radeon/uvd: add UVD implementation v5

Just everything you need for UVD with r600g and radeonsi.

v2: move UVD code to radeon subdir, clean up build system additions,
    remove an unused SI function, disable tiling on SI for now.
v3: some minor indentation fix and rebased
v4: dpb size calculation fixed
v5: implement proper fall-back in case the kernel doesn't support UVD,
    based on patches from Andreas Boll but cleaned up a bit more.

Signed-off-by: Christian König <[email protected]>
Reviewed-by: Alex Deucher <[email protected]>

1 files changed, 1111 insertions, 0 deletions

diff --git a/src/gallium/drivers/radeon/radeon_uvd.c b/src/gallium/drivers/radeon/radeon_uvd.c
new file mode 100644
index 00000000000..dae78803e7d
--- /dev/null
+++ b/src/gallium/drivers/radeon/radeon_uvd.c
@@ -0,0 +1,1111 @@
+/**************************************************************************
+ *
+ * Copyright 2011 Advanced Micro Devices, Inc.
+ * All Rights Reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT.
+ * IN NO EVENT SHALL THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR
+ * ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+ * TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+ * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ **************************************************************************/
+
+/*
+ * Authors:
+ *	Christian König <[email protected]>
+ *
+ */
+
+#include <sys/types.h>
+#include <assert.h>
+#include <errno.h>
+#include <unistd.h>
+#include <stdio.h>
+
+#include "pipe/p_video_decoder.h"
+
+#include "util/u_memory.h"
+#include "util/u_video.h"
+
+#include "vl/vl_defines.h"
+#include "vl/vl_mpeg12_decoder.h"
+
+#include "../../winsys/radeon/drm/radeon_winsys.h"
+#include "radeon_uvd.h"
+
+#define RUVD_ERR(fmt, args...) \
+	fprintf(stderr, "EE %s:%d %s UVD - "fmt, __FILE__, __LINE__, __func__, ##args)
+
+#define NUM_BUFFERS 4
+
+#define NUM_MPEG2_REFS 6
+#define NUM_H264_REFS 17
+
+/* UVD buffer representation */
+struct ruvd_buffer
+{
+	struct pb_buffer*		buf;
+	struct radeon_winsys_cs_handle*	cs_handle;
+};
+
+/* UVD decoder representation */
+struct ruvd_decoder {
+	struct pipe_video_decoder	base;
+
+	ruvd_set_dtb			set_dtb;
+
+	unsigned			stream_handle;
+	unsigned			frame_number;
+
+	struct radeon_winsys*		ws;
+	struct radeon_winsys_cs*	cs;
+
+	unsigned			cur_buffer;
+
+	struct ruvd_buffer		msg_fb_buffers[NUM_BUFFERS];
+	struct ruvd_buffer		bs_buffers[NUM_BUFFERS];
+	void*				bs_ptr;
+	unsigned			bs_size;
+
+	struct ruvd_buffer		dpb;
+};
+
+/* generate an UVD stream handle */
+static unsigned alloc_stream_handle()
+{
+	static unsigned counter = 0;
+	unsigned stream_handle = 0;
+	unsigned pid = getpid();
+	int i;
+
+	for (i = 0; i < 32; ++i)
+		stream_handle |= ((pid >> i) & 1) << (31 - i);
+
+	stream_handle ^= ++counter;
+	return stream_handle;
+}
+
+/* flush IB to the hardware */
+static void flush(struct ruvd_decoder *dec)
+{
+	uint32_t *pm4 =	dec->cs->buf;
+
+	// align IB
+	while(dec->cs->cdw % 16)
+		pm4[dec->cs->cdw++] = RUVD_PKT2();
+
+	dec->ws->cs_flush(dec->cs, 0);
+}
+
+/* add a new set register command to the IB */
+static void set_reg(struct ruvd_decoder *dec, unsigned reg, uint32_t val)
+{
+	uint32_t *pm4 =	dec->cs->buf;
+	pm4[dec->cs->cdw++] = RUVD_PKT0(reg >> 2, 0);
+	pm4[dec->cs->cdw++] = val;
+}
+
+/* send a command to the VCPU through the GPCOM registers */
+static void send_cmd(struct ruvd_decoder *dec, unsigned cmd,
+		     struct radeon_winsys_cs_handle* cs_buf, uint32_t off,
+		     enum radeon_bo_usage usage, enum radeon_bo_domain domain)
+{
+	int reloc_idx;
+
+	reloc_idx = dec->ws->cs_add_reloc(dec->cs, cs_buf, usage, domain);
+	set_reg(dec, RUVD_GPCOM_VCPU_DATA0, off);
+	set_reg(dec, RUVD_GPCOM_VCPU_DATA1, reloc_idx * 4);
+	set_reg(dec, RUVD_GPCOM_VCPU_CMD, cmd << 1);
+}
+
+/* send a message command to the VCPU */
+static void send_msg(struct ruvd_decoder *dec, struct ruvd_msg *msg)
+{
+	struct ruvd_buffer* buf;
+	void *ptr;
+
