);
+			}
+
+			q_data->rect = s->r;
+
+			coda_dbg(1, ctx, "Setting crop rectangle: %dx%d\n",
+				 s->r.width, s->r.height);
+
+			return 0;
+		}
+		fallthrough;
+	case V4L2_SEL_TGT_NATIVE_SIZE:
+	case V4L2_SEL_TGT_COMPOSE:
+		return coda_g_selection(file, fh, s);
+	default:
+		/* v4l2-compliance expects this to fail for read-only targets */
+		return -EINVAL;
+	}
+}
+
+static void coda_wake_up_capture_queue(struct coda_ctx *ctx)
+{
+	struct vb2_queue *dst_vq;
+
+	coda_dbg(1, ctx, "waking up capture queue\n");
+
+	dst_vq = v4l2_m2m_get_vq(ctx->fh.m2m_ctx, V4L2_BUF_TYPE_VIDEO_CAPTURE);
+	dst_vq->last_buffer_dequeued = true;
+	wake_up(&dst_vq->done_wq);
+}
+
+static int coda_encoder_cmd(struct file *file, void *fh,
+			    struct v4l2_encoder_cmd *ec)
+{
+	struct coda_ctx *ctx = fh_to_ctx(fh);
+	struct vb2_v4l2_buffer *buf;
+	int ret;
+
+	ret = v4l2_m2m_ioctl_try_encoder_cmd(file, fh, ec);
+	if (ret < 0)
+		return ret;
+
+	mutex_lock(&ctx->wakeup_mutex);
+	buf = v4l2_m2m_last_src_buf(ctx->fh.m2m_ctx);
+	if (buf) {
+		/*
+		 * If the last output buffer is still on the queue, make sure
+		 * that decoder finish_run will see the last flag and report it
+		 * to userspace.
+		 */
+		buf->flags |= V4L2_BUF_FLAG_LAST;
+	} else {
+		/* Set the stream-end flag on this context */
+		ctx->bit_stream_param |= CODA_BIT_STREAM_END_FLAG;
+
+		/*
+		 * If the last output buffer has already been taken from the
+		 * queue, wake up the capture queue and signal end of stream
+		 * via the -EPIPE mechanism.
+		 */
+		coda_wake_up_capture_queue(ctx);
+	}
+	mutex_unlock(&ctx->wakeup_mutex);
+
+	return 0;
+}
+
+static bool coda_mark_last_meta(struct coda_ctx *ctx)
+{
+	struct coda_buffer_meta *meta;
+
+	coda_dbg(1, ctx, "marking last meta\n");
+
+	spin_lock(&ctx->buffer_meta_lock);
+	if (list_empty(&ctx->buffer_meta_list)) {
+		spin_unlock(&ctx->buffer_meta_lock);
+		return false;
+	}
+
+	meta = list_last_entry(&ctx->buffer_meta_list, struct coda_buffer_meta,
+			       list);
+	meta->last = true;
+
+	spin_unlock(&ctx->buffer_meta_lock);
+	return true;
+}
+
+static bool coda_mark_last_dst_buf(struct coda_ctx *ctx)
+{
+	struct vb2_v4l2_buffer *buf;
+	struct vb2_buffer *dst_vb;
+	struct vb2_queue *dst_vq;
+	unsigned long flags;
+
+	coda_dbg(1, ctx, "marking last capture buffer\n");
+
+	dst_vq = v4l2_m2m_get_vq(ctx->fh.m2m_ctx, V4L2_BUF_TYPE_VIDEO_CAPTURE);
+	spin_lock_irqsave(&dst_vq->done_lock, flags);
+	if (list_empty(&dst_vq->done_list)) {
+		spin_unlock_irqrestore(&dst_vq->done_lock, flags);
+		return false;
+	}
+
+	dst_vb = list_last_entry(&dst_vq->done_list, struct vb2_buffer,
+				 done_entry);
+	buf = to_vb2_v4l2_buffer(dst_vb);
+	buf->flags |= V4L2_BUF_FLAG_LAST;
+
+	spin_unlock_irqrestore(&dst_vq->done_lock, flags);
+	return true;
+}
+
+static int coda_decoder_cmd(struct file *file, void *fh,
+			    struct v4l2_decoder_cmd *dc)
+{
+	struct coda_ctx *ctx = fh_to_ctx(fh);
+	struct coda_dev *dev = ctx->dev;
+	struct vb2_v4l2_buffer *buf;
+	struct vb2_queue *dst_vq;
+	bool stream_end;
+	bool wakeup;
+	int ret;
+
+	ret = v4l2_m2m_ioctl_try_decoder_cmd(file, fh, dc);
+	if (ret < 0)
+		return ret;
+
+	switch (dc->cmd) {
+	case V4L2_DEC_CMD_START:
+		mutex_lock(&dev->coda_mutex);
+		mutex_lock(&ctx->bitstream_mutex);
+		coda_bitstream_flush(ctx);
+		dst_vq = v4l2_m2m_get_vq(ctx->fh.m2m_ctx,
+					 V4L2_BUF_TYPE_VIDEO_CAPTURE);
+		vb2_clear_last_buffer_dequeued(dst_vq);
+		ctx->bit_stream_param &= ~CODA_BIT_STREAM_END_FLAG;
+		coda_fill_bitstream(ctx, NULL);
+		mutex_unlock(&ctx->bitstream_mutex);
+		mutex_unlock(&dev->coda_mutex);
+		break;
+	case V4L2_DEC_CMD_STOP:
+		stream_end = false;
+		wakeup = false;
+
+		mutex_lock(&ctx->wakeup_mutex);
+
+		buf = v4l2_m2m_last_src_buf(ctx->fh.m2m_ctx);
+		if (buf) {
+			coda_dbg(1, ctx, "marking last pending buffer\n");
+
+			/* Mark last buffer */
+			buf->flags |= V4L2_BUF_FLAG_LAST;
+
+			if (v4l2_m2m_num_src_bufs_ready(ctx->fh.m2m_ctx) == 0) {
+				coda_dbg(1, ctx, "all remaining buffers queued\n");
+				stream_end = true;
+			}
+		} else {
+			if (ctx->use_bit)
+				if (coda_mark_last_meta(ctx))
+					stream_end = true;
+				else
+					wakeup = true;
+			else
+				if (!coda_mark_last_dst_buf(ctx))
+					wakeup = true;
+		}
+
+		if (stream_end) {
+			coda_dbg(1, ctx, "all remaining buffers queued\n");
+
+			/* Set the stream-end flag on this context */
+			coda_bit_stream_end_flag(ctx);
+			ctx->hold = false;
+			v4l2_m2m_try_schedule(ctx->fh.m2m_ctx);
+		}
+
+		if (wakeup) {
+			/* If there is no buffer in flight, wake up */
+			coda_wake_up_capture_queue(ctx);
+		}
+
+		mutex_unlock(&ctx->wakeup_mutex);
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int coda_enum_framesizes(struct file *file, void *fh,
+				struct v4l2_frmsizeenum *fsize)
+{
+	struct coda_ctx *ctx = fh_to_ctx(fh);
+	struct coda_q_data *q_data_dst;
+	const struct coda_codec *codec;
+
+	if (fsize->index)
+		return -EINVAL;
+
+	if (coda_format_normalize_yuv(fsize->pixel_format) ==
+	    V4L2_PIX_FMT_YUV420) {
+		q_data_dst = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_CAPTURE);
+		codec = coda_find_codec(ctx->dev, fsize->pixel_format,
+					q_data_dst->fourcc);
+	} else {
+		codec = coda_find_codec(ctx->dev, V4L2_PIX_FMT_YUV420,
+					fsize->pixel_format);
+	}
+	if (!codec)
+		return -EINVAL;
+
+	fsize->type = V4L2_FRMSIZE_TYPE_CONTINUOUS;
+	fsize->stepwise.min_width = MIN_W;
+	fsize->stepwise.max_width = codec->max_w;
+	fsize->stepwise.step_width = 1;
+	fsize->stepwise.min_height = MIN_H;
+	fsize->stepwise.max_height = codec->max_h;
+	fsize->stepwise.step_height = 1;
+
+	return 0;
+}
+
+static int coda_enum_frameintervals(struct file *file, void *fh,
+				    struct v4l2_frmivalenum *f)
+{
+	struct coda_ctx *ctx = fh_to_ctx(fh);
+	struct coda_q_data *q_data;
+	const struct coda_codec *codec;
+
+	if (f->index)
+		return -EINVAL;
+
+	/* Disallow YUYV if the vdoa is not available */
+	if (!ctx->vdoa && f->pixel_format == V4L2_PIX_FMT_YUYV)
+		return -EINVAL;
+
+	if (coda_format_normalize_yuv(f->pixel_format) == V4L2_PIX_FMT_YUV420) {
+		q_data = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_CAPTURE);
+		codec = coda_find_codec(ctx->dev, f->pixel_format,
+					q_data->fourcc);
+	} else {
+		codec = coda_find_codec(ctx->dev, V4L2_PIX_FMT_YUV420,
+					f->pixel_format);
+	}
+	if (!codec)
+		return -EINVAL;
+
+	if (f->width < MIN_W || f->width > codec->max_w ||
+	    f->height < MIN_H || f->height > codec->max_h)
+		return -EINVAL;
+
+	f->type = V4L2_FRMIVAL_TYPE_CONTINUOUS;
+	f->stepwise.min.numerator = 1;
+	f->stepwise.min.denominator = 65535;
+	f->stepwise.max.numerator = 65536;
+	f->stepwise.max.denominator = 1;
+	f->stepwise.step.numerator = 1;
+	f->stepwise.step.denominator = 1;
+
+	return 0;
+}
+
+static int coda_g_parm(struct file *file, void *fh, struct v4l2_streamparm *a)
+{
+	struct coda_ctx *ctx = fh_to_ctx(fh);
+	struct v4l2_fract *tpf;
+
+	if (a->type != V4L2_BUF_TYPE_VIDEO_OUTPUT)
+		return -EINVAL;
+
+	a->parm.output.capability = V4L2_CAP_TIMEPERFRAME;
+	tpf = &a->parm.output.timeperframe;
+	tpf->denominator = ctx->params.framerate & CODA_FRATE_RES_MASK;
+	tpf->numerator = 1 + (ctx->params.framerate >>
+			      CODA_FRATE_DIV_OFFSET);
+
+	return 0;
+}
+
+/*
+ * Approximate timeperframe v4l2_fract with values that can be written
+ * into the 16-bit CODA_FRATE_DIV and CODA_FRATE_RES fields.
+ */
+static void coda_approximate_timeperframe(struct v4l2_fract *timeperframe)
+{
+	struct v4l2_fract s = *timeperframe;
+	struct v4l2_fract f0;
+	struct v4l2_fract f1 = { 1, 0 };
+	struct v4l2_fract f2 = { 0, 1 };
+	unsigned int i, div, s_denominator;
+
+	/* Lower bound is 1/65535 */
+	if (s.numerator == 0 || s.denominator / s.numerator > 65535) {
+		timeperframe->numerator = 1;
+		timeperframe->denominator = 65535;
+		return;
+	}
+
+	/* Upper bound is 65536/1 */
+	if (s.denominator == 0 || s.numerator / s.denominator > 65536) {
+		timeperframe->numerator = 65536;
+		timeperframe->denominator = 1;
+		return;
+	}
+
+	/* Reduce fraction to lowest terms */
+	div = gcd(s.numerator, s.denominator);
+	if (div > 1) {
+		s.numerator /= div;
+		s.denominator /= div;
+	}
+
+	if (s.numerator <= 65536 && s.denominator < 65536) {
+		*timeperframe = s;
+		return;
+	}
+
+	/* Find successive convergents from continued fraction expansion */
+	while (f2.numerator <= 65536 && f2.denominator < 65536) {
+		f0 = f1;
+		f1 = f2;
+
+		/* Stop when f2 exactly equals timeperframe */
+		if (s.numerator == 0)
+			break;
+
+		i = s.denominator / s.numerator;
+
+		f2.numerator = f0.numerator + i * f1.numerator;
+		f2.denominator = f0.denominator + i * f2.denominator;
+
+		s_denominator = s.numerator;
+		s.numerator = s.denominator % s.numerator;
+		s.denominator = s_denominator;
+	}
+
+	*timeperframe = f1;
+}
+
+static uint32_t coda_timeperframe_to_frate(struct v4l2_fract *timeperframe)
+{
+	return ((timeperframe->numerator - 1) << CODA_FRATE_DIV_OFFSET) |
+		timeperframe->denominator;
+}
+
+static int coda_s_parm(struct file *file, void *fh, struct v4l2_streamparm *a)
+{
+	struct coda_ctx *ctx = fh_to_ctx(fh);
+	struct v4l2_fract *tpf;
+
+	if (a->type != V4L2_BUF_TYPE_VIDEO_OUTPUT)
+		return -EINVAL;
+
+	a->parm.output.capability = V4L2_CAP_TIMEPERFRAME;
+	tpf = &a->parm.output.timeperframe;
+	coda_approximate_timeperframe(tpf);
+	ctx->params.framerate = coda_timeperframe_to_frate(tpf);
+	ctx->params.framerate_changed = true;
+
+	return 0;
+}
+
+static int coda_subscribe_event(struct v4l2_fh *fh,
+				const struct v4l2_event_subscription *sub)
+{
+	struct coda_ctx *ctx = fh_to_ctx(fh);
+
+	switch (sub->type) {
+	case V4L2_EVENT_EOS:
+		return v4l2_event_subscribe(fh, sub, 0, NULL);
+	case V4L2_EVENT_SOURCE_CHANGE:
+		if (ctx->inst_type == CODA_INST_DECODER)
+			return v4l2_event_subscribe(fh, sub, 0, NULL);
+		else
+			return -EINVAL;
+	default:
+		return v4l2_ctrl_subscribe_event(fh, sub);
+	}
+}
+
+static const struct v4l2_ioctl_ops coda_ioctl_ops = {
+	.vidioc_querycap	= coda_querycap,
+
+	.vidioc_enum_fmt_vid_cap = coda_enum_fmt,
+	.vidioc_g_fmt_vid_cap	= coda_g_fmt,
+	.vidioc_try_fmt_vid_cap	= coda_try_fmt_vid_cap,
+	.vidioc_s_fmt_vid_cap	= coda_s_fmt_vid_cap,
+
+	.vidioc_enum_fmt_vid_out = coda_enum_fmt,
+	.vidioc_g_fmt_vid_out	= coda_g_fmt,
+	.vidioc_try_fmt_vid_out	= coda_try_fmt_vid_out,
+	.vidioc_s_fmt_vid_out	= coda_s_fmt_vid_out,
+
+	.vidioc_reqbufs		= coda_reqbufs,
+	.vidioc_querybuf	= v4l2_m2m_ioctl_querybuf,
+
+	.vidioc_qbuf		= coda_qbuf,
+	.vidioc_expbuf		= v4l2_m2m_ioctl_expbuf,
+	.vidioc_dqbuf		= coda_dqbuf,
+	.vidioc_create_bufs	= v4l2_m2m_ioctl_create_bufs,
+	.vidioc_prepare_buf	= v4l2_m2m_ioctl_prepare_buf,
+
+	.vidioc_streamon	= v4l2_m2m_ioctl_streamon,
+	.vidioc_streamoff	= v4l2_m2m_ioctl_streamoff,
+
+	.vidioc_g_selection	= coda_g_selection,
+	.vidioc_s_selection	= coda_s_selection,
+
+	.vidioc_try_encoder_cmd	= v4l2_m2m_ioctl_try_encoder_cmd,
+	.vidioc_encoder_cmd	= coda_encoder_cmd,
+	.vidioc_try_decoder_cmd	= v4l2_m2m_ioctl_try_decoder_cmd,
+	.vidioc_decoder_cmd	= coda_decoder_cmd,
+
+	.vidioc_g_parm		= coda_g_parm,
+	.vidioc_s_parm		= coda_s_parm,
+
+	.vidioc_enum_framesizes	= coda_enum_framesizes,
+	.vidioc_enum_frameintervals = coda_enum_frameintervals,
+
+	.vidioc_subscribe_event = coda_subscribe_event,
+	.vidioc_unsubscribe_event = v4l2_event_unsubscribe,
+};
+
+/*
+ * Mem-to-mem operations.
+ */
+
+static void coda_device_run(void *m2m_priv)
+{
+	struct coda_ctx *ctx = m2m_priv;
+	struct coda_dev *dev = ctx->dev;
+
+	queue_work(dev->workqueue, &ctx->pic_run_work);
+}
+
+static void coda_pic_run_work(struct work_struct *work)
+{
+	struct coda_ctx *ctx = container_of(work, struct coda_ctx, pic_run_work);
+	struct coda_dev *dev = ctx->dev;
+	int ret;
+
+	mutex_lock(&ctx->buffer_mutex);
+	mutex_lock(&dev->coda_mutex);
+
+	ret = ctx->ops->prepare_run(ctx);
+	if (ret < 0 && ctx->inst_type == CODA_INST_DECODER)
+		goto out;
+
+	if (!wait_for_completion_timeout(&ctx->completion,
+					 msecs_to_jiffies(1000))) {
+		if (ctx->use_bit) {
+			dev_err(dev->dev, "CODA PIC_RUN timeout\n");
+
+			ctx->hold = true;
+
+			coda_hw_reset(ctx);
+		}
+
+		if (ctx->ops->run_timeout)
+			ctx->ops->run_timeout(ctx);
+	} else {
+		ctx->ops->finish_run(ctx);
+	}
+
+	if ((ctx->aborting || (!ctx->streamon_cap && !ctx->streamon_out)) &&
+	    ctx->ops->seq_end_work)
+		queue_work(dev->workqueue, &ctx->seq_end_work);
+
+out:
+	mutex_unlock(&dev->coda_mutex);
+	mutex_unlock(&ctx->buffer_mutex);
+
+	v4l2_m2m_job_finish(ctx->dev->m2m_dev, ctx->fh.m2m_ctx);
+}
+
+static int coda_job_ready(void *m2m_priv)
+{
+	struct coda_ctx *ctx = m2m_priv;
+	int src_bufs = v4l2_m2m_num_src_bufs_ready(ctx->fh.m2m_ctx);
+
+	/*
+	 * For both 'P' and 'key' frame cases 1 picture
+	 * and 1 frame are needed. In the decoder case,
+	 * the compressed frame can be in the bitstream.
+	 */
+	if (!src_bufs && ctx->inst_type != CODA_INST_DECODER) {
+		coda_dbg(1, ctx, "not ready: not enough vid-out buffers.\n");
+		return 0;
+	}
+
+	if (!v4l2_m2m_num_dst_bufs_ready(ctx->fh.m2m_ctx)) {
+		coda_dbg(1, ctx, "not ready: not enough vid-cap buffers.\n");
+		return 0;
+	}
+
+	if (ctx->inst_type == CODA_INST_DECODER && ctx->use_bit) {
+		bool stream_end = ctx->bit_stream_param &
+				  CODA_BIT_STREAM_END_FLAG;
+		int num_metas = ctx->num_metas;
+		struct coda_buffer_meta *meta;
+		unsigned int count;
+
+		count = hweight32(ctx->frm_dis_flg);
+		if (ctx->use_vdoa && count >= (ctx->num_internal_frames - 1)) {
+			coda_dbg(1, ctx,
+				 "not ready: all internal buffers in use: %d/%d (0x%x)",
+				 count, ctx->num_internal_frames,
+				 ctx->frm_dis_flg);
+			return 0;
+		}
+
+		if (ctx->hold && !src_bufs) {
+			coda_dbg(1, ctx,
+				 "not ready: on hold for more buffers.\n");
+			return 0;
+		}
+
+		if (!stream_end && (num_metas + src_bufs) < 2) {
+			coda_dbg(1, ctx,
+				 "not ready: need 2 buffers available (queue:%d + bitstream:%d)\n",
+				 num_metas, src_bufs);
+			return 0;
+		}
+
+		meta = list_first_entry(&ctx->buffer_meta_list,
+					struct coda_buffer_meta, list);
+		if (!coda_bitstream_can_fetch_past(ctx, meta->end) &&
+		    !stream_end) {
+			coda_dbg(1, ctx,
+				 "not ready: not enough bitstream data to read past %u (%u)\n",
+				 meta->end, ctx->bitstream_fifo.kfifo.in);
+			return 0;
+		}
+	}
+
+	if (ctx->aborting) {
+		coda_dbg(1, ctx, "not ready: aborting\n");
+		return 0;
+	}
+
+	coda_dbg(2, ctx, "job ready\n");
+
+	return 1;
+}
+
+static void coda_job_abort(void *priv)
+{
+	struct coda_ctx *ctx = priv;
+
+	ctx->aborting = 1;
+
+	coda_dbg(1, ctx, "job abort\n");
+}
+
+static const struct v4l2_m2m_ops coda_m2m_ops = {
+	.device_run	= coda_device_run,
+	.job_ready	= coda_job_ready,
+	.job_abort	= coda_job_abort,
+};
+
+static void set_default_params(struct coda_ctx *ctx)
+{
+	unsigned int max_w, max_h, usize, csize;
+
+	ctx->codec = coda_find_codec(ctx->dev, ctx->cvd->src_formats[0],
+				     ctx->cvd->dst_formats[0]);
+	max_w = min(ctx->codec->max_w, 1920U);
+	max_h = min(ctx->codec->max_h, 1088U);
+	usize = max_w * max_h * 3 / 2;
+	csize = coda_estimate_sizeimage(ctx, usize, max_w, max_h);
+
+	ctx->params.codec_mode = ctx->codec->mode;
+	if (ctx->cvd->src_formats[0] == V4L2_PIX_FMT_JPEG ||
+	    ctx->cvd->dst_formats[0] == V4L2_PIX_FMT_JPEG) {
+		ctx->colorspace = V4L2_COLORSPACE_SRGB;
+		ctx->xfer_func = V4L2_XFER_FUNC_SRGB;
+		ctx->ycbcr_enc = V4L2_YCBCR_ENC_601;
+		ctx->quantization = V4L2_QUANTIZATION_FULL_RANGE;
+	} else {
+		ctx->colorspace = V4L2_COLORSPACE_REC709;
+		ctx->xfer_func = V4L2_XFER_FUNC_DEFAULT;
+		ctx->ycbcr_enc = V4L2_YCBCR_ENC_DEFAULT;
+		ctx->quantization = V4L2_QUANTIZATION_DEFAULT;
+	}
+	ctx->params.framerate = 30;
+
+	/* Default formats for output and input queues */
+	ctx->q_data[V4L2_M2M_SRC].fourcc = ctx->cvd->src_formats[0];
+	ctx->q_data[V4L2_M2M_DST].fourcc = ctx->cvd->dst_formats[0];
+	ctx->q_data[V4L2_M2M_SRC].width = max_w;
+	ctx->q_data[V4L2_M2M_SRC].height = max_h;
+	ctx->q_data[V4L2_M2M_DST].width = max_w;
+	ctx->q_data[V4L2_M2M_DST].height = max_h;
+	if (ctx->codec->src_fourcc == V4L2_PIX_FMT_YUV420) {
+		ctx->q_data[V4L2_M2M_SRC].bytesperline = max_w;
+		ctx->q_data[V4L2_M2M_SRC].sizeimage = usize;
+		ctx->q_data[V4L2_M2M_DST].bytesperline = 0;
+		ctx->q_data[V4L2_M2M_DST].sizeimage = csize;
+	} else {
+		ctx->q_data[V4L2_M2M_SRC].bytesperline = 0;
+		ctx->q_data[V4L2_M2M_SRC].sizeimage = csize;
+		ctx->q_data[V4L2_M2M_DST].bytesperline = max_w;
+		ctx->q_data[V4L2_M2M_DST].sizeimage = usize;
+	}
+	ctx->q_data[V4L2_M2M_SRC].rect.width = max_w;
+	ctx->q_data[V4L2_M2M_SRC].rect.height = max_h;
+	ctx->q_data[V4L2_M2M_DST].rect.width = max_w;
+	ctx->q_data[V4L2_M2M_DST].rect.height = max_h;
+
+	/*
+	 * Since the RBC2AXI logic only supports a single chroma plane,
+	 * macroblock tiling only works for to NV12 pixel format.
+	 */
+	ctx->tiled_map_type = GDI_LINEAR_FRAME_MAP;
+}
+
+/*
+ * Queue operations
+ */
+static int coda_queue_setup(struct vb2_queue *vq,
+				unsigned int *nbuffers, unsigned int *nplanes,
+				unsigned int sizes[], struct device *alloc_devs[])
+{
+	struct coda_ctx *ctx = vb2_get_drv_priv(vq);
+	struct coda_q_data *q_data;
+	unsigned int size;
+
+	q_data = get_q_data(ctx, vq->type);
+	size = q_data->sizeimage;
+
+	if (*nplanes)
+		return sizes[0] < size ? -EINVAL : 0;
+
+	*nplanes = 1;
+	sizes[0] = size;
+
+	coda_dbg(1, ctx, "get %d buffer(s) of size %d each.\n", *nbuffers,
+		 size);
+
+	return 0;
+}
+
+static int coda_buf_prepare(struct vb2_buffer *vb)
+{
+	struct vb2_v4l2_buffer *vbuf = to_vb2_v4l2_buffer(vb);
+	struct coda_ctx *ctx = vb2_get_drv_priv(vb->vb2_queue);
+	struct coda_q_data *q_data;
+
+	q_data = get_q_data(ctx, vb->vb2_queue->type);
+	if (V4L2_TYPE_IS_OUTPUT(vb->vb2_queue->type)) {
+		if (vbuf->field == V4L2_FIELD_ANY)
+			vbuf->field = V4L2_FIELD_NONE;
+		if (vbuf->field != V4L2_FIELD_NONE) {
+			v4l2_warn(&ctx->dev->v4l2_dev,
+				  "%s field isn't supported\n", __func__);
+			return -EINVAL;
+		}
+	}
+
+	if (vb2_plane_size(vb, 0) < q_data->sizeimage) {
+		v4l2_warn(&ctx->dev->v4l2_dev,
+			  "%s data will not fit into plane (%lu < %lu)\n",
+			  __func__, vb2_plane_size(vb, 0),
+			  (long)q_data->sizeimage);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static void coda_update_menu_ctrl(struct v4l2_ctrl *ctrl, int value)
+{
+	if (!ctrl)
+		return;
+
+	v4l2_ctrl_lock(ctrl);
+
+	/*
+	 * Extend the control range if the parsed stream contains a known but
+	 * unsupported value or level.
+	 */
+	if (value > ctrl->maximum) {
+		__v4l2_ctrl_modify_range(ctrl, ctrl->minimum, value,
+			ctrl->menu_skip_mask & ~(1 << value),
+			ctrl->default_value);
+	} else if (value < ctrl->minimum) {
+		__v4l2_ctrl_modify_range(ctrl, value, ctrl->maximum,
+			ctrl->menu_skip_mask & ~(1 << value),
+			ctrl->default_value);
+	}
+
+	__v4l2_ctrl_s_ctrl(ctrl, value);
+
+	v4l2_ctrl_unlock(ctrl);
+}
+
+void coda_update_profile_level_ctrls(struct coda_ctx *ctx, u8 profile_idc,
+				     u8 level_idc)
+{
+	const char * const *profile_names;
+	const char * const *level_names;
+	struct v4l2_ctrl *profile_ctrl;
+	struct v4l2_ctrl *level_ctrl;
+	const char *codec_name;
+	u32 profile_cid;
+	u32 level_cid;
+	int profile;
+	int level;
+
+	switch (ctx->codec->src_fourcc) {
+	case V4L2_PIX_FMT_H264:
+		codec_name = "H264";
+		profile_cid = V4L2_CID_MPEG_VIDEO_H264_PROFILE;
+		level_cid = V4L2_CID_MPEG_VIDEO_H264_LEVEL;
+		profile_ctrl = ctx->h264_profile_ctrl;
+		level_ctrl = ctx->h264_level_ctrl;
+		profile = coda_h264_profile(profile_idc);
+		level = coda_h264_level(level_idc);
+		break;
+	case V4L2_PIX_FMT_MPEG2:
+		codec_name = "MPEG-2";
+		profile_cid = V4L2_CID_MPEG_VIDEO_MPEG2_PROFILE;
+		level_cid = V4L2_CID_MPEG_VIDEO_MPEG2_LEVEL;
+		profile_ctrl = ctx->mpeg2_profile_ctrl;
+		level_ctrl = ctx->mpeg2_level_ctrl;
+		profile = coda_mpeg2_profile(profile_idc);
+		level = coda_mpeg2_level(level_idc);
+		break;
+	case V4L2_PIX_FMT_MPEG4:
+		codec_name = "MPEG-4";
+		profile_cid = V4L2_CID_MPEG_VIDEO_MPEG4_PROFILE;
+		level_cid = V4L2_CID_MPEG_VIDEO_MPEG4_LEVEL;
+		profile_ctrl = ctx->mpeg4_profile_ctrl;
+		level_ctrl = ctx->mpeg4_level_ctrl;
+		profile = coda_mpeg4_profile(profile_idc);
+		level = coda_mpeg4_level(level_idc);
+		break;
+	default:
+		return;
+	}
+
+	profile_names = v4l2_ctrl_get_menu(profile_cid);
+	level_names = v4l2_ctrl_get_menu(level_cid);
+
+	if (profile < 0) {
+		v4l2_warn(&ctx->dev->v4l2_dev, "Invalid %s profile: %u\n",
+			  codec_name, profile_idc);
+	} else {
+		coda_dbg(1, ctx, "Parsed %s profile: %s\n", codec_name,
+			 profile_names[profile]);
+		coda_update_menu_ctrl(profile_ctrl, profile);
+	}
+
+	if (level < 0) {
+		v4l2_warn(&ctx->dev->v4l2_dev, "Invalid %s level: %u\n",
+			  codec_name, level_idc);
+	} else {
+		coda_dbg(1, ctx, "Parsed %s level: %s\n", codec_name,
+			 level_names[level]);
+		coda_update_menu_ctrl(level_ctrl, level);
+	}
+}
+
+static void coda_queue_source_change_event(struct coda_ctx *ctx)
+{
+	static const struct v4l2_event source_change_event = {
+		.type = V4L2_EVENT_SOURCE_CHANGE,
+		.u.src_change.changes = V4L2_EVENT_SRC_CH_RESOLUTION,
+	};
+
+	v4l2_event_queue_fh(&ctx->fh, &source_change_event);
+}
+
+static void coda_buf_queue(struct vb2_buffer *vb)
+{
+	struct vb2_v4l2_buffer *vbuf = to_vb2_v4l2_buffer(vb);
+	struct coda_ctx *ctx = vb2_get_drv_priv(vb->vb2_queue);
+	struct vb2_queue *vq = vb->vb2_queue;
+	struct coda_q_data *q_data;
+
+	q_data = get_q_data(ctx, vb->vb2_queue->type);
+
+	/*
+	 * In the decoder case, immediately try to copy the buffer into the
+	 * bitstream ringbuffer and mark it as ready to be dequeued.
+	 */
+	if (ctx->bitstream.size && vq->type == V4L2_BUF_TYPE_VIDEO_OUTPUT) {
+		/*
+		 * For backwards compatibility, queuing an empty buffer marks
+		 * the stream end
+		 */
+		if (vb2_get_plane_payload(vb, 0) == 0)
+			coda_bit_stream_end_flag(ctx);
+
+		if (q_data->fourcc == V4L2_PIX_FMT_H264) {
+			/*
+			 * Unless already done, try to obtain profile_idc and
+			 * level_idc from the SPS header. This allows to decide
+			 * whether to enable reordering during sequence
+			 * initialization.
+			 */
+			if (!ctx->params.h264_profile_idc) {
+				coda_sps_parse_profile(ctx, vb);
+				coda_update_profile_level_ctrls(ctx,
+						ctx->params.h264_profile_idc,
+						ctx->params.h264_level_idc);
+			}
+		}
+
+		mutex_lock(&ctx->bitstream_mutex);
+		v4l2_m2m_buf_queue(ctx->fh.m2m_ctx, vbuf);
+		if (vb2_is_streaming(vb->vb2_queue))
+			/* This set buf->sequence = ctx->qsequence++ */
+			coda_fill_bitstream(ctx, NULL);
+		mutex_unlock(&ctx->bitstream_mutex);
+
+		if (!ctx->initialized) {
+			/*
+			 * Run sequence initialization in case the queued
+			 * buffer contained headers.
