From 056194f77cd326c5db064daf90df3720b006f920 Mon Sep 17 00:00:00 2001 From: Detlev Casanova Date: Wed, 11 Jun 2025 17:04:28 -0400 Subject: [PATCH 079/113] DETLEV(v3): media: rkvdec: Add HEVC support for the VDPU383 variant The VDPU383 decoder is used on the RK3576 SoC and has support for HEVC. This patch also moves some functions to a common rkvdec-hevc-common.c file and adds a specific scaling matrix flatten function. The fluster score for JCT-VC-HEVC_V1 is 146/147. Signed-off-by: Detlev Casanova --- .../media/platform/rockchip/rkvdec/Makefile | 1 + .../rockchip/rkvdec/rkvdec-hevc-common.c | 49 +- .../rockchip/rkvdec/rkvdec-vdpu383-hevc.c | 680 ++++++++++++++++++ .../media/platform/rockchip/rkvdec/rkvdec.c | 94 ++- .../media/platform/rockchip/rkvdec/rkvdec.h | 2 + 5 files changed, 783 insertions(+), 43 deletions(-) create mode 100644 drivers/media/platform/rockchip/rkvdec/rkvdec-vdpu383-hevc.c diff --git a/drivers/media/platform/rockchip/rkvdec/Makefile b/drivers/media/platform/rockchip/rkvdec/Makefile index e30fdd7d51c3..e629d571e4d8 100644 --- a/drivers/media/platform/rockchip/rkvdec/Makefile +++ b/drivers/media/platform/rockchip/rkvdec/Makefile @@ -11,4 +11,5 @@ rockchip-vdec-y += \ rkvdec-vdpu381-h264.o \ rkvdec-vdpu381-hevc.o \ rkvdec-vdpu383-h264.o \ + rkvdec-vdpu383-hevc.o \ rkvdec-vp9.o diff --git a/drivers/media/platform/rockchip/rkvdec/rkvdec-hevc-common.c b/drivers/media/platform/rockchip/rkvdec/rkvdec-hevc-common.c index 3646b3ce4ea0..e5d240d97792 100644 --- a/drivers/media/platform/rockchip/rkvdec/rkvdec-hevc-common.c +++ b/drivers/media/platform/rockchip/rkvdec/rkvdec-hevc-common.c @@ -155,60 +155,25 @@ static void set_ref_poc(struct rkvdec_rps_short_term_ref_set *set, int poc, int } } -/* - * Flip one or more matrices along their main diagonal and flatten them - * before writing it to the memory. - * Convert: - * ABCD AEIM - * EFGH => BFJN => AEIMBFJNCGKODHLP - * IJKL CGKO - * MNOP DHLP - */ -static void transpose_and_flatten_matrices(u8 *output, const u8 *input, - int matrices, int row_length) -{ - int i, j, row, x_offset, matrix_offset, rot_index, y_offset, matrix_size, new_value; - - matrix_size = row_length * row_length; - for (i = 0; i < matrices; i++) { - row = 0; - x_offset = 0; - matrix_offset = i * matrix_size; - for (j = 0; j < matrix_size; j++) { - y_offset = j - (row * row_length); - rot_index = y_offset * row_length + x_offset; - new_value = *(input + i * matrix_size + j); - output[matrix_offset + rot_index] = new_value; - if ((j + 1) % row_length == 0) { - row += 1; - x_offset += 1; - } - } - } -} - static void assemble_scalingfactor0(struct rkvdec_dev *rkvdec, u8 *output, const struct v4l2_ctrl_hevc_scaling_matrix *input) { + const struct rkvdec_config *cfg = rkvdec->config; int offset = 0; - transpose_and_flatten_matrices(output, (const u8 *)input->scaling_list_4x4, 6, 4); + cfg->flatten_matrices(output, (const u8 *)input->scaling_list_4x4, 6, 4); offset = 6 * 16 * sizeof(u8); - transpose_and_flatten_matrices(output + offset, - (const u8 *)input->scaling_list_8x8, 6, 8); + cfg->flatten_matrices(output + offset, (const u8 *)input->scaling_list_8x8, 6, 8); offset += 6 * 64 * sizeof(u8); - transpose_and_flatten_matrices(output + offset, - (const u8 *)input->scaling_list_16x16, 6, 8); + cfg->flatten_matrices(output + offset, (const u8 *)input->scaling_list_16x16, 6, 8); offset += 6 * 64 * sizeof(u8); /* Add a 128 byte padding with 0s between the two 32x32 matrices */ - transpose_and_flatten_matrices(output + offset, - (const u8 *)input->scaling_list_32x32, 1, 8); + cfg->flatten_matrices(output + offset, (const u8 *)input->scaling_list_32x32, 1, 8); offset += 64 * sizeof(u8); memset(output + offset, 0, 128); offset += 128 * sizeof(u8); - transpose_and_flatten_matrices(output + offset, - (const u8 *)input->scaling_list_32x32 + (64 * sizeof(u8)), - 1, 8); + cfg->flatten_matrices(output + offset, + (const u8 *)input->scaling_list_32x32 + (64 * sizeof(u8)), 1, 8); offset += 64 * sizeof(u8); memset(output + offset, 0, 128); } diff --git