xref: /aosp_15_r20/external/libhevc/common/ihevc_iquant_itrans_recon.c (revision c83a76b084498d55f252f48b2e3786804cdf24b7)

/******************************************************************************
*
* Copyright (C) 2012 Ittiam Systems Pvt Ltd, Bangalore
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at:
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*
******************************************************************************/
/**
 *******************************************************************************
 * @file
 *  ihevc_iquant_itrans_recon.c
 *
 * @brief
 *  Contains function definitions for inverse  quantization, inverse
 * transform and reconstruction
 *
 * @author
 *  100470
 *
 * @par List of Functions:
 *  - ihevc_iquant_itrans_recon_4x4_ttype1()
 *  - ihevc_iquant_itrans_recon_4x4()
 *
 * @remarks
 *  None
 *
 *******************************************************************************
 */
#include <stdio.h>
#include <string.h>
#include "ihevc_typedefs.h"
#include "ihevc_macros.h"
#include "ihevc_platform_macros.h"
#include "ihevc_defs.h"
#include "ihevc_trans_tables.h"
#include "ihevc_iquant_itrans_recon.h"
#include "ihevc_func_selector.h"
#include "ihevc_trans_macros.h"

/* All the functions here are replicated from ihevc_itrans.c and modified to */
/* include reconstruction */

/**
 *******************************************************************************
 *
 * @brief
 *  This function performs inverse quantization, inverse  transform
 * type1(DST) and reconstruction for 4x4  input block
 *
 * @par Description:
 *  Performs inverse quantization , inverse transform type 1  and adds
 * prediction data and clips output to 8 bit
 *
 * @param[in] pi2_src
 *  Input 4x4 coefficients
 *
 * @param[in] pi2_tmp
 *  Temporary 4x4 buffer for storing inverse
 *  transform 1st stage output
 *
 * @param[in] pu1_pred
 *  Prediction 4x4 block
 *
 * @param[in] pi2_dequant_coeff
 *  Dequant Coeffs
 *
 * @param[out] pu1_dst
 *  Output 4x4 block
 *
 * @param[in] qp_div
 *  Quantization parameter / 6
 *
 * @param[in] qp_rem
 *  Quantization parameter % 6
 *
 * @param[in] src_strd
 *  Input stride
 *
 * @param[in] pred_strd
 *  Prediction stride
 *
 * @param[in] dst_strd
 *  Output Stride
 *
 * @param[in] zero_cols
 *  Zero columns in pi2_src
 *
 * @param[in] zero_rows
 *  Zero Rows in pi2_src
 *
 * @returns  Void
 *
 * @remarks
 *  None
 *
 *******************************************************************************
 */

void ihevc_iquant_itrans_recon_4x4_ttype1(WORD16 *pi2_src,
                                          WORD16 *pi2_tmp,
                                          UWORD8 *pu1_pred,
                                          WORD16 *pi2_dequant_coeff,
                                          UWORD8 *pu1_dst,
                                          WORD32 qp_div, /* qpscaled / 6 */
                                          WORD32 qp_rem, /* qpscaled % 6 */
                                          WORD32 src_strd,
                                          WORD32 pred_strd,
                                          WORD32 dst_strd,
                                          WORD32 zero_cols,
                                          WORD32 zero_rows)
{
    UNUSED(zero_rows);
    /* Inverse Quant and Inverse Transform and Reconstruction */
    {
        WORD32 i, c[4];
        WORD32 add;
        WORD32 shift;
        WORD16 *pi2_tmp_orig;
        WORD32 shift_iq;
        WORD32 trans_size;
        /* Inverse Quantization constants */
        {
            WORD32 log2_trans_size, bit_depth;

            log2_trans_size = 2;
            bit_depth = 8 + 0;
            shift_iq = bit_depth + log2_trans_size - 5;
        }

        trans_size = TRANS_SIZE_4;
        pi2_tmp_orig = pi2_tmp;

        /* Inverse Transform 1st stage */
        shift = IT_SHIFT_STAGE_1;
        add = 1 << (shift - 1);

        for(i = 0; i < trans_size; i++)
        {
            /* Checking for Zero Cols */
            if((zero_cols & 1) == 1)
            {
                memset(pi2_tmp, 0, trans_size * sizeof(WORD16));
            }
            else
            {
                WORD32 iq_tmp_1, iq_tmp_2, iq_tmp_3;
                // Intermediate Variables
                IQUANT_4x4(iq_tmp_1,
                           pi2_src[0 * src_strd],
                           pi2_dequant_coeff[0 * trans_size] * g_ihevc_iquant_scales[qp_rem],
                           shift_iq, qp_div);
                IQUANT_4x4(iq_tmp_2,
                           pi2_src[2 * src_strd],
                           pi2_dequant_coeff[2 * trans_size] * g_ihevc_iquant_scales[qp_rem],
                           shift_iq, qp_div);
                c[0] = iq_tmp_1 + iq_tmp_2;