+	/* grap a message buffer */
+	buf = &dec->msg_fb_buffers[dec->cur_buffer];
+
+	/* copy the message into it */
+	ptr = dec->ws->buffer_map(buf->cs_handle, dec->cs, PIPE_TRANSFER_WRITE);
+	if (!ptr)
+		return;
+
+	memcpy(ptr, msg, sizeof(*msg));
+	memset(ptr + sizeof(*msg), 0, buf->buf->size - sizeof(*msg));
+	dec->ws->buffer_unmap(buf->cs_handle);
+
+	/* and send it to the hardware */
+	send_cmd(dec, RUVD_CMD_MSG_BUFFER, buf->cs_handle, 0,
+		 RADEON_USAGE_READ, RADEON_DOMAIN_VRAM);
+}
+
+/* create a buffer in the winsys */
+static bool create_buffer(struct ruvd_decoder *dec,
+			  struct ruvd_buffer *buffer,
+			  unsigned size)
+{
+	buffer->buf = dec->ws->buffer_create(dec->ws, size, 4096, false,
+					     RADEON_DOMAIN_GTT | RADEON_DOMAIN_VRAM);
+	if (!buffer->buf)
+		return false;
+
+	buffer->cs_handle = dec->ws->buffer_get_cs_handle(buffer->buf);
+	if (!buffer->cs_handle)
+		return false;
+
+	return true;
+}
+
+/* destroy a buffer */
+static void destroy_buffer(struct ruvd_buffer *buffer)
+{
+	pb_reference(&buffer->buf, NULL);
+	buffer->cs_handle = NULL;
+}
+
+/* reallocate a buffer, preserving its content */
+static bool resize_buffer(struct ruvd_decoder *dec,
+			  struct ruvd_buffer *new_buf,
+			  unsigned new_size)
+{
+	unsigned bytes = MIN2(new_buf->buf->size, new_size);
+	struct ruvd_buffer old_buf = *new_buf;
+	void *src = NULL, *dst = NULL;
+
+	if (!create_buffer(dec, new_buf, new_size))
+		goto error;
+
+	src = dec->ws->buffer_map(old_buf.cs_handle, dec->cs, PIPE_TRANSFER_READ);
+	if (!src)
+		goto error;
+
+	dst = dec->ws->buffer_map(new_buf->cs_handle, dec->cs, PIPE_TRANSFER_WRITE);
+	if (!dst)
+		goto error;
+
+	memcpy(dst, src, bytes);
+	if (new_size > bytes) {
+		new_size -= bytes;
+		dst += bytes;
+		memset(dst, 0, new_size);
+	}
+	dec->ws->buffer_unmap(new_buf->cs_handle);
+	dec->ws->buffer_unmap(old_buf.cs_handle);
+	destroy_buffer(&old_buf);
+	return true;
+
+error:
+	if (src) dec->ws->buffer_unmap(old_buf.cs_handle);
+	destroy_buffer(new_buf);
+	*new_buf = old_buf;
+	return false;
+}
+
+/* clear the buffer with zeros */
+static void clear_buffer(struct ruvd_decoder *dec,
+			 struct ruvd_buffer* buffer)
+{
+	//TODO: let the GPU do the job
+	void *ptr = dec->ws->buffer_map(buffer->cs_handle, dec->cs,
+					PIPE_TRANSFER_WRITE);
+	if (!ptr)
+		return;
+
+	memset(ptr, 0, buffer->buf->size);
+	dec->ws->buffer_unmap(buffer->cs_handle);
+}
+
+/* cycle to the next set of buffers */
+static void next_buffer(struct ruvd_decoder *dec)
+{
+	++dec->cur_buffer;
+	dec->cur_buffer %= NUM_BUFFERS;
+}
+
+/* convert the profile into something UVD understands */
+static uint32_t profile2stream_type(enum pipe_video_profile profile)
+{
+	switch (u_reduce_video_profile(profile)) {
+	case PIPE_VIDEO_CODEC_MPEG4_AVC:
+		return RUVD_CODEC_H264;
+
+	case PIPE_VIDEO_CODEC_VC1:
+		return RUVD_CODEC_VC1;
+
+	case PIPE_VIDEO_CODEC_MPEG12:
+		return RUVD_CODEC_MPEG2;
+
+	case PIPE_VIDEO_CODEC_MPEG4:
+		return RUVD_CODEC_MPEG4;
+
+	default:
+		assert(0);
+		return 0;
+	}
+}
+
+/* calculate size of reference picture buffer */
+static unsigned calc_dpb_size(enum pipe_video_profile profile,
+			      unsigned width, unsigned height,
+			      unsigned max_references)
+{
+	unsigned width_in_mb, height_in_mb, image_size, dpb_size;
+
+	// always align them to MB size for dpb calculation
+	width = align(width, VL_MACROBLOCK_WIDTH);
+	height = align(height, VL_MACROBLOCK_HEIGHT);
+
+	// always one more for currently decoded picture
+	max_references += 1;
+
+	// aligned size of a single frame
+	image_size = width * height;
+	image_size += image_size / 2;
+	image_size = align(image_size, 1024);
+
+	// picture width & height in 16 pixel units
+	width_in_mb = width / VL_MACROBLOCK_WIDTH;
+	height_in_mb = align(height / VL_MACROBLOCK_HEIGHT, 2);
+
+	switch (u_reduce_video_profile(profile)) {
+	case PIPE_VIDEO_CODEC_MPEG4_AVC:
+		// the firmware seems to allways assume a minimum of ref frames
+		max_references = MAX2(NUM_H264_REFS, max_references);