a/drivers/media/platform/rockchip/rkvdec/rkvdec-vdpu383-hevc.c b/drivers/media/platform/rockchip/rkvdec/rkvdec-vdpu383-hevc.c new file mode 100644 index 000000000000..7d53c28e954e --- /dev/null +++ b/drivers/media/platform/rockchip/rkvdec/rkvdec-vdpu383-hevc.c @@ -0,0 +1,680 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Rockchip VDPU383 HEVC backend + * + * Copyright (C) 2025 Collabora, Ltd. + * Detlev Casanova + */ + +#include + +#include "rkvdec.h" +#include "rkvdec-rcb.h" +#include "rkvdec-hevc-common.h" +#include "rkvdec-vdpu383-regs.h" + +struct rkvdec_hevc_sps_pps { + // SPS + u16 video_parameters_set_id : 4; + u16 seq_parameters_set_id_sps : 4; + u16 chroma_format_idc : 2; + u16 width : 16; + u16 height : 16; + u16 bit_depth_luma : 3; + u16 bit_depth_chroma : 3; + u16 max_pic_order_count_lsb : 5; + u16 diff_max_min_luma_coding_block_size : 2; + u16 min_luma_coding_block_size : 3; + u16 min_transform_block_size : 3; + u16 diff_max_min_transform_block_size : 2; + u16 max_transform_hierarchy_depth_inter : 3; + u16 max_transform_hierarchy_depth_intra : 3; + u16 scaling_list_enabled_flag : 1; + u16 amp_enabled_flag : 1; + u16 sample_adaptive_offset_enabled_flag : 1; + u16 pcm_enabled_flag : 1; + u16 pcm_sample_bit_depth_luma : 4; + u16 pcm_sample_bit_depth_chroma : 4; + u16 pcm_loop_filter_disabled_flag : 1; + u16 diff_max_min_pcm_luma_coding_block_size : 3; + u16 min_pcm_luma_coding_block_size : 3; + u16 num_short_term_ref_pic_sets : 7; + u16 long_term_ref_pics_present_flag : 1; + u16 num_long_term_ref_pics_sps : 6; + u16 sps_temporal_mvp_enabled_flag : 1; + u16 strong_intra_smoothing_enabled_flag : 1; + u16 reserved0 : 7; + u16 sps_max_dec_pic_buffering_minus1 : 4; + u16 separate_colour_plane_flag : 1; + u16 high_precision_offsets_enabled_flag : 1; + u16 persistent_rice_adaptation_enabled_flag : 1; + + // PPS + u16 picture_parameters_set_id : 6; + u16 seq_parameters_set_id_pps : 4; + u16 dependent_slice_segments_enabled_flag : 1; + u16 output_flag_present_flag : 1; + u16 num_extra_slice_header_bits : 13; + u16 sign_data_hiding_enabled_flag : 1; + u16 cabac_init_present_flag : 1; + u16 num_ref_idx_l0_default_active : 4; + u16 num_ref_idx_l1_default_active : 4; + u16 init_qp_minus26 : 7; + u16 constrained_intra_pred_flag : 1; + u16 transform_skip_enabled_flag : 1; + u16 cu_qp_delta_enabled_flag : 1; + u16 log2_min_cb_size : 3; + u16 pps_cb_qp_offset : 5; + u16 pps_cr_qp_offset : 5; + u16 pps_slice_chroma_qp_offsets_present_flag : 1; + u16 weighted_pred_flag : 1; + u16 weighted_bipred_flag : 1; + u16 transquant_bypass_enabled_flag : 1; + u16 tiles_enabled_flag : 1; + u16 entropy_coding_sync_enabled_flag : 1; + u16 pps_loop_filter_across_slices_enabled_flag : 1; + u16 loop_filter_across_tiles_enabled_flag : 1; + u16 deblocking_filter_override_enabled_flag : 1; + u16 pps_deblocking_filter_disabled_flag : 1; + u16 pps_beta_offset_div2 : 4; + u16 pps_tc_offset_div2 : 4; + u16 lists_modification_present_flag : 1; + u16 log2_parallel_merge_level : 3; + u16 slice_segment_header_extension_present_flag : 1; + u16 reserved1 : 3; + + // pps extensions + u16 log2_max_transform_skip_block_size : 2; + u16 cross_component_prediction_enabled_flag : 1; + u16 chroma_qp_offset_list_enabled_flag : 1; + u16 log2_min_cu_chroma_qp_delta_size : 3; + u16 cb_qp_offset_list0 : 5; + u16 cb_qp_offset_list1 : 5; + u16 cb_qp_offset_list2 : 5; + u16 cb_qp_offset_list3 : 5; + u16 cb_qp_offset_list4 : 5; + u16 cb_qp_offset_list5 : 5; + u16 cb_cr_offset_list0 : 5; + u16 cb_cr_offset_list1 : 5; + u16 cb_cr_offset_list2 : 5; + u16 cb_cr_offset_list3 : 5; + u16 cb_cr_offset_list4 : 5; + u16 cb_cr_offset_list5 : 5; + u16 chroma_qp_offset_list_len_minus1 : 3; + + /* mvc0 && mvc1 */ + u16 mvc_ff : 16; + u16 mvc_00 : 9; + + /* poc info */ + u16 reserved2 : 3; + u32 current_poc : 32; + u32 ref_pic_poc0 : 32; + u32 ref_pic_poc1 : 32; + u32 ref_pic_poc2 : 32; + u32 ref_pic_poc3 : 32; + u32 ref_pic_poc4 : 32; + u32 ref_pic_poc5 : 32; + u32 ref_pic_poc6 : 32; + u32 ref_pic_poc7 : 32; + u32 ref_pic_poc8 : 32; + u32 ref_pic_poc9 : 