+			 */
+			if (vb2_is_streaming(vb->vb2_queue) &&
+			    ctx->ops->seq_init_work) {
+				queue_work(ctx->dev->workqueue,
+					   &ctx->seq_init_work);
+				flush_work(&ctx->seq_init_work);
+			}
+
+			if (ctx->initialized)
+				coda_queue_source_change_event(ctx);
+		}
+	} else {
+		if ((ctx->inst_type == CODA_INST_ENCODER || !ctx->use_bit) &&
+		    vq->type == V4L2_BUF_TYPE_VIDEO_OUTPUT)
+			vbuf->sequence = ctx->qsequence++;
+		v4l2_m2m_buf_queue(ctx->fh.m2m_ctx, vbuf);
+	}
+}
+
+int coda_alloc_aux_buf(struct coda_dev *dev, struct coda_aux_buf *buf,
+		       size_t size, const char *name, struct dentry *parent)
+{
+	buf->vaddr = dma_alloc_coherent(dev->dev, size, &buf->paddr,
+					GFP_KERNEL);
+	if (!buf->vaddr) {
+		v4l2_err(&dev->v4l2_dev,
+			 "Failed to allocate %s buffer of size %zu\n",
+			 name, size);
+		return -ENOMEM;
+	}
+
+	buf->size = size;
+
+	if (name && parent) {
+		buf->blob.data = buf->vaddr;
+		buf->blob.size = size;
+		buf->dentry = debugfs_create_blob(name, 0444, parent,
+						  &buf->blob);
+	}
+
+	return 0;
+}
+
+void coda_free_aux_buf(struct coda_dev *dev,
+		       struct coda_aux_buf *buf)
+{
+	if (buf->vaddr) {
+		dma_free_coherent(dev->dev, buf->size, buf->vaddr, buf->paddr);
+		buf->vaddr = NULL;
+		buf->size = 0;
+		debugfs_remove(buf->dentry);
+		buf->dentry = NULL;
+	}
+}
+
+static int coda_start_streaming(struct vb2_queue *q, unsigned int count)
+{
+	struct coda_ctx *ctx = vb2_get_drv_priv(q);
+	struct v4l2_device *v4l2_dev = &ctx->dev->v4l2_dev;
+	struct coda_q_data *q_data_src, *q_data_dst;
+	struct v4l2_m2m_buffer *m2m_buf, *tmp;
+	struct vb2_v4l2_buffer *buf;
+	struct list_head list;
+	int ret = 0;
+
+	if (count < 1)
+		return -EINVAL;
+
+	coda_dbg(1, ctx, "start streaming %s\n", v4l2_type_names[q->type]);
+
+	INIT_LIST_HEAD(&list);
+
+	q_data_src = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+	if (q->type == V4L2_BUF_TYPE_VIDEO_OUTPUT) {
+		if (ctx->inst_type == CODA_INST_DECODER && ctx->use_bit) {
+			/* copy the buffers that were queued before streamon */
+			mutex_lock(&ctx->bitstream_mutex);
+			coda_fill_bitstream(ctx, &list);
+			mutex_unlock(&ctx->bitstream_mutex);
+
+			if (ctx->dev->devtype->product != CODA_960 &&
+			    coda_get_bitstream_payload(ctx) < 512) {
+				v4l2_err(v4l2_dev, "start payload < 512\n");
+				ret = -EINVAL;
+				goto err;
+			}
+
+			if (!ctx->initialized) {
+				/* Run sequence initialization */
+				if (ctx->ops->seq_init_work) {
+					queue_work(ctx->dev->workqueue,
+						   &ctx->seq_init_work);
+					flush_work(&ctx->seq_init_work);
+				}
+			}
+		}
+
+		/*
+		 * Check the first input JPEG buffer to determine chroma
+		 * subsampling.
+		 */
+		if (q_data_src->fourcc == V4L2_PIX_FMT_JPEG) {
+			buf = v4l2_m2m_next_src_buf(ctx->fh.m2m_ctx);
+			coda_jpeg_decode_header(ctx, &buf->vb2_buf);
+			/*
+			 * We have to start streaming even if the first buffer
+			 * does not contain a valid JPEG image. The error will
+			 * be caught during device run and will be signalled
+			 * via the capture buffer error flag.
+			 */
+
+			q_data_dst = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_CAPTURE);
+			q_data_dst->width = round_up(q_data_src->width, 16);
+			q_data_dst->height = round_up(q_data_src->height, 16);
+			q_data_dst->bytesperline = q_data_dst->width;
+			if (ctx->params.jpeg_chroma_subsampling ==
+			    V4L2_JPEG_CHROMA_SUBSAMPLING_420) {
+				q_data_dst->sizeimage =
+						q_data_dst->bytesperline *
+						q_data_dst->height * 3 / 2;
+				if (q_data_dst->fourcc != V4L2_PIX_FMT_YUV420)
+					q_data_dst->fourcc = V4L2_PIX_FMT_NV12;
+			} else {
+				q_data_dst->sizeimage =
+						q_data_dst->bytesperline *
+						q_data_dst->height * 2;
+				q_data_dst->fourcc = V4L2_PIX_FMT_YUV422P;
+			}
+			q_data_dst->rect.left = 0;
+			q_data_dst->rect.top = 0;
+			q_data_dst->rect.width = q_data_src->width;
+			q_data_dst->rect.height = q_data_src->height;
+		}
+		ctx->streamon_out = 1;
+	} else {
+		ctx->streamon_cap = 1;
+	}
+
+	/* Don't start the coda unless both queues are on */
+	if (!(ctx->streamon_out && ctx->streamon_cap))
+		goto out;
+
+	q_data_dst = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_CAPTURE);
+	if ((q_data_src->rect.width != q_data_dst->width &&
+	     round_up(q_data_src->rect.width, 16) != q_data_dst->width) ||
+	    (q_data_src->rect.height != q_data_dst->height &&
+	     round_up(q_data_src->rect.height, 16) != q_data_dst->height)) {
+		v4l2_err(v4l2_dev, "can't convert %dx%d to %dx%d\n",
+			 q_data_src->rect.width, q_data_src->rect.height,
+			 q_data_dst->width, q_data_dst->height);
+		ret = -EINVAL;
+		goto err;
+	}
+
+	/* Allow BIT decoder device_run with no new buffers queued */
+	if (ctx->inst_type == CODA_INST_DECODER && ctx->use_bit)
+		v4l2_m2m_set_src_buffered(ctx->fh.m2m_ctx, true);
+
+	ctx->gopcounter = ctx->params.gop_size - 1;
+
+	if (q_data_dst->fourcc == V4L2_PIX_FMT_JPEG)
+		ctx->params.gop_size = 1;
+	ctx->gopcounter = ctx->params.gop_size - 1;
+	/* Only decoders have this control */
+	if (ctx->mb_err_cnt_ctrl)
+		v4l2_ctrl_s_ctrl(ctx->mb_err_cnt_ctrl, 0);
+
+	ret = ctx->ops->start_streaming(ctx);
+	if (ctx->inst_type == CODA_INST_DECODER) {
+		if (ret == -EAGAIN)
+			goto out;
+	}
+	if (ret < 0)
+		goto err;
+
+out:
+	if (q->type == V4L2_BUF_TYPE_VIDEO_OUTPUT) {
+		list_for_each_entry_safe(m2m_buf, tmp, &list, list) {
+			list_del(&m2m_buf->list);
+			v4l2_m2m_buf_done(&m2m_buf->vb, VB2_BUF_STATE_DONE);
+		}
+	}
+	return 0;
+
+err:
+	if (q->type == V4L2_BUF_TYPE_VIDEO_OUTPUT) {
+		list_for_each_entry_safe(m2m_buf, tmp, &list, list) {
+			list_del(&m2m_buf->list);
+			v4l2_m2m_buf_done(&m2m_buf->vb, VB2_BUF_STATE_QUEUED);
+		}
+		while ((buf = v4l2_m2m_src_buf_remove(ctx->fh.m2m_ctx)))
+			v4l2_m2m_buf_done(buf, VB2_BUF_STATE_QUEUED);
+	} else {
+		while ((buf = v4l2_m2m_dst_buf_remove(ctx->fh.m2m_ctx)))
+			v4l2_m2m_buf_done(buf, VB2_BUF_STATE_QUEUED);
+	}
+	return ret;
+}
+
+static void coda_stop_streaming(struct vb2_queue *q)
+{
+	struct coda_ctx *ctx = vb2_get_drv_priv(q);
+	struct coda_dev *dev = ctx->dev;
+	struct vb2_v4l2_buffer *buf;
+	bool stop;
+
+	stop = ctx->streamon_out && ctx->streamon_cap;
+
+	coda_dbg(1, ctx, "stop streaming %s\n", v4l2_type_names[q->type]);
+
+	if (q->type == V4L2_BUF_TYPE_VIDEO_OUTPUT) {
+		ctx->streamon_out = 0;
+
+		coda_bit_stream_end_flag(ctx);
+
+		ctx->qsequence = 0;
+
+		while ((buf = v4l2_m2m_src_buf_remove(ctx->fh.m2m_ctx)))
+			v4l2_m2m_buf_done(buf, VB2_BUF_STATE_ERROR);
+	} else {
+		ctx->streamon_cap = 0;
+
+		ctx->osequence = 0;
+		ctx->sequence_offset = 0;
+
+		while ((buf = v4l2_m2m_dst_buf_remove(ctx->fh.m2m_ctx)))
+			v4l2_m2m_buf_done(buf, VB2_BUF_STATE_ERROR);
+	}
+
+	if (stop) {
+		struct coda_buffer_meta *meta;
+
+		if (ctx->ops->seq_end_work) {
+			queue_work(dev->workqueue, &ctx->seq_end_work);
+			flush_work(&ctx->seq_end_work);
+		}
+		spin_lock(&ctx->buffer_meta_lock);
+		while (!list_empty(&ctx->buffer_meta_list)) {
+			meta = list_first_entry(&ctx->buffer_meta_list,
+						struct coda_buffer_meta, list);
+			list_del(&meta->list);
+			kfree(meta);
+		}
+		ctx->num_metas = 0;
+		spin_unlock(&ctx->buffer_meta_lock);
+		kfifo_init(&ctx->bitstream_fifo,
+			ctx->bitstream.vaddr, ctx->bitstream.size);
+		ctx->runcounter = 0;
+		ctx->aborting = 0;
+		ctx->hold = false;
+	}
+
+	if (!ctx->streamon_out && !ctx->streamon_cap)
+		ctx->bit_stream_param &= ~CODA_BIT_STREAM_END_FLAG;
+}
+
+static const struct vb2_ops coda_qops = {
+	.queue_setup		= coda_queue_setup,
+	.buf_prepare		= coda_buf_prepare,
+	.buf_queue		= coda_buf_queue,
+	.start_streaming	= coda_start_streaming,
+	.stop_streaming		= coda_stop_streaming,
+	.wait_prepare		= vb2_ops_wait_prepare,
+	.wait_finish		= vb2_ops_wait_finish,
+};
+
+static int coda_s_ctrl(struct v4l2_ctrl *ctrl)
+{
+	const char * const *val_names = v4l2_ctrl_get_menu(ctrl->id);
+	struct coda_ctx *ctx =
+			container_of(ctrl->handler, struct coda_ctx, ctrls);
+
+	if (val_names)
+		coda_dbg(2, ctx, "s_ctrl: id = 0x%x, name = \"%s\", val = %d (\"%s\")\n",
+			 ctrl->id, ctrl->name, ctrl->val, val_names[ctrl->val]);
+	else
+		coda_dbg(2, ctx, "s_ctrl: id = 0x%x, name = \"%s\", val = %d\n",
+			 ctrl->id, ctrl->name, ctrl->val);
+
+	switch (ctrl->id) {
+	case V4L2_CID_HFLIP:
+		if (ctrl->val)
+			ctx->params.rot_mode |= CODA_MIR_HOR;
+		else
+			ctx->params.rot_mode &= ~CODA_MIR_HOR;
+		break;
+	case V4L2_CID_VFLIP:
+		if (ctrl->val)
+			ctx->params.rot_mode |= CODA_MIR_VER;
+		else
+			ctx->params.rot_mode &= ~CODA_MIR_VER;
+		break;
+	case V4L2_CID_MPEG_VIDEO_BITRATE:
+		ctx->params.bitrate = ctrl->val / 1000;
+		ctx->params.bitrate_changed = true;
+		break;
+	case V4L2_CID_MPEG_VIDEO_GOP_SIZE:
+		ctx->params.gop_size = ctrl->val;
+		break;
+	case V4L2_CID_MPEG_VIDEO_H264_I_FRAME_QP:
+		ctx->params.h264_intra_qp = ctrl->val;
+		ctx->params.h264_intra_qp_changed = true;
+		break;
+	case V4L2_CID_MPEG_VIDEO_H264_P_FRAME_QP:
+		ctx->params.h264_inter_qp = ctrl->val;
+		break;
+	case V4L2_CID_MPEG_VIDEO_H264_MIN_QP:
+		ctx->params.h264_min_qp = ctrl->val;
+		break;
+	case V4L2_CID_MPEG_VIDEO_H264_MAX_QP:
+		ctx->params.h264_max_qp = ctrl->val;
+		break;
+	case V4L2_CID_MPEG_VIDEO_H264_LOOP_FILTER_ALPHA:
+		ctx->params.h264_slice_alpha_c0_offset_div2 = ctrl->val;
+		break;
+	case V4L2_CID_MPEG_VIDEO_H264_LOOP_FILTER_BETA:
+		ctx->params.h264_slice_beta_offset_div2 = ctrl->val;
+		break;
+	case V4L2_CID_MPEG_VIDEO_H264_LOOP_FILTER_MODE:
+		ctx->params.h264_disable_deblocking_filter_idc = ctrl->val;
+		break;
+	case V4L2_CID_MPEG_VIDEO_H264_CONSTRAINED_INTRA_PREDICTION:
+		ctx->params.h264_constrained_intra_pred_flag = ctrl->val;
+		break;
+	case V4L2_CID_MPEG_VIDEO_FRAME_RC_ENABLE:
+		ctx->params.frame_rc_enable = ctrl->val;
+		break;
+	case V4L2_CID_MPEG_VIDEO_MB_RC_ENABLE:
+		ctx->params.mb_rc_enable = ctrl->val;
+		break;
+	case V4L2_CID_MPEG_VIDEO_H264_CHROMA_QP_INDEX_OFFSET:
+		ctx->params.h264_chroma_qp_index_offset = ctrl->val;
+		break;
+	case V4L2_CID_MPEG_VIDEO_H264_PROFILE:
+		/* TODO: switch between baseline and constrained baseline */
+		if (ctx->inst_type == CODA_INST_ENCODER)
+			ctx->params.h264_profile_idc = 66;
+		break;
+	case V4L2_CID_MPEG_VIDEO_H264_LEVEL:
+		/* nothing to do, this is set by the encoder */
+		break;
+	case V4L2_CID_MPEG_VIDEO_MPEG4_I_FRAME_QP:
+		ctx->params.mpeg4_intra_qp = ctrl->val;
+		break;
+	case V4L2_CID_MPEG_VIDEO_MPEG4_P_FRAME_QP:
+		ctx->params.mpeg4_inter_qp = ctrl->val;
+		break;
+	case V4L2_CID_MPEG_VIDEO_MPEG2_PROFILE:
+	case V4L2_CID_MPEG_VIDEO_MPEG2_LEVEL:
+	case V4L2_CID_MPEG_VIDEO_MPEG4_PROFILE:
+	case V4L2_CID_MPEG_VIDEO_MPEG4_LEVEL:
+		/* nothing to do, these are fixed */
+		break;
+	case V4L2_CID_MPEG_VIDEO_MULTI_SLICE_MODE:
+		ctx->params.slice_mode = ctrl->val;
+		ctx->params.slice_mode_changed = true;
+		break;
+	case V4L2_CID_MPEG_VIDEO_MULTI_SLICE_MAX_MB:
+		ctx->params.slice_max_mb = ctrl->val;
+		ctx->params.slice_mode_changed = true;
+		break;
+	case V4L2_CID_MPEG_VIDEO_MULTI_SLICE_MAX_BYTES:
+		ctx->params.slice_max_bits = ctrl->val * 8;
+		ctx->params.slice_mode_changed = true;
+		break;
+	case V4L2_CID_MPEG_VIDEO_HEADER_MODE:
+		break;
+	case V4L2_CID_MPEG_VIDEO_CYCLIC_INTRA_REFRESH_MB:
+		ctx->params.intra_refresh = ctrl->val;
+		ctx->params.intra_refresh_changed = true;
+		break;
+	case V4L2_CID_MPEG_VIDEO_FORCE_KEY_FRAME:
+		ctx->params.force_ipicture = true;
+		break;
+	case V4L2_CID_JPEG_COMPRESSION_QUALITY:
+		coda_set_jpeg_compression_quality(ctx, ctrl->val);
+		break;
+	case V4L2_CID_JPEG_RESTART_INTERVAL:
+		ctx->params.jpeg_restart_interval = ctrl->val;
+		break;
+	case V4L2_CID_MPEG_VIDEO_VBV_DELAY:
+		ctx->params.vbv_delay = ctrl->val;
+		break;
+	case V4L2_CID_MPEG_VIDEO_VBV_SIZE:
+		ctx->params.vbv_size = min(ctrl->val * 8192, 0x7fffffff);
+		break;
+	default:
+		coda_dbg(1, ctx, "Invalid control, id=%d, val=%d\n",
+			 ctrl->id, ctrl->val);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static const struct v4l2_ctrl_ops coda_ctrl_ops = {
+	.s_ctrl = coda_s_ctrl,
+};
+
+static void coda_encode_ctrls(struct coda_ctx *ctx)
+{
+	int max_gop_size = (ctx->dev->devtype->product == CODA_DX6) ? 60 : 99;
+
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_BITRATE, 0, 32767000, 1000, 0);
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_GOP_SIZE, 0, max_gop_size, 1, 16);
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_H264_I_FRAME_QP, 0, 51, 1, 25);
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_H264_P_FRAME_QP, 0, 51, 1, 25);
+	if (ctx->dev->devtype->product != CODA_960) {
+		v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+			V4L2_CID_MPEG_VIDEO_H264_MIN_QP, 0, 51, 1, 12);
+	}
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_H264_MAX_QP, 0, 51, 1, 51);
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_H264_LOOP_FILTER_ALPHA, -6, 6, 1, 0);
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_H264_LOOP_FILTER_BETA, -6, 6, 1, 0);
+	v4l2_ctrl_new_std_menu(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_H264_LOOP_FILTER_MODE,
+		V4L2_MPEG_VIDEO_H264_LOOP_FILTER_MODE_DISABLED_AT_SLICE_BOUNDARY,
+		0x0, V4L2_MPEG_VIDEO_H264_LOOP_FILTER_MODE_ENABLED);
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_H264_CONSTRAINED_INTRA_PREDICTION, 0, 1, 1,
+		0);
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_FRAME_RC_ENABLE, 0, 1, 1, 1);
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_MB_RC_ENABLE, 0, 1, 1, 1);
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_H264_CHROMA_QP_INDEX_OFFSET, -12, 12, 1, 0);
+	v4l2_ctrl_new_std_menu(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_H264_PROFILE,
+		V4L2_MPEG_VIDEO_H264_PROFILE_CONSTRAINED_BASELINE, 0x0,
+		V4L2_MPEG_VIDEO_H264_PROFILE_CONSTRAINED_BASELINE);
+	if (ctx->dev->devtype->product == CODA_HX4 ||
+	    ctx->dev->devtype->product == CODA_7541) {
+		v4l2_ctrl_new_std_menu(&ctx->ctrls, &coda_ctrl_ops,
+			V4L2_CID_MPEG_VIDEO_H264_LEVEL,
+			V4L2_MPEG_VIDEO_H264_LEVEL_3_1,
+			~((1 << V4L2_MPEG_VIDEO_H264_LEVEL_2_0) |
+			  (1 << V4L2_MPEG_VIDEO_H264_LEVEL_3_0) |
+			  (1 << V4L2_MPEG_VIDEO_H264_LEVEL_3_1)),
+			V4L2_MPEG_VIDEO_H264_LEVEL_3_1);
+	}
+	if (ctx->dev->devtype->product == CODA_960) {
+		v4l2_ctrl_new_std_menu(&ctx->ctrls, &coda_ctrl_ops,
+			V4L2_CID_MPEG_VIDEO_H264_LEVEL,
+			V4L2_MPEG_VIDEO_H264_LEVEL_4_2,
+			~((1 << V4L2_MPEG_VIDEO_H264_LEVEL_1_0) |
+			  (1 << V4L2_MPEG_VIDEO_H264_LEVEL_2_0) |
+			  (1 << V4L2_MPEG_VIDEO_H264_LEVEL_3_0) |
+			  (1 << V4L2_MPEG_VIDEO_H264_LEVEL_3_1) |
+			  (1 << V4L2_MPEG_VIDEO_H264_LEVEL_3_2) |
+			  (1 << V4L2_MPEG_VIDEO_H264_LEVEL_4_0) |
+			  (1 << V4L2_MPEG_VIDEO_H264_LEVEL_4_1) |
+			  (1 << V4L2_MPEG_VIDEO_H264_LEVEL_4_2)),
+			V4L2_MPEG_VIDEO_H264_LEVEL_4_0);
+	}
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_MPEG4_I_FRAME_QP, 1, 31, 1, 2);
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_MPEG4_P_FRAME_QP, 1, 31, 1, 2);
+	v4l2_ctrl_new_std_menu(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_MPEG4_PROFILE,
+		V4L2_MPEG_VIDEO_MPEG4_PROFILE_SIMPLE, 0x0,
+		V4L2_MPEG_VIDEO_MPEG4_PROFILE_SIMPLE);
+	if (ctx->dev->devtype->product == CODA_HX4 ||
+	    ctx->dev->devtype->product == CODA_7541 ||
+	    ctx->dev->devtype->product == CODA_960) {
+		v4l2_ctrl_new_std_menu(&ctx->ctrls, &coda_ctrl_ops,
+			V4L2_CID_MPEG_VIDEO_MPEG4_LEVEL,
+			V4L2_MPEG_VIDEO_MPEG4_LEVEL_5,
+			~(1 << V4L2_MPEG_VIDEO_MPEG4_LEVEL_5),
+			V4L2_MPEG_VIDEO_MPEG4_LEVEL_5);
+	}
+	v4l2_ctrl_new_std_menu(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_MULTI_SLICE_MODE,
+		V4L2_MPEG_VIDEO_MULTI_SLICE_MODE_MAX_BYTES, 0x0,
+		V4L2_MPEG_VIDEO_MULTI_SLICE_MODE_SINGLE);
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_MULTI_SLICE_MAX_MB, 1, 0x3fffffff, 1, 1);
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_MULTI_SLICE_MAX_BYTES, 1, 0x3fffffff, 1,
+		500);
+	v4l2_ctrl_new_std_menu(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_HEADER_MODE,
+		V4L2_MPEG_VIDEO_HEADER_MODE_JOINED_WITH_1ST_FRAME,
+		(1 << V4L2_MPEG_VIDEO_HEADER_MODE_SEPARATE),
+		V4L2_MPEG_VIDEO_HEADER_MODE_JOINED_WITH_1ST_FRAME);
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_CYCLIC_INTRA_REFRESH_MB, 0,
+		1920 * 1088 / 256, 1, 0);
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_VBV_DELAY, 0, 0x7fff, 1, 0);
+	/*
+	 * The maximum VBV size value is 0x7fffffff bits,
+	 * one bit less than 262144 KiB
+	 */
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_MPEG_VIDEO_VBV_SIZE, 0, 262144, 1, 0);
+}
+
+static void coda_jpeg_encode_ctrls(struct coda_ctx *ctx)
+{
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_JPEG_COMPRESSION_QUALITY, 5, 100, 1, 50);
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_JPEG_RESTART_INTERVAL, 0, 100, 1, 0);
+}
+
+static void coda_decode_ctrls(struct coda_ctx *ctx)
+{
+	u8 max;
+
+	ctx->h264_profile_ctrl = v4l2_ctrl_new_std_menu(&ctx->ctrls,
+		&coda_ctrl_ops, V4L2_CID_MPEG_VIDEO_H264_PROFILE,
+		V4L2_MPEG_VIDEO_H264_PROFILE_HIGH,
+		~((1 << V4L2_MPEG_VIDEO_H264_PROFILE_CONSTRAINED_BASELINE) |
+		  (1 << V4L2_MPEG_VIDEO_H264_PROFILE_MAIN) |
+		  (1 << V4L2_MPEG_VIDEO_H264_PROFILE_HIGH)),
+		V4L2_MPEG_VIDEO_H264_PROFILE_HIGH);
+	if (ctx->h264_profile_ctrl)
+		ctx->h264_profile_ctrl->flags |= V4L2_CTRL_FLAG_READ_ONLY;
+
+	if (ctx->dev->devtype->product == CODA_HX4 ||
+	    ctx->dev->devtype->product == CODA_7541)
+		max = V4L2_MPEG_VIDEO_H264_LEVEL_4_0;
+	else if (ctx->dev->devtype->product == CODA_960)
+		max = V4L2_MPEG_VIDEO_H264_LEVEL_4_1;
+	else
+		return;
+	ctx->h264_level_ctrl = v4l2_ctrl_new_std_menu(&ctx->ctrls,
+		&coda_ctrl_ops, V4L2_CID_MPEG_VIDEO_H264_LEVEL, max, 0, max);
+	if (ctx->h264_level_ctrl)
+		ctx->h264_level_ctrl->flags |= V4L2_CTRL_FLAG_READ_ONLY;
+
+	ctx->mpeg2_profile_ctrl = v4l2_ctrl_new_std_menu(&ctx->ctrls,
+		&coda_ctrl_ops, V4L2_CID_MPEG_VIDEO_MPEG2_PROFILE,
+		V4L2_MPEG_VIDEO_MPEG2_PROFILE_HIGH, 0,
+		V4L2_MPEG_VIDEO_MPEG2_PROFILE_HIGH);
+	if (ctx->mpeg2_profile_ctrl)
+		ctx->mpeg2_profile_ctrl->flags |= V4L2_CTRL_FLAG_READ_ONLY;
+
+	ctx->mpeg2_level_ctrl = v4l2_ctrl_new_std_menu(&ctx->ctrls,
+		&coda_ctrl_ops, V4L2_CID_MPEG_VIDEO_MPEG2_LEVEL,
+		V4L2_MPEG_VIDEO_MPEG2_LEVEL_HIGH, 0,
+		V4L2_MPEG_VIDEO_MPEG2_LEVEL_HIGH);
+	if (ctx->mpeg2_level_ctrl)
+		ctx->mpeg2_level_ctrl->flags |= V4L2_CTRL_FLAG_READ_ONLY;
+
+	ctx->mpeg4_profile_ctrl = v4l2_ctrl_new_std_menu(&ctx->ctrls,
+		&coda_ctrl_ops, V4L2_CID_MPEG_VIDEO_MPEG4_PROFILE,
+		V4L2_MPEG_VIDEO_MPEG4_PROFILE_ADVANCED_CODING_EFFICIENCY, 0,
+		V4L2_MPEG_VIDEO_MPEG4_PROFILE_ADVANCED_CODING_EFFICIENCY);
+	if (ctx->mpeg4_profile_ctrl)
+		ctx->mpeg4_profile_ctrl->flags |= V4L2_CTRL_FLAG_READ_ONLY;
+
+	ctx->mpeg4_level_ctrl = v4l2_ctrl_new_std_menu(&ctx->ctrls,
+		&coda_ctrl_ops, V4L2_CID_MPEG_VIDEO_MPEG4_LEVEL,
+		V4L2_MPEG_VIDEO_MPEG4_LEVEL_5, 0,
+		V4L2_MPEG_VIDEO_MPEG4_LEVEL_5);
+	if (ctx->mpeg4_level_ctrl)
+		ctx->mpeg4_level_ctrl->flags |= V4L2_CTRL_FLAG_READ_ONLY;
+}
+
+static const struct v4l2_ctrl_config coda_mb_err_cnt_ctrl_config = {
+	.id	= V4L2_CID_CODA_MB_ERR_CNT,
+	.name	= "Macroblocks Error Count",
+	.type	= V4L2_CTRL_TYPE_INTEGER,
+	.min	= 0,
+	.max	= 0x7fffffff,
+	.step	= 1,
+};
+
+static int coda_ctrls_setup(struct coda_ctx *ctx)
+{
+	v4l2_ctrl_handler_init(&ctx->ctrls, 2);
+
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_HFLIP, 0, 1, 1, 0);
+	v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+		V4L2_CID_VFLIP, 0, 1, 1, 0);
+	if (ctx->inst_type == CODA_INST_ENCODER) {
+		v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+				  V4L2_CID_MIN_BUFFERS_FOR_OUTPUT,
+				  1, 1, 1, 1);
+		if (ctx->cvd->dst_formats[0] == V4L2_PIX_FMT_JPEG)
+			coda_jpeg_encode_ctrls(ctx);
+		else
+			coda_encode_ctrls(ctx);
+	} else {
+		v4l2_ctrl_new_std(&ctx->ctrls, &coda_ctrl_ops,
+				  V4L2_CID_MIN_BUFFERS_FOR_CAPTURE,
+				  1, 1, 1, 1);
+		if (ctx->cvd->src_formats[0] == V4L2_PIX_FMT_H264)
+			coda_decode_ctrls(ctx);
+
+		ctx->mb_err_cnt_ctrl = v4l2_ctrl_new_custom(&ctx->ctrls,
+						&coda_mb_err_cnt_ctrl_config,
+						NULL);
+		if (ctx->mb_err_cnt_ctrl)
+			ctx->mb_err_cnt_ctrl->flags |= V4L2_CTRL_FLAG_READ_ONLY;
+	}
+
+	if (ctx->ctrls.error) {
+		v4l2_err(&ctx->dev->v4l2_dev,
+			"control initialization error (%d)",
+			ctx->ctrls.error);
+		return -EINVAL;
+	}
+
+	return v4l2_ctrl_handler_setup(&ctx->ctrls);
+}
+
+static int coda_queue_init(struct coda_ctx *ctx, struct vb2_queue *vq)
+{
+	vq->drv_priv = ctx;
+	vq->ops = &coda_qops;
+	vq->buf_struct_size = sizeof(struct v4l2_m2m_buffer);
+	vq->timestamp_flags = V4L2_BUF_FLAG_TIMESTAMP_COPY;
+	vq->lock = &ctx->dev->dev_mutex;
+	/* One way to indicate end-of-stream for coda is to set the
+	 * bytesused == 0. However by default videobuf2 handles bytesused
+	 * equal to 0 as a special case and changes its value to the size
+	 * of the buffer. Set the allow_zero_bytesused flag, so
+	 * that videobuf2 will keep the value of bytesused intact.
+	 */
+	vq->allow_zero_bytesused = 1;
+	/*
+	 * We might be fine with no buffers on some of the queues, but that
+	 * would need to be reflected in job_ready(). Currently we expect all
+	 * queues to have at least one buffer queued.
+	 */
+	vq->min_buffers_needed = 1;
+	vq->dev = ctx->dev->dev;
+
+	return vb2_queue_init(vq);
+}
+
+int coda_encoder_queue_init(void *priv, struct vb2_queue *src_vq,
+			    struct vb2_queue *dst_vq)
+{
+	int ret;
+
+	src_vq->type = V4L2_BUF_TYPE_VIDEO_OUTPUT;
+	src_vq->io_modes = VB2_DMABUF | VB2_MMAP;
+	src_vq->mem_ops = &vb2_dma_contig_memops;
+
+	ret = coda_queue_init(priv, src_vq);
+	if (ret)
+		return ret;
+
+	dst_vq->type = V4L2_BUF_TYPE_VIDEO_CAPTURE;
+	dst_vq->io_modes = VB2_DMABUF | VB2_MMAP;
+	dst_vq->mem_ops = &vb2_dma_contig_memops;
+
+	return coda_queue_init(priv, dst_vq);
+}
+
+int coda_decoder_queue_init(void *priv, struct vb2_queue *src_vq,
+			    struct vb2_queue *dst_vq)
+{
+	int ret;
+
+	src_vq->type = V4L2_BUF_TYPE_VIDEO_OUTPUT;
+	src_vq->io_modes = VB2_DMABUF | VB2_MMAP | VB2_USERPTR;
+	src_vq->mem_ops = &vb2_vmalloc_memops;
+
+	ret = coda_queue_init(priv, src_vq);
+	if (ret)
+		return ret;
+
+	dst_vq->type = V4L2_BUF_TYPE_VIDEO_CAPTURE;
+	dst_vq->io_modes = VB2_DMABUF | VB2_MMAP;
+	dst_vq->dma_attrs = DMA_ATTR_NO_KERNEL_MAPPING;
+	dst_vq->mem_ops = &vb2_dma_contig_memops;
+
+	return coda_queue_init(priv, dst_vq);
+}
+
+/*
+ * File operations
+ */
+
+static int coda_open(struct file *file)
+{
+	struct video_device *vdev = video_devdata(file);
+	struct coda_dev *dev = video_get_drvdata(vdev);
+	struct coda_ctx *ctx;
+	unsigned int max = ~0;
+	char *name;
+	int ret;
+	int idx;
+
+	ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
+	if (!ctx)
+		return -ENOMEM;
+
+	if (dev->devtype->product == CODA_DX6)
+		max = CODADX6_MAX_INSTANCES - 1;
+	idx = ida_alloc_max(&dev->ida, max, GFP_KERNEL);
+	if (idx < 0) {
+		ret = idx;
+		goto err_coda_max;
+	}
+
+	name = kasprintf(GFP_KERNEL, "context%d", idx);
+	if (!name) {
+		ret = -ENOMEM;
+		goto err_coda_name_init;
+	}
+
+	ctx->debugfs_entry = debugfs_create_dir(name, dev->debugfs_root);
+	kfree(name);
+
+	ctx->cvd = to_coda_video_device(vdev);
+	ctx->inst_type = ctx->cvd->type;
+	ctx->ops = ctx->cvd->ops;
+	ctx->use_bit = !ctx->cvd->direct;
+	init_completion(&ctx->completion);
+	INIT_WORK(&ctx->pic_run_work, coda_pic_run_work);
+	if (ctx->ops->seq_init_work)
+		INIT_WORK(&ctx->seq_init_work, ctx->ops->seq_init_work);
+	if (ctx->ops->seq_end_work)
+		INIT_WORK(&ctx->seq_end_work, ctx->ops->seq_end_work);
+	v4l2_fh_init(&ctx->fh, video_devdata(file));
+	file->private_data = &ctx->fh;
+	v4l2_fh_add(&ctx->fh);
+	ctx->dev = dev;
+	ctx->idx = idx;
+
+	coda_dbg(1, ctx, "open instance (%p)\n", ctx);
+
+	switch (dev->devtype->product) {
+	case CODA_960:
+		/*
+		 * Enabling the BWB when decoding can hang the firmware with
+		 * certain streams. The issue was tracked as ENGR00293425 by
+		 * Freescale. As a workaround, disable BWB for all decoders.
+		 * The enable_bwb module parameter allows to override this.