+
+		// reference picture buffer
+		dpb_size = image_size * max_references;
+
+		// macroblock context buffer
+		dpb_size += width_in_mb * height_in_mb * max_references * 192;
+
+		// IT surface buffer
+		dpb_size += width_in_mb * height_in_mb * 32;
+		break;
+
+	case PIPE_VIDEO_CODEC_VC1:
+		// reference picture buffer
+		dpb_size = image_size * max_references;
+
+		// CONTEXT_BUFFER
+		dpb_size += width_in_mb * height_in_mb * 128;
+
+		// IT surface buffer
+		dpb_size += width_in_mb * 64;
+
+		// DB surface buffer
+		dpb_size += width_in_mb * 128;
+
+		// BP
+		dpb_size += align(MAX2(width_in_mb, height_in_mb) * 7 * 16, 64);
+		break;
+
+	case PIPE_VIDEO_CODEC_MPEG12:
+		// reference picture buffer, must be big enough for all frames
+		dpb_size = image_size * NUM_MPEG2_REFS;
+		break;
+
+	case PIPE_VIDEO_CODEC_MPEG4:
+		// reference picture buffer
+		dpb_size = image_size * max_references;
+
+		// CM
+		dpb_size += width_in_mb * height_in_mb * 64;
+
+		// IT surface buffer
+		dpb_size += align(width_in_mb * height_in_mb * 32, 64);
+		break;
+
+	default:
+		// something is missing here
+		assert(0);
+
+		// at least use a sane default value
+		dpb_size = 32 * 1024 * 1024;
+		break;
+	}
+	return dpb_size;
+}
+
+/* get h264 specific message bits */
+static struct ruvd_h264 get_h264_msg(struct ruvd_decoder *dec, struct pipe_h264_picture_desc *pic)
+{
+	struct ruvd_h264 result;
+
+	memset(&result, 0, sizeof(result));
+	switch (pic->base.profile) {
+	case PIPE_VIDEO_PROFILE_MPEG4_AVC_BASELINE:
+		result.profile = RUVD_H264_PROFILE_BASELINE;
+		break;
+
+	case PIPE_VIDEO_PROFILE_MPEG4_AVC_MAIN:
+		result.profile = RUVD_H264_PROFILE_MAIN;
+		break;
+
+	case PIPE_VIDEO_PROFILE_MPEG4_AVC_HIGH:
+		result.profile = RUVD_H264_PROFILE_HIGH;
+		break;
+
+	default:
+		assert(0);
+		break;
+	}
+	if (((dec->base.width * dec->base.height) >> 8) <= 1620)
+		result.level = 30;
+	else
+		result.level = 41;
+
+	result.sps_info_flags = 0;
+	result.sps_info_flags |= pic->direct_8x8_inference_flag << 0;
+	result.sps_info_flags |= pic->mb_adaptive_frame_field_flag << 1;
+	result.sps_info_flags |= pic->frame_mbs_only_flag << 2;
+	result.sps_info_flags |= pic->delta_pic_order_always_zero_flag << 3;
+
+	result.pps_info_flags = 0;
+	result.pps_info_flags |= pic->transform_8x8_mode_flag << 0;
+	result.pps_info_flags |= pic->redundant_pic_cnt_present_flag << 1;
+	result.pps_info_flags |= pic->constrained_intra_pred_flag << 2;
+	result.pps_info_flags |= pic->deblocking_filter_control_present_flag << 3;
+	result.pps_info_flags |= pic->weighted_bipred_idc << 4;
+	result.pps_info_flags |= pic->weighted_pred_flag << 6;
+	result.pps_info_flags |= pic->pic_order_present_flag << 7;
+	result.pps_info_flags |= pic->entropy_coding_mode_flag << 8;
+
+	result.chroma_format = 0x1;
+	result.bit_depth_luma_minus8 = 0;
+	result.bit_depth_chroma_minus8 = 0;
+
+	result.log2_max_frame_num_minus4 = pic->log2_max_frame_num_minus4;
+	result.pic_order_cnt_type = pic->pic_order_cnt_type;
+	result.log2_max_pic_order_cnt_lsb_minus4 = pic->log2_max_pic_order_cnt_lsb_minus4;
+	result.num_ref_frames = pic->num_ref_frames;
+	result.pic_init_qp_minus26 = pic->pic_init_qp_minus26;
+	result.chroma_qp_index_offset = pic->chroma_qp_index_offset;
+	result.second_chroma_qp_index_offset = pic->second_chroma_qp_index_offset;
+
+	result.num_slice_groups_minus1 = 0;
+	result.slice_group_map_type = 0;
+
+	result.num_ref_idx_l0_active_minus1 = pic->num_ref_idx_l0_active_minus1;
+	result.num_ref_idx_l1_active_minus1 = pic->num_ref_idx_l1_active_minus1;
+
+	result.slice_group_change_rate_minus1 = 0;
+
+	memcpy(result.scaling_list_4x4, pic->scaling_lists_4x4, 6*64);
+	memcpy(result.scaling_list_8x8, pic->scaling_lists_8x8, 2*64);
+
+	result.frame_num = pic->frame_num;
+	memcpy(result.frame_num_list, pic->frame_num_list, 4*16);
+	result.curr_field_order_cnt_list[0] = pic->field_order_cnt[0];
+	result.curr_field_order_cnt_list[1] = pic->field_order_cnt[1];