32; + u32 ref_pic_poc10 : 32; + u32 ref_pic_poc11 : 32; + u32 ref_pic_poc12 : 32; + u32 ref_pic_poc13 : 32; + u32 ref_pic_poc14 : 32; + u32 reserved3 : 32; + u32 ref_is_valid : 15; + u32 reserved4 : 1; + + /* tile info*/ + u16 num_tile_columns : 5; + u16 num_tile_rows : 5; + u32 column_width0 : 24; + u32 column_width1 : 24; + u32 column_width2 : 24; + u32 column_width3 : 24; + u32 column_width4 : 24; + u32 column_width5 : 24; + u32 column_width6 : 24; + u32 column_width7 : 24; + u32 column_width8 : 24; + u32 column_width9 : 24; + u32 row_height0 : 24; + u32 row_height1 : 24; + u32 row_height2 : 24; + u32 row_height3 : 24; + u32 row_height4 : 24; + u32 row_height5 : 24; + u32 row_height6 : 24; + u32 row_height7 : 24; + u32 row_height8 : 24; + u32 row_height9 : 24; + u32 row_height10 : 24; + u32 reserved5 : 2; + u32 padding; +} __packed; + +struct rkvdec_hevc_priv_tbl { + struct rkvdec_hevc_sps_pps param_set; + struct rkvdec_rps rps; + struct scaling_factor scaling_list; + u8 cabac_table[27456]; +} __packed; + +struct rkvdec_hevc_ctx { + struct rkvdec_aux_buf priv_tbl; + struct v4l2_ctrl_hevc_scaling_matrix scaling_matrix_cache; + struct v4l2_ctrl_hevc_ext_sps_st_rps st_cache; + struct vdpu383_regs_h26x regs; +}; + +static void set_column_row(struct rkvdec_hevc_sps_pps *hw_ps, u16 column, u16 row, int i) +{ + int shift = (i & 1) ? 12 : 0; + + switch (i / 2) { + case 0: + hw_ps->column_width0 |= column << shift; + hw_ps->row_height0 |= row << shift; + break; + case 1: + hw_ps->column_width1 |= column << shift; + hw_ps->row_height1 |= row << shift; + break; + case 2: + hw_ps->column_width2 |= column << shift; + hw_ps->row_height2 |= row << shift; + break; + case 3: + hw_ps->column_width3 |= column << shift; + hw_ps->row_height3 |= row << shift; + break; + case 4: + hw_ps->column_width4 |= column << shift; + hw_ps->row_height4 |= row << shift; + break; + case 5: + hw_ps->column_width5 |= column << shift; + hw_ps->row_height5 |= row << shift; + break; + case 6: + hw_ps->column_width6 |= column << shift; + hw_ps->row_height6 |= row << shift; + break; + case 7: + hw_ps->column_width7 |= column << shift; + hw_ps->row_height7 |= row << shift; + break; + case 8: + hw_ps->column_width8 |= column << shift; + hw_ps->row_height8 |= row << shift; + break; + case 9: + hw_ps->column_width9 |= column << shift; + hw_ps->row_height9 |= row << shift; + break; + case 10: + hw_ps->row_height10 |= row << shift; + break; + } +} + +static void set_pps_ref_pic_poc(struct rkvdec_hevc_sps_pps *hw_ps, u32 poc, int i) +{ + switch (i) { + case 0: + hw_ps->ref_pic_poc0 = poc; + break; + case 1: + hw_ps->ref_pic_poc1 = poc; + break; + case 2: + hw_ps->ref_pic_poc2 = poc; + break; + case 3: + hw_ps->ref_pic_poc3 = poc; + break; + case 4: + hw_ps->ref_pic_poc4 = poc; + break; + case 5: + hw_ps->ref_pic_poc5 = poc; + break; + case 6: + hw_ps->ref_pic_poc6 = poc; + break; + case 7: + hw_ps->ref_pic_poc7 = poc; + break; + case 8: + hw_ps->ref_pic_poc8 = poc; + break; + case 9: + hw_ps->ref_pic_poc9 = poc; + break; + case 10: + hw_ps->ref_pic_poc10 = poc; + break; + case 11: + hw_ps->ref_pic_poc11 = poc; + break; + case 12: + hw_ps->ref_pic_poc12 = poc; + break; + case 13: + hw_ps->ref_pic_poc13 = poc; + break; + case 14: + hw_ps->ref_pic_poc14 = poc; + break; + } +} + +static void assemble_hw_pps(struct rkvdec_ctx *ctx, + struct rkvdec_hevc_run *run) +{ + struct rkvdec_hevc_ctx *h264_ctx = ctx->priv; + const struct v4l2_ctrl_hevc_sps *sps = run->sps; + const struct v4l2_ctrl_hevc_pps *pps = run->pps; + const struct v4l2_ctrl_hevc_decode_params *dec_params = run->decode_params; + struct rkvdec_hevc_priv_tbl *priv_tbl = h264_ctx->priv_tbl.cpu; + struct rkvdec_hevc_sps_pps *hw_ps; + bool tiles_enabled; + s32 max_cu_width; + s32 pic_in_cts_width; + s32 pic_in_cts_height; + u16 log2_min_cb_size, width, height; + u16 column_width[22]; + u16 row_height[22]; + u8 pcm_enabled; + u32 i; + + /* + * HW read the SPS/PPS information from PPS packet index by PPS id. + * offset from the base can be calculated by PPS_id * 32 (size