                IQUANT_4x4(iq_tmp_1,
                           pi2_src[2 * src_strd],
                           pi2_dequant_coeff[2 * trans_size] * g_ihevc_iquant_scales[qp_rem],
                           shift_iq, qp_div);
                IQUANT_4x4(iq_tmp_2,
                           pi2_src[3 * src_strd],
                           pi2_dequant_coeff[3 * trans_size] * g_ihevc_iquant_scales[qp_rem],
                           shift_iq, qp_div);
                c[1] = iq_tmp_1 + iq_tmp_2;

                IQUANT_4x4(iq_tmp_1,
                           pi2_src[0 * src_strd],
                           pi2_dequant_coeff[0 * trans_size] * g_ihevc_iquant_scales[qp_rem],
                           shift_iq, qp_div);
                IQUANT_4x4(iq_tmp_2,
                           pi2_src[3 * src_strd],
                           pi2_dequant_coeff[3 * trans_size] * g_ihevc_iquant_scales[qp_rem],
                           shift_iq, qp_div);
                c[2] = iq_tmp_1 - iq_tmp_2;

                IQUANT_4x4(iq_tmp_1,
                           pi2_src[1 * src_strd],
                           pi2_dequant_coeff[1 * trans_size] * g_ihevc_iquant_scales[qp_rem],
                           shift_iq, qp_div);
                c[3] = 74 * iq_tmp_1;

                pi2_tmp[0] =
                                CLIP_S16((29 * c[0] + 55 * c[1] + c[3] + add) >> shift);
                pi2_tmp[1] =
                                CLIP_S16((55 * c[2] - 29 * c[1] + c[3] + add) >> shift);

                IQUANT_4x4(iq_tmp_1,
                           pi2_src[0 * src_strd],
                           pi2_dequant_coeff[0 * trans_size] * g_ihevc_iquant_scales[qp_rem],
                           shift_iq, qp_div);
                IQUANT_4x4(iq_tmp_2,
                           pi2_src[2 * src_strd],
                           pi2_dequant_coeff[2 * trans_size] * g_ihevc_iquant_scales[qp_rem],
                           shift_iq, qp_div);
                IQUANT_4x4(iq_tmp_3,
                           pi2_src[3 * src_strd],
                           pi2_dequant_coeff[3 * trans_size] * g_ihevc_iquant_scales[qp_rem],
                           shift_iq, qp_div);

                pi2_tmp[2] =
                                CLIP_S16((74 * (iq_tmp_1 - iq_tmp_2 + iq_tmp_3) + add) >> shift);
                pi2_tmp[3] =
                                CLIP_S16((55 * c[0] + 29 * c[2] - c[3] + add) >> shift);
            }
            pi2_src++;
            pi2_dequant_coeff++;
            pi2_tmp += trans_size;
            zero_cols = zero_cols >> 1;
        }

        pi2_tmp = pi2_tmp_orig;

        /* Inverse Transform 2nd stage */
        shift = IT_SHIFT_STAGE_2;
        add = 1 << (shift - 1);

        for(i = 0; i < trans_size; i++)
        {
            WORD32 itrans_out;

            // Intermediate Variables
            c[0] = pi2_tmp[0] + pi2_tmp[2 * trans_size];
            c[1] = pi2_tmp[2 * trans_size] + pi2_tmp[3 * trans_size];
            c[2] = pi2_tmp[0] - pi2_tmp[3 * trans_size];
            c[3] = 74 * pi2_tmp[trans_size];

            itrans_out =
                            CLIP_S16((29 * c[0] + 55 * c[1] + c[3] + add) >> shift);
            pu1_dst[0] = CLIP_U8((itrans_out + pu1_pred[0]));

            itrans_out =
                            CLIP_S16((55 * c[2] - 29 * c[1] + c[3] + add) >> shift);
            pu1_dst[1] = CLIP_U8((itrans_out + pu1_pred[1]));

            itrans_out =
                            CLIP_S16((74 * (pi2_tmp[0] - pi2_tmp[2 * trans_size] + pi2_tmp[3 * trans_size]) + add) >> shift);
            pu1_dst[2] = CLIP_U8((itrans_out + pu1_pred[2]));

            itrans_out =
                            CLIP_S16((55 * c[0] + 29 * c[2] - c[3] + add) >> shift);
            pu1_dst[3] = CLIP_U8((itrans_out + pu1_pred[3]));
            pi2_tmp++;
            pu1_pred += pred_strd;
            pu1_dst += dst_strd;
        }
    }
}