+	memcpy(result.field_order_cnt_list, pic->field_order_cnt_list, 4*16*2);
+
+	result.decoded_pic_idx = pic->frame_num;
+
+	return result;
+}
+
+/* get vc1 specific message bits */
+static struct ruvd_vc1 get_vc1_msg(struct pipe_vc1_picture_desc *pic)
+{
+	struct ruvd_vc1 result;
+
+	memset(&result, 0, sizeof(result));
+	switch(pic->base.profile) {
+	case PIPE_VIDEO_PROFILE_VC1_SIMPLE:
+		result.profile = RUVD_VC1_PROFILE_SIMPLE;
+		break;
+
+	case PIPE_VIDEO_PROFILE_VC1_MAIN:
+		result.profile = RUVD_VC1_PROFILE_MAIN;
+		break;
+		
+	case PIPE_VIDEO_PROFILE_VC1_ADVANCED:
+		result.profile = RUVD_VC1_PROFILE_ADVANCED;
+		break;
+	default:
+		assert(0);
+	}
+
+	if (pic->base.profile == PIPE_VIDEO_PROFILE_VC1_ADVANCED) {
+		result.level = 0;
+
+		result.sps_info_flags |= pic->postprocflag << 7;
+		result.sps_info_flags |= pic->pulldown << 6;
+		result.sps_info_flags |= pic->interlace << 5;
+		result.sps_info_flags |= pic->tfcntrflag << 4;
+		result.sps_info_flags |= pic->psf << 1;
+
+		result.pps_info_flags |= pic->panscan_flag << 7;
+		result.pps_info_flags |= pic->refdist_flag << 6;
+		result.pps_info_flags |= pic->extended_dmv << 8;
+		result.pps_info_flags |= pic->range_mapy_flag << 31;
+		result.pps_info_flags |= pic->range_mapy << 28;
+		result.pps_info_flags |= pic->range_mapuv_flag << 27;
+		result.pps_info_flags |= pic->range_mapuv << 24;
+
+	} else {
+		result.level = 0;
+		result.pps_info_flags |= pic->multires << 21;
+		result.pps_info_flags |= pic->syncmarker << 20;
+		result.pps_info_flags |= pic->rangered << 19;
+		result.pps_info_flags |= pic->maxbframes << 16;
+	}
+
+	result.sps_info_flags |= pic->finterpflag << 3;
+	//(((unsigned int)(pPicParams->advance.reserved1))        << SPS_INFO_VC1_RESERVED_SHIFT)
+
+	result.pps_info_flags |= pic->loopfilter << 5;
+	result.pps_info_flags |= pic->fastuvmc << 4;
+	result.pps_info_flags |= pic->extended_mv << 3;
+	result.pps_info_flags |= pic->dquant << 1;
+	result.pps_info_flags |= pic->vstransform << 0;
+	result.pps_info_flags |= pic->overlap << 11;
+	result.pps_info_flags |= pic->quantizer << 9;
+
+
+#if 0
+uint32_t 	slice_count
+uint8_t 	picture_type
+uint8_t 	frame_coding_mode
+uint8_t 	deblockEnable
+uint8_t 	pquant
+#endif
+
+        result.chroma_format  = 1;
+	return result;
+}
+
+/* extract the frame number from a referenced video buffer */
+static uint32_t get_ref_pic_idx(struct ruvd_decoder *dec, struct pipe_video_buffer *ref)
+{
+	uint32_t min = dec->frame_number - NUM_MPEG2_REFS;
+	uint32_t max = dec->frame_number - 1;
+	uintptr_t frame;
+
+	/* seems to be the most sane fallback */
+	if (!ref)
+		return max;
+
+	/* get the frame number from the associated data */
+	frame = (uintptr_t)vl_video_buffer_get_associated_data(ref, &dec->base);
+
+	/* limit the frame number to a valid range */
+	return MAX2(MIN2(frame, max), min);
+}
+
+/* get mpeg2 specific msg bits */
+static struct ruvd_mpeg2 get_mpeg2_msg(struct ruvd_decoder *dec,
+				       struct pipe_mpeg12_picture_desc *pic)
+{
+	struct ruvd_mpeg2 result;
+	unsigned i;
+
+	memset(&result, 0, sizeof(result));
+	result.decoded_pic_idx = dec->frame_number;
+	for (i = 0; i < 2; ++i)
+		result.ref_pic_idx[i] = get_ref_pic_idx(dec, pic->ref[i]);
+
+	result.load_intra_quantiser_matrix = 1;
+	result.load_nonintra_quantiser_matrix = 1;
+	memcpy(&result.intra_quantiser_matrix, pic->intra_matrix, 64);
+	memcpy(&result.nonintra_quantiser_matrix, pic->non_intra_matrix, 64);
+
+	result.profile_and_level_indication = 0;
+	result.chroma_format = 0x1;
+
+	result.picture_coding_type = pic->picture_coding_type;
+	result.f_code[0][0] = pic->f_code[0][0] + 1;
+	result.f_code[0][1] = pic->f_code[0][1] + 1;
+	result.f_code[1][0] = pic->f_code[1][0] + 1;
+	result.f_code[1][1] = pic->f_code[1][1] + 1;
+	result.intra_dc_precision = pic->intra_dc_precision;
+	result.pic_structure = pic->picture_structure;
+	result.top_field_first = pic->top_field_first;
+	result.frame_pred_frame_dct = pic->frame_pred_frame_dct;