per PPS + * packet unit). so the driver copy SPS/PPS information to the exact PPS + * packet unit for HW accessing. + */ + hw_ps = &priv_tbl->param_set; + memset(hw_ps, 0, sizeof(*hw_ps)); + + /* write sps */ + hw_ps->video_parameters_set_id = sps->video_parameter_set_id; + hw_ps->seq_parameters_set_id_sps = sps->seq_parameter_set_id; + hw_ps->chroma_format_idc = sps->chroma_format_idc; + + log2_min_cb_size = sps->log2_min_luma_coding_block_size_minus3 + 3; + width = sps->pic_width_in_luma_samples; + height = sps->pic_height_in_luma_samples; + hw_ps->width = width; + hw_ps->height = height; + hw_ps->bit_depth_luma = sps->bit_depth_luma_minus8 + 8; + hw_ps->bit_depth_chroma = sps->bit_depth_chroma_minus8 + 8; + hw_ps->max_pic_order_count_lsb = sps->log2_max_pic_order_cnt_lsb_minus4 + 4; + hw_ps->diff_max_min_luma_coding_block_size = sps->log2_diff_max_min_luma_coding_block_size; + hw_ps->min_luma_coding_block_size = sps->log2_min_luma_coding_block_size_minus3 + 3; + hw_ps->min_transform_block_size = sps->log2_min_luma_transform_block_size_minus2 + 2; + hw_ps->diff_max_min_transform_block_size = + sps->log2_diff_max_min_luma_transform_block_size; + hw_ps->max_transform_hierarchy_depth_inter = sps->max_transform_hierarchy_depth_inter; + hw_ps->max_transform_hierarchy_depth_intra = sps->max_transform_hierarchy_depth_intra; + hw_ps->scaling_list_enabled_flag = + !!(sps->flags & V4L2_HEVC_SPS_FLAG_SCALING_LIST_ENABLED); + hw_ps->amp_enabled_flag = !!(sps->flags & V4L2_HEVC_SPS_FLAG_AMP_ENABLED); + hw_ps->sample_adaptive_offset_enabled_flag = + !!(sps->flags & V4L2_HEVC_SPS_FLAG_SAMPLE_ADAPTIVE_OFFSET); + + pcm_enabled = !!(sps->flags & V4L2_HEVC_SPS_FLAG_PCM_ENABLED); + hw_ps->pcm_enabled_flag = pcm_enabled; + hw_ps->pcm_sample_bit_depth_luma = + pcm_enabled ? sps->pcm_sample_bit_depth_luma_minus1 + 1 : 0; + hw_ps->pcm_sample_bit_depth_chroma = + pcm_enabled ? sps->pcm_sample_bit_depth_chroma_minus1 + 1 : 0; + hw_ps->pcm_loop_filter_disabled_flag = + !!(sps->flags & V4L2_HEVC_SPS_FLAG_PCM_LOOP_FILTER_DISABLED); + hw_ps->diff_max_min_pcm_luma_coding_block_size = + sps->log2_diff_max_min_pcm_luma_coding_block_size; + hw_ps->min_pcm_luma_coding_block_size = + pcm_enabled ? sps->log2_min_pcm_luma_coding_block_size_minus3 + 3 : 0; + hw_ps->num_short_term_ref_pic_sets = sps->num_short_term_ref_pic_sets; + hw_ps->long_term_ref_pics_present_flag = + !!(sps->flags & V4L2_HEVC_SPS_FLAG_LONG_TERM_REF_PICS_PRESENT); + hw_ps->num_long_term_ref_pics_sps = sps->num_long_term_ref_pics_sps; + hw_ps->sps_temporal_mvp_enabled_flag = + !!(sps->flags & V4L2_HEVC_SPS_FLAG_SPS_TEMPORAL_MVP_ENABLED); + hw_ps->strong_intra_smoothing_enabled_flag = + !!(sps->flags & V4L2_HEVC_SPS_FLAG_STRONG_INTRA_SMOOTHING_ENABLED); + hw_ps->sps_max_dec_pic_buffering_minus1 = sps->sps_max_dec_pic_buffering_minus1; + + /* write pps */ + hw_ps->picture_parameters_set_id = pps->pic_parameter_set_id; + hw_ps->seq_parameters_set_id_pps = sps->seq_parameter_set_id; + hw_ps->dependent_slice_segments_enabled_flag = + !!(pps->flags & V4L2_HEVC_PPS_FLAG_DEPENDENT_SLICE_SEGMENT_ENABLED); + hw_ps->output_flag_present_flag = !!(pps->flags & V4L2_HEVC_PPS_FLAG_OUTPUT_FLAG_PRESENT); + hw_ps->num_extra_slice_header_bits = pps->num_extra_slice_header_bits; + hw_ps->sign_data_hiding_enabled_flag = + !!(pps->flags & V4L2_HEVC_PPS_FLAG_SIGN_DATA_HIDING_ENABLED); + hw_ps->cabac_init_present_flag = !!(pps->flags & V4L2_HEVC_PPS_FLAG_CABAC_INIT_PRESENT); + hw_ps->num_ref_idx_l0_default_active = pps->num_ref_idx_l0_default_active_minus1 + 1; + hw_ps->num_ref_idx_l1_default_active = pps->num_ref_idx_l1_default_active_minus1 + 1; + hw_ps->init_qp_minus26 = pps->init_qp_minus26; + hw_ps->constrained_intra_pred_flag = + !!(pps->flags & V4L2_HEVC_PPS_FLAG_CONSTRAINED_INTRA_PRED); + hw_ps->transform_skip_enabled_flag = + !!(pps->flags & V4L2_HEVC_PPS_FLAG_TRANSFORM_SKIP_ENABLED); + hw_ps->cu_qp_delta_enabled_flag = !!