+		 */
+		if (enable_bwb || ctx->inst_type == CODA_INST_ENCODER)
+			ctx->frame_mem_ctrl = CODA9_FRAME_ENABLE_BWB;
+		fallthrough;
+	case CODA_HX4:
+	case CODA_7541:
+		ctx->reg_idx = 0;
+		break;
+	default:
+		ctx->reg_idx = idx;
+	}
+	if (ctx->dev->vdoa && !disable_vdoa) {
+		ctx->vdoa = vdoa_context_create(dev->vdoa);
+		if (!ctx->vdoa)
+			v4l2_warn(&dev->v4l2_dev,
+				  "Failed to create vdoa context: not using vdoa");
+	}
+	ctx->use_vdoa = false;
+
+	/* Power up and upload firmware if necessary */
+	ret = pm_runtime_resume_and_get(dev->dev);
+	if (ret < 0) {
+		v4l2_err(&dev->v4l2_dev, "failed to power up: %d\n", ret);
+		goto err_pm_get;
+	}
+
+	ret = clk_prepare_enable(dev->clk_per);
+	if (ret)
+		goto err_clk_enable;
+
+	ret = clk_prepare_enable(dev->clk_ahb);
+	if (ret)
+		goto err_clk_ahb;
+
+	set_default_params(ctx);
+	ctx->fh.m2m_ctx = v4l2_m2m_ctx_init(dev->m2m_dev, ctx,
+					    ctx->ops->queue_init);
+	if (IS_ERR(ctx->fh.m2m_ctx)) {
+		ret = PTR_ERR(ctx->fh.m2m_ctx);
+
+		v4l2_err(&dev->v4l2_dev, "%s return error (%d)\n",
+			 __func__, ret);
+		goto err_ctx_init;
+	}
+
+	ret = coda_ctrls_setup(ctx);
+	if (ret) {
+		v4l2_err(&dev->v4l2_dev, "failed to setup coda controls\n");
+		goto err_ctrls_setup;
+	}
+
+	ctx->fh.ctrl_handler = &ctx->ctrls;
+
+	mutex_init(&ctx->bitstream_mutex);
+	mutex_init(&ctx->buffer_mutex);
+	mutex_init(&ctx->wakeup_mutex);
+	INIT_LIST_HEAD(&ctx->buffer_meta_list);
+	spin_lock_init(&ctx->buffer_meta_lock);
+
+	return 0;
+
+err_ctrls_setup:
+	v4l2_m2m_ctx_release(ctx->fh.m2m_ctx);
+err_ctx_init:
+	clk_disable_unprepare(dev->clk_ahb);
+err_clk_ahb:
+	clk_disable_unprepare(dev->clk_per);
+err_clk_enable:
+	pm_runtime_put_sync(dev->dev);
+err_pm_get:
+	v4l2_fh_del(&ctx->fh);
+	v4l2_fh_exit(&ctx->fh);
+err_coda_name_init:
+	ida_free(&dev->ida, ctx->idx);
+err_coda_max:
+	kfree(ctx);
+	return ret;
+}
+
+static int coda_release(struct file *file)
+{
+	struct coda_dev *dev = video_drvdata(file);
+	struct coda_ctx *ctx = fh_to_ctx(file->private_data);
+
+	coda_dbg(1, ctx, "release instance (%p)\n", ctx);
+
+	if (ctx->inst_type == CODA_INST_DECODER && ctx->use_bit)
+		coda_bit_stream_end_flag(ctx);
+
+	/* If this instance is running, call .job_abort and wait for it to end */
+	v4l2_m2m_ctx_release(ctx->fh.m2m_ctx);
+
+	if (ctx->vdoa)
+		vdoa_context_destroy(ctx->vdoa);
+
+	/* In case the instance was not running, we still need to call SEQ_END */
+	if (ctx->ops->seq_end_work) {
+		queue_work(dev->workqueue, &ctx->seq_end_work);
+		flush_work(&ctx->seq_end_work);
+	}
+
+	if (ctx->dev->devtype->product == CODA_DX6)
+		coda_free_aux_buf(dev, &ctx->workbuf);
+
+	v4l2_ctrl_handler_free(&ctx->ctrls);
+	clk_disable_unprepare(dev->clk_ahb);
+	clk_disable_unprepare(dev->clk_per);
+	pm_runtime_put_sync(dev->dev);
+	v4l2_fh_del(&ctx->fh);
+	v4l2_fh_exit(&ctx->fh);
+	ida_free(&dev->ida, ctx->idx);
+	if (ctx->ops->release)
+		ctx->ops->release(ctx);
+	debugfs_remove_recursive(ctx->debugfs_entry);
+	kfree(ctx);
+
+	return 0;
+}
+
+static const struct v4l2_file_operations coda_fops = {
+	.owner		= THIS_MODULE,
+	.open		= coda_open,
+	.release	= coda_release,
+	.poll		= v4l2_m2m_fop_poll,
+	.unlocked_ioctl	= video_ioctl2,
+	.mmap		= v4l2_m2m_fop_mmap,
+};
+
+static int coda_hw_init(struct coda_dev *dev)
+{
+	u32 data;
+	u16 *p;
+	int i, ret;
+
+	ret = clk_prepare_enable(dev->clk_per);
+	if (ret)
+		goto err_clk_per;
+
+	ret = clk_prepare_enable(dev->clk_ahb);
+	if (ret)
+		goto err_clk_ahb;
+
+	reset_control_reset(dev->rstc);
+
+	/*
+	 * Copy the first CODA_ISRAM_SIZE in the internal SRAM.
+	 * The 16-bit chars in the code buffer are in memory access
+	 * order, re-sort them to CODA order for register download.
+	 * Data in this SRAM survives a reboot.
+	 */
+	p = (u16 *)dev->codebuf.vaddr;
+	if (dev->devtype->product == CODA_DX6) {
+		for (i = 0; i < (CODA_ISRAM_SIZE / 2); i++)  {
+			data = CODA_DOWN_ADDRESS_SET(i) |
+				CODA_DOWN_DATA_SET(p[i ^ 1]);
+			coda_write(dev, data, CODA_REG_BIT_CODE_DOWN);
+		}
+	} else {
+		for (i = 0; i < (CODA_ISRAM_SIZE / 2); i++) {
+			data = CODA_DOWN_ADDRESS_SET(i) |
+				CODA_DOWN_DATA_SET(p[round_down(i, 4) +
+							3 - (i % 4)]);
+			coda_write(dev, data, CODA_REG_BIT_CODE_DOWN);
+		}
+	}
+
+	/* Clear registers */
+	for (i = 0; i < 64; i++)
+		coda_write(dev, 0, CODA_REG_BIT_CODE_BUF_ADDR + i * 4);
+
+	/* Tell the BIT where to find everything it needs */
+	if (dev->devtype->product == CODA_960 ||
+	    dev->devtype->product == CODA_7541 ||
+	    dev->devtype->product == CODA_HX4) {
+		coda_write(dev, dev->tempbuf.paddr,
+				CODA_REG_BIT_TEMP_BUF_ADDR);
+		coda_write(dev, 0, CODA_REG_BIT_BIT_STREAM_PARAM);
+	} else {
+		coda_write(dev, dev->workbuf.paddr,
+			      CODA_REG_BIT_WORK_BUF_ADDR);
+	}
+	coda_write(dev, dev->codebuf.paddr,
+		      CODA_REG_BIT_CODE_BUF_ADDR);
+	coda_write(dev, 0, CODA_REG_BIT_CODE_RUN);
+
+	/* Set default values */
+	switch (dev->devtype->product) {
+	case CODA_DX6:
+		coda_write(dev, CODADX6_STREAM_BUF_PIC_FLUSH,
+			   CODA_REG_BIT_STREAM_CTRL);
+		break;
+	default:
+		coda_write(dev, CODA7_STREAM_BUF_PIC_FLUSH,
+			   CODA_REG_BIT_STREAM_CTRL);
+	}
+	if (dev->devtype->product == CODA_960)
+		coda_write(dev, CODA9_FRAME_ENABLE_BWB,
+				CODA_REG_BIT_FRAME_MEM_CTRL);
+	else
+		coda_write(dev, 0, CODA_REG_BIT_FRAME_MEM_CTRL);
+
+	if (dev->devtype->product != CODA_DX6)
+		coda_write(dev, 0, CODA7_REG_BIT_AXI_SRAM_USE);
+
+	coda_write(dev, CODA_INT_INTERRUPT_ENABLE,
+		      CODA_REG_BIT_INT_ENABLE);
+
+	/* Reset VPU and start processor */
+	data = coda_read(dev, CODA_REG_BIT_CODE_RESET);
+	data |= CODA_REG_RESET_ENABLE;
+	coda_write(dev, data, CODA_REG_BIT_CODE_RESET);
+	udelay(10);
+	data &= ~CODA_REG_RESET_ENABLE;
+	coda_write(dev, data, CODA_REG_BIT_CODE_RESET);
+	coda_write(dev, CODA_REG_RUN_ENABLE, CODA_REG_BIT_CODE_RUN);
+
+	clk_disable_unprepare(dev->clk_ahb);
+	clk_disable_unprepare(dev->clk_per);
+
+	return 0;
+
+err_clk_ahb:
+	clk_disable_unprepare(dev->clk_per);
+err_clk_per:
+	return ret;
+}
+
+static int coda_register_device(struct coda_dev *dev, int i)
+{
+	struct video_device *vfd = &dev->vfd[i];
+	const char *name;
+	int ret;
+
+	if (i >= dev->devtype->num_vdevs)
+		return -EINVAL;
+	name = dev->devtype->vdevs[i]->name;
+
+	strscpy(vfd->name, dev->devtype->vdevs[i]->name, sizeof(vfd->name));
+	vfd->fops	= &coda_fops;
+	vfd->ioctl_ops	= &coda_ioctl_ops;
+	vfd->release	= video_device_release_empty;
+	vfd->lock	= &dev->dev_mutex;
+	vfd->v4l2_dev	= &dev->v4l2_dev;
+	vfd->vfl_dir	= VFL_DIR_M2M;
+	vfd->device_caps = V4L2_CAP_VIDEO_M2M | V4L2_CAP_STREAMING;
+	video_set_drvdata(vfd, dev);
+
+	/* Not applicable, use the selection API instead */
+	v4l2_disable_ioctl(vfd, VIDIOC_CROPCAP);
+	v4l2_disable_ioctl(vfd, VIDIOC_G_CROP);
+	v4l2_disable_ioctl(vfd, VIDIOC_S_CROP);
+
+	if (dev->devtype->vdevs[i]->type == CODA_INST_ENCODER) {
+		v4l2_disable_ioctl(vfd, VIDIOC_DECODER_CMD);
+		v4l2_disable_ioctl(vfd, VIDIOC_TRY_DECODER_CMD);
+		if (dev->devtype->vdevs[i]->dst_formats[0] == V4L2_PIX_FMT_JPEG) {
+			v4l2_disable_ioctl(vfd, VIDIOC_ENUM_FRAMEINTERVALS);
+			v4l2_disable_ioctl(vfd, VIDIOC_G_PARM);
+			v4l2_disable_ioctl(vfd, VIDIOC_S_PARM);
+		}
+	} else {
+		v4l2_disable_ioctl(vfd, VIDIOC_ENCODER_CMD);
+		v4l2_disable_ioctl(vfd, VIDIOC_TRY_ENCODER_CMD);
+		v4l2_disable_ioctl(vfd, VIDIOC_ENUM_FRAMESIZES);
+		v4l2_disable_ioctl(vfd, VIDIOC_ENUM_FRAMEINTERVALS);
+		v4l2_disable_ioctl(vfd, VIDIOC_G_PARM);
+		v4l2_disable_ioctl(vfd, VIDIOC_S_PARM);
+	}
+
+	ret = video_register_device(vfd, VFL_TYPE_VIDEO, 0);
+	if (!ret)
+		v4l2_info(&dev->v4l2_dev, "%s registered as %s\n",
+			  name, video_device_node_name(vfd));
+	return ret;
+}
+
+static void coda_copy_firmware(struct coda_dev *dev, const u8 * const buf,
+			       size_t size)
+{
+	u32 *src = (u32 *)buf;
+
+	/* Check if the firmware has a 16-byte Freescale header, skip it */
+	if (buf[0] == 'M' && buf[1] == 'X')
+		src += 4;
+	/*
+	 * Check whether the firmware is in native order or pre-reordered for
+	 * memory access. The first instruction opcode always is 0xe40e.
+	 */
+	if (__le16_to_cpup((__le16 *)src) == 0xe40e) {
+		u32 *dst = dev->codebuf.vaddr;
+		int i;
+
+		/* Firmware in native order, reorder while copying */
+		if (dev->devtype->product == CODA_DX6) {
+			for (i = 0; i < (size - 16) / 4; i++)
+				dst[i] = (src[i] << 16) | (src[i] >> 16);
+		} else {
+			for (i = 0; i < (size - 16) / 4; i += 2) {
+				dst[i] = (src[i + 1] << 16) | (src[i + 1] >> 16);
+				dst[i + 1] = (src[i] << 16) | (src[i] >> 16);
+			}
+		}
+	} else {
+		/* Copy the already reordered firmware image */
+		memcpy(dev->codebuf.vaddr, src, size);
+	}
+}
+
+static void coda_fw_callback(const struct firmware *fw, void *context);
+
+static int coda_firmware_request(struct coda_dev *dev)
+{
+	char *fw;
+
+	if (dev->firmware >= ARRAY_SIZE(dev->devtype->firmware))
+		return -EINVAL;
+
+	fw = dev->devtype->firmware[dev->firmware];
+
+	dev_dbg(dev->dev, "requesting firmware '%s' for %s\n", fw,
+		coda_product_name(dev->devtype->product));
+
+	return request_firmware_nowait(THIS_MODULE, true, fw, dev->dev,
+				       GFP_KERNEL, dev, coda_fw_callback);
+}
+
+static void coda_fw_callback(const struct firmware *fw, void *context)
+{
+	struct coda_dev *dev = context;
+	int i, ret;
+
+	if (!fw) {
+		dev->firmware++;
+		ret = coda_firmware_request(dev);
+		if (ret < 0) {
+			v4l2_err(&dev->v4l2_dev, "firmware request failed\n");
+			goto put_pm;
+		}
+		return;
+	}
+	if (dev->firmware > 0) {
+		/*
+		 * Since we can't suppress warnings for failed asynchronous
+		 * firmware requests, report that the fallback firmware was
+		 * found.
+		 */
+		dev_info(dev->dev, "Using fallback firmware %s\n",
+			 dev->devtype->firmware[dev->firmware]);
+	}
+
+	/* allocate auxiliary per-device code buffer for the BIT processor */
+	ret = coda_alloc_aux_buf(dev, &dev->codebuf, fw->size, "codebuf",
+				 dev->debugfs_root);
+	if (ret < 0)
+		goto put_pm;
+
+	coda_copy_firmware(dev, fw->data, fw->size);
+	release_firmware(fw);
+
+	ret = coda_hw_init(dev);
+	if (ret < 0) {
+		v4l2_err(&dev->v4l2_dev, "HW initialization failed\n");
+		goto put_pm;
+	}
+
+	ret = coda_check_firmware(dev);
+	if (ret < 0)
+		goto put_pm;
+
+	dev->m2m_dev = v4l2_m2m_init(&coda_m2m_ops);
+	if (IS_ERR(dev->m2m_dev)) {
+		v4l2_err(&dev->v4l2_dev, "Failed to init mem2mem device\n");
+		goto put_pm;
+	}
+
+	for (i = 0; i < dev->devtype->num_vdevs; i++) {
+		ret = coda_register_device(dev, i);
+		if (ret) {
+			v4l2_err(&dev->v4l2_dev,
+				 "Failed to register %s video device: %d\n",
+				 dev->devtype->vdevs[i]->name, ret);
+			goto rel_vfd;
+		}
+	}
+
+	pm_runtime_put_sync(dev->dev);
+	return;
+
+rel_vfd:
+	while (--i >= 0)
+		video_unregister_device(&dev->vfd[i]);
+	v4l2_m2m_release(dev->m2m_dev);
+put_pm:
+	pm_runtime_put_sync(dev->dev);
+}
+
+enum coda_platform {
+	CODA_IMX27,
+	CODA_IMX51,
+	CODA_IMX53,
+	CODA_IMX6Q,
+	CODA_IMX6DL,
+};
+
+static const struct coda_devtype coda_devdata[] = {
+	[CODA_IMX27] = {
+		.firmware     = {
+			"vpu_fw_imx27_TO2.bin",
+			"vpu/vpu_fw_imx27_TO2.bin",
+			"v4l-codadx6-imx27.bin"
+		},
+		.product      = CODA_DX6,
+		.codecs       = codadx6_codecs,
+		.num_codecs   = ARRAY_SIZE(codadx6_codecs),
+		.vdevs        = codadx6_video_devices,
+		.num_vdevs    = ARRAY_SIZE(codadx6_video_devices),
+		.workbuf_size = 288 * 1024 + FMO_SLICE_SAVE_BUF_SIZE * 8 * 1024,
+		.iram_size    = 0xb000,
+	},
+	[CODA_IMX51] = {
+		.firmware     = {
+			"vpu_fw_imx51.bin",
+			"vpu/vpu_fw_imx51.bin",
+			"v4l-codahx4-imx51.bin"
+		},
+		.product      = CODA_HX4,
+		.codecs       = codahx4_codecs,
+		.num_codecs   = ARRAY_SIZE(codahx4_codecs),
+		.vdevs        = codahx4_video_devices,
+		.num_vdevs    = ARRAY_SIZE(codahx4_video_devices),
+		.workbuf_size = 128 * 1024,
+		.tempbuf_size = 304 * 1024,
+		.iram_size    = 0x14000,
+	},
+	[CODA_IMX53] = {
+		.firmware     = {
+			"vpu_fw_imx53.bin",
+			"vpu/vpu_fw_imx53.bin",
+			"v4l-coda7541-imx53.bin"
+		},
+		.product      = CODA_7541,
+		.codecs       = coda7_codecs,
+		.num_codecs   = ARRAY_SIZE(coda7_codecs),
+		.vdevs        = coda7_video_devices,
+		.num_vdevs    = ARRAY_SIZE(coda7_video_devices),
+		.workbuf_size = 128 * 1024,
+		.tempbuf_size = 304 * 1024,
+		.iram_size    = 0x14000,
+	},
+	[CODA_IMX6Q] = {
+		.firmware     = {
+			"vpu_fw_imx6q.bin",
+			"vpu/vpu_fw_imx6q.bin",
+			"v4l-coda960-imx6q.bin"
+		},
+		.product      = CODA_960,
+		.codecs       = coda9_codecs,
+		.num_codecs   = ARRAY_SIZE(coda9_codecs),
+		.vdevs        = coda9_video_devices,
+		.num_vdevs    = ARRAY_SIZE(coda9_video_devices),
+		.workbuf_size = 80 * 1024,
+		.tempbuf_size = 204 * 1024,
+		.iram_size    = 0x21000,
+	},
+	[CODA_IMX6DL] = {
+		.firmware     = {
+			"vpu_fw_imx6d.bin",
+			"vpu/vpu_fw_imx6d.bin",
+			"v4l-coda960-imx6dl.bin"
+		},
+		.product      = CODA_960,
+		.codecs       = coda9_codecs,
+		.num_codecs   = ARRAY_SIZE(coda9_codecs),
+		.vdevs        = coda9_video_devices,
+		.num_vdevs    = ARRAY_SIZE(coda9_video_devices),
+		.workbuf_size = 80 * 1024,
+		.tempbuf_size = 204 * 1024,
+		.iram_size    = 0x1f000, /* leave 4k for suspend code */
+	},
+};
+
+static const struct of_device_id coda_dt_ids[] = {
+	{ .compatible = "fsl,imx27-vpu", .data = &coda_devdata[CODA_IMX27] },
+	{ .compatible = "fsl,imx51-vpu", .data = &coda_devdata[CODA_IMX51] },
+	{ .compatible = "fsl,imx53-vpu", .data = &coda_devdata[CODA_IMX53] },
+	{ .compatible = "fsl,imx6q-vpu", .data = &coda_devdata[CODA_IMX6Q] },
+	{ .compatible = "fsl,imx6dl-vpu", .data = &coda_devdata[CODA_IMX6DL] },
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, coda_dt_ids);
+
+static int coda_probe(struct platform_device *pdev)
+{
+	struct device_node *np = pdev->dev.of_node;
+	struct gen_pool *pool;
+	struct coda_dev *dev;
+	int ret, irq;
+
+	dev = devm_kzalloc(&pdev->dev, sizeof(*dev), GFP_KERNEL);
+	if (!dev)
+		return -ENOMEM;
+
+	dev->devtype = of_device_get_match_data(&pdev->dev);
+
+	dev->dev = &pdev->dev;
+	dev->clk_per = devm_clk_get(&pdev->dev, "per");
+	if (IS_ERR(dev->clk_per)) {
+		dev_err(&pdev->dev, "Could not get per clock\n");
+		return PTR_ERR(dev->clk_per);
+	}
+
+	dev->clk_ahb = devm_clk_get(&pdev->dev, "ahb");
+	if (IS_ERR(dev->clk_ahb)) {
+		dev_err(&pdev->dev, "Could not get ahb clock\n");
+		return PTR_ERR(dev->clk_ahb);
+	}
+
+	/* Get  memory for physical registers */
+	dev->regs_base = devm_platform_ioremap_resource(pdev, 0);
+	if (IS_ERR(dev->regs_base))
+		return PTR_ERR(dev->regs_base);
+
+	/* IRQ */
+	irq = platform_get_irq_byname(pdev, "bit");
+	if (irq < 0)
+		irq = platform_get_irq(pdev, 0);
+	if (irq < 0)
+		return irq;
+
+	ret = devm_request_irq(&pdev->dev, irq, coda_irq_handler, 0,
+			       CODA_NAME "-video", dev);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to request irq: %d\n", ret);
+		return ret;
+	}
+
+	/* JPEG IRQ */
+	if (dev->devtype->product == CODA_960) {
+		irq = platform_get_irq_byname(pdev, "jpeg");
+		if (irq < 0)
+			return irq;
+
+		ret = devm_request_threaded_irq(&pdev->dev, irq, NULL,
+						coda9_jpeg_irq_handler,
+						IRQF_ONESHOT, CODA_NAME "-jpeg",
+						dev);
+		if (ret < 0) {
+			dev_err(&pdev->dev, "failed to request jpeg irq\n");
+			return ret;
+		}
+	}
+
+	dev->rstc = devm_reset_control_get_optional_exclusive(&pdev->dev,
+							      NULL);
+	if (IS_ERR(dev->rstc)) {
+		ret = PTR_ERR(dev->rstc);
+		dev_err(&pdev->dev, "failed get reset control: %d\n", ret);
+		return ret;
+	}
+
+	/* Get IRAM pool from device tree */
+	pool = of_gen_pool_get(np, "iram", 0);
+	if (!pool) {
+		dev_err(&pdev->dev, "iram pool not available\n");
+		return -ENOMEM;
+	}
+	dev->iram_pool = pool;
+
+	/* Get vdoa_data if supported by the platform */
+	dev->vdoa = coda_get_vdoa_data();
+	if (PTR_ERR(dev->vdoa) == -EPROBE_DEFER)
+		return -EPROBE_DEFER;
+
+	ret = v4l2_device_register(&pdev->dev, &dev->v4l2_dev);
+	if (ret)
+		return ret;
+
+	ratelimit_default_init(&dev->mb_err_rs);
+	mutex_init(&dev->dev_mutex);
+	mutex_init(&dev->coda_mutex);
+	ida_init(&dev->ida);
+
+	dev->debugfs_root = debugfs_create_dir("coda", NULL);
+
+	/* allocate auxiliary per-device buffers for the BIT processor */
+	if (dev->devtype->product == CODA_DX6) {
+		ret = coda_alloc_aux_buf(dev, &dev->workbuf,
+					 dev->devtype->workbuf_size, "workbuf",
+					 dev->debugfs_root);
+		if (ret < 0)
+			goto err_v4l2_register;
+	}
+
+	if (dev->devtype->tempbuf_size) {
+		ret = coda_alloc_aux_buf(dev, &dev->tempbuf,
+					 dev->devtype->tempbuf_size, "tempbuf",
+					 dev->debugfs_root);
+		if (ret < 0)
+			goto err_v4l2_register;
+	}
+
+	dev->iram.size = dev->devtype->iram_size;
+	dev->iram.vaddr = gen_pool_dma_alloc(dev->iram_pool, dev->iram.size,
+					     &dev->iram.paddr);
+	if (!dev->iram.vaddr) {
+		dev_warn(&pdev->dev, "unable to alloc iram\n");
+	} else {
+		memset(dev->iram.vaddr, 0, dev->iram.size);
+		dev->iram.blob.data = dev->iram.vaddr;
+		dev->iram.blob.size = dev->iram.size;
+		dev->iram.dentry = debugfs_create_blob("iram", 0444,
+						       dev->debugfs_root,
+						       &dev->iram.blob);
+	}
+
+	dev->workqueue = alloc_workqueue("coda", WQ_UNBOUND | WQ_MEM_RECLAIM, 1);
+	if (!dev->workqueue) {
+		dev_err(&pdev->dev, "unable to alloc workqueue\n");
+		ret = -ENOMEM;
+		goto err_v4l2_register;
+	}
+
+	platform_set_drvdata(pdev, dev);
+
+	/*
+	 * Start activated so we can directly call coda_hw_init in
+	 * coda_fw_callback regardless of whether CONFIG_PM is
+	 * enabled or whether the device is associated with a PM domain.
+	 */
+	pm_runtime_get_noresume(&pdev->dev);
+	pm_runtime_set_active(&pdev->dev);
+	pm_runtime_enable(&pdev->dev);
+
+	ret = coda_firmware_request(dev);
+	if (ret)
+		goto err_alloc_workqueue;
+	return 0;
+
+err_alloc_workqueue:
+	pm_runtime_disable(&pdev->dev);
+	pm_runtime_put_noidle(&pdev->dev);
+	destroy_workqueue(dev->workqueue);
+err_v4l2_register:
+	v4l2_device_unregister(&dev->v4l2_dev);
+	return ret;
+}
+
+static int coda_remove(struct platform_device *pdev)
+{
+	struct coda_dev *dev = platform_get_drvdata(pdev);
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(dev->vfd); i++) {
+		if (video_get_drvdata(&dev->vfd[i]))
+			video_unregister_device(&dev->vfd[i]);
+	}
+	if (dev->m2m_dev)
+		v4l2_m2m_release(dev->m2m_dev);
+	pm_runtime_disable(&pdev->dev);
+	v4l2_device_unregister(&dev->v4l2_dev);
+	destroy_workqueue(dev->workqueue);
+	if (dev->iram.vaddr)
+		gen_pool_free(dev->iram_pool, (unsigned long)dev->iram.vaddr,
+			      dev->iram.size);
+	coda_free_aux_buf(dev, &dev->codebuf);
+	coda_free_aux_buf(dev, &dev->tempbuf);
+	coda_free_aux_buf(dev, &dev->workbuf);
+	debugfs_remove_recursive(dev->debugfs_root);
+	ida_destroy(&dev->ida);
+	return 0;
+}
+
+#ifdef CONFIG_PM
+static int coda_runtime_resume(struct device *dev)
+{
+	struct coda_dev *cdev = dev_get_drvdata(dev);
+	int ret = 0;
+
+	if (dev->pm_domain && cdev->codebuf.vaddr) {
+		ret = coda_hw_init(cdev);
+		if (ret)
+			v4l2_err(&cdev->v4l2_dev, "HW initialization failed\n");
+	}
+
+	return ret;
+}
+#endif
+
+static const struct dev_pm_ops coda_pm_ops = {
+	SET_RUNTIME_PM_OPS(NULL, coda_runtime_resume, NULL)
+};
+
+static struct platform_driver coda_driver = {
+	.probe	= coda_probe,
+	.remove	= coda_remove,
+	.driver	= {
+		.name	= CODA_NAME,
+		.of_match_table = coda_dt_ids,
+		.pm	= &coda_pm_ops,
+	},
+};
+
+module_platform_driver(coda_driver);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Javier Martin <javier.martin@vista-silicon.com>");
+MODULE_DESCRIPTION("Coda multi-standard codec V4L2 driver");
diff --git a/drivers/media/platform/chips-media/coda-gdi.c b/drivers/media/platform/chips-media/coda-gdi.c
new file mode 100644
index 000000000..59d65daca
--- /dev/null
+++ b/drivers/media/platform/chips-media/coda-gdi.c
@@ -0,0 +1,146 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Coda multi-standard codec IP
+ *
+ * Copyright (C) 2014 Philipp Zabel, Pengutronix
+ */
+
+#include <linux/bitops.h>
+#include "coda.h"
+
+#define XY2_INVERT	BIT(7)
+#define XY2_ZERO	BIT(6)
+#define XY2_TB_XOR	BIT(5)
+#define XY2_XYSEL	BIT(4)
+#define XY2_Y		(1 << 4)
+#define XY2_X		(0 << 4)
+
+#define XY2(luma_sel, luma_bit, chroma_sel, chroma_bit) \
+	(((XY2_##luma_sel) | (luma_bit)) << 8 | \
+	 (XY2_##chroma_sel) | (chroma_bit))
+
+static const u16 xy2ca_zero_map[16] = {
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+};
+
+static const u16 xy2ca_tiled_map[16] = {
+	XY2(Y,    0, Y,    0),
+	XY2(Y,    1, Y,    1),
+	XY2(Y,    2, Y,    2),
+	XY2(Y,    3, X,    3),
+	XY2(X,    3, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+	XY2(ZERO, 0, ZERO, 0),
+};
+
+/*
+ * RA[15:0], CA[15:8] are hardwired to contain the 24-bit macroblock
+ * start offset (macroblock size is 16x16 for luma, 16x8 for chroma).
+ * Bits CA[4:0] are set using XY2CA above. BA[3:0] seems to be unused.
+ */
+
+#define RBC_CA		(0 << 4)
+#define RBC_BA		(1 << 4)
+#define RBC_RA		(2 << 4)
+#define RBC_ZERO	(3 << 4)
+
+#define RBC(luma_sel, luma_bit, chroma_sel, chroma_bit) \
+	(((RBC_##luma_sel) | (luma_bit)) << 6 | \
+	 (RBC_##chroma_sel) | (chroma_bit))
+
+static const u16 rbc2axi_tiled_map[32] = {
+	RBC(ZERO, 0, ZERO, 0),
+	RBC(ZERO, 0, ZERO, 0),
+	RBC(ZERO, 0, ZERO, 0),
+	RBC(CA,   0, CA,   0),
+	RBC(CA,   1, CA,   1),
+	RBC(CA,   2, CA,   2),
+	RBC(CA,   3, CA,   3),
+	RBC(CA,   4, CA,   8),
+	RBC(CA,   8, CA,   9),
+	RBC(CA,   9, CA,  10),
+	RBC(CA,  10, CA,  11),
+	RBC(CA,  11, CA,  12),
+	RBC(CA,  12, CA,  13),
+	RBC(CA,  13, CA,  14),
+	RBC(CA,  14, CA,  15),
+	RBC(CA,  15, RA,   0),
+	RBC(RA,   0, RA,   1),
+	RBC(RA,   1, RA,   2),
+	RBC(RA,   2, RA,   3),
+	RBC(RA,   3, RA,   4),
+	RBC(RA,   4, RA,   5),
+	RBC(RA,   5, RA,   6),
+	RBC(RA,   6, RA,   7),
+	RBC(RA,   7, RA,   8),
+	RBC(RA,   8, RA,   9),
+	RBC(RA,   9, RA,  10),
+	RBC(RA,  10, RA,  11),
+	RBC(RA,  11, RA,  12),
+	RBC(RA,  12, RA,  13),
+	RBC(RA,  13, RA,  14),
+	RBC(RA,  14, RA,  15),
+	RBC(RA,  15, ZERO, 0),
+};
+
+void coda_set_gdi_regs(struct coda_ctx *ctx)
+{
+	struct coda_dev *dev = ctx->dev;
+	const u16 *xy2ca_map;
+	u32 xy2rbc_config;
+	int i;
+
+	switch (ctx->tiled_map_type) {
+	case GDI_LINEAR_FRAME_MAP:
+	default:
+		xy2ca_map = xy2ca_zero_map;
+		xy2rbc_config = 0;
+		break;
+	case GDI_TILED_FRAME_MB_RASTER_MAP:
+		xy2ca_map = xy2ca_tiled_map;
+		xy2rbc_config = CODA9_XY2RBC_TILED_MAP |
+				CODA9_XY2RBC_CA_INC_HOR |
+				(16 - 1) << 12 | (8 - 1) << 4;
+		break;
+	}
+
+	for (i = 0; i < 16; i++)
+		coda_write(dev, xy2ca_map[i],
+				CODA9_GDI_XY2_CAS_0 + 4 * i);
+	for (i = 0; i < 4; i++)
+		coda_write(dev, XY2(ZERO, 0, ZERO, 0),
+				CODA9_GDI_XY2_BA_0 + 4 * i);
+	for (i = 0; i < 16; i++)
+		coda_write(dev, XY2(ZERO, 0, ZERO, 0),
+				CODA9_GDI_XY2_RAS_0 + 4 * i);
+	coda_write(dev, xy2rbc_config, CODA9_GDI_XY2_RBC_CONFIG);
+	if (xy2rbc_config) {
+		for (i = 0; i < 32; i++)
+			coda_write(dev, rbc2axi_tiled_map[i],
+					CODA9_GDI_RBC2_AXI_0 + 4 * i);
+	}
+}
diff --git a/drivers/media/platform/chips-media/coda-h264.c b/drivers/media/platform/chips-media/coda-h264.c
new file mode 100644
index 000000000..8bd0aa8af
--- /dev/null
+++ b/drivers/media/platform/chips-media/coda-h264.c
@@ -0,0 +1,429 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Coda multi-standard codec IP - H.264 helper functions
+ *
+ * Copyright (C) 2012 Vista Silicon S.L.