+	result.concealment_motion_vectors = pic->concealment_motion_vectors;
+	result.q_scale_type = pic->q_scale_type;
+	result.intra_vlc_format = pic->intra_vlc_format;
+	result.alternate_scan = pic->alternate_scan;
+
+	return result;
+}
+
+/* get mpeg4 specific msg bits */
+static struct ruvd_mpeg4 get_mpeg4_msg(struct ruvd_decoder *dec,
+				       struct pipe_mpeg4_picture_desc *pic)
+{
+	struct ruvd_mpeg4 result;
+	unsigned i;
+	memset(&result, 0, sizeof(result));
+	result.decoded_pic_idx = dec->frame_number;
+	for (i = 0; i < 2; ++i)
+		result.ref_pic_idx[i] = get_ref_pic_idx(dec, pic->ref[i]);
+
+	result.video_object_layer_width = dec->base.width;
+        result.video_object_layer_height = dec->base.height;
+
+	result.vop_time_increment_resolution = pic->vop_time_increment_resolution;
+	result.quant_type = pic->quant_type;
+
+	result.flags |= pic->short_video_header << 0;
+	//result.flags |= obmc_disable << 1;
+	result.flags |= pic->interlaced << 2;
+        result.flags |= 1 << 3; // load_intra_quant_mat
+	result.flags |= 1 << 4; // load_nonintra_quant_mat
+	result.flags |= pic->quarter_sample << 5;
+	//result.flags |= complexity_estimation_disable << 6
+	result.flags |= pic->resync_marker_disable << 7;
+	//result.flags |= data_partitioned << 8;
+	//result.flags |= reversible_vlc << 9;
+	//result.flags |= newpred_enable << 10;
+	//result.flags |= reduced_resolution_vop_enable << 11;
+	//result.flags |= scalability << 12;
+	//result.flags |= is_object_layer_identifier << 13;
+	//result.flags |= fixed_vop_rate << 14;
+	//result.flags |= newpred_segment_type << 15;
+
+	memcpy(&result.intra_quant_mat, pic->intra_matrix, 64);
+	memcpy(&result.nonintra_quant_mat, pic->non_intra_matrix, 64);
+
+	/*
+	int32_t 	trd [2]
+	int32_t 	trb [2]
+	uint8_t 	vop_coding_type
+	uint8_t 	vop_fcode_forward
+	uint8_t 	vop_fcode_backward
+	uint8_t 	rounding_control
+	uint8_t 	alternate_vertical_scan_flag
+	uint8_t 	top_field_first
+	*/
+
+	return result;
+}
+
+/**
+ * destroy this video decoder
+ */
+static void ruvd_destroy(struct pipe_video_decoder *decoder)
+{
+	struct ruvd_decoder *dec = (struct ruvd_decoder*)decoder;
+	struct ruvd_msg msg;
+	unsigned i;
+
+	assert(decoder);
+
+	memset(&msg, 0, sizeof(msg));
+	msg.size = sizeof(msg);
+	msg.msg_type = RUVD_MSG_DESTROY;
+	msg.stream_handle = dec->stream_handle;
+	send_msg(dec, &msg);
+
+	flush(dec);
+
+	dec->ws->cs_destroy(dec->cs);
+
+	for (i = 0; i < NUM_BUFFERS; ++i) {
+		destroy_buffer(&dec->msg_fb_buffers[i]);
+		destroy_buffer(&dec->bs_buffers[i]);
+	}
+
+	destroy_buffer(&dec->dpb);
+
+	FREE(dec);
+}
+
+/* free associated data in the video buffer callback */
+static void ruvd_destroy_associated_data(void *data)
+{
+	/* NOOP, since we only use an intptr */
+}
+
+/**
+ * start decoding of a new frame
+ */
+static void ruvd_begin_frame(struct pipe_video_decoder *decoder,
+			     struct pipe_video_buffer *target,
+			     struct pipe_picture_desc *picture)
+{
+	struct ruvd_decoder *dec = (struct ruvd_decoder*)decoder;
+	uintptr_t frame;
+
+	assert(decoder);
+
+	frame = ++dec->frame_number;
+	vl_video_buffer_set_associated_data(target, decoder, (void *)frame,
+					    &ruvd_destroy_associated_data);
+
+	dec->bs_size = 0;
+	dec->bs_ptr = dec->ws->buffer_map(
+		dec->bs_buffers[dec->cur_buffer].cs_handle,
+		dec->cs, PIPE_TRANSFER_WRITE);
+}
+
+/**
+ * decode a macroblock
+ */
+static void ruvd_decode_macroblock(struct pipe_video_decoder *decoder,
+				   struct pipe_video_buffer *target,
+				   struct pipe_picture_desc *picture,
+				   const struct pipe_macroblock *macroblocks,
+				   unsigned num_macroblocks)
+{
+	/* not supported (yet) */
+	assert(0);
+}
+
+/**
+ * decode a bitstream
+ */
+static void ruvd_decode_bitstream(struct pipe_video_decoder *decoder,
+				  struct pipe_video_buffer *target,
+				  struct pipe_picture_desc *picture,
+				  unsigned num_buffers,
+				  const void * const *buffers,
+				  const unsigned *sizes)
+{
+	struct ruvd_decoder *dec = (struct ruvd_decoder*)decoder;