(pps->flags & V4L2_HEVC_PPS_FLAG_CU_QP_DELTA_ENABLED); + hw_ps->log2_min_cb_size = log2_min_cb_size + + sps->log2_diff_max_min_luma_coding_block_size - + pps->diff_cu_qp_delta_depth; + hw_ps->pps_cb_qp_offset = pps->pps_cb_qp_offset; + hw_ps->pps_cr_qp_offset = pps->pps_cr_qp_offset; + hw_ps->pps_slice_chroma_qp_offsets_present_flag = + !!(pps->flags & V4L2_HEVC_PPS_FLAG_PPS_SLICE_CHROMA_QP_OFFSETS_PRESENT); + hw_ps->weighted_pred_flag = !!(pps->flags & V4L2_HEVC_PPS_FLAG_WEIGHTED_PRED); + hw_ps->weighted_bipred_flag = !!(pps->flags & V4L2_HEVC_PPS_FLAG_WEIGHTED_BIPRED); + hw_ps->transquant_bypass_enabled_flag = + !!(pps->flags & V4L2_HEVC_PPS_FLAG_TRANSQUANT_BYPASS_ENABLED); + tiles_enabled = !!(pps->flags & V4L2_HEVC_PPS_FLAG_TILES_ENABLED); + hw_ps->tiles_enabled_flag = tiles_enabled; + hw_ps->entropy_coding_sync_enabled_flag = + !!(pps->flags & V4L2_HEVC_PPS_FLAG_ENTROPY_CODING_SYNC_ENABLED); + hw_ps->pps_loop_filter_across_slices_enabled_flag = + !!(pps->flags & V4L2_HEVC_PPS_FLAG_PPS_LOOP_FILTER_ACROSS_SLICES_ENABLED); + hw_ps->loop_filter_across_tiles_enabled_flag = + !!(pps->flags & V4L2_HEVC_PPS_FLAG_LOOP_FILTER_ACROSS_TILES_ENABLED); + hw_ps->deblocking_filter_override_enabled_flag = + !!(pps->flags & V4L2_HEVC_PPS_FLAG_DEBLOCKING_FILTER_OVERRIDE_ENABLED); + hw_ps->pps_deblocking_filter_disabled_flag = + !!(pps->flags & V4L2_HEVC_PPS_FLAG_PPS_DISABLE_DEBLOCKING_FILTER); + hw_ps->pps_beta_offset_div2 = pps->pps_beta_offset_div2; + hw_ps->pps_tc_offset_div2 = pps->pps_tc_offset_div2; + hw_ps->lists_modification_present_flag = + !!(pps->flags & V4L2_HEVC_PPS_FLAG_LISTS_MODIFICATION_PRESENT); + hw_ps->log2_parallel_merge_level = pps->log2_parallel_merge_level_minus2 + 2; + hw_ps->slice_segment_header_extension_present_flag = + !!(pps->flags & V4L2_HEVC_PPS_FLAG_SLICE_SEGMENT_HEADER_EXTENSION_PRESENT); + hw_ps->num_tile_columns = tiles_enabled ? pps->num_tile_columns_minus1 + 1 : 1; + hw_ps->num_tile_rows = tiles_enabled ? pps->num_tile_rows_minus1 + 1 : 1; + hw_ps->mvc_ff = 0xffff; + + // Setup tiles information + memset(column_width, 0, sizeof(column_width)); + memset(row_height, 0, sizeof(row_height)); + + max_cu_width = 1 << (sps->log2_diff_max_min_luma_coding_block_size + log2_min_cb_size); + pic_in_cts_width = (width + max_cu_width - 1) / max_cu_width; + pic_in_cts_height = (height + max_cu_width - 1) / max_cu_width; + + if (tiles_enabled) { + if (pps->flags & V4L2_HEVC_PPS_FLAG_UNIFORM_SPACING) { + compute_tiles_uniform(run, log2_min_cb_size, width, height, + pic_in_cts_width, pic_in_cts_height, + column_width, row_height); + } else { + compute_tiles_non_uniform(run, log2_min_cb_size, width, height, + pic_in_cts_width, pic_in_cts_height, + column_width, row_height); + } + } else { + column_width[0] = (width + max_cu_width - 1) / max_cu_width; + row_height[0] = (height + max_cu_width - 1) / max_cu_width; + } + + for (i = 0; i < 22; i++) + set_column_row(hw_ps, column_width[i], row_height[i], i); + + // Setup POC information + hw_ps->current_poc = dec_params->pic_order_cnt_val; + + for (i = 0; i < ARRAY_SIZE(dec_params->dpb); i++) { + u32 valid = !!(dec_params->num_active_dpb_entries > i); + + set_pps_ref_pic_poc(hw_ps, dec_params->dpb[i].pic_order_cnt_val, i); + hw_ps->ref_is_valid |= valid << i; + } +} + +static void rkvdec_write_regs(struct rkvdec_ctx *ctx) +{ + struct rkvdec_dev *rkvdec = ctx->dev; + struct rkvdec_hevc_ctx *h265_ctx = ctx->priv; + + rkvdec_memcpy_toio(rkvdec->regs + VDPU383_OFFSET_COMMON_REGS, + &h265_ctx->regs.common, + sizeof(h265_ctx->regs.common)); + rkvdec_memcpy_toio(rkvdec->regs + VDPU383_OFFSET_COMMON_ADDR_REGS, + &h265_ctx->regs.common_addr, + sizeof(h265_ctx->regs.common_addr)); + rkvdec_memcpy_toio(rkvdec->regs + VDPU383_OFFSET_CODEC_PARAMS_REGS, + &h265_ctx->regs.h26x_params, + sizeof(h265_ctx->regs.h26x_params)); + rkvdec_memcpy_toio(rkvdec->regs + VDPU383_OFFSET_CODEC_ADDR_REGS, + &h265_ctx->regs.h26x_addr, + sizeof(h265_ctx->regs.h26x_addr)); +} + +static void config_registers(struct rkvdec_ctx *ctx, + struct rkvdec_hevc_run *run) +{ + const struct v4l2_ctrl_hevc_decode_params *dec_params = run->decode_params; + struct rkvdec_hevc_ctx *h265_ctx = ctx->priv; + const