+ *    Javier Martin, <javier.martin@vista-silicon.com>
+ *    Xavier Duret
+ */
+
+#include <linux/kernel.h>
+#include <linux/string.h>
+#include <linux/videodev2.h>
+
+#include "coda.h"
+
+static const u8 coda_filler_size[8] = { 0, 7, 14, 13, 12, 11, 10, 9 };
+
+static const u8 *coda_find_nal_header(const u8 *buf, const u8 *end)
+{
+	u32 val = 0xffffffff;
+
+	do {
+		val = val << 8 | *buf++;
+		if (buf >= end)
+			return NULL;
+	} while (val != 0x00000001);
+
+	return buf;
+}
+
+int coda_sps_parse_profile(struct coda_ctx *ctx, struct vb2_buffer *vb)
+{
+	const u8 *buf = vb2_plane_vaddr(vb, 0);
+	const u8 *end = buf + vb2_get_plane_payload(vb, 0);
+
+	/* Find SPS header */
+	do {
+		buf = coda_find_nal_header(buf, end);
+		if (!buf)
+			return -EINVAL;
+	} while ((*buf++ & 0x1f) != 0x7);
+
+	ctx->params.h264_profile_idc = buf[0];
+	ctx->params.h264_level_idc = buf[2];
+
+	return 0;
+}
+
+int coda_h264_filler_nal(int size, char *p)
+{
+	if (size < 6)
+		return -EINVAL;
+
+	p[0] = 0x00;
+	p[1] = 0x00;
+	p[2] = 0x00;
+	p[3] = 0x01;
+	p[4] = 0x0c;
+	memset(p + 5, 0xff, size - 6);
+	/* Add rbsp stop bit and trailing at the end */
+	p[size - 1] = 0x80;
+
+	return 0;
+}
+
+int coda_h264_padding(int size, char *p)
+{
+	int nal_size;
+	int diff;
+
+	diff = size - (size & ~0x7);
+	if (diff == 0)
+		return 0;
+
+	nal_size = coda_filler_size[diff];
+	coda_h264_filler_nal(nal_size, p);
+
+	return nal_size;
+}
+
+int coda_h264_profile(int profile_idc)
+{
+	switch (profile_idc) {
+	case 66: return V4L2_MPEG_VIDEO_H264_PROFILE_BASELINE;
+	case 77: return V4L2_MPEG_VIDEO_H264_PROFILE_MAIN;
+	case 88: return V4L2_MPEG_VIDEO_H264_PROFILE_EXTENDED;
+	case 100: return V4L2_MPEG_VIDEO_H264_PROFILE_HIGH;
+	default: return -EINVAL;
+	}
+}
+
+int coda_h264_level(int level_idc)
+{
+	switch (level_idc) {
+	case 10: return V4L2_MPEG_VIDEO_H264_LEVEL_1_0;
+	case 9:  return V4L2_MPEG_VIDEO_H264_LEVEL_1B;
+	case 11: return V4L2_MPEG_VIDEO_H264_LEVEL_1_1;
+	case 12: return V4L2_MPEG_VIDEO_H264_LEVEL_1_2;
+	case 13: return V4L2_MPEG_VIDEO_H264_LEVEL_1_3;
+	case 20: return V4L2_MPEG_VIDEO_H264_LEVEL_2_0;
+	case 21: return V4L2_MPEG_VIDEO_H264_LEVEL_2_1;
+	case 22: return V4L2_MPEG_VIDEO_H264_LEVEL_2_2;
+	case 30: return V4L2_MPEG_VIDEO_H264_LEVEL_3_0;
+	case 31: return V4L2_MPEG_VIDEO_H264_LEVEL_3_1;
+	case 32: return V4L2_MPEG_VIDEO_H264_LEVEL_3_2;
+	case 40: return V4L2_MPEG_VIDEO_H264_LEVEL_4_0;
+	case 41: return V4L2_MPEG_VIDEO_H264_LEVEL_4_1;
+	case 42: return V4L2_MPEG_VIDEO_H264_LEVEL_4_2;
+	case 50: return V4L2_MPEG_VIDEO_H264_LEVEL_5_0;
+	case 51: return V4L2_MPEG_VIDEO_H264_LEVEL_5_1;
+	default: return -EINVAL;
+	}
+}
+
+struct rbsp {
+	char *buf;
+	int size;
+	int pos;
+};
+
+static inline int rbsp_read_bit(struct rbsp *rbsp)
+{
+	int shift = 7 - (rbsp->pos % 8);
+	int ofs = rbsp->pos++ / 8;
+
+	if (ofs >= rbsp->size)
+		return -EINVAL;
+
+	return (rbsp->buf[ofs] >> shift) & 1;
+}
+
+static inline int rbsp_write_bit(struct rbsp *rbsp, int bit)
+{
+	int shift = 7 - (rbsp->pos % 8);
+	int ofs = rbsp->pos++ / 8;
+
+	if (ofs >= rbsp->size)
+		return -EINVAL;
+
+	rbsp->buf[ofs] &= ~(1 << shift);
+	rbsp->buf[ofs] |= bit << shift;
+
+	return 0;
+}
+
+static inline int rbsp_read_bits(struct rbsp *rbsp, int num, int *val)
+{
+	int i, ret;
+	int tmp = 0;
+
+	if (num > 32)
+		return -EINVAL;
+
+	for (i = 0; i < num; i++) {
+		ret = rbsp_read_bit(rbsp);
+		if (ret < 0)
+			return ret;
+		tmp |= ret << (num - i - 1);
+	}
+
+	if (val)
+		*val = tmp;
+
+	return 0;
+}
+
+static int rbsp_write_bits(struct rbsp *rbsp, int num, int value)
+{
+	int ret;
+
+	while (num--) {
+		ret = rbsp_write_bit(rbsp, (value >> num) & 1);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+static int rbsp_read_uev(struct rbsp *rbsp, unsigned int *val)
+{
+	int leading_zero_bits = 0;
+	unsigned int tmp = 0;
+	int ret;
+
+	while ((ret = rbsp_read_bit(rbsp)) == 0)
+		leading_zero_bits++;
+	if (ret < 0)
+		return ret;
+
+	if (leading_zero_bits > 0) {
+		ret = rbsp_read_bits(rbsp, leading_zero_bits, &tmp);
+		if (ret)
+			return ret;
+	}
+
+	if (val)
+		*val = (1 << leading_zero_bits) - 1 + tmp;
+
+	return 0;
+}
+
+static int rbsp_write_uev(struct rbsp *rbsp, unsigned int value)
+{
+	int i;
+	int ret;
+	int tmp = value + 1;
+	int leading_zero_bits = fls(tmp) - 1;
+
+	for (i = 0; i < leading_zero_bits; i++) {
+		ret = rbsp_write_bit(rbsp, 0);
+		if (ret)
+			return ret;
+	}
+
+	return rbsp_write_bits(rbsp, leading_zero_bits + 1, tmp);
+}
+
+static int rbsp_read_sev(struct rbsp *rbsp, int *val)
+{
+	unsigned int tmp;
+	int ret;
+
+	ret = rbsp_read_uev(rbsp, &tmp);
+	if (ret)
+		return ret;
+
+	if (val) {
+		if (tmp & 1)
+			*val = (tmp + 1) / 2;
+		else
+			*val = -(tmp / 2);
+	}
+
+	return 0;
+}
+
+/**
+ * coda_h264_sps_fixup - fixes frame cropping values in h.264 SPS
+ * @ctx: encoder context
+ * @width: visible width
+ * @height: visible height
+ * @buf: buffer containing h.264 SPS RBSP, starting with NAL header
+ * @size: modified RBSP size return value
+ * @max_size: available size in buf
+ *
+ * Rewrites the frame cropping values in an h.264 SPS RBSP correctly for the
+ * given visible width and height.
+ */
+int coda_h264_sps_fixup(struct coda_ctx *ctx, int width, int height, char *buf,
+			int *size, int max_size)
+{
+	int profile_idc;
+	unsigned int pic_order_cnt_type;
+	int pic_width_in_mbs_minus1, pic_height_in_map_units_minus1;
+	int frame_mbs_only_flag, frame_cropping_flag;
+	int vui_parameters_present_flag;
+	unsigned int crop_right, crop_bottom;
+	struct rbsp sps;
+	int pos;
+	int ret;
+
+	if (*size < 8 || *size >= max_size)
+		return -EINVAL;
+
+	sps.buf = buf + 5; /* Skip NAL header */
+	sps.size = *size - 5;
+
+	profile_idc = sps.buf[0];
+	/* Skip constraint_set[0-5]_flag, reserved_zero_2bits */
+	/* Skip level_idc */
+	sps.pos = 24;
+
+	/* seq_parameter_set_id */
+	ret = rbsp_read_uev(&sps, NULL);
+	if (ret)
+		return ret;
+
+	if (profile_idc == 100 || profile_idc == 110 || profile_idc == 122 ||
+	    profile_idc == 244 || profile_idc == 44 || profile_idc == 83 ||
+	    profile_idc == 86 || profile_idc == 118 || profile_idc == 128 ||
+	    profile_idc == 138 || profile_idc == 139 || profile_idc == 134 ||
+	    profile_idc == 135) {
+		dev_err(ctx->fh.vdev->dev_parent,
+			"%s: Handling profile_idc %d not implemented\n",
+			__func__, profile_idc);
+		return -EINVAL;
+	}
+
+	/* log2_max_frame_num_minus4 */
+	ret = rbsp_read_uev(&sps, NULL);
+	if (ret)
+		return ret;
+
+	ret = rbsp_read_uev(&sps, &pic_order_cnt_type);
+	if (ret)
+		return ret;
+
+	if (pic_order_cnt_type == 0) {
+		/* log2_max_pic_order_cnt_lsb_minus4 */
+		ret = rbsp_read_uev(&sps, NULL);
+		if (ret)
+			return ret;
+	} else if (pic_order_cnt_type == 1) {
+		unsigned int i, num_ref_frames_in_pic_order_cnt_cycle;
+
+		/* delta_pic_order_always_zero_flag */
+		ret = rbsp_read_bit(&sps);
+		if (ret < 0)
+			return ret;
+		/* offset_for_non_ref_pic */
+		ret = rbsp_read_sev(&sps, NULL);
+		if (ret)
+			return ret;
+		/* offset_for_top_to_bottom_field */
+		ret = rbsp_read_sev(&sps, NULL);
+		if (ret)
+			return ret;
+
+		ret = rbsp_read_uev(&sps,
+				    &num_ref_frames_in_pic_order_cnt_cycle);
+		if (ret)
+			return ret;
+		for (i = 0; i < num_ref_frames_in_pic_order_cnt_cycle; i++) {
+			/* offset_for_ref_frame */
+			ret = rbsp_read_sev(&sps, NULL);
+			if (ret)
+				return ret;
+		}
+	}
+
+	/* max_num_ref_frames */
+	ret = rbsp_read_uev(&sps, NULL);
+	if (ret)
+		return ret;
+
+	/* gaps_in_frame_num_value_allowed_flag */
+	ret = rbsp_read_bit(&sps);
+	if (ret < 0)
+		return ret;
+	ret = rbsp_read_uev(&sps, &pic_width_in_mbs_minus1);
+	if (ret)
+		return ret;
+	ret = rbsp_read_uev(&sps, &pic_height_in_map_units_minus1);
+	if (ret)
+		return ret;
+	frame_mbs_only_flag = ret = rbsp_read_bit(&sps);
+	if (ret < 0)
+		return ret;
+	if (!frame_mbs_only_flag) {
+		/* mb_adaptive_frame_field_flag */
+		ret = rbsp_read_bit(&sps);
+		if (ret < 0)
+			return ret;
+	}
+	/* direct_8x8_inference_flag */
+	ret = rbsp_read_bit(&sps);
+	if (ret < 0)
+		return ret;
+
+	/* Mark position of the frame cropping flag */
+	pos = sps.pos;
+	frame_cropping_flag = ret = rbsp_read_bit(&sps);
+	if (ret < 0)
+		return ret;
+	if (frame_cropping_flag) {
+		unsigned int crop_left, crop_top;
+
+		ret = rbsp_read_uev(&sps, &crop_left);
+		if (ret)
+			return ret;
+		ret = rbsp_read_uev(&sps, &crop_right);
+		if (ret)
+			return ret;
+		ret = rbsp_read_uev(&sps, &crop_top);
+		if (ret)
+			return ret;
+		ret = rbsp_read_uev(&sps, &crop_bottom);
+		if (ret)
+			return ret;
+	}
+	vui_parameters_present_flag = ret = rbsp_read_bit(&sps);
+	if (ret < 0)
+		return ret;
+	if (vui_parameters_present_flag) {
+		dev_err(ctx->fh.vdev->dev_parent,
+			"%s: Handling vui_parameters not implemented\n",
+			__func__);
+		return -EINVAL;
+	}
+
+	crop_right = round_up(width, 16) - width;
+	crop_bottom = round_up(height, 16) - height;
+	crop_right /= 2;
+	if (frame_mbs_only_flag)
+		crop_bottom /= 2;
+	else
+		crop_bottom /= 4;
+
+
+	sps.size = max_size - 5;
+	sps.pos = pos;
+	frame_cropping_flag = 1;
+	ret = rbsp_write_bit(&sps, frame_cropping_flag);
+	if (ret)
+		return ret;
+	ret = rbsp_write_uev(&sps, 0); /* crop_left */
+	if (ret)
+		return ret;
+	ret = rbsp_write_uev(&sps, crop_right);
+	if (ret)
+		return ret;
+	ret = rbsp_write_uev(&sps, 0); /* crop_top */
+	if (ret)
+		return ret;
+	ret = rbsp_write_uev(&sps, crop_bottom);
+	if (ret)
+		return ret;
+	ret = rbsp_write_bit(&sps, 0); /* vui_parameters_present_flag */
+	if (ret)
+		return ret;
+	ret = rbsp_write_bit(&sps, 1);
+	if (ret)
+		return ret;
+
+	*size = 5 + DIV_ROUND_UP(sps.pos, 8);
+
+	return 0;
+}
diff --git a/drivers/media/platform/chips-media/coda-jpeg.c b/drivers/media/platform/chips-media/coda-jpeg.c
new file mode 100644
index 000000000..ba8f41002
--- /dev/null
+++ b/drivers/media/platform/chips-media/coda-jpeg.c
@@ -0,0 +1,1547 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Coda multi-standard codec IP - JPEG support functions
+ *
+ * Copyright (C) 2014 Philipp Zabel, Pengutronix
+ */
+
+#include <asm/unaligned.h>
+#include <linux/irqreturn.h>
+#include <linux/kernel.h>
+#include <linux/ktime.h>
+#include <linux/slab.h>
+#include <linux/swab.h>
+#include <linux/videodev2.h>
+
+#include <media/v4l2-common.h>
+#include <media/v4l2-fh.h>
+#include <media/v4l2-jpeg.h>
+#include <media/v4l2-mem2mem.h>
+#include <media/videobuf2-core.h>
+#include <media/videobuf2-dma-contig.h>
+
+#include "coda.h"
+#include "trace.h"
+
+#define SOI_MARKER	0xffd8
+#define APP9_MARKER	0xffe9
+#define DRI_MARKER	0xffdd
+#define DQT_MARKER	0xffdb
+#define DHT_MARKER	0xffc4
+#define SOF_MARKER	0xffc0
+#define SOS_MARKER	0xffda
+#define EOI_MARKER	0xffd9
+
+enum {
+	CODA9_JPEG_FORMAT_420,
+	CODA9_JPEG_FORMAT_422,
+	CODA9_JPEG_FORMAT_224,
+	CODA9_JPEG_FORMAT_444,
+	CODA9_JPEG_FORMAT_400,
+};
+
+struct coda_huff_tab {
+	u8 luma_dc[16 + 12];
+	u8 chroma_dc[16 + 12];
+	u8 luma_ac[16 + 162];
+	u8 chroma_ac[16 + 162];
+
+	/* DC Luma, DC Chroma, AC Luma, AC Chroma */
+	s16	min[4 * 16];
+	s16	max[4 * 16];
+	s8	ptr[4 * 16];
+};
+
+#define CODA9_JPEG_ENC_HUFF_DATA_SIZE	(256 + 256 + 16 + 16)
+
+/*
+ * Typical Huffman tables for 8-bit precision luminance and
+ * chrominance from JPEG ITU-T.81 (ISO/IEC 10918-1) Annex K.3
+ */
+
+static const unsigned char luma_dc[16 + 12] = {
+	/* bits */
+	0x00, 0x01, 0x05, 0x01, 0x01, 0x01, 0x01, 0x01,
+	0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+	/* values */
+	0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
+	0x08, 0x09, 0x0a, 0x0b,
+};
+
+static const unsigned char chroma_dc[16 + 12] = {
+	/* bits */
+	0x00, 0x03, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01,
+	0x01, 0x01, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00,
+	/* values */
+	0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
+	0x08, 0x09, 0x0a, 0x0b,
+};
+
+static const unsigned char luma_ac[16 + 162 + 2] = {
+	/* bits */
+	0x00, 0x02, 0x01, 0x03, 0x03, 0x02, 0x04, 0x03,
+	0x05, 0x05, 0x04, 0x04, 0x00, 0x00, 0x01, 0x7d,
+	/* values */
+	0x01, 0x02, 0x03, 0x00, 0x04, 0x11, 0x05, 0x12,
+	0x21, 0x31, 0x41, 0x06, 0x13, 0x51, 0x61, 0x07,
+	0x22, 0x71, 0x14, 0x32, 0x81, 0x91, 0xa1, 0x08,
+	0x23, 0x42, 0xb1, 0xc1, 0x15, 0x52, 0xd1, 0xf0,
+	0x24, 0x33, 0x62, 0x72, 0x82, 0x09, 0x0a, 0x16,
+	0x17, 0x18, 0x19, 0x1a, 0x25, 0x26, 0x27, 0x28,
+	0x29, 0x2a, 0x34, 0x35, 0x36, 0x37, 0x38, 0x39,
+	0x3a, 0x43, 0x44, 0x45, 0x46, 0x47, 0x48, 0x49,
+	0x4a, 0x53, 0x54, 0x55, 0x56, 0x57, 0x58, 0x59,
+	0x5a, 0x63, 0x64, 0x65, 0x66, 0x67, 0x68, 0x69,
+	0x6a, 0x73, 0x74, 0x75, 0x76, 0x77, 0x78, 0x79,
+	0x7a, 0x83, 0x84, 0x85, 0x86, 0x87, 0x88, 0x89,
+	0x8a, 0x92, 0x93, 0x94, 0x95, 0x96, 0x97, 0x98,
+	0x99, 0x9a, 0xa2, 0xa3, 0xa4, 0xa5, 0xa6, 0xa7,
+	0xa8, 0xa9, 0xaa, 0xb2, 0xb3, 0xb4, 0xb5, 0xb6,
+	0xb7, 0xb8, 0xb9, 0xba, 0xc2, 0xc3, 0xc4, 0xc5,
+	0xc6, 0xc7, 0xc8, 0xc9, 0xca, 0xd2, 0xd3, 0xd4,
+	0xd5, 0xd6, 0xd7, 0xd8, 0xd9, 0xda, 0xe1, 0xe2,
+	0xe3, 0xe4, 0xe5, 0xe6, 0xe7, 0xe8, 0xe9, 0xea,
+	0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7, 0xf8,
+	0xf9, 0xfa, /* padded to 32-bit */
+};
+
+static const unsigned char chroma_ac[16 + 162 + 2] = {
+	/* bits */
+	0x00, 0x02, 0x01, 0x02, 0x04, 0x04, 0x03, 0x04,
+	0x07, 0x05, 0x04, 0x04, 0x00, 0x01, 0x02, 0x77,
+	/* values */
+	0x00, 0x01, 0x02, 0x03, 0x11, 0x04, 0x05, 0x21,
+	0x31, 0x06, 0x12, 0x41, 0x51, 0x07, 0x61, 0x71,
+	0x13, 0x22, 0x32, 0x81, 0x08, 0x14, 0x42, 0x91,
+	0xa1, 0xb1, 0xc1, 0x09, 0x23, 0x33, 0x52, 0xf0,
+	0x15, 0x62, 0x72, 0xd1, 0x0a, 0x16, 0x24, 0x34,
+	0xe1, 0x25, 0xf1, 0x17, 0x18, 0x19, 0x1a, 0x26,
+	0x27, 0x28, 0x29, 0x2a, 0x35, 0x36, 0x37, 0x38,
+	0x39, 0x3a, 0x43, 0x44, 0x45, 0x46, 0x47, 0x48,
+	0x49, 0x4a, 0x53, 0x54, 0x55, 0x56, 0x57, 0x58,
+	0x59, 0x5a, 0x63, 0x64, 0x65, 0x66, 0x67, 0x68,
+	0x69, 0x6a, 0x73, 0x74, 0x75, 0x76, 0x77, 0x78,
+	0x79, 0x7a, 0x82, 0x83, 0x84, 0x85, 0x86, 0x87,
+	0x88, 0x89, 0x8a, 0x92, 0x93, 0x94, 0x95, 0x96,
+	0x97, 0x98, 0x99, 0x9a, 0xa2, 0xa3, 0xa4, 0xa5,
+	0xa6, 0xa7, 0xa8, 0xa9, 0xaa, 0xb2, 0xb3, 0xb4,
+	0xb5, 0xb6, 0xb7, 0xb8, 0xb9, 0xba, 0xc2, 0xc3,
+	0xc4, 0xc5, 0xc6, 0xc7, 0xc8, 0xc9, 0xca, 0xd2,
+	0xd3, 0xd4, 0xd5, 0xd6, 0xd7, 0xd8, 0xd9, 0xda,
+	0xe2, 0xe3, 0xe4, 0xe5, 0xe6, 0xe7, 0xe8, 0xe9,
+	0xea, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7, 0xf8,
+	0xf9, 0xfa, /* padded to 32-bit */
+};
+
+/*
+ * Quantization tables for luminance and chrominance components in
+ * zig-zag scan order from the Freescale i.MX VPU libraries
+ */
+
+static unsigned char luma_q[64] = {
+	0x06, 0x04, 0x04, 0x04, 0x05, 0x04, 0x06, 0x05,
+	0x05, 0x06, 0x09, 0x06, 0x05, 0x06, 0x09, 0x0b,
+	0x08, 0x06, 0x06, 0x08, 0x0b, 0x0c, 0x0a, 0x0a,
+	0x0b, 0x0a, 0x0a, 0x0c, 0x10, 0x0c, 0x0c, 0x0c,
+	0x0c, 0x0c, 0x0c, 0x10, 0x0c, 0x0c, 0x0c, 0x0c,
+	0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c,
+	0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c,
+	0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c,
+};
+
+static unsigned char chroma_q[64] = {
+	0x07, 0x07, 0x07, 0x0d, 0x0c, 0x0d, 0x18, 0x10,
+	0x10, 0x18, 0x14, 0x0e, 0x0e, 0x0e, 0x14, 0x14,
+	0x0e, 0x0e, 0x0e, 0x0e, 0x14, 0x11, 0x0c, 0x0c,
+	0x0c, 0x0c, 0x0c, 0x11, 0x11, 0x0c, 0x0c, 0x0c,
+	0x0c, 0x0c, 0x0c, 0x11, 0x0c, 0x0c, 0x0c, 0x0c,
+	0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c,
+	0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c,
+	0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c,
+};
+
+static const unsigned char width_align[] = {
+	[CODA9_JPEG_FORMAT_420] = 16,
+	[CODA9_JPEG_FORMAT_422] = 16,
+	[CODA9_JPEG_FORMAT_224] = 8,
+	[CODA9_JPEG_FORMAT_444] = 8,
+	[CODA9_JPEG_FORMAT_400] = 8,
+};
+
+static const unsigned char height_align[] = {
+	[CODA9_JPEG_FORMAT_420] = 16,
+	[CODA9_JPEG_FORMAT_422] = 8,
+	[CODA9_JPEG_FORMAT_224] = 16,
+	[CODA9_JPEG_FORMAT_444] = 8,
+	[CODA9_JPEG_FORMAT_400] = 8,
+};
+
+static int coda9_jpeg_chroma_format(u32 pixfmt)
+{
+	switch (pixfmt) {
+	case V4L2_PIX_FMT_YUV420:
+	case V4L2_PIX_FMT_NV12:
+		return CODA9_JPEG_FORMAT_420;
+	case V4L2_PIX_FMT_YUV422P:
+		return CODA9_JPEG_FORMAT_422;
+	case V4L2_PIX_FMT_YUV444:
+		return CODA9_JPEG_FORMAT_444;
+	case V4L2_PIX_FMT_GREY:
+		return CODA9_JPEG_FORMAT_400;
+	}
+	return -EINVAL;
+}
+
+struct coda_memcpy_desc {
+	int offset;
+	const void *src;
+	size_t len;
+};
+
+static void coda_memcpy_parabuf(void *parabuf,
+				const struct coda_memcpy_desc *desc)
+{
+	u32 *dst = parabuf + desc->offset;
+	const u32 *src = desc->src;
+	int len = desc->len / 4;
+	int i;
+
+	for (i = 0; i < len; i += 2) {
+		dst[i + 1] = swab32(src[i]);
+		dst[i] = swab32(src[i + 1]);
+	}
+}
+
+int coda_jpeg_write_tables(struct coda_ctx *ctx)
+{
+	int i;
+	static const struct coda_memcpy_desc huff[8] = {
+		{ 0,   luma_dc,    sizeof(luma_dc)    },
+		{ 32,  luma_ac,    sizeof(luma_ac)    },
+		{ 216, chroma_dc,  sizeof(chroma_dc)  },
+		{ 248, chroma_ac,  sizeof(chroma_ac)  },
+	};
+	struct coda_memcpy_desc qmat[3] = {
+		{ 512, ctx->params.jpeg_qmat_tab[0], 64 },
+		{ 576, ctx->params.jpeg_qmat_tab[1], 64 },
+		{ 640, ctx->params.jpeg_qmat_tab[1], 64 },
+	};
+
+	/* Write huffman tables to parameter memory */
+	for (i = 0; i < ARRAY_SIZE(huff); i++)
+		coda_memcpy_parabuf(ctx->parabuf.vaddr, huff + i);
+
+	/* Write Q-matrix to parameter memory */
+	for (i = 0; i < ARRAY_SIZE(qmat); i++)
+		coda_memcpy_parabuf(ctx->parabuf.vaddr, qmat + i);
+
+	return 0;
+}
+
+bool coda_jpeg_check_buffer(struct coda_ctx *ctx, struct vb2_buffer *vb)
+{
+	void *vaddr = vb2_plane_vaddr(vb, 0);
+	u16 soi, eoi;
+	int len, i;
+
+	soi = be16_to_cpup((__be16 *)vaddr);
+	if (soi != SOI_MARKER)
+		return false;
+
+	len = vb2_get_plane_payload(vb, 0);
+	vaddr += len - 2;
+	for (i = 0; i < 32; i++) {
+		eoi = be16_to_cpup((__be16 *)(vaddr - i));
+		if (eoi == EOI_MARKER) {
+			if (i > 0)
+				vb2_set_plane_payload(vb, 0, len - i);
+			return true;
+		}
+	}
+
+	return false;
+}
+
+static int coda9_jpeg_gen_dec_huff_tab(struct coda_ctx *ctx, int tab_num);
+
+int coda_jpeg_decode_header(struct coda_ctx *ctx, struct vb2_buffer *vb)
+{
+	struct coda_dev *dev = ctx->dev;
+	u8 *buf = vb2_plane_vaddr(vb, 0);
+	size_t len = vb2_get_plane_payload(vb, 0);
+	struct v4l2_jpeg_scan_header scan_header;
+	struct v4l2_jpeg_reference quantization_tables[4] = { };
+	struct v4l2_jpeg_reference huffman_tables[4] = { };
+	struct v4l2_jpeg_header header = {
+		.scan = &scan_header,
+		.quantization_tables = quantization_tables,
+		.huffman_tables = huffman_tables,
+	};
+	struct coda_q_data *q_data_src;
+	struct coda_huff_tab *huff_tab;
+	int i, j, ret;
+
+	ret = v4l2_jpeg_parse_header(buf, len, &header);
+	if (ret < 0) {
+		v4l2_err(&dev->v4l2_dev, "failed to parse JPEG header: %pe\n",
+			 ERR_PTR(ret));
+		return ret;
+	}
+
+	ctx->params.jpeg_restart_interval = header.restart_interval;
+
+	/* check frame header */
+	if (header.frame.height > ctx->codec->max_h ||
+	    header.frame.width > ctx->codec->max_w) {
+		v4l2_err(&dev->v4l2_dev, "invalid dimensions: %dx%d\n",
+			 header.frame.width, header.frame.height);
+		return -EINVAL;
+	}
+
+	q_data_src = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+	if (header.frame.height != q_data_src->height ||
+	    header.frame.width != q_data_src->width) {
+		v4l2_err(&dev->v4l2_dev,
+			 "dimensions don't match format: %dx%d\n",
+			 header.frame.width, header.frame.height);
+		return -EINVAL;
+	}
+
+	if (header.frame.num_components != 3) {
+		v4l2_err(&dev->v4l2_dev,
+			 "unsupported number of components: %d\n",
+			 header.frame.num_components);
+		return -EINVAL;
+	}
+
+	/* install quantization tables */
+	if (quantization_tables[3].start) {
+		v4l2_err(&dev->v4l2_dev,
+			 "only 3 quantization tables supported\n");
+		return -EINVAL;
+	}
+	for (i = 0; i < 3; i++) {
+		if (!quantization_tables[i].start)
+			continue;
+		if (quantization_tables[i].length != 64) {
+			v4l2_err(&dev->v4l2_dev,
+				 "only 8-bit quantization tables supported\n");
+			continue;
+		}
+		if (!ctx->params.jpeg_qmat_tab[i]) {
+			ctx->params.jpeg_qmat_tab[i] = kmalloc(64, GFP_KERNEL);
+			if (!ctx->params.jpeg_qmat_tab[i])
+				return -ENOMEM;
+		}
+		memcpy(ctx->params.jpeg_qmat_tab[i],
+		       quantization_tables[i].start, 64);
+	}
+
+	/* install Huffman tables */
+	for (i = 0; i < 4; i++) {
+		if (!huffman_tables[i].start) {
+			v4l2_err(&dev->v4l2_dev, "missing Huffman table\n");
+			return -EINVAL;
+		}
+		/* AC tables should be between 17 -> 178, DC between 17 -> 28 */
+		if (huffman_tables[i].length < 17 ||
+		    huffman_tables[i].length > 178 ||
+		    ((i & 2) == 0 && huffman_tables[i].length > 28)) {
+			v4l2_err(&dev->v4l2_dev,
+				 "invalid Huffman table %d length: %zu\n",
+				 i, huffman_tables[i].length);
+			return -EINVAL;
+		}
+	}
+	huff_tab = ctx->params.jpeg_huff_tab;
+	if (!huff_tab) {
+		huff_tab = kzalloc(sizeof(struct coda_huff_tab), GFP_KERNEL);
+		if (!huff_tab)
+			return -ENOMEM;
+		ctx->params.jpeg_huff_tab = huff_tab;
+	}
+
+	memset(huff_tab, 0, sizeof(*huff_tab));
+	memcpy(huff_tab->luma_dc, huffman_tables[0].start, huffman_tables[0].length);
+	memcpy(huff_tab->chroma_dc, huffman_tables[1].start, huffman_tables[1].length);
+	memcpy(huff_tab->luma_ac, huffman_tables[2].start, huffman_tables[2].length);
+	memcpy(huff_tab->chroma_ac, huffman_tables[3].start, huffman_tables[3].length);
+
+	/* check scan header */
+	for (i = 0; i < scan_header.num_components; i++) {
+		struct v4l2_jpeg_scan_component_spec *scan_component;
+
+		scan_component = &scan_header.component[i];
+		for (j = 0; j < header.frame.num_components; j++) {
+			if (header.frame.component[j].component_identifier ==
+			    scan_component->component_selector)
+				break;
+		}
+		if (j == header.frame.num_components)
+			continue;
+
+		ctx->params.jpeg_huff_dc_index[j] =
+			scan_component->dc_entropy_coding_table_selector;
+		ctx->params.jpeg_huff_ac_index[j] =
+			scan_component->ac_entropy_coding_table_selector;
+	}
+
+	/* Generate Huffman table information */
+	for (i = 0; i < 4; i++)
+		coda9_jpeg_gen_dec_huff_tab(ctx, i);
+
+	/* start of entropy coded segment */
+	ctx->jpeg_ecs_offset = header.ecs_offset;
+
+	switch (header.frame.subsampling) {
+	case V4L2_JPEG_CHROMA_SUBSAMPLING_420:
+	case V4L2_JPEG_CHROMA_SUBSAMPLING_422:
+		ctx->params.jpeg_chroma_subsampling = header.frame.subsampling;
+		break;
+	default:
+		v4l2_err(&dev->v4l2_dev, "chroma subsampling not supported: %d",
+			 header.frame.subsampling);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static inline void coda9_jpeg_write_huff_values(struct coda_dev *dev, u8 *bits,
+						int num_values)
+{
+	s8 *values = (s8 *)(bits + 16);
+	int huff_length, i;
+
+	for (huff_length = 0, i = 0; i < 16; i++)
+		huff_length += bits[i];
+	for (i = huff_length; i < num_values; i++)
+		values[i] = -1;
+	for (i = 0; i < num_values; i++)
+		coda_write(dev, (s32)values[i], CODA9_REG_JPEG_HUFF_DATA);
+}
+
+static void coda9_jpeg_dec_huff_setup(struct coda_ctx *ctx)
+{
+	struct coda_huff_tab *huff_tab = ctx->params.jpeg_huff_tab;
+	struct coda_dev *dev = ctx->dev;
+	s16 *huff_min = huff_tab->min;
+	s16 *huff_max = huff_tab->max;
+	s8 *huff_ptr = huff_tab->ptr;
+	int i;
+
+	/* MIN Tables */
+	coda_write(dev, 0x003, CODA9_REG_JPEG_HUFF_CTRL);
+	coda_write(dev, 0x000, CODA9_REG_JPEG_HUFF_ADDR);
+	for (i = 0; i < 4 * 16; i++)
+		coda_write(dev, (s32)huff_min[i], CODA9_REG_JPEG_HUFF_DATA);
+
+	/* MAX Tables */
+	coda_write(dev, 0x403, CODA9_REG_JPEG_HUFF_CTRL);
+	coda_write(dev, 0x440, CODA9_REG_JPEG_HUFF_ADDR);
+	for (i = 0; i < 4 * 16; i++)
+		coda_write(dev, (s32)huff_max[i], CODA9_REG_JPEG_HUFF_DATA);
+
+	/* PTR Tables */
+	coda_write(dev, 0x803, CODA9_REG_JPEG_HUFF_CTRL);
+	coda_write(dev, 0x880, CODA9_REG_JPEG_HUFF_ADDR);
+	for (i = 0; i < 4 * 16; i++)
+		coda_write(dev, (s32)huff_ptr[i], CODA9_REG_JPEG_HUFF_DATA);
+
+	/* VAL Tables: DC Luma, DC Chroma, AC Luma, AC Chroma */
+	coda_write(dev, 0xc03, CODA9_REG_JPEG_HUFF_CTRL);
+	coda9_jpeg_write_huff_values(dev, huff_tab->luma_dc, 12);
+	coda9_jpeg_write_huff_values(dev, huff_tab->chroma_dc, 12);
+	coda9_jpeg_write_huff_values(dev, huff_tab->luma_ac, 162);
+	coda9_jpeg_write_huff_values(dev, huff_tab->chroma_ac, 162);
+	coda_write(dev, 0x000, CODA9_REG_JPEG_HUFF_CTRL);
+}
+
+static inline void coda9_jpeg_write_qmat_tab(struct coda_dev *dev,
+					     u8 *qmat, int index)
+{
+	int i;
+
+	coda_write(dev, index | 0x3, CODA9_REG_JPEG_QMAT_CTRL);
+	for (i = 0; i < 64; i++)
+		coda_write(dev, qmat[i], CODA9_REG_JPEG_QMAT_DATA);
+	coda_write(dev, 0, CODA9_REG_JPEG_QMAT_CTRL);
+}
+
+static void coda9_jpeg_qmat_setup(struct coda_ctx *ctx)
+{
+	struct coda_dev *dev = ctx->dev;
+	int *qmat_index = ctx->params.jpeg_qmat_index;
+	u8 **qmat_tab = ctx->params.jpeg_qmat_tab;
+
+	coda9_jpeg_write_qmat_tab(dev, qmat_tab[qmat_index[0]], 0x00);
+	coda9_jpeg_write_qmat_tab(dev, qmat_tab[qmat_index[1]], 0x40);
+	coda9_jpeg_write_qmat_tab(dev, qmat_tab[qmat_index[2]], 0x80);
+}
+
+static void coda9_jpeg_dec_bbc_gbu_setup(struct coda_ctx *ctx,
+					 struct vb2_buffer *buf, u32 ecs_offset)
+{
+	struct coda_dev *dev = ctx->dev;
+	int page_ptr, word_ptr, bit_ptr;
+	u32 bbc_base_addr, end_addr;
+	int bbc_cur_pos;
+	int ret, val;
+
+	bbc_base_addr = vb2_dma_contig_plane_dma_addr(buf, 0);
+	end_addr = bbc_base_addr + vb2_get_plane_payload(buf, 0);
+
+	page_ptr = ecs_offset / 256;
+	word_ptr = (ecs_offset % 256) / 4;
+	if (page_ptr & 1)
+		word_ptr += 64;
+	bit_ptr = (ecs_offset % 4) * 8;
+	if (word_ptr & 1)
+		bit_ptr += 32;
+	word_ptr &= ~0x1;
+
+	coda_write(dev, end_addr, CODA9_REG_JPEG_BBC_WR_PTR);
+	coda_write(dev, bbc_base_addr, CODA9_REG_JPEG_BBC_BAS_ADDR);
+
+	/* Leave 3 256-byte page margin to avoid a BBC interrupt */
+	coda_write(dev, end_addr + 256 * 3 + 256, CODA9_REG_JPEG_BBC_END_ADDR);
+	val = DIV_ROUND_UP(vb2_plane_size(buf, 0), 256) + 3;
+	coda_write(dev, BIT(31) | val, CODA9_REG_JPEG_BBC_STRM_CTRL);
+
+	bbc_cur_pos = page_ptr;
+	coda_write(dev, bbc_cur_pos, CODA9_REG_JPEG_BBC_CUR_POS);
+	coda_write(dev, bbc_base_addr + (bbc_cur_pos << 8),
+			CODA9_REG_JPEG_BBC_EXT_ADDR);
+	coda_write(dev, (bbc_cur_pos & 1) << 6, CODA9_REG_JPEG_BBC_INT_ADDR);
+	coda_write(dev, 64, CODA9_REG_JPEG_BBC_DATA_CNT);
+	coda_write(dev, 0, CODA9_REG_JPEG_BBC_COMMAND);
+	do {
+		ret = coda_read(dev, CODA9_REG_JPEG_BBC_BUSY);
+	} while (ret == 1);
+
+	bbc_cur_pos++;
+	coda_write(dev, bbc_cur_pos, CODA9_REG_JPEG_BBC_CUR_POS);
+	coda_write(dev, bbc_base_addr + (bbc_cur_pos << 8),
+			CODA9_REG_JPEG_BBC_EXT_ADDR);
+	coda_write(dev, (bbc_cur_pos & 1) << 6, CODA9_REG_JPEG_BBC_INT_ADDR);
+	coda_write(dev, 64, CODA9_REG_JPEG_BBC_DATA_CNT);
+	coda_write(dev, 0, CODA9_REG_JPEG_BBC_COMMAND);
+	do {
+		ret = coda_read(dev, CODA9_REG_JPEG_BBC_BUSY);
+	} while (ret == 1);
+
+	bbc_cur_pos++;
+	coda_write(dev, bbc_cur_pos, CODA9_REG_JPEG_BBC_CUR_POS);
+	coda_write(dev, 1, CODA9_REG_JPEG_BBC_CTRL);
+
+	coda_write(dev, 0, CODA9_REG_JPEG_GBU_TT_CNT);
+	coda_write(dev, word_ptr, CODA9_REG_JPEG_GBU_WD_PTR);
+	coda_write(dev, 0, CODA9_REG_JPEG_GBU_BBSR);
+	coda_write(dev, 127, CODA9_REG_JPEG_GBU_BBER);
+	if (page_ptr & 1) {
+		coda_write(dev, 0, CODA9_REG_JPEG_GBU_BBIR);
+		coda_write(dev, 0, CODA9_REG_JPEG_GBU_BBHR);
+	} else {
+		coda_write(dev, 64, CODA9_REG_JPEG_GBU_BBIR);
+		coda_write(dev, 64, CODA9_REG_JPEG_GBU_BBHR);
+	}
+	coda_write(dev, 4, CODA9_REG_JPEG_GBU_CTRL);
+	coda_write(dev, bit_ptr, CODA9_REG_JPEG_GBU_FF_RPTR);
+	coda_write(dev, 3, CODA9_REG_JPEG_GBU_CTRL);
+}
+
+static const int bus_req_num[] = {
+	[CODA9_JPEG_FORMAT_420] = 2,
+	[CODA9_JPEG_FORMAT_422] = 3,
+	[CODA9_JPEG_FORMAT_224] = 3,
+	[CODA9_JPEG_FORMAT_444] = 4,
+	[CODA9_JPEG_FORMAT_400] = 4,
+};
+
+#define MCU_INFO(mcu_block_num, comp_num, comp0_info, comp1_info, comp2_info) \
+	(((mcu_block_num) << CODA9_JPEG_MCU_BLOCK_NUM_OFFSET) | \
+	 ((comp_num) << CODA9_JPEG_COMP_NUM_OFFSET) | \
+	 ((comp0_info) << CODA9_JPEG_COMP0_INFO_OFFSET) | \
+	 ((comp1_info) << CODA9_JPEG_COMP1_INFO_OFFSET) | \
+	 ((comp2_info) << CODA9_JPEG_COMP2_INFO_OFFSET))
+
+static const u32 mcu_info[] = {
+	[CODA9_JPEG_FORMAT_420] = MCU_INFO(6, 3, 10, 5, 5),
+	[CODA9_JPEG_FORMAT_422] = MCU_INFO(4, 3, 9, 5, 5),
+	[CODA9_JPEG_FORMAT_224] = MCU_INFO(4, 3, 6, 5, 5),
+	[CODA9_JPEG_FORMAT_444] = MCU_INFO(3, 3, 5, 5, 5),
+	[CODA9_JPEG_FORMAT_400] = MCU_INFO(1, 1, 5, 0, 0),
+};
+
+/*
+ * Convert Huffman table specifcations to tables of codes and code lengths.