+	unsigned i;
+
+	assert(decoder);
+
+	if (!dec->bs_ptr)
+		return;
+
+	for (i = 0; i < num_buffers; ++i) {
+		struct ruvd_buffer *buf = &dec->bs_buffers[dec->cur_buffer];
+		unsigned new_size = dec->bs_size + sizes[i];
+
+		if (new_size > buf->buf->size) {
+			dec->ws->buffer_unmap(buf->cs_handle);
+			if (!resize_buffer(dec, buf, new_size)) {
+				RUVD_ERR("Can't resize bitstream buffer!");
+				return;
+			}
+
+			dec->bs_ptr = dec->ws->buffer_map(buf->cs_handle, dec->cs,
+							  PIPE_TRANSFER_WRITE);
+			if (!dec->bs_ptr)
+				return;
+
+			dec->bs_ptr += dec->bs_size;
+		}
+
+		memcpy(dec->bs_ptr, buffers[i], sizes[i]);
+		dec->bs_size += sizes[i];
+		dec->bs_ptr += sizes[i];
+	}
+}
+
+/**
+ * end decoding of the current frame
+ */
+static void ruvd_end_frame(struct pipe_video_decoder *decoder,
+			   struct pipe_video_buffer *target,
+			   struct pipe_picture_desc *picture)
+{
+	struct ruvd_decoder *dec = (struct ruvd_decoder*)decoder;
+	struct radeon_winsys_cs_handle *dt;
+	struct ruvd_buffer *msg_fb_buf, *bs_buf;
+	struct ruvd_msg msg;
+	unsigned bs_size;
+
+	assert(decoder);
+
+	if (!dec->bs_ptr)
+		return;
+
+	msg_fb_buf = &dec->msg_fb_buffers[dec->cur_buffer];
+	bs_buf = &dec->bs_buffers[dec->cur_buffer];
+
+	bs_size = align(dec->bs_size, 128);
+	memset(dec->bs_ptr, 0, bs_size - dec->bs_size);
+	dec->ws->buffer_unmap(bs_buf->cs_handle);
+
+	memset(&msg, 0, sizeof(msg));
+	msg.size = sizeof(msg);
+	msg.msg_type = RUVD_MSG_DECODE;
+	msg.stream_handle = dec->stream_handle;
+	msg.status_report_feedback_number = dec->frame_number;
+
+	msg.decode.stream_type = profile2stream_type(dec->base.profile);
+	msg.decode.decode_flags = 0x1;
+	msg.decode.width_in_samples = dec->base.width;
+	msg.decode.height_in_samples = dec->base.height;
+
+	msg.decode.dpb_size = dec->dpb.buf->size;
+	msg.decode.bsd_size = bs_size;
+
+	dt = dec->set_dtb(&msg, (struct vl_video_buffer *)target);
+
+	switch (u_reduce_video_profile(picture->profile)) {
+	case PIPE_VIDEO_CODEC_MPEG4_AVC:
+		msg.decode.h264 = get_h264_msg(dec, (struct pipe_h264_picture_desc*)picture);
+		break;
+
+	case PIPE_VIDEO_CODEC_VC1:
+		msg.decode.vc1 = get_vc1_msg((struct pipe_vc1_picture_desc*)picture);
+		break;
+
+	case PIPE_VIDEO_CODEC_MPEG12:
+		msg.decode.mpeg2 = get_mpeg2_msg(dec, (struct pipe_mpeg12_picture_desc*)picture);
+		break;
+
+	case PIPE_VIDEO_CODEC_MPEG4:
+		msg.decode.mpeg4 = get_mpeg4_msg(dec, (struct pipe_mpeg4_picture_desc*)picture);
+		break;
+
+	default:
+		assert(0);
+		return;
+	}
+
+	msg.decode.db_surf_tile_config = msg.decode.dt_surf_tile_config;
+	msg.decode.extension_support = 0x1;
+
+	send_msg(dec, &msg);
+	send_cmd(dec, RUVD_CMD_DPB_BUFFER, dec->dpb.cs_handle, 0,
+		 RADEON_USAGE_READWRITE, RADEON_DOMAIN_VRAM);
+	send_cmd(dec, RUVD_CMD_BITSTREAM_BUFFER, bs_buf->cs_handle,
+		 0, RADEON_USAGE_READ, RADEON_DOMAIN_GTT);
+	send_cmd(dec, RUVD_CMD_DECODING_TARGET_BUFFER, dt, 0,
+		 RADEON_USAGE_WRITE, RADEON_DOMAIN_VRAM);
+	send_cmd(dec, RUVD_CMD_FEEDBACK_BUFFER, msg_fb_buf->cs_handle,
+		 0x1000, RADEON_USAGE_WRITE, RADEON_DOMAIN_VRAM);
+	set_reg(dec, RUVD_ENGINE_CNTL, 1);
+
+	flush(dec);
+	next_buffer(dec);
+}
+
+/**
+ * flush any outstanding command buffers to the hardware
+ */
+static void ruvd_flush(struct pipe_video_decoder *decoder)
+{
+}
+
+/**
+ * create and UVD decoder
+ */
+struct pipe_video_decoder *ruvd_create_decoder(struct pipe_context *context,
+					       enum pipe_video_profile profile,
+					       enum pipe_video_entrypoint entrypoint,
+					       enum pipe_video_chroma_format chroma_format,
+					       unsigned width, unsigned height,
+					       unsigned max_references, bool expect_chunked_decode,
+					       struct radeon_winsys* ws,
+					       ruvd_set_dtb set_dtb)
+{
+	unsigned dpb_size = calc_dpb_size(profile, width, height, max_references);
+	struct ruvd_decoder *dec;
+	struct ruvd_msg msg;
+	int i;
+
+	switch(u_reduce_video_profile(profile)) {
+	case PIPE_VIDEO_CODEC_MPEG12:
+		if (entrypoint > PIPE_VIDEO_ENTRYPOINT_BITSTREAM)
+			return vl_create_mpeg12_decoder(context, profile, entrypoint,
+							chroma_format, width,
+							height, max_references, expect_chunked_decode);
+
+		/* fall through */
+	case PIPE_VIDEO_CODEC_MPEG4:
+	case PIPE_VIDEO_CODEC_MPEG4_AVC:
+		width = align(width, VL_MACROBLOCK_WIDTH);
+		height = align(height, VL_MACROBLOCK_HEIGHT);
+		break;
+
+	default:
+		break;
+	}
+
+
+	dec = CALLOC_STRUCT(ruvd_decoder);
+
+	if (!dec)
+		return NULL;
+
+	dec->base.context = context;
+	dec->base.profile = profile;
+	dec->base.entrypoint = entrypoint;
+	dec->base.chroma_format = chroma_format;
+	dec->base.width = width;
+	dec->base.height = height;
+
+	dec->base.destroy = ruvd_destroy;
+	dec->base.begin_frame = ruvd_begin_frame;
+	dec->base.decode_macroblock = ruvd_decode_macroblock;
+	dec->base.decode_bitstream = ruvd_decode_bitstream;
+	dec->base.end_frame = ruvd_end_frame;
+	dec->base.flush = ruvd_flush;
+
+	dec->set_dtb = set_dtb;
+	dec->stream_handle = alloc_stream_handle();
+	dec->ws = ws;
+	dec->cs = ws->cs_create(ws, RING_UVD);
+	if (!dec->cs) {
+		RUVD_ERR("Can't get command submission context.\n");
+		goto error;
+	}
+
+	for (i = 0; i < NUM_BUFFERS; ++i) {
+		unsigned msg_fb_size = align(sizeof(struct ruvd_msg), 0x1000) + 0x1000;
+		if (!create_buffer(dec, &dec->msg_fb_buffers[i], msg_fb_size)) {
+			RUVD_ERR("Can't allocated message buffers.\n");
+			goto error;
+		}
+
+		if (!create_buffer(dec, &dec->bs_buffers[i], 4096)) {
+			RUVD_ERR("Can't allocated bitstream buffers.\n");
+			goto error;
+		}
+
+		clear_buffer(dec, &dec->msg_fb_buffers[i]);
+		clear_buffer(dec, &dec->bs_buffers[i]);
+	}
+
+	if (!create_buffer(dec, &dec->dpb, dpb_size)) {
+		RUVD_ERR("Can't allocated dpb.\n");
+		goto error;
+	}
+
+	clear_buffer(dec, &dec->dpb);
+
+	memset(&msg, 0, sizeof(msg));
+	msg.size = sizeof(msg);
+	msg.msg_type = RUVD_MSG_CREATE;
+	msg.stream_handle = dec->stream_handle;
+	msg.create.stream_type = profile2stream_type(dec->base.profile);
+	msg.create.width_in_samples = dec->base.width;
+	msg.create.height_in_samples = dec->base.height;
+	msg.create.dpb_size = dec->dpb.buf->size;
+	send_msg(dec, &msg);
+	flush(dec);
+	next_buffer(dec);
+
+	return &dec->base;
+
+error:
+	if (dec->cs) dec->ws->cs_destroy(dec->cs);
+
+	for (i = 0; i < NUM_BUFFERS; ++i) {
+		destroy_buffer(&dec->msg_fb_buffers[i]);
+		destroy_buffer(&dec->bs_buffers[i]);
+	}
+
+	destroy_buffer(&dec->dpb);
+
+	FREE(dec);
+
+	return NULL;
+}
+
+/**
+ * join surfaces into the same buffer with identical tiling params
+ * sumup their sizes and replace the backend buffers with a single bo
+ */
+void ruvd_join_surfaces(struct radeon_winsys* ws, unsigned bind,
+			struct pb_buffer** buffers[VL_NUM_COMPONENTS],
+			struct radeon_surface *surfaces[VL_NUM_COMPONENTS])
+{
+	unsigned best_tiling, best_wh, off;
+	unsigned size, alignment;
+	struct pb_buffer *pb;
+	unsigned i, j;
+
+	for (i = 0, best_tiling = 0, best_wh = ~0; i < VL_NUM_COMPONENTS; ++i) {
+		unsigned wh;
+
+		if (!surfaces[i])
+			continue;
+
+		/* choose the smallest bank w/h for now */
+		wh = surfaces[i]->bankw * surfaces[i]->bankh;
+		if (wh < best_wh) {
+			best_wh = wh;
+			best_tiling = i;
+		}
+	}
+
+	for (i = 0, off = 0; i < VL_NUM_COMPONENTS; ++i) {
+		if (!surfaces[i])
+			continue;
+
+		/* copy the tiling parameters */
+		surfaces[i]->bankw = surfaces[best_tiling]->bankw;
+		surfaces[i]->bankh = surfaces[best_tiling]->bankh;
+		surfaces[i]->mtilea = surfaces[best_tiling]->mtilea;
+		surfaces[i]->tile_split = surfaces[best_tiling]->tile_split;
+
+		/* adjust the texture layer offsets */
+		off = align(off, surfaces[i]->bo_alignment);
+		for (j = 0; j < Elements(surfaces[i]->level); ++j)
+			surfaces[i]->level[j].offset += off;
+		off += surfaces[i]->bo_size;
+	}
+
+	for (i = 0, size = 0, alignment = 0; i < VL_NUM_COMPONENTS; ++i) {
+		if (!buffers[i] || !*buffers[i])
+			continue;
+
+		size = align(size, (*buffers[i])->alignment);