struct v4l2_ctrl_hevc_sps *sps = run->sps; + dma_addr_t priv_start_addr = h265_ctx->priv_tbl.dma; + const struct v4l2_pix_format_mplane *dst_fmt; + struct vb2_v4l2_buffer *src_buf = run->base.bufs.src; + struct vb2_v4l2_buffer *dst_buf = run->base.bufs.dst; + struct vdpu383_regs_h26x *regs = &h265_ctx->regs; + const struct v4l2_format *f; + dma_addr_t rlc_addr; + dma_addr_t dst_addr; + u32 hor_virstride; + u32 ver_virstride; + u32 y_virstride; + u32 offset; + u32 pixels; + u32 i; + + memset(regs, 0, sizeof(*regs)); + + /* Set H264 mode */ + regs->common.reg008_dec_mode = VDPU383_MODE_HEVC; + + /* Set input stream length */ + regs->h26x_params.reg066_stream_len = vb2_get_plane_payload(&src_buf->vb2_buf, 0); + + /* Set strides */ + f = &ctx->decoded_fmt; + dst_fmt = &f->fmt.pix_mp; + hor_virstride = dst_fmt->plane_fmt[0].bytesperline; + ver_virstride = dst_fmt->height; + y_virstride = hor_virstride * ver_virstride; + + pixels = dst_fmt->height * dst_fmt->width; + + regs->h26x_params.reg068_hor_virstride = hor_virstride / 16; + regs->h26x_params.reg069_raster_uv_hor_virstride = hor_virstride / 16; + regs->h26x_params.reg070_y_virstride = y_virstride / 16; + + /* Activate block gating */ + regs->common.reg010.strmd_auto_gating_e = 1; + regs->common.reg010.inter_auto_gating_e = 1; + regs->common.reg010.intra_auto_gating_e = 1; + regs->common.reg010.transd_auto_gating_e = 1; + regs->common.reg010.recon_auto_gating_e = 1; + regs->common.reg010.filterd_auto_gating_e = 1; + regs->common.reg010.bus_auto_gating_e = 1; + regs->common.reg010.ctrl_auto_gating_e = 1; + regs->common.reg010.rcb_auto_gating_e = 1; + regs->common.reg010.err_prc_auto_gating_e = 1; + + /* Set timeout threshold */ + if (pixels < VDPU383_1080P_PIXELS) + regs->common.reg013_core_timeout_threshold = VDPU383_TIMEOUT_1080p; + else if (pixels < VDPU383_4K_PIXELS) + regs->common.reg013_core_timeout_threshold = VDPU383_TIMEOUT_4K; + else if (pixels < VDPU383_8K_PIXELS) + regs->common.reg013_core_timeout_threshold = VDPU383_TIMEOUT_8K; + else + regs->common.reg013_core_timeout_threshold = VDPU383_TIMEOUT_MAX; + + regs->common.reg016.error_proc_disable = 1; + + /* Set ref pic address & poc */ + for (i = 0; i < ARRAY_SIZE(dec_params->dpb) - 1; i++) { + struct vb2_buffer *vb_buf = get_ref_buf(ctx, run, i); + dma_addr_t buf_dma; + + buf_dma = vb2_dma_contig_plane_dma_addr(vb_buf, 0); + + /* Set reference addresses */ + regs->h26x_addr.reg170_185_ref_base[i] = buf_dma; + regs->h26x_addr.reg195_210_payload_st_ref_base[i] = buf_dma; + + /* Set COLMV addresses */ + regs->h26x_addr.reg217_232_colmv_ref_base[i] = buf_dma + ctx->colmv_offset; + } + + /* Set rlc base address (input stream) */ + rlc_addr = vb2_dma_contig_plane_dma_addr(&src_buf->vb2_buf, 0); + regs->common_addr.reg128_strm_base = rlc_addr; + + /* Set output base address */ + dst_addr = vb2_dma_contig_plane_dma_addr(&dst_buf->vb2_buf, 0); + regs->h26x_addr.reg168_decout_base = dst_addr; + regs->h26x_addr.reg169_error_ref_base = dst_addr; + regs->h26x_addr.reg192_payload_st_cur_base = dst_addr; + + /* Set colmv address */ + regs->h26x_addr.reg216_colmv_cur_base = dst_addr + ctx->colmv_offset; + + /* Set RCB addresses */ + for (i = 0; i < rkvdec_rcb_buf_count(ctx); i++) { + regs->common_addr.reg140_162_rcb_info[i].offset = rkvdec_rcb_buf_dma_addr(ctx, i); + regs->common_addr.reg140_162_rcb_info[i].size = rkvdec_rcb_buf_size(ctx, i); + } + + if (sps->flags & V4L2_HEVC_SPS_FLAG_SCALING_LIST_ENABLED) { + /* Set scaling matrix */ + offset = offsetof(struct rkvdec_hevc_priv_tbl, scaling_list); + regs->common_addr.reg132_scanlist_addr = priv_start_addr + offset; + } + + /* Set hw pps address */ + offset = offsetof(struct rkvdec_hevc_priv_tbl, param_set); + regs->common_addr.reg131_gbl_base = priv_start_addr + offset; + regs->h26x_params.reg067_global_len = sizeof(struct rkvdec_hevc_sps_pps) / 16; + + /* Set hw rps address */ + offset = offsetof(struct rkvdec_hevc_priv_tbl, rps); + regs->common_addr.reg129_rps_base = priv_start_addr + offset; + + /* Set cabac table */ + offset = offsetof(struct rkvdec_hevc_priv_tbl, cabac_table); + regs->common_addr.reg130_cabactbl_base = priv_start_addr + offset; + + rkvdec_write_regs(ctx); +} + +static int rkvdec_hevc_start(struct rkvdec_ctx *ctx) +{ + struct rkvdec_dev *rkvdec = ctx->dev; + struct rkvdec_hevc_priv_tbl *priv_tbl; + struct rkvdec_hevc_ctx *hevc_ctx; + struct v4l2_ctrl *ctrl; + int ret; + + ctrl = v4l2_ctrl_find(&ctx->ctrl_hdl, + V4L2_CID_STATELESS_HEVC_SPS); + if (!ctrl) + return -EINVAL; + + ret = rkvdec_hevc_validate_sps(ctx, ctrl->p_new.p_hevc_sps); + if (ret) + return ret; + + hevc_ctx = kzalloc(sizeof(*hevc_ctx), GFP_KERNEL); + if (!hevc_ctx) + return -ENOMEM; + + priv_tbl = dma_alloc_coherent(rkvdec->dev, sizeof(*priv_tbl), + &hevc_ctx->priv_tbl.dma, GFP_KERNEL); + if (!priv_tbl) { + ret = -ENOMEM; + goto err_free_ctx; + } + + hevc_ctx->priv_tbl.size = sizeof(*priv_tbl); + hevc_ctx->priv_tbl.cpu = priv_tbl; + memcpy(priv_tbl->cabac_table, rkvdec_hevc_cabac_table, + sizeof(rkvdec_hevc_cabac_table)); + + ctx->priv = hevc_ctx; + return 0; + +err_free_ctx: + kfree(hevc_ctx); + return ret; +} + +static void rkvdec_hevc_stop(struct rkvdec_ctx *ctx) +{ + struct rkvdec_hevc_ctx *hevc_ctx = ctx->priv; + struct rkvdec_dev *rkvdec = ctx->dev; + + dma_free_coherent(rkvdec->dev, hevc_ctx->priv_tbl.size, + hevc_ctx->priv_tbl.cpu, hevc_ctx->priv_tbl.dma); + kfree(hevc_ctx); +} + +static int rkvdec_hevc_run(struct rkvdec_ctx *ctx) +{ + struct rkvdec_dev *rkvdec = ctx->dev; + struct rkvdec_hevc_run run; + struct rkvdec_hevc_ctx *hevc_ctx = ctx->priv; + struct rkvdec_hevc_priv_tbl *tbl = hevc_ctx->priv_tbl.cpu; + u32 watchdog_time; + u64 timeout_threshold; + unsigned long axi_rate; + + rkvdec_hevc_run_preamble(ctx, &run); + + rkvdec_hevc_assemble_hw_scaling_list(rkvdec, &run, + &tbl->scaling_list, + &hevc_ctx->scaling_matrix_cache); + assemble_hw_pps(ctx, &run); + rkvdec_hevc_assemble_hw_rps(&run, &tbl->rps, &hevc_ctx->st_cache); + + config_registers(ctx, &run); + + rkvdec_run_postamble(ctx, &run.base); + + /* Set watchdog at 2 times the hardware timeout threshold */ + timeout_threshold = hevc_ctx->regs.common.reg013_core_timeout_threshold; + axi_rate = clk_get_rate(rkvdec->axi_clk); + + if (axi_rate) + watchdog_time = 2 * (1000 * timeout_threshold) / axi_rate; + else + watchdog_time = 2000; + schedule_delayed_work(&rkvdec->watchdog_work, + msecs_to_jiffies(watchdog_time)); + + /* Start decoding! */ + writel(timeout_threshold, rkvdec->link + VDPU383_LINK_TIMEOUT_THRESHOLD); + writel(VDPU383_IP_CRU_MODE, rkvdec->link + VDPU383_LINK_IP_ENABLE); + writel(VDPU383_DEC_E_BIT, rkvdec->link + VDPU383_LINK_DEC_ENABLE); + + return 0; +} + +const struct rkvdec_coded_fmt_ops rkvdec_vdpu383_hevc_fmt_ops = { + .adjust_fmt = rkvdec_hevc_adjust_fmt, + .start = rkvdec_hevc_start, + .stop = rkvdec_hevc_stop, + .run = rkvdec_hevc_run, + .try_ctrl = rkvdec_hevc_try_ctrl, + .get_image_fmt = rkvdec_hevc_get_image_fmt, +}; diff --git a/drivers/media/platform/rockchip/rkvdec/rkvdec.c b/drivers/media/platform/rockchip/rkvdec/rkvdec.c index ad8ab9d37add..1e263f6d73b8 100644 --- a/drivers/media/platform/rockchip/rkvdec/rkvdec.c +++ b/drivers/media/platform/rockchip/rkvdec/rkvdec.c @@ -526,6 +526,22 @@ static const struct rkvdec_coded_fmt_desc vdpu383_coded_fmts[] = { .subsystem_flags = VB2_V4L2_FL_SUPPORTS_M2M_HOLD_CAPTURE_BUF, .capability = RKVDEC_CAPABILITY_H264, }, + { + .fourcc = V4L2_PIX_FMT_HEVC_SLICE, + .frmsize = { + .min_width = 16, + .max_width = 65472, + .step_width = 16, + .min_height = 16, + .max_height = 65472, + .step_height = 16, + }, + .ctrls = &vdpu38x_hevc_ctrls, + .ops = &rkvdec_vdpu383_hevc_fmt_ops, + .num_decoded_fmts = ARRAY_SIZE(rkvdec_hevc_decoded_fmts), + .decoded_fmts = rkvdec_hevc_decoded_fmts, + .subsystem_flags = VB2_V4L2_FL_SUPPORTS_M2M_HOLD_CAPTURE_BUF, + }, }; static bool rkvdec_is_capable(struct rkvdec_ctx *ctx, unsigned int capability) @@ -1469,6 +1485,78 @@ static irqreturn_t rkvdec_irq_handler(int irq, void *priv) return cfg->irq_handler(ctx); } +/* + * Flip one or more matrices along their main