+ * For reference, see JPEG ITU-T.81 (ISO/IEC 10918-1) [1]
+ *
+ * [1] https://www.w3.org/Graphics/JPEG/itu-t81.pdf
+ */
+static int coda9_jpeg_gen_enc_huff_tab(struct coda_ctx *ctx, int tab_num,
+				       int *ehufsi, int *ehufco)
+{
+	int i, j, k, lastk, si, code, maxsymbol;
+	const u8 *bits, *huffval;
+	struct {
+		int size[256];
+		int code[256];
+	} *huff;
+	static const unsigned char *huff_tabs[4] = {
+		luma_dc, luma_ac, chroma_dc, chroma_ac,
+	};
+	int ret = -EINVAL;
+
+	huff = kzalloc(sizeof(*huff), GFP_KERNEL);
+	if (!huff)
+		return -ENOMEM;
+
+	bits = huff_tabs[tab_num];
+	huffval = huff_tabs[tab_num] + 16;
+
+	maxsymbol = tab_num & 1 ? 256 : 16;
+
+	/* Figure C.1 - Generation of table of Huffman code sizes */
+	k = 0;
+	for (i = 1; i <= 16; i++) {
+		j = bits[i - 1];
+		if (k + j > maxsymbol)
+			goto out;
+		while (j--)
+			huff->size[k++] = i;
+	}
+	lastk = k;
+
+	/* Figure C.2 - Generation of table of Huffman codes */
+	k = 0;
+	code = 0;
+	si = huff->size[0];
+	while (k < lastk) {
+		while (huff->size[k] == si) {
+			huff->code[k++] = code;
+			code++;
+		}
+		if (code >= (1 << si))
+			goto out;
+		code <<= 1;
+		si++;
+	}
+
+	/* Figure C.3 - Ordering procedure for encoding procedure code tables */
+	for (k = 0; k < lastk; k++) {
+		i = huffval[k];
+		if (i >= maxsymbol || ehufsi[i])
+			goto out;
+		ehufco[i] = huff->code[k];
+		ehufsi[i] = huff->size[k];
+	}
+
+	ret = 0;
+out:
+	kfree(huff);
+	return ret;
+}
+
+#define DC_TABLE_INDEX0		    0
+#define AC_TABLE_INDEX0		    1
+#define DC_TABLE_INDEX1		    2
+#define AC_TABLE_INDEX1		    3
+
+static u8 *coda9_jpeg_get_huff_bits(struct coda_ctx *ctx, int tab_num)
+{
+	struct coda_huff_tab *huff_tab = ctx->params.jpeg_huff_tab;
+
+	if (!huff_tab)
+		return NULL;
+
+	switch (tab_num) {
+	case DC_TABLE_INDEX0: return huff_tab->luma_dc;
+	case AC_TABLE_INDEX0: return huff_tab->luma_ac;
+	case DC_TABLE_INDEX1: return huff_tab->chroma_dc;
+	case AC_TABLE_INDEX1: return huff_tab->chroma_ac;
+	}
+
+	return NULL;
+}
+
+static int coda9_jpeg_gen_dec_huff_tab(struct coda_ctx *ctx, int tab_num)
+{
+	int ptr_cnt = 0, huff_code = 0, zero_flag = 0, data_flag = 0;
+	u8 *huff_bits;
+	s16 *huff_max;
+	s16 *huff_min;
+	s8 *huff_ptr;
+	int ofs;
+	int i;
+
+	huff_bits = coda9_jpeg_get_huff_bits(ctx, tab_num);
+	if (!huff_bits)
+		return -EINVAL;
+
+	/* DC/AC Luma, DC/AC Chroma -> DC Luma/Chroma, AC Luma/Chroma */
+	ofs = ((tab_num & 1) << 1) | ((tab_num >> 1) & 1);
+	ofs *= 16;
+
+	huff_ptr = ctx->params.jpeg_huff_tab->ptr + ofs;
+	huff_max = ctx->params.jpeg_huff_tab->max + ofs;
+	huff_min = ctx->params.jpeg_huff_tab->min + ofs;
+
+	for (i = 0; i < 16; i++) {
+		if (huff_bits[i]) {
+			huff_ptr[i] = ptr_cnt;
+			ptr_cnt += huff_bits[i];
+			huff_min[i] = huff_code;
+			huff_max[i] = huff_code + (huff_bits[i] - 1);
+			data_flag = 1;
+			zero_flag = 0;
+		} else {
+			huff_ptr[i] = -1;
+			huff_min[i] = -1;
+			huff_max[i] = -1;
+			zero_flag = 1;
+		}
+
+		if (data_flag == 1) {
+			if (zero_flag == 1)
+				huff_code <<= 1;
+			else
+				huff_code = (huff_max[i] + 1) << 1;
+		}
+	}
+
+	return 0;
+}
+
+static int coda9_jpeg_load_huff_tab(struct coda_ctx *ctx)
+{
+	struct {
+		int size[4][256];
+		int code[4][256];
+	} *huff;
+	u32 *huff_data;
+	int i, j;
+	int ret;
+
+	huff = kzalloc(sizeof(*huff), GFP_KERNEL);
+	if (!huff)
+		return -ENOMEM;
+
+	/* Generate all four (luma/chroma DC/AC) code/size lookup tables */
+	for (i = 0; i < 4; i++) {
+		ret = coda9_jpeg_gen_enc_huff_tab(ctx, i, huff->size[i],
+						  huff->code[i]);
+		if (ret)
+			goto out;
+	}
+
+	if (!ctx->params.jpeg_huff_data) {
+		ctx->params.jpeg_huff_data =
+			kzalloc(sizeof(u32) * CODA9_JPEG_ENC_HUFF_DATA_SIZE,
+				GFP_KERNEL);
+		if (!ctx->params.jpeg_huff_data) {
+			ret = -ENOMEM;
+			goto out;
+		}
+	}
+	huff_data = ctx->params.jpeg_huff_data;
+
+	for (j = 0; j < 4; j++) {
+		/* Store Huffman lookup tables in AC0, AC1, DC0, DC1 order */
+		int t = (j == 0) ? AC_TABLE_INDEX0 :
+			(j == 1) ? AC_TABLE_INDEX1 :
+			(j == 2) ? DC_TABLE_INDEX0 :
+				   DC_TABLE_INDEX1;
+		/* DC tables only have 16 entries */
+		int len = (j < 2) ? 256 : 16;
+
+		for (i = 0; i < len; i++) {
+			if (huff->size[t][i] == 0 && huff->code[t][i] == 0)
+				*(huff_data++) = 0;
+			else
+				*(huff_data++) =
+					((huff->size[t][i] - 1) << 16) |
+					huff->code[t][i];
+		}
+	}
+
+	ret = 0;
+out:
+	kfree(huff);
+	return ret;
+}
+
+static void coda9_jpeg_write_huff_tab(struct coda_ctx *ctx)
+{
+	struct coda_dev *dev = ctx->dev;
+	u32 *huff_data = ctx->params.jpeg_huff_data;
+	int i;
+
+	/* Write Huffman size/code lookup tables in AC0, AC1, DC0, DC1 order */
+	coda_write(dev, 0x3, CODA9_REG_JPEG_HUFF_CTRL);
+	for (i = 0; i < CODA9_JPEG_ENC_HUFF_DATA_SIZE; i++)
+		coda_write(dev, *(huff_data++), CODA9_REG_JPEG_HUFF_DATA);
+	coda_write(dev, 0x0, CODA9_REG_JPEG_HUFF_CTRL);
+}
+
+static inline void coda9_jpeg_write_qmat_quotients(struct coda_dev *dev,
+						   u8 *qmat, int index)
+{
+	int i;
+
+	coda_write(dev, index | 0x3, CODA9_REG_JPEG_QMAT_CTRL);
+	for (i = 0; i < 64; i++)
+		coda_write(dev, 0x80000 / qmat[i], CODA9_REG_JPEG_QMAT_DATA);
+	coda_write(dev, index, CODA9_REG_JPEG_QMAT_CTRL);
+}
+
+static void coda9_jpeg_load_qmat_tab(struct coda_ctx *ctx)
+{
+	struct coda_dev *dev = ctx->dev;
+	u8 *luma_tab;
+	u8 *chroma_tab;
+
+	luma_tab = ctx->params.jpeg_qmat_tab[0];
+	if (!luma_tab)
+		luma_tab = luma_q;
+
+	chroma_tab = ctx->params.jpeg_qmat_tab[1];
+	if (!chroma_tab)
+		chroma_tab = chroma_q;
+
+	coda9_jpeg_write_qmat_quotients(dev, luma_tab, 0x00);
+	coda9_jpeg_write_qmat_quotients(dev, chroma_tab, 0x40);
+	coda9_jpeg_write_qmat_quotients(dev, chroma_tab, 0x80);
+}
+
+struct coda_jpeg_stream {
+	u8 *curr;
+	u8 *end;
+};
+
+static inline int coda_jpeg_put_byte(u8 byte, struct coda_jpeg_stream *stream)
+{
+	if (stream->curr >= stream->end)
+		return -EINVAL;
+
+	*stream->curr++ = byte;
+
+	return 0;
+}
+
+static inline int coda_jpeg_put_word(u16 word, struct coda_jpeg_stream *stream)
+{
+	if (stream->curr + sizeof(__be16) > stream->end)
+		return -EINVAL;
+
+	put_unaligned_be16(word, stream->curr);
+	stream->curr += sizeof(__be16);
+
+	return 0;
+}
+
+static int coda_jpeg_put_table(u16 marker, u8 index, const u8 *table,
+			       size_t len, struct coda_jpeg_stream *stream)
+{
+	int i, ret;
+
+	ret = coda_jpeg_put_word(marker, stream);
+	if (ret < 0)
+		return ret;
+	ret = coda_jpeg_put_word(3 + len, stream);
+	if (ret < 0)
+		return ret;
+	ret = coda_jpeg_put_byte(index, stream);
+	for (i = 0; i < len && ret == 0; i++)
+		ret = coda_jpeg_put_byte(table[i], stream);
+
+	return ret;
+}
+
+static int coda_jpeg_define_quantization_table(struct coda_ctx *ctx, u8 index,
+					       struct coda_jpeg_stream *stream)
+{
+	return coda_jpeg_put_table(DQT_MARKER, index,
+				   ctx->params.jpeg_qmat_tab[index], 64,
+				   stream);
+}
+
+static int coda_jpeg_define_huffman_table(u8 index, const u8 *table, size_t len,
+					  struct coda_jpeg_stream *stream)
+{
+	return coda_jpeg_put_table(DHT_MARKER, index, table, len, stream);
+}
+
+static int coda9_jpeg_encode_header(struct coda_ctx *ctx, int len, u8 *buf)
+{
+	struct coda_jpeg_stream stream = { buf, buf + len };
+	struct coda_q_data *q_data_src;
+	int chroma_format, comp_num;
+	int i, ret, pad;
+
+	q_data_src = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+	chroma_format = coda9_jpeg_chroma_format(q_data_src->fourcc);
+	if (chroma_format < 0)
+		return 0;
+
+	/* Start Of Image */
+	ret = coda_jpeg_put_word(SOI_MARKER, &stream);
+	if (ret < 0)
+		return ret;
+
+	/* Define Restart Interval */
+	if (ctx->params.jpeg_restart_interval) {
+		ret = coda_jpeg_put_word(DRI_MARKER, &stream);
+		if (ret < 0)
+			return ret;
+		ret = coda_jpeg_put_word(4, &stream);
+		if (ret < 0)
+			return ret;
+		ret = coda_jpeg_put_word(ctx->params.jpeg_restart_interval,
+					 &stream);
+		if (ret < 0)
+			return ret;
+	}
+
+	/* Define Quantization Tables */
+	ret = coda_jpeg_define_quantization_table(ctx, 0x00, &stream);
+	if (ret < 0)
+		return ret;
+	if (chroma_format != CODA9_JPEG_FORMAT_400) {
+		ret = coda_jpeg_define_quantization_table(ctx, 0x01, &stream);
+		if (ret < 0)
+			return ret;
+	}
+
+	/* Define Huffman Tables */
+	ret = coda_jpeg_define_huffman_table(0x00, luma_dc, 16 + 12, &stream);
+	if (ret < 0)
+		return ret;
+	ret = coda_jpeg_define_huffman_table(0x10, luma_ac, 16 + 162, &stream);
+	if (ret < 0)
+		return ret;
+	if (chroma_format != CODA9_JPEG_FORMAT_400) {
+		ret = coda_jpeg_define_huffman_table(0x01, chroma_dc, 16 + 12,
+						     &stream);
+		if (ret < 0)
+			return ret;
+		ret = coda_jpeg_define_huffman_table(0x11, chroma_ac, 16 + 162,
+						     &stream);
+		if (ret < 0)
+			return ret;
+	}
+
+	/* Start Of Frame */
+	ret = coda_jpeg_put_word(SOF_MARKER, &stream);
+	if (ret < 0)
+		return ret;
+	comp_num = (chroma_format == CODA9_JPEG_FORMAT_400) ? 1 : 3;
+	ret = coda_jpeg_put_word(8 + comp_num * 3, &stream);
+	if (ret < 0)
+		return ret;
+	ret = coda_jpeg_put_byte(0x08, &stream);
+	if (ret < 0)
+		return ret;
+	ret = coda_jpeg_put_word(q_data_src->height, &stream);
+	if (ret < 0)
+		return ret;
+	ret = coda_jpeg_put_word(q_data_src->width, &stream);
+	if (ret < 0)
+		return ret;
+	ret = coda_jpeg_put_byte(comp_num, &stream);
+	if (ret < 0)
+		return ret;
+	for (i = 0; i < comp_num; i++) {
+		static unsigned char subsampling[5][3] = {
+			[CODA9_JPEG_FORMAT_420] = { 0x22, 0x11, 0x11 },
+			[CODA9_JPEG_FORMAT_422] = { 0x21, 0x11, 0x11 },
+			[CODA9_JPEG_FORMAT_224] = { 0x12, 0x11, 0x11 },
+			[CODA9_JPEG_FORMAT_444] = { 0x11, 0x11, 0x11 },
+			[CODA9_JPEG_FORMAT_400] = { 0x11 },
+		};
+
+		/* Component identifier, matches SOS */
+		ret = coda_jpeg_put_byte(i + 1, &stream);
+		if (ret < 0)
+			return ret;
+		ret = coda_jpeg_put_byte(subsampling[chroma_format][i],
+					 &stream);
+		if (ret < 0)
+			return ret;
+		/* Chroma table index */
+		ret = coda_jpeg_put_byte((i == 0) ? 0 : 1, &stream);
+		if (ret < 0)
+			return ret;
+	}
+
+	/* Pad to multiple of 8 bytes */
+	pad = (stream.curr - buf) % 8;
+	if (pad) {
+		pad = 8 - pad;
+		while (pad--) {
+			ret = coda_jpeg_put_byte(0x00, &stream);
+			if (ret < 0)
+				return ret;
+		}
+	}
+
+	return stream.curr - buf;
+}
+
+/*
+ * Scale quantization table using nonlinear scaling factor
+ * u8 qtab[64], scale [50,190]
+ */
+static void coda_scale_quant_table(u8 *q_tab, int scale)
+{
+	unsigned int temp;
+	int i;
+
+	for (i = 0; i < 64; i++) {
+		temp = DIV_ROUND_CLOSEST((unsigned int)q_tab[i] * scale, 100);
+		if (temp <= 0)
+			temp = 1;
+		if (temp > 255)
+			temp = 255;
+		q_tab[i] = (unsigned char)temp;
+	}
+}
+
+void coda_set_jpeg_compression_quality(struct coda_ctx *ctx, int quality)
+{
+	unsigned int scale;
+
+	ctx->params.jpeg_quality = quality;
+
+	/* Clip quality setting to [5,100] interval */
+	if (quality > 100)
+		quality = 100;
+	if (quality < 5)
+		quality = 5;
+
+	/*
+	 * Non-linear scaling factor:
+	 * [5,50] -> [1000..100], [51,100] -> [98..0]
+	 */
+	if (quality < 50)
+		scale = 5000 / quality;
+	else
+		scale = 200 - 2 * quality;
+
+	if (ctx->params.jpeg_qmat_tab[0]) {
+		memcpy(ctx->params.jpeg_qmat_tab[0], luma_q, 64);
+		coda_scale_quant_table(ctx->params.jpeg_qmat_tab[0], scale);
+	}
+	if (ctx->params.jpeg_qmat_tab[1]) {
+		memcpy(ctx->params.jpeg_qmat_tab[1], chroma_q, 64);
+		coda_scale_quant_table(ctx->params.jpeg_qmat_tab[1], scale);
+	}
+}
+
+/*
+ * Encoder context operations
+ */
+
+static int coda9_jpeg_start_encoding(struct coda_ctx *ctx)
+{
+	struct coda_dev *dev = ctx->dev;
+	int ret;
+
+	ret = coda9_jpeg_load_huff_tab(ctx);
+	if (ret < 0) {
+		v4l2_err(&dev->v4l2_dev, "error loading Huffman tables\n");
+		return ret;
+	}
+	if (!ctx->params.jpeg_qmat_tab[0]) {
+		ctx->params.jpeg_qmat_tab[0] = kmalloc(64, GFP_KERNEL);
+		if (!ctx->params.jpeg_qmat_tab[0])
+			return -ENOMEM;
+	}
+	if (!ctx->params.jpeg_qmat_tab[1]) {
+		ctx->params.jpeg_qmat_tab[1] = kmalloc(64, GFP_KERNEL);
+		if (!ctx->params.jpeg_qmat_tab[1])
+			return -ENOMEM;
+	}
+	coda_set_jpeg_compression_quality(ctx, ctx->params.jpeg_quality);
+
+	return 0;
+}
+
+static int coda9_jpeg_prepare_encode(struct coda_ctx *ctx)
+{
+	struct coda_q_data *q_data_src;
+	struct vb2_v4l2_buffer *src_buf, *dst_buf;
+	struct coda_dev *dev = ctx->dev;
+	u32 start_addr, end_addr;
+	u16 aligned_width, aligned_height;
+	bool chroma_interleave;
+	int chroma_format;
+	int header_len;
+	int ret;
+	ktime_t timeout;
+
+	src_buf = v4l2_m2m_next_src_buf(ctx->fh.m2m_ctx);
+	dst_buf = v4l2_m2m_next_dst_buf(ctx->fh.m2m_ctx);
+	q_data_src = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+
+	if (vb2_get_plane_payload(&src_buf->vb2_buf, 0) == 0)
+		vb2_set_plane_payload(&src_buf->vb2_buf, 0,
+				      vb2_plane_size(&src_buf->vb2_buf, 0));
+
+	src_buf->sequence = ctx->osequence;
+	dst_buf->sequence = ctx->osequence;
+	ctx->osequence++;
+
+	src_buf->flags |= V4L2_BUF_FLAG_KEYFRAME;
+	src_buf->flags &= ~V4L2_BUF_FLAG_PFRAME;
+
+	coda_set_gdi_regs(ctx);
+
+	start_addr = vb2_dma_contig_plane_dma_addr(&dst_buf->vb2_buf, 0);
+	end_addr = start_addr + vb2_plane_size(&dst_buf->vb2_buf, 0);
+
+	chroma_format = coda9_jpeg_chroma_format(q_data_src->fourcc);
+	if (chroma_format < 0)
+		return chroma_format;
+
+	/* Round image dimensions to multiple of MCU size */
+	aligned_width = round_up(q_data_src->width, width_align[chroma_format]);
+	aligned_height = round_up(q_data_src->height,
+				  height_align[chroma_format]);
+	if (aligned_width != q_data_src->bytesperline) {
+		v4l2_err(&dev->v4l2_dev, "wrong stride: %d instead of %d\n",
+			 aligned_width, q_data_src->bytesperline);
+	}
+
+	header_len =
+		coda9_jpeg_encode_header(ctx,
+					 vb2_plane_size(&dst_buf->vb2_buf, 0),
+					 vb2_plane_vaddr(&dst_buf->vb2_buf, 0));
+	if (header_len < 0)
+		return header_len;
+
+	coda_write(dev, start_addr + header_len, CODA9_REG_JPEG_BBC_BAS_ADDR);
+	coda_write(dev, end_addr, CODA9_REG_JPEG_BBC_END_ADDR);
+	coda_write(dev, start_addr + header_len, CODA9_REG_JPEG_BBC_WR_PTR);
+	coda_write(dev, start_addr + header_len, CODA9_REG_JPEG_BBC_RD_PTR);
+	coda_write(dev, 0, CODA9_REG_JPEG_BBC_CUR_POS);
+	/* 64 words per 256-byte page */
+	coda_write(dev, 64, CODA9_REG_JPEG_BBC_DATA_CNT);
+	coda_write(dev, start_addr, CODA9_REG_JPEG_BBC_EXT_ADDR);
+	coda_write(dev, 0, CODA9_REG_JPEG_BBC_INT_ADDR);
+
+	coda_write(dev, 0, CODA9_REG_JPEG_GBU_BT_PTR);
+	coda_write(dev, 0, CODA9_REG_JPEG_GBU_WD_PTR);
+	coda_write(dev, 0, CODA9_REG_JPEG_GBU_BBSR);
+	coda_write(dev, BIT(31) | ((end_addr - start_addr - header_len) / 256),
+		   CODA9_REG_JPEG_BBC_STRM_CTRL);
+	coda_write(dev, 0, CODA9_REG_JPEG_GBU_CTRL);
+	coda_write(dev, 0, CODA9_REG_JPEG_GBU_FF_RPTR);
+	coda_write(dev, 127, CODA9_REG_JPEG_GBU_BBER);
+	coda_write(dev, 64, CODA9_REG_JPEG_GBU_BBIR);
+	coda_write(dev, 64, CODA9_REG_JPEG_GBU_BBHR);
+
+	chroma_interleave = (q_data_src->fourcc == V4L2_PIX_FMT_NV12);
+	coda_write(dev, CODA9_JPEG_PIC_CTRL_TC_DIRECTION |
+		   CODA9_JPEG_PIC_CTRL_ENCODER_EN, CODA9_REG_JPEG_PIC_CTRL);
+	coda_write(dev, 0, CODA9_REG_JPEG_SCL_INFO);
+	coda_write(dev, chroma_interleave, CODA9_REG_JPEG_DPB_CONFIG);
+	coda_write(dev, ctx->params.jpeg_restart_interval,
+		   CODA9_REG_JPEG_RST_INTVAL);
+	coda_write(dev, 1, CODA9_REG_JPEG_BBC_CTRL);
+
+	coda_write(dev, bus_req_num[chroma_format], CODA9_REG_JPEG_OP_INFO);
+
+	coda9_jpeg_write_huff_tab(ctx);
+	coda9_jpeg_load_qmat_tab(ctx);
+
+	if (ctx->params.rot_mode & CODA_ROT_90) {
+		aligned_width = aligned_height;
+		aligned_height = q_data_src->bytesperline;
+		if (chroma_format == CODA9_JPEG_FORMAT_422)
+			chroma_format = CODA9_JPEG_FORMAT_224;
+		else if (chroma_format == CODA9_JPEG_FORMAT_224)
+			chroma_format = CODA9_JPEG_FORMAT_422;
+	}
+	/* These need to be multiples of MCU size */
+	coda_write(dev, aligned_width << 16 | aligned_height,
+		   CODA9_REG_JPEG_PIC_SIZE);
+	coda_write(dev, ctx->params.rot_mode ?
+		   (CODA_ROT_MIR_ENABLE | ctx->params.rot_mode) : 0,
+		   CODA9_REG_JPEG_ROT_INFO);
+
+	coda_write(dev, mcu_info[chroma_format], CODA9_REG_JPEG_MCU_INFO);
+
+	coda_write(dev, 1, CODA9_GDI_CONTROL);
+	timeout = ktime_add_us(ktime_get(), 100000);
+	do {
+		ret = coda_read(dev, CODA9_GDI_STATUS);
+		if (ktime_compare(ktime_get(), timeout) > 0) {
+			v4l2_err(&dev->v4l2_dev, "timeout waiting for GDI\n");
+			return -ETIMEDOUT;
+		}
+	} while (!ret);
+
+	coda_write(dev, (chroma_format << 17) | (chroma_interleave << 16) |
+		   q_data_src->bytesperline, CODA9_GDI_INFO_CONTROL);
+	/* The content of this register seems to be irrelevant: */
+	coda_write(dev, aligned_width << 16 | aligned_height,
+		   CODA9_GDI_INFO_PIC_SIZE);
+
+	coda_write_base(ctx, q_data_src, src_buf, CODA9_GDI_INFO_BASE_Y);
+
+	coda_write(dev, 0, CODA9_REG_JPEG_DPB_BASE00);
+	coda_write(dev, 0, CODA9_GDI_CONTROL);
+	coda_write(dev, 1, CODA9_GDI_PIC_INIT_HOST);
+
+	coda_write(dev, 1, CODA9_GDI_WPROT_ERR_CLR);
+	coda_write(dev, 0, CODA9_GDI_WPROT_RGN_EN);
+
+	trace_coda_jpeg_run(ctx, src_buf);
+
+	coda_write(dev, 1, CODA9_REG_JPEG_PIC_START);
+
+	return 0;
+}
+
+static void coda9_jpeg_finish_encode(struct coda_ctx *ctx)
+{
+	struct vb2_v4l2_buffer *src_buf, *dst_buf;
+	struct coda_dev *dev = ctx->dev;
+	u32 wr_ptr, start_ptr;
+	u32 err_mb;
+
+	if (ctx->aborting) {
+		coda_write(ctx->dev, 0, CODA9_REG_JPEG_BBC_FLUSH_CMD);
+		return;
+	}
+
+	/*
+	 * Lock to make sure that an encoder stop command running in parallel
+	 * will either already have marked src_buf as last, or it will wake up
+	 * the capture queue after the buffers are returned.