+		size += (*buffers[i])->size;
+		alignment = MAX2(alignment, (*buffers[i])->alignment * 1);
+	}
+
+	if (!size)
+		return;
+
+	/* TODO: 2D tiling workaround */
+	alignment *= 2;
+
+	pb = ws->buffer_create(ws, size, alignment, bind, RADEON_DOMAIN_VRAM);
+	if (!pb)
+		return;
+
+	for (i = 0; i < VL_NUM_COMPONENTS; ++i) {
+		if (!buffers[i] || !*buffers[i])
+			continue;
+
+		pb_reference(buffers[i], pb);
+	}
+
+	pb_reference(&pb, NULL);
+}
+
+/* calculate top/bottom offset */
+static unsigned texture_offset(struct radeon_surface *surface, unsigned layer)
+{
+	return surface->level[0].offset +
+		layer * surface->level[0].slice_size;
+}
+
+/* hw encode the aspect of macro tiles */
+static unsigned macro_tile_aspect(unsigned macro_tile_aspect)
+{
+	switch (macro_tile_aspect) {
+	default:
+	case 1: macro_tile_aspect = 0;  break;
+	case 2: macro_tile_aspect = 1;  break;
+	case 4: macro_tile_aspect = 2;  break;
+	case 8: macro_tile_aspect = 3;  break;
+	}
+	return macro_tile_aspect;
+}
+
+/* hw encode the bank width and height */
+static unsigned bank_wh(unsigned bankwh)
+{
+	switch (bankwh) {
+	default:
+	case 1: bankwh = 0;     break;
+	case 2: bankwh = 1;     break;
+	case 4: bankwh = 2;     break;
+	case 8: bankwh = 3;     break;
+	}
+	return bankwh;
+}
+
+/**
+ * fill decoding target field from the luma and chroma surfaces
+ */
+void ruvd_set_dt_surfaces(struct ruvd_msg *msg, struct radeon_surface *luma,
+			  struct radeon_surface *chroma)
+{
+	msg->decode.dt_pitch = luma->level[0].pitch_bytes;
+	switch (luma->level[0].mode) {
+	case RADEON_SURF_MODE_LINEAR_ALIGNED:
+		msg->decode.dt_tiling_mode = RUVD_TILE_LINEAR;
+		msg->decode.dt_array_mode = RUVD_ARRAY_MODE_LINEAR;
+		break;
+	case RADEON_SURF_MODE_1D:
+		msg->decode.dt_tiling_mode = RUVD_TILE_8X8;
+		msg->decode.dt_array_mode = RUVD_ARRAY_MODE_1D_THIN;
+		break;
+	case RADEON_SURF_MODE_2D:
+		msg->decode.dt_tiling_mode = RUVD_TILE_8X8;
+		msg->decode.dt_array_mode = RUVD_ARRAY_MODE_2D_THIN;
+		break;
+	default:
+		assert(0);
+		break;
+	}
+
+	msg->decode.dt_luma_top_offset = texture_offset(luma, 0);
+	msg->decode.dt_chroma_top_offset = texture_offset(chroma, 0);
+	if (msg->decode.dt_field_mode) {
+		msg->decode.dt_luma_bottom_offset = texture_offset(luma, 1);
+		msg->decode.dt_chroma_bottom_offset = texture_offset(chroma, 1);
+	} else {
+		msg->decode.dt_luma_bottom_offset = msg->decode.dt_luma_top_offset;
+		msg->decode.dt_chroma_bottom_offset = msg->decode.dt_chroma_top_offset;
+	}
+
+	assert(luma->bankw == chroma->bankw);
+	assert(luma->bankh == chroma->bankh);
+	assert(luma->mtilea == chroma->mtilea);
+
+	msg->decode.dt_surf_tile_config |= RUVD_BANK_WIDTH(bank_wh(luma->bankw));
+	msg->decode.dt_surf_tile_config |= RUVD_BANK_HEIGHT(bank_wh(luma->bankh));
+	msg->decode.dt_surf_tile_config |= RUVD_MACRO_TILE_ASPECT_RATIO(macro_tile_aspect(luma->mtilea));
+}
+
+int ruvd_get_video_param(struct pipe_screen *screen,
+			 enum pipe_video_profile profile,
+			 enum pipe_video_cap param)
+{
+	switch (param) {
+	case PIPE_VIDEO_CAP_SUPPORTED:
+		switch (u_reduce_video_profile(profile)) {
+		case PIPE_VIDEO_CODEC_MPEG12:
+		case PIPE_VIDEO_CODEC_MPEG4:
+			/* TODO not all hw families support MPEG4 */
+		case PIPE_VIDEO_CODEC_MPEG4_AVC:
+		case PIPE_VIDEO_CODEC_VC1:
+			return true;
+		default:
+			return false;
+		}
+	case PIPE_VIDEO_CAP_NPOT_TEXTURES:
+		return 1;
+	case PIPE_VIDEO_CAP_MAX_WIDTH:
+		return 2048;
+	case PIPE_VIDEO_CAP_MAX_HEIGHT:
+		return 1152;
+	case PIPE_VIDEO_CAP_PREFERED_FORMAT:
+		return PIPE_FORMAT_NV12;
+	case PIPE_VIDEO_CAP_PREFERS_INTERLACED:
+		return false;
+	case PIPE_VIDEO_CAP_SUPPORTS_INTERLACED:
+		return false; /* TODO: enable this */
+	case PIPE_VIDEO_CAP_SUPPORTS_PROGRESSIVE:
+		return true;
+	default:
+		return 0;
+	}
+}
+
+boolean ruvd_is_format_supported(struct pipe_screen *screen,
+				 enum pipe_format format,
+				 enum pipe_video_profile profile)
+{
+	/* we can only handle this one anyway */
+	return format == PIPE_FORMAT_NV12;
+}