diagonal and flatten them + * before writing it to the memory. + * Convert: + * ABCD AEIM + * EFGH => BFJN => AEIMBFJNCGKODHLP + * IJKL CGKO + * MNOP DHLP + */ +static void transpose_and_flatten_matrices(u8 *output, const u8 *input, + int matrices, int row_length) +{ + int i, j, row, x_offset, matrix_offset, rot_index, y_offset, matrix_size, new_value; + + matrix_size = row_length * row_length; + for (i = 0; i < matrices; i++) { + row = 0; + x_offset = 0; + matrix_offset = i * matrix_size; + for (j = 0; j < matrix_size; j++) { + y_offset = j - (row * row_length); + rot_index = y_offset * row_length + x_offset; + new_value = *(input + i * matrix_size + j); + output[matrix_offset + rot_index] = new_value; + if ((j + 1) % row_length == 0) { + row += 1; + x_offset += 1; + } + } + } +} + +/* + * VDPU383 needs a specific order: + * The 8x8 flatten matrix is based on 4x4 blocks. + * Each 4x4 block is written separately in order. + * + * Base data => Transposed VDPU383 transposed + * + * ABCDEFGH AIQYaiqy AIQYBJRZ + * IJKLMNOP BJRZbjrz CKS0DLT1 + * QRSTUVWX CKS0cks6 aiqybjrz + * YZ012345 => DLT1dlt7 cks6dlt7 + * abcdefgh EMU2emu8 EMU2FNV3 + * ijklmnop FNV3fnv9 GOW4HPX5 + * qrstuvwx GOW4gow# emu8fnv9 + * yz6789#$ HPX5hpx$ gow#hpx$ + * + * As the function reads block of 4x4 it can be used for both 4x4 and 8x8 matrices. + * + */ +static void vdpu383_flatten_matrices(u8 *output, const u8 *input, int matrices, int row_length) +{ + u8 block; + int i, j, matrix_offset, matrix_size, new_value, input_idx, line_offset, block_offset; + + matrix_size = row_length * row_length; + for (i = 0; i < matrices; i++) { + matrix_offset = i * matrix_size; + for (j = 0; j < matrix_size; j++) { + block = j / 16; + line_offset = (j % 16) / 4; + block_offset = (block & 1) * 32 + (block & 2) * 2; + input_idx = ((j % 4) * row_length) + line_offset + block_offset; + + new_value = *(input + i * matrix_size + input_idx); + + output[matrix_offset + j] = new_value; + } + } +} + static void rkvdec_watchdog_func(struct work_struct *work) { struct rkvdec_dev *rkvdec; @@ -1532,6 +1620,7 @@ static const struct rkvdec_config config_rkvdec = { .coded_fmts_num = ARRAY_SIZE(rkvdec_coded_fmts), .irq_handler = rk3399_irq_handler, .fill_pixfmt_mp = v4l2_fill_pixfmt_mp, + .flatten_matrices = transpose_and_flatten_matrices, .colmv_size = rkvdec_colmv_size, }; @@ -1556,6 +1645,7 @@ static const struct rkvdec_config config_vdpu381 = { .irq_handler = vdpu381_irq_handler, .fill_pixfmt_mp = vdpu38x_fill_pixfmt_mp, .colmv_size = rkvdec_colmv_size, + .flatten_matrices = transpose_and_flatten_matrices, .named_regs = true, }; @@ -1581,6 +1671,7 @@ const struct rkvdec_config config_vdpu383 = { .fill_pixfmt_mp = vdpu38x_fill_pixfmt_mp, .colmv_size = rkvdec_vdpu383_colmv_size, .fill_pixfmt_mp = vdpu38x_fill_pixfmt_mp, + .flatten_matrices = vdpu383_flatten_matrices, .named_regs = true, }; @@ -1615,7 +1706,8 @@ static const struct rkvdec_variant rk3588_vdpu381_variant = { static const struct rkvdec_variant rk3576_vdpu383_variant = { .config = &config_vdpu383, - .capabilities = RKVDEC_CAPABILITY_H264, + .capabilities = RKVDEC_CAPABILITY_H264 | + RKVDEC_CAPABILITY_HEVC, }; static const struct of_device_id of_rkvdec_match[] = { diff --git a/drivers/media/platform/rockchip/rkvdec/rkvdec.h b/drivers/media/platform/rockchip/rkvdec/rkvdec.h index 33cd3406b5ea..aa77b6f69d62 100644 --- a/drivers/media/platform/rockchip/rkvdec/rkvdec.h +++ b/drivers/media/platform/rockchip/rkvdec/rkvdec.h @@ -125,6 +125,7 @@ struct rkvdec_config { int (*fill_pixfmt_mp)(struct v4l2_pix_format_mplane *pix_mp, u32 pixelformat, u32 width, u32 height); u32 (*colmv_size)(u16 width, u16 height); + void (*flatten_matrices)(u8 *output, const u8 *input, int matrices, int row_length); bool named_regs; }; @@ -196,5 +197,6 @@ extern const struct rkvdec_coded_fmt_ops rkvdec_vdpu381_hevc_fmt_ops; /* VDPU383 ops */ extern const struct rkvdec_coded_fmt_ops rkvdec_vdpu383_h264_fmt_ops; +extern const struct rkvdec_coded_fmt_ops rkvdec_vdpu383_hevc_fmt_ops; #endif /* RKVDEC_H_ */ -- 2.34.1