+	 */
+	mutex_lock(&ctx->wakeup_mutex);
+	src_buf = v4l2_m2m_src_buf_remove(ctx->fh.m2m_ctx);
+	dst_buf = v4l2_m2m_dst_buf_remove(ctx->fh.m2m_ctx);
+
+	trace_coda_jpeg_done(ctx, dst_buf);
+
+	/*
+	 * Set plane payload to the number of bytes written out
+	 * by the JPEG processing unit
+	 */
+	start_ptr = vb2_dma_contig_plane_dma_addr(&dst_buf->vb2_buf, 0);
+	wr_ptr = coda_read(dev, CODA9_REG_JPEG_BBC_WR_PTR);
+	vb2_set_plane_payload(&dst_buf->vb2_buf, 0, wr_ptr - start_ptr);
+
+	err_mb = coda_read(dev, CODA9_REG_JPEG_PIC_ERRMB);
+	if (err_mb)
+		coda_dbg(1, ctx, "ERRMB: 0x%x\n", err_mb);
+
+	coda_write(dev, 0, CODA9_REG_JPEG_BBC_FLUSH_CMD);
+
+	dst_buf->flags &= ~(V4L2_BUF_FLAG_PFRAME | V4L2_BUF_FLAG_LAST);
+	dst_buf->flags |= V4L2_BUF_FLAG_KEYFRAME;
+	dst_buf->flags |= src_buf->flags & V4L2_BUF_FLAG_LAST;
+
+	v4l2_m2m_buf_copy_metadata(src_buf, dst_buf, false);
+
+	v4l2_m2m_buf_done(src_buf, VB2_BUF_STATE_DONE);
+	coda_m2m_buf_done(ctx, dst_buf, err_mb ? VB2_BUF_STATE_ERROR :
+						 VB2_BUF_STATE_DONE);
+	mutex_unlock(&ctx->wakeup_mutex);
+
+	coda_dbg(1, ctx, "job finished: encoded frame (%u)%s\n",
+		 dst_buf->sequence,
+		 (dst_buf->flags & V4L2_BUF_FLAG_LAST) ? " (last)" : "");
+
+	/*
+	 * Reset JPEG processing unit after each encode run to work
+	 * around hangups when switching context between encoder and
+	 * decoder.
+	 */
+	coda_hw_reset(ctx);
+}
+
+static void coda9_jpeg_encode_timeout(struct coda_ctx *ctx)
+{
+	struct coda_dev *dev = ctx->dev;
+	u32 end_addr, wr_ptr;
+
+	/* Handle missing BBC overflow interrupt via timeout */
+	end_addr = coda_read(dev, CODA9_REG_JPEG_BBC_END_ADDR);
+	wr_ptr = coda_read(dev, CODA9_REG_JPEG_BBC_WR_PTR);
+	if (wr_ptr >= end_addr - 256) {
+		v4l2_err(&dev->v4l2_dev, "JPEG too large for capture buffer\n");
+		coda9_jpeg_finish_encode(ctx);
+		return;
+	}
+
+	coda_hw_reset(ctx);
+}
+
+static void coda9_jpeg_release(struct coda_ctx *ctx)
+{
+	int i;
+
+	if (ctx->params.jpeg_qmat_tab[0] == luma_q)
+		ctx->params.jpeg_qmat_tab[0] = NULL;
+	if (ctx->params.jpeg_qmat_tab[1] == chroma_q)
+		ctx->params.jpeg_qmat_tab[1] = NULL;
+	for (i = 0; i < 3; i++)
+		kfree(ctx->params.jpeg_qmat_tab[i]);
+	kfree(ctx->params.jpeg_huff_data);
+	kfree(ctx->params.jpeg_huff_tab);
+}
+
+const struct coda_context_ops coda9_jpeg_encode_ops = {
+	.queue_init = coda_encoder_queue_init,
+	.start_streaming = coda9_jpeg_start_encoding,
+	.prepare_run = coda9_jpeg_prepare_encode,
+	.finish_run = coda9_jpeg_finish_encode,
+	.run_timeout = coda9_jpeg_encode_timeout,
+	.release = coda9_jpeg_release,
+};
+
+/*
+ * Decoder context operations
+ */
+
+static int coda9_jpeg_start_decoding(struct coda_ctx *ctx)
+{
+	ctx->params.jpeg_qmat_index[0] = 0;
+	ctx->params.jpeg_qmat_index[1] = 1;
+	ctx->params.jpeg_qmat_index[2] = 1;
+	ctx->params.jpeg_qmat_tab[0] = luma_q;
+	ctx->params.jpeg_qmat_tab[1] = chroma_q;
+	/* nothing more to do here */
+
+	/* TODO: we could already scan the first header to get the chroma
+	 * format.
+	 */
+
+	return 0;
+}
+
+static int coda9_jpeg_prepare_decode(struct coda_ctx *ctx)
+{
+	struct coda_dev *dev = ctx->dev;
+	int aligned_width, aligned_height;
+	int chroma_format;
+	int ret;
+	u32 val, dst_fourcc;
+	struct coda_q_data *q_data_src, *q_data_dst;
+	struct vb2_v4l2_buffer *src_buf, *dst_buf;
+	int chroma_interleave;
+	int scl_hor_mode, scl_ver_mode;
+
+	src_buf = v4l2_m2m_next_src_buf(ctx->fh.m2m_ctx);
+	dst_buf = v4l2_m2m_next_dst_buf(ctx->fh.m2m_ctx);
+	q_data_src = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_OUTPUT);
+	q_data_dst = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_CAPTURE);
+	dst_fourcc = q_data_dst->fourcc;
+
+	scl_hor_mode = coda_jpeg_scale(q_data_src->width, q_data_dst->width);
+	scl_ver_mode = coda_jpeg_scale(q_data_src->height, q_data_dst->height);
+
+	if (vb2_get_plane_payload(&src_buf->vb2_buf, 0) == 0)
+		vb2_set_plane_payload(&src_buf->vb2_buf, 0,
+				      vb2_plane_size(&src_buf->vb2_buf, 0));
+
+	chroma_format = coda9_jpeg_chroma_format(q_data_dst->fourcc);
+	if (chroma_format < 0)
+		return chroma_format;
+
+	ret = coda_jpeg_decode_header(ctx, &src_buf->vb2_buf);
+	if (ret < 0) {
+		src_buf = v4l2_m2m_src_buf_remove(ctx->fh.m2m_ctx);
+		dst_buf = v4l2_m2m_dst_buf_remove(ctx->fh.m2m_ctx);
+		v4l2_m2m_buf_done(src_buf, VB2_BUF_STATE_DONE);
+		v4l2_m2m_buf_done(dst_buf, VB2_BUF_STATE_ERROR);
+
+		return ret;
+	}
+
+	/* Round image dimensions to multiple of MCU size */
+	aligned_width = round_up(q_data_src->width, width_align[chroma_format]);
+	aligned_height = round_up(q_data_src->height, height_align[chroma_format]);
+	if (aligned_width != q_data_dst->bytesperline) {
+		v4l2_err(&dev->v4l2_dev, "stride mismatch: %d != %d\n",
+			 aligned_width, q_data_dst->bytesperline);
+	}
+
+	coda_set_gdi_regs(ctx);
+
+	val = ctx->params.jpeg_huff_ac_index[0] << 12 |
+	      ctx->params.jpeg_huff_ac_index[1] << 11 |
+	      ctx->params.jpeg_huff_ac_index[2] << 10 |
+	      ctx->params.jpeg_huff_dc_index[0] << 9 |
+	      ctx->params.jpeg_huff_dc_index[1] << 8 |
+	      ctx->params.jpeg_huff_dc_index[2] << 7;
+	if (ctx->params.jpeg_huff_tab)
+		val |= CODA9_JPEG_PIC_CTRL_USER_HUFFMAN_EN;
+	coda_write(dev, val, CODA9_REG_JPEG_PIC_CTRL);
+
+	coda_write(dev, aligned_width << 16 | aligned_height,
+			CODA9_REG_JPEG_PIC_SIZE);
+
+	chroma_interleave = (dst_fourcc == V4L2_PIX_FMT_NV12);
+	coda_write(dev, 0, CODA9_REG_JPEG_ROT_INFO);
+	coda_write(dev, bus_req_num[chroma_format], CODA9_REG_JPEG_OP_INFO);
+	coda_write(dev, mcu_info[chroma_format], CODA9_REG_JPEG_MCU_INFO);
+	if (scl_hor_mode || scl_ver_mode)
+		val = CODA9_JPEG_SCL_ENABLE | (scl_hor_mode << 2) | scl_ver_mode;
+	else
+		val = 0;
+	coda_write(dev, val, CODA9_REG_JPEG_SCL_INFO);
+	coda_write(dev, chroma_interleave, CODA9_REG_JPEG_DPB_CONFIG);
+	coda_write(dev, ctx->params.jpeg_restart_interval,
+			CODA9_REG_JPEG_RST_INTVAL);
+
+	if (ctx->params.jpeg_huff_tab)
+		coda9_jpeg_dec_huff_setup(ctx);
+
+	coda9_jpeg_qmat_setup(ctx);
+
+	coda9_jpeg_dec_bbc_gbu_setup(ctx, &src_buf->vb2_buf,
+				     ctx->jpeg_ecs_offset);
+
+	coda_write(dev, 0, CODA9_REG_JPEG_RST_INDEX);
+	coda_write(dev, 0, CODA9_REG_JPEG_RST_COUNT);
+
+	coda_write(dev, 0, CODA9_REG_JPEG_DPCM_DIFF_Y);
+	coda_write(dev, 0, CODA9_REG_JPEG_DPCM_DIFF_CB);
+	coda_write(dev, 0, CODA9_REG_JPEG_DPCM_DIFF_CR);
+
+	coda_write(dev, 0, CODA9_REG_JPEG_ROT_INFO);
+
+	coda_write(dev, 1, CODA9_GDI_CONTROL);
+	do {
+		ret = coda_read(dev, CODA9_GDI_STATUS);
+	} while (!ret);
+
+	val = (chroma_format << 17) | (chroma_interleave << 16) |
+	      q_data_dst->bytesperline;
+	if (ctx->tiled_map_type == GDI_TILED_FRAME_MB_RASTER_MAP)
+		val |= 3 << 20;
+	coda_write(dev, val, CODA9_GDI_INFO_CONTROL);
+
+	coda_write(dev, aligned_width << 16 | aligned_height,
+			CODA9_GDI_INFO_PIC_SIZE);
+
+	coda_write_base(ctx, q_data_dst, dst_buf, CODA9_GDI_INFO_BASE_Y);
+
+	coda_write(dev, 0, CODA9_REG_JPEG_DPB_BASE00);
+	coda_write(dev, 0, CODA9_GDI_CONTROL);
+	coda_write(dev, 1, CODA9_GDI_PIC_INIT_HOST);
+
+	trace_coda_jpeg_run(ctx, src_buf);
+
+	coda_write(dev, 1, CODA9_REG_JPEG_PIC_START);
+
+	return 0;
+}
+
+static void coda9_jpeg_finish_decode(struct coda_ctx *ctx)
+{
+	struct coda_dev *dev = ctx->dev;
+	struct vb2_v4l2_buffer *dst_buf, *src_buf;
+	struct coda_q_data *q_data_dst;
+	u32 err_mb;
+
+	err_mb = coda_read(dev, CODA9_REG_JPEG_PIC_ERRMB);
+	if (err_mb)
+		v4l2_err(&dev->v4l2_dev, "ERRMB: 0x%x\n", err_mb);
+
+	coda_write(dev, 0, CODA9_REG_JPEG_BBC_FLUSH_CMD);
+
+	/*
+	 * Lock to make sure that a decoder stop command running in parallel
+	 * will either already have marked src_buf as last, or it will wake up
+	 * the capture queue after the buffers are returned.
+	 */
+	mutex_lock(&ctx->wakeup_mutex);
+	src_buf = v4l2_m2m_src_buf_remove(ctx->fh.m2m_ctx);
+	dst_buf = v4l2_m2m_dst_buf_remove(ctx->fh.m2m_ctx);
+	dst_buf->sequence = ctx->osequence++;
+
+	trace_coda_jpeg_done(ctx, dst_buf);
+
+	dst_buf->flags &= ~(V4L2_BUF_FLAG_PFRAME | V4L2_BUF_FLAG_LAST);
+	dst_buf->flags |= V4L2_BUF_FLAG_KEYFRAME;
+	dst_buf->flags |= src_buf->flags & V4L2_BUF_FLAG_LAST;
+
+	v4l2_m2m_buf_copy_metadata(src_buf, dst_buf, false);
+
+	q_data_dst = get_q_data(ctx, V4L2_BUF_TYPE_VIDEO_CAPTURE);
+	vb2_set_plane_payload(&dst_buf->vb2_buf, 0, q_data_dst->sizeimage);
+
+	v4l2_m2m_buf_done(src_buf, VB2_BUF_STATE_DONE);
+	coda_m2m_buf_done(ctx, dst_buf, err_mb ? VB2_BUF_STATE_ERROR :
+						 VB2_BUF_STATE_DONE);
+
+	mutex_unlock(&ctx->wakeup_mutex);
+
+	coda_dbg(1, ctx, "job finished: decoded frame (%u)%s\n",
+		 dst_buf->sequence,
+		 (dst_buf->flags & V4L2_BUF_FLAG_LAST) ? " (last)" : "");
+
+	/*
+	 * Reset JPEG processing unit after each decode run to work
+	 * around hangups when switching context between encoder and
+	 * decoder.
+	 */
+	coda_hw_reset(ctx);
+}
+
+const struct coda_context_ops coda9_jpeg_decode_ops = {
+	.queue_init = coda_encoder_queue_init, /* non-bitstream operation */
+	.start_streaming = coda9_jpeg_start_decoding,
+	.prepare_run = coda9_jpeg_prepare_decode,
+	.finish_run = coda9_jpeg_finish_decode,
+	.release = coda9_jpeg_release,
+};
+
+irqreturn_t coda9_jpeg_irq_handler(int irq, void *data)
+{
+	struct coda_dev *dev = data;
+	struct coda_ctx *ctx;
+	int status;
+	int err_mb;
+
+	status = coda_read(dev, CODA9_REG_JPEG_PIC_STATUS);
+	if (status == 0)
+		return IRQ_HANDLED;
+	coda_write(dev, status, CODA9_REG_JPEG_PIC_STATUS);
+
+	if (status & CODA9_JPEG_STATUS_OVERFLOW)
+		v4l2_err(&dev->v4l2_dev, "JPEG overflow\n");
+
+	if (status & CODA9_JPEG_STATUS_BBC_INT)
+		v4l2_err(&dev->v4l2_dev, "JPEG BBC interrupt\n");
+
+	if (status & CODA9_JPEG_STATUS_ERROR) {
+		v4l2_err(&dev->v4l2_dev, "JPEG error\n");
+
+		err_mb = coda_read(dev, CODA9_REG_JPEG_PIC_ERRMB);
+		if (err_mb) {
+			v4l2_err(&dev->v4l2_dev,
+				 "ERRMB: 0x%x: rst idx %d, mcu pos (%d,%d)\n",
+				 err_mb, err_mb >> 24, (err_mb >> 12) & 0xfff,
+				 err_mb & 0xfff);
+		}
+	}
+
+	ctx = v4l2_m2m_get_curr_priv(dev->m2m_dev);
+	if (!ctx) {
+		v4l2_err(&dev->v4l2_dev,
+			 "Instance released before the end of transaction\n");
+		mutex_unlock(&dev->coda_mutex);
+		return IRQ_HANDLED;
+	}
+
+	complete(&ctx->completion);
+
+	return IRQ_HANDLED;
+}
diff --git a/drivers/media/platform/chips-media/coda-mpeg2.c b/drivers/media/platform/chips-media/coda-mpeg2.c
new file mode 100644
index 000000000..6f3f6721d
--- /dev/null
+++ b/drivers/media/platform/chips-media/coda-mpeg2.c
@@ -0,0 +1,87 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Coda multi-standard codec IP - MPEG-2 helper functions
+ *
+ * Copyright (C) 2019 Pengutronix, Philipp Zabel
+ */
+
+#include <linux/kernel.h>
+#include <linux/videodev2.h>
+#include "coda.h"
+
+int coda_mpeg2_profile(int profile_idc)
+{
+	switch (profile_idc) {
+	case 5:
+		return V4L2_MPEG_VIDEO_MPEG2_PROFILE_SIMPLE;
+	case 4:
+		return V4L2_MPEG_VIDEO_MPEG2_PROFILE_MAIN;
+	case 3:
+		return V4L2_MPEG_VIDEO_MPEG2_PROFILE_SNR_SCALABLE;
+	case 2:
+		return V4L2_MPEG_VIDEO_MPEG2_PROFILE_SPATIALLY_SCALABLE;
+	case 1:
+		return V4L2_MPEG_VIDEO_MPEG2_PROFILE_HIGH;
+	default:
+		return -EINVAL;
+	}
+}
+
+int coda_mpeg2_level(int level_idc)
+{
+	switch (level_idc) {
+	case 10:
+		return V4L2_MPEG_VIDEO_MPEG2_LEVEL_LOW;
+	case 8:
+		return V4L2_MPEG_VIDEO_MPEG2_LEVEL_MAIN;
+	case 6:
+		return V4L2_MPEG_VIDEO_MPEG2_LEVEL_HIGH_1440;
+	case 4:
+		return V4L2_MPEG_VIDEO_MPEG2_LEVEL_HIGH;
+	default:
+		return -EINVAL;
+	}
+}
+
+/*
+ * Check if the buffer starts with the MPEG-2 sequence header (with or without
+ * quantization matrix) and extension header, for example:
+ *
+ *   00 00 01 b3 2d 01 e0 34 08 8b a3 81
+ *               10 11 11 12 12 12 13 13 13 13 14 14 14 14 14 15
+ *               15 15 15 15 15 16 16 16 16 16 16 16 17 17 17 17
+ *               17 17 17 17 18 18 18 19 18 18 18 19 1a 1a 1a 1a
+ *               19 1b 1b 1b 1b 1b 1c 1c 1c 1c 1e 1e 1e 1f 1f 21
+ *   00 00 01 b5 14 8a 00 01 00 00
+ *
+ * or:
+ *
+ *   00 00 01 b3 08 00 40 15 ff ff e0 28
+ *   00 00 01 b5 14 8a 00 01 00 00
+ *
+ * Returns the detected header size in bytes or 0.
+ */
+u32 coda_mpeg2_parse_headers(struct coda_ctx *ctx, u8 *buf, u32 size)
+{
+	static const u8 sequence_header_start[4] = { 0x00, 0x00, 0x01, 0xb3 };
+	static const union {
+		u8 extension_start[4];
+		u8 start_code_prefix[3];
+	} u = { { 0x00, 0x00, 0x01, 0xb5 } };
+
+	if (size < 22 ||
+	    memcmp(buf, sequence_header_start, 4) != 0)
+		return 0;
+
+	if ((size == 22 ||
+	     (size >= 25 && memcmp(buf + 22, u.start_code_prefix, 3) == 0)) &&
+	    memcmp(buf + 12, u.extension_start, 4) == 0)
+		return 22;
+
+	if ((size == 86 ||
+	     (size > 89 && memcmp(buf + 86, u.start_code_prefix, 3) == 0)) &&
+	    memcmp(buf + 76, u.extension_start, 4) == 0)
+		return 86;
+
+	return 0;
+}
diff --git a/drivers/media/platform/chips-media/coda-mpeg4.c b/drivers/media/platform/chips-media/coda-mpeg4.c
new file mode 100644
index 000000000..483a4fba1
--- /dev/null
+++ b/drivers/media/platform/chips-media/coda-mpeg4.c
@@ -0,0 +1,87 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Coda multi-standard codec IP - MPEG-4 helper functions
+ *
+ * Copyright (C) 2019 Pengutronix, Philipp Zabel
+ */
+
+#include <linux/kernel.h>
+#include <linux/videodev2.h>
+
+#include "coda.h"
+
+int coda_mpeg4_profile(int profile_idc)
+{
+	switch (profile_idc) {
+	case 0:
+		return V4L2_MPEG_VIDEO_MPEG4_PROFILE_SIMPLE;
+	case 15:
+		return V4L2_MPEG_VIDEO_MPEG4_PROFILE_ADVANCED_SIMPLE;
+	case 2:
+		return V4L2_MPEG_VIDEO_MPEG4_PROFILE_CORE;
+	case 1:
+		return V4L2_MPEG_VIDEO_MPEG4_PROFILE_SIMPLE_SCALABLE;
+	case 11:
+		return V4L2_MPEG_VIDEO_MPEG4_PROFILE_ADVANCED_CODING_EFFICIENCY;
+	default:
+		return -EINVAL;
+	}
+}
+
+int coda_mpeg4_level(int level_idc)
+{
+	switch (level_idc) {
+	case 0:
+		return V4L2_MPEG_VIDEO_MPEG4_LEVEL_0;
+	case 1:
+		return V4L2_MPEG_VIDEO_MPEG4_LEVEL_1;
+	case 2:
+		return V4L2_MPEG_VIDEO_MPEG4_LEVEL_2;
+	case 3:
+		return V4L2_MPEG_VIDEO_MPEG4_LEVEL_3;
+	case 4:
+		return V4L2_MPEG_VIDEO_MPEG4_LEVEL_4;
+	case 5:
+		return V4L2_MPEG_VIDEO_MPEG4_LEVEL_5;
+	default:
+		return -EINVAL;
+	}
+}
+
+/*
+ * Check if the buffer starts with the MPEG-4 visual object sequence and visual
+ * object headers, for example:
+ *
+ *   00 00 01 b0 f1
+ *   00 00 01 b5 a9 13 00 00 01 00 00 00 01 20 08
+ *               d4 8d 88 00 f5 04 04 08 14 30 3f
+ *
+ * Returns the detected header size in bytes or 0.
+ */
+u32 coda_mpeg4_parse_headers(struct coda_ctx *ctx, u8 *buf, u32 size)
+{
+	static const u8 vos_start[4] = { 0x00, 0x00, 0x01, 0xb0 };
+	static const union {
+		u8 vo_start[4];
+		u8 start_code_prefix[3];
+	} u = { { 0x00, 0x00, 0x01, 0xb5 } };
+
+	if (size < 30 ||
+	    memcmp(buf, vos_start, 4) != 0 ||
+	    memcmp(buf + 5, u.vo_start, 4) != 0)
+		return 0;
+
+	if (size == 30 ||
+	    (size >= 33 && memcmp(buf + 30, u.start_code_prefix, 3) == 0))
+		return 30;
+
+	if (size == 31 ||
+	    (size >= 34 && memcmp(buf + 31, u.start_code_prefix, 3) == 0))
+		return 31;
+
+	if (size == 32 ||
+	    (size >= 35 && memcmp(buf + 32, u.start_code_prefix, 3) == 0))
+		return 32;
+
+	return 0;
+}
diff --git a/drivers/media/platform/chips-media/coda.h b/drivers/media/platform/chips-media/coda.h
new file mode 100644
index 000000000..ddfd0a32c
--- /dev/null
+++ b/drivers/media/platform/chips-media/coda.h
@@ -0,0 +1,403 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * Coda multi-standard codec IP
+ *
+ * Copyright (C) 2012 Vista Silicon S.L.
+ *    Javier Martin, <javier.martin@vista-silicon.com>
+ *    Xavier Duret
+ * Copyright (C) 2012-2014 Philipp Zabel, Pengutronix
+ */
+
+#ifndef __CODA_H__
+#define __CODA_H__
+
+#include <linux/debugfs.h>
+#include <linux/idr.h>
+#include <linux/irqreturn.h>
+#include <linux/mutex.h>
+#include <linux/kfifo.h>
+#include <linux/videodev2.h>
+#include <linux/ratelimit.h>
+
+#include <media/v4l2-ctrls.h>
+#include <media/v4l2-device.h>
+#include <media/v4l2-fh.h>
+#include <media/videobuf2-v4l2.h>
+
+#include "coda_regs.h"
+
+#define CODA_MAX_FRAMEBUFFERS	19
+#define FMO_SLICE_SAVE_BUF_SIZE	(32)
+
+/*
+ * This control allows applications to read the per-stream
+ * (i.e. per-context) Macroblocks Error Count. This value
+ * is CODA specific.
+ */
+#define V4L2_CID_CODA_MB_ERR_CNT (V4L2_CID_USER_CODA_BASE + 0)
+
+enum {
+	V4L2_M2M_SRC = 0,
+	V4L2_M2M_DST = 1,
+};
+
+enum coda_inst_type {
+	CODA_INST_ENCODER,
+	CODA_INST_DECODER,
+};
+
+enum coda_product {
+	CODA_DX6 = 0xf001,
+	CODA_HX4 = 0xf00a,
+	CODA_7541 = 0xf012,
+	CODA_960 = 0xf020,
+};
+
+struct coda_video_device;
+
+struct coda_devtype {
+	char			*firmware[3];
+	enum coda_product	product;
+	const struct coda_codec	*codecs;
+	unsigned int		num_codecs;
+	const struct coda_video_device **vdevs;
+	unsigned int		num_vdevs;
+	size_t			workbuf_size;
+	size_t			tempbuf_size;
+	size_t			iram_size;
+};
+
+struct coda_aux_buf {
+	void			*vaddr;
+	dma_addr_t		paddr;
+	u32			size;
+	struct debugfs_blob_wrapper blob;
+	struct dentry		*dentry;
+};
+
+struct coda_dev {
+	struct v4l2_device	v4l2_dev;
+	struct video_device	vfd[6];
+	struct device		*dev;
+	const struct coda_devtype *devtype;
+	int			firmware;
+	struct vdoa_data	*vdoa;
+
+	void __iomem		*regs_base;
+	struct clk		*clk_per;
+	struct clk		*clk_ahb;
+	struct reset_control	*rstc;
+
+	struct coda_aux_buf	codebuf;
+	struct coda_aux_buf	tempbuf;
+	struct coda_aux_buf	workbuf;
+	struct gen_pool		*iram_pool;
+	struct coda_aux_buf	iram;
+
+	struct mutex		dev_mutex;
+	struct mutex		coda_mutex;
+	struct workqueue_struct	*workqueue;
+	struct v4l2_m2m_dev	*m2m_dev;
+	struct ida		ida;
+	struct dentry		*debugfs_root;
+	struct ratelimit_state	mb_err_rs;
+};
+
+struct coda_codec {
+	u32 mode;
+	u32 src_fourcc;
+	u32 dst_fourcc;
+	u32 max_w;
+	u32 max_h;
+};
+
+struct coda_huff_tab;
+
+struct coda_params {
+	u8			rot_mode;
+	u8			h264_intra_qp;
+	u8			h264_inter_qp;
+	u8			h264_min_qp;
+	u8			h264_max_qp;
+	u8			h264_disable_deblocking_filter_idc;
+	s8			h264_slice_alpha_c0_offset_div2;
+	s8			h264_slice_beta_offset_div2;
+	bool			h264_constrained_intra_pred_flag;
+	s8			h264_chroma_qp_index_offset;
+	u8			h264_profile_idc;
+	u8			h264_level_idc;
+	u8			mpeg2_profile_idc;
+	u8			mpeg2_level_idc;
+	u8			mpeg4_intra_qp;
+	u8			mpeg4_inter_qp;
+	u8			gop_size;
+	int			intra_refresh;
+	enum v4l2_jpeg_chroma_subsampling jpeg_chroma_subsampling;
+	u8			jpeg_quality;
+	u8			jpeg_restart_interval;
+	u8			*jpeg_qmat_tab[3];
+	int			jpeg_qmat_index[3];
+	int			jpeg_huff_dc_index[3];
+	int			jpeg_huff_ac_index[3];
+	u32			*jpeg_huff_data;
+	struct coda_huff_tab	*jpeg_huff_tab;
+	int			codec_mode;
+	int			codec_mode_aux;
+	enum v4l2_mpeg_video_multi_slice_mode slice_mode;
+	u32			framerate;
+	u16			bitrate;
+	u16			vbv_delay;
+	u32			vbv_size;
+	u32			slice_max_bits;
+	u32			slice_max_mb;
+	bool			force_ipicture;
+	bool			gop_size_changed;
+	bool			bitrate_changed;
+	bool			framerate_changed;
+	bool			h264_intra_qp_changed;
+	bool			intra_refresh_changed;
+	bool			slice_mode_changed;
+	bool			frame_rc_enable;
+	bool			mb_rc_enable;
+};
+
+struct coda_buffer_meta {
+	struct list_head	list;
+	u32			sequence;
+	struct v4l2_timecode	timecode;
+	u64			timestamp;
+	unsigned int		start;
+	unsigned int		end;
+	bool			last;
+};
+
+/* Per-queue, driver-specific private data */
+struct coda_q_data {
+	unsigned int		width;
+	unsigned int		height;
+	unsigned int		bytesperline;
+	unsigned int		sizeimage;
+	unsigned int		fourcc;
+	struct v4l2_rect	rect;
+};
+
+struct coda_iram_info {
+	u32		axi_sram_use;
+	phys_addr_t	buf_bit_use;
+	phys_addr_t	buf_ip_ac_dc_use;
+	phys_addr_t	buf_dbk_y_use;
+	phys_addr_t	buf_dbk_c_use;
+	phys_addr_t	buf_ovl_use;
+	phys_addr_t	buf_btp_use;
+	phys_addr_t	search_ram_paddr;
+	int		search_ram_size;
+	int		remaining;
+	phys_addr_t	next_paddr;
+};
+
+#define GDI_LINEAR_FRAME_MAP 0
+#define GDI_TILED_FRAME_MB_RASTER_MAP 1
+
+struct coda_ctx;
+
+struct coda_context_ops {
+	int (*queue_init)(void *priv, struct vb2_queue *src_vq,
+			  struct vb2_queue *dst_vq);
+	int (*reqbufs)(struct coda_ctx *ctx, struct v4l2_requestbuffers *rb);
+	int (*start_streaming)(struct coda_ctx *ctx);
+	int (*prepare_run)(struct coda_ctx *ctx);
+	void (*finish_run)(struct coda_ctx *ctx);
+	void (*run_timeout)(struct coda_ctx *ctx);
+	void (*seq_init_work)(struct work_struct *work);
+	void (*seq_end_work)(struct work_struct *work);
+	void (*release)(struct coda_ctx *ctx);
+};
+
+struct coda_internal_frame {
+	struct coda_aux_buf		buf;
+	struct coda_buffer_meta		meta;
+	u32				type;
+	u32				error;
+};
+
+struct coda_ctx {
+	struct coda_dev			*dev;
+	struct mutex			buffer_mutex;
+	struct work_struct		pic_run_work;
+	struct work_struct		seq_init_work;
+	struct work_struct		seq_end_work;
+	struct completion		completion;
+	const struct coda_video_device	*cvd;
+	const struct coda_context_ops	*ops;
+	int				aborting;
+	int				initialized;
+	int				streamon_out;
+	int				streamon_cap;
+	u32				qsequence;
+	u32				osequence;
+	u32				sequence_offset;
+	struct coda_q_data		q_data[2];
+	enum coda_inst_type		inst_type;
+	const struct coda_codec		*codec;
+	enum v4l2_colorspace		colorspace;
+	enum v4l2_xfer_func		xfer_func;
+	enum v4l2_ycbcr_encoding	ycbcr_enc;
+	enum v4l2_quantization		quantization;
+	struct coda_params		params;
+	struct v4l2_ctrl_handler	ctrls;
+	struct v4l2_ctrl		*h264_profile_ctrl;
+	struct v4l2_ctrl		*h264_level_ctrl;
+	struct v4l2_ctrl		*mpeg2_profile_ctrl;
+	struct v4l2_ctrl		*mpeg2_level_ctrl;
+	struct v4l2_ctrl		*mpeg4_profile_ctrl;
+	struct v4l2_ctrl		*mpeg4_level_ctrl;
+	struct v4l2_ctrl		*mb_err_cnt_ctrl;
+	struct v4l2_fh			fh;
+	int				gopcounter;
+	int				runcounter;
+	int				jpeg_ecs_offset;
+	char				vpu_header[3][64];
+	int				vpu_header_size[3];
+	struct kfifo			bitstream_fifo;
+	struct mutex			bitstream_mutex;
+	struct coda_aux_buf		bitstream;
+	bool				hold;
+	struct coda_aux_buf		parabuf;
+	struct coda_aux_buf		psbuf;
+	struct coda_aux_buf		slicebuf;
+	struct coda_internal_frame	internal_frames[CODA_MAX_FRAMEBUFFERS];
+	struct list_head		buffer_meta_list;
+	spinlock_t			buffer_meta_lock;
+	int				num_metas;
+	unsigned int			first_frame_sequence;
+	struct coda_aux_buf		workbuf;
+	int				num_internal_frames;
+	int				idx;
+	int				reg_idx;
+	struct coda_iram_info		iram_info;
+	int				tiled_map_type;
+	u32				bit_stream_param;
+	u32				frm_dis_flg;
+	u32				frame_mem_ctrl;
+	u32				para_change;
+	int				display_idx;
+	struct dentry			*debugfs_entry;
+	bool				use_bit;
+	bool				use_vdoa;
+	struct vdoa_ctx			*vdoa;
+	/*
+	 * wakeup mutex used to serialize encoder stop command and finish_run,
+	 * ensures that finish_run always either flags the last returned buffer
+	 * or wakes up the capture queue to signal EOS afterwards.
+	 */
+	struct mutex			wakeup_mutex;
+};
+
+extern int coda_debug;
+
+#define coda_dbg(level, ctx, fmt, arg...)				\
+	do {								\
+		if (coda_debug >= (level))				\
+			v4l2_dbg((level), coda_debug, &(ctx)->dev->v4l2_dev, \
+			 "%u: " fmt, (ctx)->idx, ##arg);		\
+	} while (0)
+
+void coda_write(struct coda_dev *dev, u32 data, u32 reg);
+unsigned int coda_read(struct coda_dev *dev, u32 reg);
+void coda_write_base(struct coda_ctx *ctx, struct coda_q_data *q_data,
+		     struct vb2_v4l2_buffer *buf, unsigned int reg_y);
+
+int coda_alloc_aux_buf(struct coda_dev *dev, struct coda_aux_buf *buf,
+		       size_t size, const char *name, struct dentry *parent);
+void coda_free_aux_buf(struct coda_dev *dev, struct coda_aux_buf *buf);
+
+int coda_encoder_queue_init(void *priv, struct vb2_queue *src_vq,
+			    struct vb2_queue *dst_vq);
+int coda_decoder_queue_init(void *priv, struct vb2_queue *src_vq,
+			    struct vb2_queue *dst_vq);
+
+int coda_hw_reset(struct coda_ctx *ctx);
+
+void coda_fill_bitstream(struct coda_ctx *ctx, struct list_head *buffer_list);
+
+void coda_set_gdi_regs(struct coda_ctx *ctx);
+
+static inline struct coda_q_data *get_q_data(struct coda_ctx *ctx,
+					     enum v4l2_buf_type type)
+{
+	switch (type) {
+	case V4L2_BUF_TYPE_VIDEO_OUTPUT:
+		return &(ctx->q_data[V4L2_M2M_SRC]);
+	case V4L2_BUF_TYPE_VIDEO_CAPTURE:
+		return &(ctx->q_data[V4L2_M2M_DST]);
+	default:
+		return NULL;
+	}
+}
+
+const char *coda_product_name(int product);
+
+int coda_check_firmware(struct coda_dev *dev);
+
+static inline unsigned int coda_get_bitstream_payload(struct coda_ctx *ctx)
+{
+	return kfifo_len(&ctx->bitstream_fifo);
+}
+
+/*
+ * The bitstream prefetcher needs to read at least 2 256 byte periods past
+ * the desired bitstream position for all data to reach the decoder.
+ */
+static inline bool coda_bitstream_can_fetch_past(struct coda_ctx *ctx,
+						 unsigned int pos)
+{
+	return (int)(ctx->bitstream_fifo.kfifo.in - ALIGN(pos, 256)) > 512;
+}
+
+bool coda_bitstream_can_fetch_past(struct coda_ctx *ctx, unsigned int pos);
+int coda_bitstream_flush(struct coda_ctx *ctx);
+
+void coda_bit_stream_end_flag(struct coda_ctx *ctx);
+
+void coda_m2m_buf_done(struct coda_ctx *ctx, struct vb2_v4l2_buffer *buf,
+		       enum vb2_buffer_state state);
+
+int coda_h264_filler_nal(int size, char *p);
+int coda_h264_padding(int size, char *p);
+int coda_h264_profile(int profile_idc);
+int coda_h264_level(int level_idc);
+int coda_sps_parse_profile(struct coda_ctx *ctx, struct vb2_buffer *vb);
+int coda_h264_sps_fixup(struct coda_ctx *ctx, int width, int height, char *buf,
+			int *size, int max_size);
+
+int coda_mpeg2_profile(int profile_idc);
+int coda_mpeg2_level(int level_idc);
+u32 coda_mpeg2_parse_headers(struct coda_ctx *ctx, u8 *buf, u32 size);
+int coda_mpeg4_profile(int profile_idc);
+int coda_mpeg4_level(int level_idc);
+u32 coda_mpeg4_parse_headers(struct coda_ctx *ctx, u8 *buf, u32 size);
+
+void coda_update_profile_level_ctrls(struct coda_ctx *ctx, u8 profile_idc,
+				     u8 level_idc);
+
+static inline int coda_jpeg_scale(int src, int dst)
+{
+	return (dst <= src / 8) ? 3 :
+	       (dst <= src / 4) ? 2 :
+	       (dst <= src / 2) ? 1 : 0;
+}
+
+bool coda_jpeg_check_buffer(struct coda_ctx *ctx, struct vb2_buffer *vb);
+int coda_jpeg_decode_header(struct coda_ctx *ctx, struct vb2_buffer *vb);
+int coda_jpeg_write_tables(struct coda_ctx *ctx);
+void coda_set_jpeg_compression_quality(struct coda_ctx *ctx, int quality);
+
+extern const struct coda_context_ops coda_bit_encode_ops;
+extern const struct coda_context_ops coda_bit_decode_ops;
+extern const struct coda_context_ops coda9_jpeg_encode_ops;
+extern const struct coda_context_ops coda9_jpeg_decode_ops;
+
+irqreturn_t coda_irq_handler(int irq, void *data);
+irqreturn_t coda9_jpeg_irq_handler(int irq, void *data);
+
+#endif /* __CODA_H__ */
diff --git a/drivers/media/platform/chips-media/coda_regs.h b/drivers/media/platform/chips-media/coda_regs.h
new file mode 100644
index 000000000..db81a904c
--- /dev/null
+++ b/drivers/media/platform/chips-media/coda_regs.h
@@ -0,0 +1,563 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * linux/drivers/media/platform/chips-media/coda_regs.h
+ *
+ * Copyright (C) 2012 Vista Silicon SL
+ *    Javier Martin <javier.martin@vista-silicon.com>
+ *    Xavier Duret
+ */
+
+#ifndef _REGS_CODA_H_
+#define _REGS_CODA_H_
+
+/* HW registers */
+#define CODA_REG_BIT_CODE_RUN			0x000
+#define		CODA_REG_RUN_ENABLE		(1 << 0)
+#define CODA_REG_BIT_CODE_DOWN			0x004
+#define		CODA_DOWN_ADDRESS_SET(x)	(((x) & 0xffff) << 16)
+#define		CODA_DOWN_DATA_SET(x)		((x) & 0xffff)
+#define CODA_REG_BIT_HOST_IN_REQ		0x008
+#define CODA_REG_BIT_INT_CLEAR			0x00c
+#define		CODA_REG_BIT_INT_CLEAR_SET	0x1
+#define CODA_REG_BIT_INT_STATUS		0x010
+#define CODA_REG_BIT_CODE_RESET		0x014
+#define		CODA_REG_RESET_ENABLE		(1 << 0)
+#define CODA_REG_BIT_CUR_PC			0x018
+#define CODA9_REG_BIT_SW_RESET			0x024
+#define		CODA9_SW_RESET_BPU_CORE   0x008
+#define		CODA9_SW_RESET_BPU_BUS    0x010
+#define		CODA9_SW_RESET_VCE_CORE   0x020
+#define		CODA9_SW_RESET_VCE_BUS    0x040
+#define		CODA9_SW_RESET_GDI_CORE   0x080
+#define		CODA9_SW_RESET_GDI_BUS    0x100
+#define CODA9_REG_BIT_SW_RESET_STATUS		0x034
+
+/* Static SW registers */
+#define CODA_REG_BIT_CODE_BUF_ADDR		0x100
+#define CODA_REG_BIT_WORK_BUF_ADDR		0x104
+#define CODA_REG_BIT_PARA_BUF_ADDR		0x108
+#define CODA_REG_BIT_STREAM_CTRL		0x10c
+#define		CODA7_STREAM_BUF_PIC_RESET	(1 << 4)
+#define		CODADX6_STREAM_BUF_PIC_RESET	(1 << 3)
+#define		CODA7_STREAM_BUF_PIC_FLUSH	(1 << 3)
+#define		CODADX6_STREAM_BUF_PIC_FLUSH	(1 << 2)
+#define		CODA7_STREAM_BUF_DYNALLOC_EN	(1 << 5)
+#define		CODADX6_STREAM_BUF_DYNALLOC_EN	(1 << 4)
+#define		CODADX6_STREAM_CHKDIS_OFFSET	(1 << 1)
+#define		CODA7_STREAM_SEL_64BITS_ENDIAN	(1 << 1)
+#define		CODA_STREAM_ENDIAN_SELECT	(1 << 0)
+#define CODA_REG_BIT_FRAME_MEM_CTRL		0x110
+#define		CODA9_FRAME_ENABLE_BWB		(1 << 12)
+#define		CODA9_FRAME_TILED2LINEAR	(1 << 11)
+#define		CODA_FRAME_CHROMA_INTERLEAVE	(1 << 2)
+#define		CODA_IMAGE_ENDIAN_SELECT	(1 << 0)
+#define CODA_REG_BIT_BIT_STREAM_PARAM		0x114
+#define		CODA_BIT_STREAM_END_FLAG	(1 << 2)
+#define		CODA_BIT_DEC_SEQ_INIT_ESCAPE	(1 << 0)
+#define CODA_REG_BIT_TEMP_BUF_ADDR		0x118
+#define CODA_REG_BIT_RD_PTR(x)			(0x120 + 8 * (x))
+#define CODA_REG_BIT_WR_PTR(x)			(0x124 + 8 * (x))
+#define CODA_REG_BIT_FRM_DIS_FLG(x)		(0x150 + 4 * (x))
+#define CODADX6_REG_BIT_SEARCH_RAM_BASE_ADDR	0x140
+#define CODA7_REG_BIT_AXI_SRAM_USE		0x140
+#define		CODA9_USE_HOST_BTP_ENABLE	(1 << 13)
+#define		CODA9_USE_HOST_OVL_ENABLE	(1 << 12)
+#define		CODA7_USE_HOST_ME_ENABLE	(1 << 11)
+#define		CODA9_USE_HOST_DBK_ENABLE	(3 << 10)
+#define		CODA7_USE_HOST_OVL_ENABLE	(1 << 10)
+#define		CODA7_USE_HOST_DBK_ENABLE	(1 << 9)
+#define		CODA9_USE_HOST_IP_ENABLE	(1 << 9)
+#define		CODA7_USE_HOST_IP_ENABLE	(1 << 8)
+#define		CODA9_USE_HOST_BIT_ENABLE	(1 << 8)
+#define		CODA7_USE_HOST_BIT_ENABLE	(1 << 7)
+#define		CODA9_USE_BTP_ENABLE		(1 << 5)
+#define		CODA7_USE_ME_ENABLE		(1 << 4)
+#define		CODA9_USE_OVL_ENABLE		(1 << 4)
+#define		CODA7_USE_OVL_ENABLE		(1 << 3)
+#define		CODA9_USE_DBK_ENABLE		(3 << 2)
+#define		CODA7_USE_DBK_ENABLE		(1 << 2)
+#define		CODA7_USE_IP_ENABLE		(1 << 1)
+#define		CODA7_USE_BIT_ENABLE		(1 << 0)
+
+#define CODA_REG_BIT_BUSY			0x160
+#define		CODA_REG_BIT_BUSY_FLAG		1
+#define CODA_REG_BIT_RUN_COMMAND		0x164
+#define		CODA_COMMAND_SEQ_INIT		1
+#define		CODA_COMMAND_SEQ_END		2
+#define		CODA_COMMAND_PIC_RUN		3
+#define		CODA_COMMAND_SET_FRAME_BUF	4
+#define		CODA_COMMAND_ENCODE_HEADER	5
+#define		CODA_COMMAND_ENC_PARA_SET	6
+#define		CODA_COMMAND_DEC_PARA_SET	7
+#define		CODA_COMMAND_DEC_BUF_FLUSH	8
+#define		CODA_COMMAND_RC_CHANGE_PARAMETER 9
+#define		CODA_COMMAND_FIRMWARE_GET	0xf
+#define CODA_REG_BIT_RUN_INDEX			0x168
+#define		CODA_INDEX_SET(x)		((x) & 0x3)
+#define CODA_REG_BIT_RUN_COD_STD		0x16c
+#define		CODADX6_MODE_DECODE_MP4		0
+#define		CODADX6_MODE_ENCODE_MP4		1
+#define		CODADX6_MODE_DECODE_H264	2
+#define		CODADX6_MODE_ENCODE_H264	3
+#define		CODA7_MODE_DECODE_H264		0
+#define		CODA7_MODE_DECODE_VC1		1
+#define		CODA7_MODE_DECODE_MP2		2
+#define		CODA7_MODE_DECODE_MP4		3
+#define		CODA7_MODE_DECODE_DV3		3
+#define		CODA7_MODE_DECODE_RV		4
+#define		CODA7_MODE_DECODE_MJPG		5
+#define		CODA7_MODE_ENCODE_H264		8
+#define		CODA7_MODE_ENCODE_MP4		11
+#define		CODA7_MODE_ENCODE_MJPG		13
+#define		CODA9_MODE_DECODE_H264		0
+#define		CODA9_MODE_DECODE_VC1		1
+#define		CODA9_MODE_DECODE_MP2		2
+#define		CODA9_MODE_DECODE_MP4		3
+#define		CODA9_MODE_DECODE_DV3		3
+#define		CODA9_MODE_DECODE_RV		4
+#define		CODA9_MODE_DECODE_AVS		5
+#define		CODA9_MODE_DECODE_MJPG		6
+#define		CODA9_MODE_DECODE_VPX		7
+#define		CODA9_MODE_ENCODE_H264		8
+#define		CODA9_MODE_ENCODE_MP4		11
+#define		CODA9_MODE_ENCODE_MJPG		13
+#define		CODA_MODE_INVALID		0xffff
+#define CODA_REG_BIT_INT_ENABLE		0x170
+#define		CODA_INT_INTERRUPT_ENABLE	(1 << 3)
+#define CODA_REG_BIT_INT_REASON			0x174
+#define CODA7_REG_BIT_RUN_AUX_STD		0x178
+#define		CODA_MP4_AUX_MPEG4		0
+#define		CODA_MP4_AUX_DIVX3		1
+#define		CODA_VPX_AUX_THO		0
+#define		CODA_VPX_AUX_VP6		1
+#define		CODA_VPX_AUX_VP8		2
+#define		CODA_H264_AUX_AVC		0
+#define		CODA_H264_AUX_MVC		1
+
+/*
+ * Commands' mailbox:
+ * registers with offsets in the range 0x180-0x1d0
+ * have different meaning depending on the command being
+ * issued.
+ */
+
+/* Decoder Sequence Initialization */
+#define CODA_CMD_DEC_SEQ_BB_START		0x180
+#define CODA_CMD_DEC_SEQ_BB_SIZE		0x184
+#define CODA_CMD_DEC_SEQ_OPTION			0x188
+#define		CODA_NO_INT_ENABLE			(1 << 10)
+#define		CODA_REORDER_ENABLE			(1 << 1)
+#define		CODADX6_QP_REPORT			(1 << 0)
+#define		CODA7_MP4_DEBLK_ENABLE			(1 << 0)
+#define CODA_CMD_DEC_SEQ_SRC_SIZE		0x18c
+#define CODA_CMD_DEC_SEQ_START_BYTE		0x190
+#define CODA_CMD_DEC_SEQ_PS_BB_START		0x194
+#define CODA_CMD_DEC_SEQ_PS_BB_SIZE		0x198
+#define CODA_CMD_DEC_SEQ_JPG_THUMB_EN		0x19c
+#define CODA_CMD_DEC_SEQ_MP4_ASP_CLASS		0x19c
+#define		CODA_MP4_CLASS_MPEG4			0
+#define CODA_CMD_DEC_SEQ_X264_MV_EN		0x19c
+#define CODA_CMD_DEC_SEQ_SPP_CHUNK_SIZE		0x1a0
+
+#define CODA7_RET_DEC_SEQ_ASPECT		0x1b0
+#define CODA9_RET_DEC_SEQ_BITRATE		0x1b4
+#define CODA_RET_DEC_SEQ_SUCCESS		0x1c0
+#define CODA_RET_DEC_SEQ_SRC_FMT		0x1c4 /* SRC_SIZE on CODA7 */
+#define CODA_RET_DEC_SEQ_SRC_SIZE		0x1c4
+#define CODA_RET_DEC_SEQ_SRC_F_RATE		0x1c8
+#define CODA9_RET_DEC_SEQ_ASPECT		0x1c8
+#define CODA_RET_DEC_SEQ_FRAME_NEED		0x1cc
+#define CODA_RET_DEC_SEQ_FRAME_DELAY		0x1d0
+#define CODA_RET_DEC_SEQ_INFO			0x1d4
+#define CODA_RET_DEC_SEQ_CROP_LEFT_RIGHT	0x1d8
+#define CODA_RET_DEC_SEQ_CROP_TOP_BOTTOM	0x1dc
+#define CODA_RET_DEC_SEQ_NEXT_FRAME_NUM		0x1e0
+#define CODA_RET_DEC_SEQ_ERR_REASON		0x1e0
+#define CODA_RET_DEC_SEQ_FRATE_NR		0x1e4
+#define CODA_RET_DEC_SEQ_FRATE_DR		0x1e8
+#define CODA_RET_DEC_SEQ_JPG_PARA		0x1e4
+#define CODA_RET_DEC_SEQ_JPG_THUMB_IND		0x1e8
+#define CODA7_RET_DEC_SEQ_HEADER_REPORT		0x1ec
+
+/* Decoder Picture Run */
+#define CODA_CMD_DEC_PIC_ROT_MODE		0x180
+#define CODA_CMD_DEC_PIC_ROT_ADDR_Y		0x184
+#define CODA9_CMD_DEC_PIC_ROT_INDEX		0x184
+#define CODA_CMD_DEC_PIC_ROT_ADDR_CB		0x188
+#define CODA9_CMD_DEC_PIC_ROT_ADDR_Y		0x188
+#define CODA_CMD_DEC_PIC_ROT_ADDR_CR		0x18c
+#define CODA9_CMD_DEC_PIC_ROT_ADDR_CB		0x18c
+#define CODA_CMD_DEC_PIC_ROT_STRIDE		0x190
+#define CODA9_CMD_DEC_PIC_ROT_ADDR_CR		0x190
+#define CODA9_CMD_DEC_PIC_ROT_STRIDE		0x1b8
+
+#define CODA_CMD_DEC_PIC_OPTION			0x194
+#define		CODA_PRE_SCAN_EN			(1 << 0)
+#define		CODA_PRE_SCAN_MODE_DECODE		(0 << 1)
+#define		CODA_PRE_SCAN_MODE_RETURN		(1 << 1)
+#define		CODA_IFRAME_SEARCH_EN			(1 << 2)
+#define		CODA_SKIP_FRAME_MODE			(0x3 << 3)
+#define CODA_CMD_DEC_PIC_SKIP_NUM		0x198
+#define CODA_CMD_DEC_PIC_CHUNK_SIZE		0x19c
+#define CODA_CMD_DEC_PIC_BB_START		0x1a0
+#define CODA_CMD_DEC_PIC_START_BYTE		0x1a4
+#define CODA_RET_DEC_PIC_SIZE			0x1bc
+#define CODA_RET_DEC_PIC_FRAME_NUM		0x1c0
+#define CODA_RET_DEC_PIC_FRAME_IDX		0x1c4
+#define CODA_RET_DEC_PIC_ERR_MB			0x1c8
+#define CODA_RET_DEC_PIC_TYPE			0x1cc
+#define		CODA_PIC_TYPE_MASK			0x7
+#define		CODA_PIC_TYPE_MASK_VC1			0x3f
+#define		CODA9_PIC_TYPE_FIRST_MASK		(0x7 << 3)
+#define		CODA9_PIC_TYPE_IDR_MASK			(0x3 << 6)
+#define		CODA7_PIC_TYPE_H264_NPF_MASK		(0x3 << 16)
+#define		CODA7_PIC_TYPE_INTERLACED		(1 << 18)
+#define CODA_RET_DEC_PIC_POST			0x1d0
+#define CODA_RET_DEC_PIC_MVC_REPORT		0x1d0
+#define CODA_RET_DEC_PIC_OPTION			0x1d4
+#define CODA_RET_DEC_PIC_SUCCESS		0x1d8
+#define CODA_RET_DEC_PIC_CUR_IDX		0x1dc
+#define CODA_RET_DEC_PIC_CROP_LEFT_RIGHT	0x1e0
+#define CODA_RET_DEC_PIC_CROP_TOP_BOTTOM	0x1e4
+#define CODA_RET_DEC_PIC_FRAME_NEED		0x1ec
+
+#define CODA9_RET_DEC_PIC_VP8_PIC_REPORT	0x1e8
+#define CODA9_RET_DEC_PIC_ASPECT		0x1f0
+#define CODA9_RET_DEC_PIC_VP8_SCALE_INFO	0x1f0
+#define CODA9_RET_DEC_PIC_FRATE_NR		0x1f4
+#define CODA9_RET_DEC_PIC_FRATE_DR		0x1f8
+
+/* Encoder Sequence Initialization */
+#define CODA_CMD_ENC_SEQ_BB_START				0x180
+#define CODA_CMD_ENC_SEQ_BB_SIZE				0x184
+#define CODA_CMD_ENC_SEQ_OPTION				0x188
+#define		CODA7_OPTION_AVCINTRA16X16ONLY_OFFSET		9
+#define		CODA9_OPTION_MVC_PREFIX_NAL_OFFSET		9
+#define		CODA7_OPTION_GAMMA_OFFSET			8
+#define		CODA9_OPTION_MVC_PARASET_REFRESH_OFFSET		8
+#define		CODA7_OPTION_RCQPMAX_OFFSET			7
+#define		CODA9_OPTION_GAMMA_OFFSET			7
+#define		CODADX6_OPTION_GAMMA_OFFSET			7
+#define		CODA7_OPTION_RCQPMIN_OFFSET			6
+#define		CODA9_OPTION_RCQPMAX_OFFSET			6
+#define		CODA_OPTION_LIMITQP_OFFSET			6
+#define		CODA_OPTION_RCINTRAQP_OFFSET			5
+#define		CODA_OPTION_FMO_OFFSET				4
+#define		CODA9_OPTION_MVC_INTERVIEW_OFFSET		4
+#define		CODA_OPTION_AVC_AUD_OFFSET			2
+#define		CODA_OPTION_SLICEREPORT_OFFSET			1
+#define CODA_CMD_ENC_SEQ_COD_STD				0x18c
+#define		CODA_STD_MPEG4					0
+#define		CODA9_STD_H264					0
+#define		CODA_STD_H263					1
+#define		CODA_STD_H264					2
+#define		CODA9_STD_MPEG4					3
+
+#define CODA_CMD_ENC_SEQ_SRC_SIZE				0x190
+#define		CODA7_PICWIDTH_OFFSET				16
+#define		CODA7_PICWIDTH_MASK				0xffff
+#define		CODADX6_PICWIDTH_OFFSET				10
+#define		CODADX6_PICWIDTH_MASK				0x3ff
+#define		CODA_PICHEIGHT_OFFSET				0
+#define		CODADX6_PICHEIGHT_MASK				0x3ff
+#define		CODA7_PICHEIGHT_MASK				0xffff
+#define CODA_CMD_ENC_SEQ_SRC_F_RATE				0x194
+#define		CODA_FRATE_RES_OFFSET				0
+#define		CODA_FRATE_RES_MASK				0xffff
+#define		CODA_FRATE_DIV_OFFSET				16
+#define		CODA_FRATE_DIV_MASK				0xffff
+#define CODA_CMD_ENC_SEQ_MP4_PARA				0x198
+#define		CODA_MP4PARAM_VERID_OFFSET			6
+#define		CODA_MP4PARAM_VERID_MASK			0x01
+#define		CODA_MP4PARAM_INTRADCVLCTHR_OFFSET		2
+#define		CODA_MP4PARAM_INTRADCVLCTHR_MASK		0x07
+#define		CODA_MP4PARAM_REVERSIBLEVLCENABLE_OFFSET	1
+#define		CODA_MP4PARAM_REVERSIBLEVLCENABLE_MASK		0x01
+#define		CODA_MP4PARAM_DATAPARTITIONENABLE_OFFSET	0
+#define		CODA_MP4PARAM_DATAPARTITIONENABLE_MASK		0x01
+#define CODA_CMD_ENC_SEQ_263_PARA				0x19c
+#define		CODA_263PARAM_ANNEXJENABLE_OFFSET		2
+#define		CODA_263PARAM_ANNEXJENABLE_MASK		0x01
+#define		CODA_263PARAM_ANNEXKENABLE_OFFSET		1
+#define		CODA_263PARAM_ANNEXKENABLE_MASK		0x01
+#define		CODA_263PARAM_ANNEXTENABLE_OFFSET		0
+#define		CODA_263PARAM_ANNEXTENABLE_MASK		0x01
+#define CODA_CMD_ENC_SEQ_264_PARA				0x1a0
+#define		CODA_264PARAM_DEBLKFILTEROFFSETBETA_OFFSET	12
+#define		CODA_264PARAM_DEBLKFILTEROFFSETBETA_MASK	0x0f
+#define		CODA_264PARAM_DEBLKFILTEROFFSETALPHA_OFFSET	8
+#define		CODA_264PARAM_DEBLKFILTEROFFSETALPHA_MASK	0x0f
+#define		CODA_264PARAM_DISABLEDEBLK_OFFSET		6
+#define		CODA_264PARAM_DISABLEDEBLK_MASK		0x03
+#define		CODA_264PARAM_CONSTRAINEDINTRAPREDFLAG_OFFSET	5
+#define		CODA_264PARAM_CONSTRAINEDINTRAPREDFLAG_MASK	0x01
+#define		CODA_264PARAM_CHROMAQPOFFSET_OFFSET		0
+#define		CODA_264PARAM_CHROMAQPOFFSET_MASK		0x1f
+#define CODA_CMD_ENC_SEQ_SLICE_MODE				0x1a4
+#define		CODA_SLICING_SIZE_OFFSET			2
+#define		CODA_SLICING_SIZE_MASK				0x3fffffff
+#define		CODA_SLICING_UNIT_OFFSET			1
+#define		CODA_SLICING_UNIT_MASK				0x01
+#define		CODA_SLICING_MODE_OFFSET			0
+#define		CODA_SLICING_MODE_MASK				0x01
+#define CODA_CMD_ENC_SEQ_GOP_SIZE				0x1a8
+#define		CODA_GOP_SIZE_OFFSET				0
+#define		CODA_GOP_SIZE_MASK				0x3f
+#define CODA_CMD_ENC_SEQ_RC_PARA				0x1ac
+#define		CODA_RATECONTROL_AUTOSKIP_OFFSET		31
+#define		CODA_RATECONTROL_AUTOSKIP_MASK			0x01
+#define		CODA_RATECONTROL_INITIALDELAY_OFFSET		16
+#define		CODA_RATECONTROL_INITIALDELAY_MASK		0x7fff
+#define		CODA_RATECONTROL_BITRATE_OFFSET		1
+#define		CODA_RATECONTROL_BITRATE_MASK			0x7fff
+#define		CODA_RATECONTROL_ENABLE_OFFSET			0
+#define		CODA_RATECONTROL_ENABLE_MASK			0x01
+#define CODA_CMD_ENC_SEQ_RC_BUF_SIZE				0x1b0
+#define CODA_CMD_ENC_SEQ_INTRA_REFRESH				0x1b4
+#define CODADX6_CMD_ENC_SEQ_FMO					0x1b8
+#define		CODA_FMOPARAM_TYPE_OFFSET			4
+#define		CODA_FMOPARAM_TYPE_MASK				1
+#define		CODA_FMOPARAM_SLICENUM_OFFSET			0
+#define		CODA_FMOPARAM_SLICENUM_MASK			0x0f
+#define CODADX6_CMD_ENC_SEQ_INTRA_QP				0x1bc
+#define CODA7_CMD_ENC_SEQ_SEARCH_BASE				0x1b8
+#define CODA7_CMD_ENC_SEQ_SEARCH_SIZE				0x1bc
+#define CODA7_CMD_ENC_SEQ_INTRA_QP				0x1c4
+#define CODA_CMD_ENC_SEQ_RC_QP_MIN_MAX				0x1c8
+#define		CODA_QPMIN_OFFSET				8
+#define		CODA_QPMIN_MASK					0x3f
+#define		CODA_QPMAX_OFFSET				0
+#define		CODA_QPMAX_MASK					0x3f
+#define CODA_CMD_ENC_SEQ_RC_GAMMA				0x1cc
+#define		CODA_GAMMA_OFFSET				0
+#define		CODA_GAMMA_MASK					0xffff
+#define CODA_CMD_ENC_SEQ_RC_INTERVAL_MODE			0x1d0
+#define CODA9_CMD_ENC_SEQ_INTRA_WEIGHT				0x1d4
+#define CODA9_CMD_ENC_SEQ_ME_OPTION				0x1d8
+#define CODA_RET_ENC_SEQ_SUCCESS				0x1c0
+
+#define CODA_CMD_ENC_SEQ_JPG_PARA				0x198
+#define CODA_CMD_ENC_SEQ_JPG_RST_INTERVAL			0x19C
+#define CODA_CMD_ENC_SEQ_JPG_THUMB_EN				0x1a0
+#define CODA_CMD_ENC_SEQ_JPG_THUMB_SIZE				0x1a4
+#define CODA_CMD_ENC_SEQ_JPG_THUMB_OFFSET			0x1a8
+
+/* Encoder Parameter Change */
+#define CODA_CMD_ENC_PARAM_CHANGE_ENABLE	0x180
+#define		CODA_PARAM_CHANGE_RC_GOP			BIT(0)
+#define		CODA_PARAM_CHANGE_RC_INTRA_QP			BIT(1)
+#define		CODA_PARAM_CHANGE_RC_BITRATE			BIT(2)
+#define		CODA_PARAM_CHANGE_RC_FRAME_RATE			BIT(3)
+#define		CODA_PARAM_CHANGE_INTRA_MB_NUM			BIT(4)
+#define		CODA_PARAM_CHANGE_SLICE_MODE			BIT(5)
+#define		CODA_PARAM_CHANGE_HEC_MODE			BIT(6)
+#define CODA_CMD_ENC_PARAM_RC_GOP		0x184
+#define CODA_CMD_ENC_PARAM_RC_INTRA_QP		0x188
+#define CODA_CMD_ENC_PARAM_RC_BITRATE		0x18c
+#define CODA_CMD_ENC_PARAM_RC_FRAME_RATE	0x190
+#define CODA_CMD_ENC_PARAM_INTRA_MB_NUM		0x194
+#define CODA_CMD_ENC_PARAM_SLICE_MODE		0x198
+#define CODA_CMD_ENC_PARAM_HEC_MODE		0x19c
+#define CODA_RET_ENC_PARAM_CHANGE_SUCCESS	0x1c0
+
+/* Encoder Picture Run */
+#define CODA9_CMD_ENC_PIC_SRC_INDEX		0x180
+#define CODA9_CMD_ENC_PIC_SRC_STRIDE		0x184
+#define CODA9_CMD_ENC_PIC_SUB_FRAME_SYNC	0x1a4
+#define CODA9_CMD_ENC_PIC_SRC_ADDR_Y		0x1a8
+#define CODA9_CMD_ENC_PIC_SRC_ADDR_CB		0x1ac
+#define CODA9_CMD_ENC_PIC_SRC_ADDR_CR		0x1b0
+#define CODA_CMD_ENC_PIC_SRC_ADDR_Y	0x180
+#define CODA_CMD_ENC_PIC_SRC_ADDR_CB	0x184
+#define CODA_CMD_ENC_PIC_SRC_ADDR_CR	0x188
+#define CODA_CMD_ENC_PIC_QS		0x18c
+#define CODA_CMD_ENC_PIC_ROT_MODE	0x190
+#define		CODA_ROT_MIR_ENABLE				(1 << 4)
+#define		CODA_ROT_0					(0x0 << 0)
+#define		CODA_ROT_90					(0x1 << 0)
+#define		CODA_ROT_180					(0x2 << 0)
+#define		CODA_ROT_270					(0x3 << 0)
+#define		CODA_MIR_NONE					(0x0 << 2)
+#define		CODA_MIR_VER					(0x1 << 2)
+#define		CODA_MIR_HOR					(0x2 << 2)
+#define		CODA_MIR_VER_HOR				(0x3 << 2)
+#define CODA_CMD_ENC_PIC_OPTION		0x194
+#define		CODA_FORCE_IPICTURE				BIT(1)
+#define		CODA_REPORT_MB_INFO				BIT(3)
+#define		CODA_REPORT_MV_INFO				BIT(4)
+#define		CODA_REPORT_SLICE_INFO				BIT(5)
+#define CODA_CMD_ENC_PIC_BB_START	0x198
+#define CODA_CMD_ENC_PIC_BB_SIZE	0x19c
+#define CODA_RET_ENC_FRAME_NUM		0x1c0
+#define CODA_RET_ENC_PIC_TYPE		0x1c4
+#define CODA_RET_ENC_PIC_FRAME_IDX	0x1c8
+#define CODA_RET_ENC_PIC_SLICE_NUM	0x1cc
+#define CODA_RET_ENC_PIC_FLAG		0x1d0
+#define CODA_RET_ENC_PIC_SUCCESS	0x1d8
+
+/* Set Frame Buffer */
+#define CODA_CMD_SET_FRAME_BUF_NUM		0x180
+#define CODA_CMD_SET_FRAME_BUF_STRIDE		0x184
+#define CODA_CMD_SET_FRAME_SLICE_BB_START	0x188
+#define CODA_CMD_SET_FRAME_SLICE_BB_SIZE	0x18c
+#define CODA9_CMD_SET_FRAME_SUBSAMP_A		0x188
+#define CODA9_CMD_SET_FRAME_SUBSAMP_B		0x18c
+#define CODA7_CMD_SET_FRAME_AXI_BIT_ADDR	0x190
+#define CODA7_CMD_SET_FRAME_AXI_IPACDC_ADDR	0x194
+#define CODA7_CMD_SET_FRAME_AXI_DBKY_ADDR	0x198
+#define CODA7_CMD_SET_FRAME_AXI_DBKC_ADDR	0x19c
+#define CODA7_CMD_SET_FRAME_AXI_OVL_ADDR	0x1a0
+#define CODA7_CMD_SET_FRAME_MAX_DEC_SIZE	0x1a4
+#define CODA9_CMD_SET_FRAME_AXI_BTP_ADDR	0x1a4
+#define CODA7_CMD_SET_FRAME_SOURCE_BUF_STRIDE	0x1a8
+#define CODA9_CMD_SET_FRAME_CACHE_SIZE		0x1a8
+#define CODA9_CMD_SET_FRAME_CACHE_CONFIG	0x1ac
+#define		CODA9_CACHE_BYPASS_OFFSET		28
+#define		CODA9_CACHE_DUALCONF_OFFSET		26
+#define		CODA9_CACHE_PAGEMERGE_OFFSET		24
+#define		CODA9_CACHE_LUMA_BUFFER_SIZE_OFFSET	16
+#define		CODA9_CACHE_CB_BUFFER_SIZE_OFFSET	8
+#define		CODA9_CACHE_CR_BUFFER_SIZE_OFFSET	0
+#define CODA9_CMD_SET_FRAME_SUBSAMP_A_MVC	0x1b0
+#define CODA9_CMD_SET_FRAME_SUBSAMP_B_MVC	0x1b4
+#define CODA9_CMD_SET_FRAME_DP_BUF_BASE		0x1b0
+#define CODA9_CMD_SET_FRAME_DP_BUF_SIZE		0x1b4
+#define CODA9_CMD_SET_FRAME_MAX_DEC_SIZE	0x1b8
+#define CODA9_CMD_SET_FRAME_DELAY		0x1bc
+
+/* Encoder Header */
+#define CODA_CMD_ENC_HEADER_CODE	0x180
+#define		CODA_GAMMA_OFFSET	0
+#define		CODA_HEADER_H264_SPS	0
+#define		CODA_HEADER_H264_PPS	1
+#define		CODA_HEADER_MP4V_VOL	0
+#define		CODA_HEADER_MP4V_VOS	1
+#define		CODA_HEADER_MP4V_VIS	2
+#define		CODA9_HEADER_FRAME_CROP	(1 << 3)
+#define CODA_CMD_ENC_HEADER_BB_START	0x184
+#define CODA_CMD_ENC_HEADER_BB_SIZE	0x188
+#define CODA9_CMD_ENC_HEADER_FRAME_CROP_H	0x18c
+#define CODA9_CMD_ENC_HEADER_FRAME_CROP_V	0x190
+
+/* Get Version */
+#define CODA_CMD_FIRMWARE_VERNUM		0x1c0
+#define		CODA_FIRMWARE_PRODUCT(x)	(((x) >> 16) & 0xffff)
+#define		CODA_FIRMWARE_MAJOR(x)		(((x) >> 12) & 0x0f)
+#define		CODA_FIRMWARE_MINOR(x)		(((x) >> 8) & 0x0f)
+#define		CODA_FIRMWARE_RELEASE(x)	((x) & 0xff)
+#define		CODA_FIRMWARE_VERNUM(product, major, minor, release)	\
+			((product) << 16 | ((major) << 12) |		\
+			((minor) << 8) | (release))
+#define CODA9_CMD_FIRMWARE_CODE_REV		0x1c4
+
+#define CODA9_GDMA_BASE				0x1000
+#define CODA9_GDI_CONTROL			(CODA9_GDMA_BASE + 0x034)
+#define CODA9_GDI_PIC_INIT_HOST			(CODA9_GDMA_BASE + 0x038)
+#define CODA9_GDI_STATUS			(CODA9_GDMA_BASE + 0x080)
+#define CODA9_GDI_WPROT_ERR_CLR			(CODA9_GDMA_BASE + 0x0a0)
+#define CODA9_GDI_WPROT_RGN_EN			(CODA9_GDMA_BASE + 0x0ac)
+
+#define CODA9_GDI_BUS_CTRL			(CODA9_GDMA_BASE + 0x0f0)
+#define CODA9_GDI_BUS_STATUS			(CODA9_GDMA_BASE + 0x0f4)
+
+#define CODA9_GDI_INFO_CONTROL			(CODA9_GDMA_BASE + 0x400)
+#define CODA9_GDI_INFO_PIC_SIZE			(CODA9_GDMA_BASE + 0x404)
+#define CODA9_GDI_INFO_BASE_Y			(CODA9_GDMA_BASE + 0x408)
+#define CODA9_GDI_INFO_BASE_CB			(CODA9_GDMA_BASE + 0x40c)
+#define CODA9_GDI_INFO_BASE_CR			(CODA9_GDMA_BASE + 0x410)
+
+#define CODA9_GDI_XY2_CAS_0			(CODA9_GDMA_BASE + 0x800)
+#define CODA9_GDI_XY2_CAS_F			(CODA9_GDMA_BASE + 0x83c)
+
+#define CODA9_GDI_XY2_BA_0			(CODA9_GDMA_BASE + 0x840)
+#define CODA9_GDI_XY2_BA_1			(CODA9_GDMA_BASE + 0x844)
+#define CODA9_GDI_XY2_BA_2			(CODA9_GDMA_BASE + 0x848)
+#define CODA9_GDI_XY2_BA_3			(CODA9_GDMA_BASE + 0x84c)
+
+#define CODA9_GDI_XY2_RAS_0			(CODA9_GDMA_BASE + 0x850)
+#define CODA9_GDI_XY2_RAS_F			(CODA9_GDMA_BASE + 0x88c)
+
+#define CODA9_GDI_XY2_RBC_CONFIG		(CODA9_GDMA_BASE + 0x890)
+#define		CODA9_XY2RBC_SEPARATE_MAP		BIT(19)
+#define		CODA9_XY2RBC_TOP_BOT_SPLIT		BIT(18)
+#define		CODA9_XY2RBC_TILED_MAP			BIT(17)
+#define		CODA9_XY2RBC_CA_INC_HOR			BIT(16)
+#define CODA9_GDI_RBC2_AXI_0			(CODA9_GDMA_BASE + 0x8a0)
+#define CODA9_GDI_RBC2_AXI_1F			(CODA9_GDMA_BASE + 0x91c)
+#define	CODA9_GDI_TILEDBUF_BASE			(CODA9_GDMA_BASE + 0x920)
+
+#define CODA9_JPEG_BASE				0x3000
+#define CODA9_REG_JPEG_PIC_START		(CODA9_JPEG_BASE + 0x000)
+#define CODA9_REG_JPEG_PIC_STATUS		(CODA9_JPEG_BASE + 0x004)
+#define		CODA9_JPEG_STATUS_OVERFLOW		BIT(3)
+#define		CODA9_JPEG_STATUS_BBC_INT		BIT(2)
+#define		CODA9_JPEG_STATUS_ERROR			BIT(1)
+#define		CODA9_JPEG_STATUS_DONE			BIT(0)
+#define CODA9_REG_JPEG_PIC_ERRMB		(CODA9_JPEG_BASE + 0x008)
+#define		CODA9_JPEG_ERRMB_RESTART_IDX_MASK	(0xf << 24)
+#define		CODA9_JPEG_ERRMB_MCU_POS_X_MASK		(0xfff << 12)
+#define		CODA9_JPEG_ERRMB_MCU_POS_Y_MASK		0xfff
+#define CODA9_REG_JPEG_PIC_CTRL			(CODA9_JPEG_BASE + 0x010)
+#define		CODA9_JPEG_PIC_CTRL_USER_HUFFMAN_EN	BIT(6)
+#define		CODA9_JPEG_PIC_CTRL_TC_DIRECTION	BIT(4)
+#define		CODA9_JPEG_PIC_CTRL_ENCODER_EN		BIT(3)
+#define CODA9_REG_JPEG_PIC_SIZE			(CODA9_JPEG_BASE + 0x014)
+#define CODA9_REG_JPEG_MCU_INFO			(CODA9_JPEG_BASE + 0x018)
+#define		CODA9_JPEG_MCU_BLOCK_NUM_OFFSET		16
+#define		CODA9_JPEG_COMP_NUM_OFFSET		12
+#define		CODA9_JPEG_COMP0_INFO_OFFSET		8
+#define		CODA9_JPEG_COMP1_INFO_OFFSET		4
+#define		CODA9_JPEG_COMP2_INFO_OFFSET		0
+#define CODA9_REG_JPEG_ROT_INFO			(CODA9_JPEG_BASE + 0x01c)
+#define		CODA9_JPEG_ROT_MIR_ENABLE		BIT(4)
+#define		CODA9_JPEG_ROT_MIR_MODE_MASK		0xf
+#define CODA9_REG_JPEG_SCL_INFO			(CODA9_JPEG_BASE + 0x020)
+#define		CODA9_JPEG_SCL_ENABLE			BIT(4)
+#define		CODA9_JPEG_SCL_HOR_MODE_MASK		(0x3 << 2)
+#define		CODA9_JPEG_SCL_VER_MODE_MASK		(0x3 << 0)
+#define CODA9_REG_JPEG_IF_INFO			(CODA9_JPEG_BASE + 0x024)
+#define		CODA9_JPEG_SENS_IF_CLR			BIT(1)
+#define		CODA9_JPEG_DISP_IF_CLR			BIT(0)
+#define CODA9_REG_JPEG_OP_INFO			(CODA9_JPEG_BASE + 0x02c)
+#define		CODA9_JPEG_BUS_REQ_NUM_OFFSET		0
+#define		CODA9_JPEG_BUS_REQ_NUM_MASK		0x7
+#define CODA9_REG_JPEG_DPB_CONFIG		(CODA9_JPEG_BASE + 0x030)
+#define CODA9_REG_JPEG_DPB_BASE00		(CODA9_JPEG_BASE + 0x040)
+#define CODA9_REG_JPEG_HUFF_CTRL		(CODA9_JPEG_BASE + 0x080)
+#define CODA9_REG_JPEG_HUFF_ADDR		(CODA9_JPEG_BASE + 0x084)
+#define CODA9_REG_JPEG_HUFF_DATA		(CODA9_JPEG_BASE + 0x088)
+#define CODA9_REG_JPEG_QMAT_CTRL		(CODA9_JPEG_BASE + 0x090)
+#define CODA9_REG_JPEG_QMAT_ADDR		(CODA9_JPEG_BASE + 0x094)
+#define CODA9_REG_JPEG_QMAT_DATA		(CODA9_JPEG_BASE + 0x098)
+#define CODA9_REG_JPEG_RST_INTVAL		(CODA9_JPEG_BASE + 0x0b0)
+#define CODA9_REG_JPEG_RST_INDEX		(CODA9_JPEG_BASE + 0x0b4)
+#define CODA9_REG_JPEG_RST_COUNT		(CODA9_JPEG_BASE + 0x0b8)
+#define CODA9_REG_JPEG_DPCM_DIFF_Y		(CODA9_JPEG_BASE + 0x0f0)
+#define CODA9_REG_JPEG_DPCM_DIFF_CB		(CODA9_JPEG_BASE + 0x0f4)
+#define CODA9_REG_JPEG_DPCM_DIFF_CR		(CODA9_JPEG_BASE + 0x0f8)
+#define CODA9_REG_JPEG_GBU_CTRL			(CODA9_JPEG_BASE + 0x100)
+#define CODA9_REG_JPEG_GBU_BT_PTR		(CODA9_JPEG_BASE + 0x110)
+#define CODA9_REG_JPEG_GBU_WD_PTR		(CODA9_JPEG_BASE + 0x114)
+#define CODA9_REG_JPEG_GBU_TT_CNT		(CODA9_JPEG_BASE + 0x118)
+#define CODA9_REG_JPEG_GBU_BBSR			(CODA9_JPEG_BASE + 0x140)
+#define CODA9_REG_JPEG_GBU_BBER			(CODA9_JPEG_BASE + 0x144)
+#define CODA9_REG_JPEG_GBU_BBIR			(CODA9_JPEG_BASE + 0x148)
+#define CODA9_REG_JPEG_GBU_BBHR			(CODA9_JPEG_BASE + 0x14c)
+#define CODA9_REG_JPEG_GBU_BCNT			(CODA9_JPEG_BASE + 0x158)
+#define CODA9_REG_JPEG_GBU_FF_RPTR		(CODA9_JPEG_BASE + 0x160)
+#define CODA9_REG_JPEG_GBU_FF_WPTR		(CODA9_JPEG_BASE + 0x164)
+#define CODA9_REG_JPEG_BBC_END_ADDR		(CODA9_JPEG_BASE + 0x208)
+#define CODA9_REG_JPEG_BBC_WR_PTR		(CODA9_JPEG_BASE + 0x20c)
+#define CODA9_REG_JPEG_BBC_RD_PTR		(CODA9_JPEG_BASE + 0x210)
+#define CODA9_REG_JPEG_BBC_EXT_ADDR		(CODA9_JPEG_BASE + 0x214)
+#define CODA9_REG_JPEG_BBC_INT_ADDR		(CODA9_JPEG_BASE + 0x218)
+#define CODA9_REG_JPEG_BBC_DATA_CNT		(CODA9_JPEG_BASE + 0x21c)
+#define CODA9_REG_JPEG_BBC_COMMAND		(CODA9_JPEG_BASE + 0x220)
+#define CODA9_REG_JPEG_BBC_BUSY			(CODA9_JPEG_BASE + 0x224)
+#define CODA9_REG_JPEG_BBC_CTRL			(CODA9_JPEG_BASE + 0x228)
+#define CODA9_REG_JPEG_BBC_CUR_POS		(CODA9_JPEG_BASE + 0x22c)
+#define CODA9_REG_JPEG_BBC_BAS_ADDR		(CODA9_JPEG_BASE + 0x230)
+#define CODA9_REG_JPEG_BBC_STRM_CTRL		(CODA9_JPEG_BASE + 0x234)
+#define CODA9_REG_JPEG_BBC_FLUSH_CMD		(CODA9_JPEG_BASE + 0x238)
+
+#endif
diff --git a/drivers/media/platform/chips-media/imx-vdoa.c b/drivers/media/platform/chips-media/imx-vdoa.c
new file mode 100644
index 000000000..c70871bae
--- /dev/null
+++ b/drivers/media/platform/chips-media/imx-vdoa.c
@@ -0,0 +1,352 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * i.MX6 Video Data Order Adapter (VDOA)
+ *
+ * Copyright (C) 2014 Philipp Zabel
+ * Copyright (C) 2016 Pengutronix, Michael Tretter <kernel@pengutronix.de>
+ */
+
+#include <linux/clk.h>
+#include <linux/device.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/mod_devicetable.h>
+#include <linux/dma-mapping.h>
+#include <linux/platform_device.h>
+#include <linux/videodev2.h>
+#include <linux/slab.h>
+
+#include "imx-vdoa.h"
+
+#define VDOA_NAME "imx-vdoa"
+
+#define VDOAC		0x00
+#define VDOASRR		0x04
+#define VDOAIE		0x08
+#define VDOAIST		0x0c
+#define VDOAFP		0x10
+#define VDOAIEBA00	0x14
+#define VDOAIEBA01	0x18
+#define VDOAIEBA02	0x1c
+#define VDOAIEBA10	0x20
+#define VDOAIEBA11	0x24
+#define VDOAIEBA12	0x28
+#define VDOASL		0x2c
+#define VDOAIUBO	0x30
+#define VDOAVEBA0	0x34
+#define VDOAVEBA1	0x38
+#define VDOAVEBA2	0x3c
+#define VDOAVUBO	0x40
+#define VDOASR		0x44
+
+#define VDOAC_ISEL		BIT(6)
+#define VDOAC_PFS		BIT(5)
+#define VDOAC_SO		BIT(4)
+#define VDOAC_SYNC		BIT(3)
+#define VDOAC_NF		BIT(2)
+#define VDOAC_BNDM_MASK		0x3
+#define VDOAC_BAND_HEIGHT_8	0x0
+#define VDOAC_BAND_HEIGHT_16	0x1
+#define VDOAC_BAND_HEIGHT_32	0x2
+
+#define VDOASRR_START		BIT(1)
+#define VDOASRR_SWRST		BIT(0)
+
+#define VDOAIE_EITERR		BIT(1)
+#define VDOAIE_EIEOT		BIT(0)
+
+#define VDOAIST_TERR		BIT(1)
+#define VDOAIST_EOT		BIT(0)
+
+#define VDOAFP_FH_MASK		(0x1fff << 16)
+#define VDOAFP_FW_MASK		(0x3fff)
+
+#define VDOASL_VSLY_MASK	(0x3fff << 16)
+#define VDOASL_ISLY_MASK	(0x7fff)
+
+#define VDOASR_ERRW		BIT(4)
+#define VDOASR_EOB		BIT(3)
+#define VDOASR_CURRENT_FRAME	(0x3 << 1)
+#define VDOASR_CURRENT_BUFFER	BIT(1)
+
+enum {
+	V4L2_M2M_SRC = 0,
+	V4L2_M2M_DST = 1,
+};
+
+struct vdoa_data {
+	struct vdoa_ctx		*curr_ctx;
+	struct device		*dev;
+	struct clk		*vdoa_clk;
+	void __iomem		*regs;
+};
+
+struct vdoa_q_data {
+	unsigned int	width;
+	unsigned int	height;
+	unsigned int	bytesperline;
+	unsigned int	sizeimage;
+	u32		pixelformat;
+};
+
+struct vdoa_ctx {
+	struct vdoa_data	*vdoa;
+	struct completion	completion;
+	struct vdoa_q_data	q_data[2];
+	unsigned int		submitted_job;
+	unsigned int		completed_job;
+};
+
+static irqreturn_t vdoa_irq_handler(int irq, void *data)
+{
+	struct vdoa_data *vdoa = data;
+	struct vdoa_ctx *curr_ctx;
+	u32 val;
+
+	/* Disable interrupts */
+	writel(0, vdoa->regs + VDOAIE);
+
+	curr_ctx = vdoa->curr_ctx;
+	if (!curr_ctx) {
+		dev_warn(vdoa->dev,
+			"Instance released before the end of transaction\n");
+		return IRQ_HANDLED;
+	}
+
+	val = readl(vdoa->regs + VDOAIST);
+	writel(val, vdoa->regs + VDOAIST);
+	if (val & VDOAIST_TERR) {
+		val = readl(vdoa->regs + VDOASR) & VDOASR_ERRW;
+		dev_err(vdoa->dev, "AXI %s error\n", val ? "write" : "read");
+	} else if (!(val & VDOAIST_EOT)) {
+		dev_warn(vdoa->dev, "Spurious interrupt\n");
+	}
+	curr_ctx->completed_job++;
+	complete(&curr_ctx->completion);
+
+	return IRQ_HANDLED;
+}
+
+int vdoa_wait_for_completion(struct vdoa_ctx *ctx)
+{
+	struct vdoa_data *vdoa = ctx->vdoa;
+
+	if (ctx->submitted_job == ctx->completed_job)
+		return 0;
+
+	if (!wait_for_completion_timeout(&ctx->completion,
+					 msecs_to_jiffies(300))) {
+		dev_err(vdoa->dev,
+			"Timeout waiting for transfer result\n");
+		return -ETIMEDOUT;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL(vdoa_wait_for_completion);
+
+void vdoa_device_run(struct vdoa_ctx *ctx, dma_addr_t dst, dma_addr_t src)
+{
+	struct vdoa_q_data *src_q_data, *dst_q_data;
+	struct vdoa_data *vdoa = ctx->vdoa;
+	u32 val;
+
+	if (vdoa->curr_ctx)
+		vdoa_wait_for_completion(vdoa->curr_ctx);
+
+	vdoa->curr_ctx = ctx;
+
+	reinit_completion(&ctx->completion);
+	ctx->submitted_job++;
+
+	src_q_data = &ctx->q_data[V4L2_M2M_SRC];
+	dst_q_data = &ctx->q_data[V4L2_M2M_DST];
+
+	/* Progressive, no sync, 1 frame per run */
+	if (dst_q_data->pixelformat == V4L2_PIX_FMT_YUYV)
+		val = VDOAC_PFS;
+	else
+		val = 0;
+	writel(val, vdoa->regs + VDOAC);
+
+	writel(dst_q_data->height << 16 | dst_q_data->width,
+	       vdoa->regs + VDOAFP);
+
+	val = dst;
+	writel(val, vdoa->regs + VDOAIEBA00);
+
+	writel(src_q_data->bytesperline << 16 | dst_q_data->bytesperline,
+	       vdoa->regs + VDOASL);
+
+	if (dst_q_data->pixelformat == V4L2_PIX_FMT_NV12 ||
+	    dst_q_data->pixelformat == V4L2_PIX_FMT_NV21)
+		val = dst_q_data->bytesperline * dst_q_data->height;
+	else
+		val = 0;
+	writel(val, vdoa->regs + VDOAIUBO);
+
+	val = src;
+	writel(val, vdoa->regs + VDOAVEBA0);
+	val = round_up(src_q_data->bytesperline * src_q_data->height, 4096);
+	writel(val, vdoa->regs + VDOAVUBO);
+
+	/* Enable interrupts and start transfer */
+	writel(VDOAIE_EITERR | VDOAIE_EIEOT, vdoa->regs + VDOAIE);
+	writel(VDOASRR_START, vdoa->regs + VDOASRR);
+}
+EXPORT_SYMBOL(vdoa_device_run);
+
+struct vdoa_ctx *vdoa_context_create(struct vdoa_data *vdoa)
+{
+	struct vdoa_ctx *ctx;
+	int err;
+
+	ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
+	if (!ctx)
+		return NULL;
+
+	err = clk_prepare_enable(vdoa->vdoa_clk);
+	if (err) {
+		kfree(ctx);
+		return NULL;
+	}
+
+	init_completion(&ctx->completion);
+	ctx->vdoa = vdoa;
+
+	return ctx;
+}
+EXPORT_SYMBOL(vdoa_context_create);
+
+void vdoa_context_destroy(struct vdoa_ctx *ctx)
+{
+	struct vdoa_data *vdoa = ctx->vdoa;
+
+	if (vdoa->curr_ctx == ctx) {
+		vdoa_wait_for_completion(vdoa->curr_ctx);
+		vdoa->curr_ctx = NULL;
+	}
+
+	clk_disable_unprepare(vdoa->vdoa_clk);
+	kfree(ctx);
+}
+EXPORT_SYMBOL(vdoa_context_destroy);
+
+int vdoa_context_configure(struct vdoa_ctx *ctx,
+			   unsigned int width, unsigned int height,
+			   u32 pixelformat)
+{
+	struct vdoa_q_data *src_q_data;
+	struct vdoa_q_data *dst_q_data;
+
+	if (width < 16 || width  > 8192 || width % 16 != 0 ||
+	    height < 16 || height > 4096 || height % 16 != 0)
+		return -EINVAL;
+
+	if (pixelformat != V4L2_PIX_FMT_YUYV &&
+	    pixelformat != V4L2_PIX_FMT_NV12)
+		return -EINVAL;
+
+	/* If no context is passed, only check if the format is valid */
+	if (!ctx)
+		return 0;
+
+	src_q_data = &ctx->q_data[V4L2_M2M_SRC];
+	dst_q_data = &ctx->q_data[V4L2_M2M_DST];
+
+	src_q_data->width = width;
+	src_q_data->height = height;
+	src_q_data->bytesperline = width;
+	src_q_data->sizeimage =
+		round_up(src_q_data->bytesperline * height, 4096) +
+		src_q_data->bytesperline * height / 2;
+
+	dst_q_data->width = width;
+	dst_q_data->height = height;
+	dst_q_data->pixelformat = pixelformat;
+	switch (pixelformat) {
+	case V4L2_PIX_FMT_YUYV:
+		dst_q_data->bytesperline = width * 2;
+		dst_q_data->sizeimage = dst_q_data->bytesperline * height;
+		break;
+	case V4L2_PIX_FMT_NV12:
+	default:
+		dst_q_data->bytesperline = width;
+		dst_q_data->sizeimage =
+			dst_q_data->bytesperline * height * 3 / 2;
+		break;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL(vdoa_context_configure);
+
+static int vdoa_probe(struct platform_device *pdev)
+{
+	struct vdoa_data *vdoa;
+	int ret;
+
+	ret = dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(32));
+	if (ret) {
+		dev_err(&pdev->dev, "DMA enable failed\n");
+		return ret;
+	}
+
+	vdoa = devm_kzalloc(&pdev->dev, sizeof(*vdoa), GFP_KERNEL);
+	if (!vdoa)
+		return -ENOMEM;
+
+	vdoa->dev = &pdev->dev;
+
+	vdoa->vdoa_clk = devm_clk_get(vdoa->dev, NULL);
+	if (IS_ERR(vdoa->vdoa_clk)) {
+		dev_err(vdoa->dev, "Failed to get clock\n");
+		return PTR_ERR(vdoa->vdoa_clk);
+	}
+
+	vdoa->regs = devm_platform_ioremap_resource(pdev, 0);
+	if (IS_ERR(vdoa->regs))
+		return PTR_ERR(vdoa->regs);
+
+	ret = platform_get_irq(pdev, 0);
+	if (ret < 0)
+		return ret;
+	ret = devm_request_threaded_irq(&pdev->dev, ret, NULL,
+					vdoa_irq_handler, IRQF_ONESHOT,
+					"vdoa", vdoa);
+	if (ret < 0) {
+		dev_err(vdoa->dev, "Failed to get irq\n");
+		return ret;
+	}
+
+	platform_set_drvdata(pdev, vdoa);
+
+	return 0;
+}
+
+static int vdoa_remove(struct platform_device *pdev)
+{
+	return 0;
+}
+
+static const struct of_device_id vdoa_dt_ids[] = {
+	{ .compatible = "fsl,imx6q-vdoa" },
+	{}
+};
+MODULE_DEVICE_TABLE(of, vdoa_dt_ids);
+
+static struct platform_driver vdoa_driver = {
+	.probe		= vdoa_probe,
+	.remove		= vdoa_remove,
+	.driver		= {
+		.name	= VDOA_NAME,
+		.of_match_table = vdoa_dt_ids,
+	},
+};
+
+module_platform_driver(vdoa_driver);
+
+MODULE_DESCRIPTION("Video Data Order Adapter");
+MODULE_AUTHOR("Philipp Zabel <philipp.zabel@gmail.com>");
+MODULE_ALIAS("platform:imx-vdoa");
+MODULE_LICENSE("GPL");
diff --git a/drivers/media/platform/chips-media/imx-vdoa.h b/drivers/media/platform/chips-media/imx-vdoa.h
new file mode 100644
index 000000000..a62eab476
--- /dev/null
+++ b/drivers/media/platform/chips-media/imx-vdoa.h
@@ -0,0 +1,50 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * Copyright (C) 2016 Pengutronix
+ */
+
+#ifndef IMX_VDOA_H
+#define IMX_VDOA_H
+
+struct vdoa_data;
+struct vdoa_ctx;
+
+#if (defined CONFIG_VIDEO_IMX_VDOA || defined CONFIG_VIDEO_IMX_VDOA_MODULE)
+
+struct vdoa_ctx *vdoa_context_create(struct vdoa_data *vdoa);
+int vdoa_context_configure(struct vdoa_ctx *ctx,
+			   unsigned int width, unsigned int height,
+			   u32 pixelformat);
+void vdoa_context_destroy(struct vdoa_ctx *ctx);
+
+void vdoa_device_run(struct vdoa_ctx *ctx, dma_addr_t dst, dma_addr_t src);
+int vdoa_wait_for_completion(struct vdoa_ctx *ctx);
+
+#else
+
+static inline struct vdoa_ctx *vdoa_context_create(struct vdoa_data *vdoa)
+{
+	return NULL;
+}
+
+static inline int vdoa_context_configure(struct vdoa_ctx *ctx,
+					 unsigned int width,
+					 unsigned int height,
+					 u32 pixelformat)
+{
+	return 0;
+}
+
+static inline void vdoa_context_destroy(struct vdoa_ctx *ctx) { };
+
+static inline void vdoa_device_run(struct vdoa_ctx *ctx,
+				   dma_addr_t dst, dma_addr_t src) { };
+
+static inline int vdoa_wait_for_completion(struct vdoa_ctx *ctx)
+{
+	return 0;
+};
+
+#endif
+
+#endif /* IMX_VDOA_H */
diff --git a/drivers/media/platform/chips-media/trace.h b/drivers/media/platform/chips-media/trace.h
new file mode 100644
index 000000000..19f98e6da
--- /dev/null
+++ b/drivers/media/platform/chips-media/trace.h
@@ -0,0 +1,175 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM coda
+
+#if !defined(__CODA_TRACE_H__) || defined(TRACE_HEADER_MULTI_READ)
+#define __CODA_TRACE_H__
+
+#include <linux/tracepoint.h>
+#include <media/videobuf2-v4l2.h>
+
+#include "coda.h"
+
+TRACE_EVENT(coda_bit_run,
+	TP_PROTO(struct coda_ctx *ctx, int cmd),
+
+	TP_ARGS(ctx, cmd),
+
+	TP_STRUCT__entry(
+		__field(int, minor)
+		__field(int, ctx)
+		__field(int, cmd)
+	),
+
+	TP_fast_assign(
+		__entry->minor = ctx->fh.vdev->minor;
+		__entry->ctx = ctx->idx;
+		__entry->cmd = cmd;
+	),
+
+	TP_printk("minor = %d, ctx = %d, cmd = %d",
+		  __entry->minor, __entry->ctx, __entry->cmd)
+);
+
+TRACE_EVENT(coda_bit_done,
+	TP_PROTO(struct coda_ctx *ctx),
+
+	TP_ARGS(ctx),
+
+	TP_STRUCT__entry(
+		__field(int, minor)
+		__field(int, ctx)
+	),
+
+	TP_fast_assign(
+		__entry->minor = ctx->fh.vdev->minor;
+		__entry->ctx = ctx->idx;
+	),
+
+	TP_printk("minor = %d, ctx = %d", __entry->minor, __entry->ctx)
+);
+
+DECLARE_EVENT_CLASS(coda_buf_class,
+	TP_PROTO(struct coda_ctx *ctx, struct vb2_v4l2_buffer *buf),
+
+	TP_ARGS(ctx, buf),
+
+	TP_STRUCT__entry(
+		__field(int, minor)
+		__field(int, index)
+		__field(int, ctx)
+	),
+
+	TP_fast_assign(
+		__entry->minor = ctx->fh.vdev->minor;
+		__entry->index = buf->vb2_buf.index;
+		__entry->ctx = ctx->idx;
+	),
+
+	TP_printk("minor = %d, index = %d, ctx = %d",
+		  __entry->minor, __entry->index, __entry->ctx)
+);
+
+DEFINE_EVENT(coda_buf_class, coda_enc_pic_run,
+	TP_PROTO(struct coda_ctx *ctx, struct vb2_v4l2_buffer *buf),
+	TP_ARGS(ctx, buf)
+);
+
+DEFINE_EVENT(coda_buf_class, coda_enc_pic_done,
+	TP_PROTO(struct coda_ctx *ctx, struct vb2_v4l2_buffer *buf),
+	TP_ARGS(ctx, buf)
+);
+
+DECLARE_EVENT_CLASS(coda_buf_meta_class,
+	TP_PROTO(struct coda_ctx *ctx, struct vb2_v4l2_buffer *buf,
+		 struct coda_buffer_meta *meta),
+
+	TP_ARGS(ctx, buf, meta),
+
+	TP_STRUCT__entry(
+		__field(int, minor)
+		__field(int, index)
+		__field(int, start)
+		__field(int, end)
+		__field(int, ctx)
+	),
+
+	TP_fast_assign(
+		__entry->minor = ctx->fh.vdev->minor;
+		__entry->index = buf->vb2_buf.index;
+		__entry->start = meta->start & ctx->bitstream_fifo.kfifo.mask;
+		__entry->end = meta->end & ctx->bitstream_fifo.kfifo.mask;
+		__entry->ctx = ctx->idx;
+	),
+
+	TP_printk("minor = %d, index = %d, start = 0x%x, end = 0x%x, ctx = %d",
+		  __entry->minor, __entry->index, __entry->start, __entry->end,
+		  __entry->ctx)
+);
+
+DEFINE_EVENT(coda_buf_meta_class, coda_bit_queue,
+	TP_PROTO(struct coda_ctx *ctx, struct vb2_v4l2_buffer *buf,
+		 struct coda_buffer_meta *meta),
+	TP_ARGS(ctx, buf, meta)
+);
+
+DECLARE_EVENT_CLASS(coda_meta_class,
+	TP_PROTO(struct coda_ctx *ctx, struct coda_buffer_meta *meta),
+
+	TP_ARGS(ctx, meta),
+
+	TP_STRUCT__entry(
+		__field(int, minor)
+		__field(int, start)
+		__field(int, end)
+		__field(int, ctx)
+	),
+
+	TP_fast_assign(
+		__entry->minor = ctx->fh.vdev->minor;
+		__entry->start = meta ? (meta->start &
+					 ctx->bitstream_fifo.kfifo.mask) : 0;
+		__entry->end = meta ? (meta->end &
+				       ctx->bitstream_fifo.kfifo.mask) : 0;
+		__entry->ctx = ctx->idx;
+	),
+
+	TP_printk("minor = %d, start = 0x%x, end = 0x%x, ctx = %d",
+		  __entry->minor, __entry->start, __entry->end, __entry->ctx)
+);
+
+DEFINE_EVENT(coda_meta_class, coda_dec_pic_run,
+	TP_PROTO(struct coda_ctx *ctx, struct coda_buffer_meta *meta),
+	TP_ARGS(ctx, meta)
+);
+
+DEFINE_EVENT(coda_meta_class, coda_dec_pic_done,
+	TP_PROTO(struct coda_ctx *ctx, struct coda_buffer_meta *meta),
+	TP_ARGS(ctx, meta)
+);
+
+DEFINE_EVENT(coda_buf_meta_class, coda_dec_rot_done,
+	TP_PROTO(struct coda_ctx *ctx, struct vb2_v4l2_buffer *buf,
+		 struct coda_buffer_meta *meta),
+	TP_ARGS(ctx, buf, meta)
+);
+
+DEFINE_EVENT(coda_buf_class, coda_jpeg_run,
+	TP_PROTO(struct coda_ctx *ctx, struct vb2_v4l2_buffer *buf),
+	TP_ARGS(ctx, buf)
+);
+
+DEFINE_EVENT(coda_buf_class, coda_jpeg_done,
+	TP_PROTO(struct coda_ctx *ctx, struct vb2_v4l2_buffer *buf),
+	TP_ARGS(ctx, buf)
+);
+
+#endif /* __CODA_TRACE_H__ */
+
+#undef TRACE_INCLUDE_PATH
+#define TRACE_INCLUDE_PATH ../../drivers/media/platform/chips-media
+#undef TRACE_INCLUDE_FILE
+#define TRACE_INCLUDE_FILE trace
+
+/* This part must be outside protection */
+#include <trace/define_trace.h>