/**
 *******************************************************************************
 *
 * @brief
 *  This function performs inverse quantization, inverse  transform and
 * reconstruction for 4x4 input block
 *
 * @par Description:
 *  Performs inverse quantization , inverse transform  and adds the
 * prediction data and clips output to 8 bit
 *
 * @param[in] pi2_src
 *  Input 4x4 coefficients
 *
 * @param[in] pi2_tmp
 *  Temporary 4x4 buffer for storing inverse
 *  transform 1st stage output
 *
 * @param[in] pu1_pred
 *  Prediction 4x4 block
 *
 * @param[in] pi2_dequant_coeff
 *  Dequant Coeffs
 *
 * @param[out] pu1_dst
 *  Output 4x4 block
 *
 * @param[in] qp_div
 *  Quantization parameter / 6
 *
 * @param[in] qp_rem
 *  Quantization parameter % 6
 *
 * @param[in] src_strd
 *  Input stride
 *
 * @param[in] pred_strd
 *  Prediction stride
 *
 * @param[in] dst_strd
 *  Output Stride
 *
 * @param[in] zero_cols
 *  Zero columns in pi2_src
 *
 * @param[in] zero_rows
 *  Zero Rows in pi2_src
 *
 * @returns  Void
 *
 * @remarks
 *  None
 *
 *******************************************************************************
 */

void ihevc_iquant_itrans_recon_4x4(WORD16 *pi2_src,
                                   WORD16 *pi2_tmp,
                                   UWORD8 *pu1_pred,
                                   WORD16 *pi2_dequant_coeff,
                                   UWORD8 *pu1_dst,
                                   WORD32 qp_div, /* qpscaled / 6 */
                                   WORD32 qp_rem, /* qpscaled % 6 */
                                   WORD32 src_strd,
                                   WORD32 pred_strd,
                                   WORD32 dst_strd,
                                   WORD32 zero_cols,
                                   WORD32 zero_rows)
{
    UNUSED(zero_rows);
    /* Inverse Transform */
    {
        WORD32 j;
        WORD32 e[2], o[2];
        WORD32 add;
        WORD32 shift;
        WORD16 *pi2_tmp_orig;
        WORD32 shift_iq;
        WORD32 trans_size;
        /* Inverse Quantization constants */
        {
            WORD32 log2_trans_size, bit_depth;

            log2_trans_size = 2;
            bit_depth = 8 + 0;
            shift_iq = bit_depth + log2_trans_size - 5;
        }

        trans_size = TRANS_SIZE_4;
        pi2_tmp_orig = pi2_tmp;

        /* Inverse Transform 1st stage */
        shift = IT_SHIFT_STAGE_1;
        add = 1 << (shift - 1);

        for(j = 0; j < trans_size; j++)
        {
            /* Checking for Zero Cols */
            if((zero_cols & 1) == 1)
            {
                memset(pi2_tmp, 0, trans_size * sizeof(WORD16));
            }
            else
            {
                WORD32 iq_tmp_1, iq_tmp_2;
                /* Utilizing symmetry properties to the maximum to minimize the number of multiplications */
                IQUANT_4x4(iq_tmp_1,
                           pi2_src[1 * src_strd],
                           pi2_dequant_coeff[1 * trans_size] * g_ihevc_iquant_scales[qp_rem],
                           shift_iq, qp_div);
                IQUANT_4x4(iq_tmp_2,
                           pi2_src[3 * src_strd],
                           pi2_dequant_coeff[3 * trans_size] * g_ihevc_iquant_scales[qp_rem],
                           shift_iq, qp_div);

                o[0] = g_ai2_ihevc_trans_4[1][0] * iq_tmp_1
                                + g_ai2_ihevc_trans_4[3][0] * iq_tmp_2;
                o[1] = g_ai2_ihevc_trans_4[1][1] * iq_tmp_1
                                + g_ai2_ihevc_trans_4[3][1] * iq_tmp_2;

                IQUANT_4x4(iq_tmp_1,
                           pi2_src[0 * src_strd],
                           pi2_dequant_coeff[0 * trans_size] * g_ihevc_iquant_scales[qp_rem],
                           shift_iq, qp_div);
                IQUANT_4x4(iq_tmp_2,
                           pi2_src[2 * src_strd],
                           pi2_dequant_coeff[2 * trans_size] * g_ihevc_iquant_scales[qp_rem],
                           shift_iq, qp_div);

                e[0] = g_ai2_ihevc_trans_4[0][0] * iq_tmp_1
                                + g_ai2_ihevc_trans_4[2][0] * iq_tmp_2;
                e[1] = g_ai2_ihevc_trans_4[0][1] * iq_tmp_1
                                + g_ai2_ihevc_trans_4[2][1] * iq_tmp_2;

                pi2_tmp[0] =
                                CLIP_S16(((e[0] + o[0] + add) >> shift));
                pi2_tmp[1] =
                                CLIP_S16(((e[1] + o[1] + add) >> shift));
                pi2_tmp[2] =
                                CLIP_S16(((e[1] - o[1] + add) >> shift));
                pi2_tmp[3] =
                                CLIP_S16(((e[0] - o[0] + add) >> shift));
            }
            pi2_src++;
            pi2_dequant_coeff++;
            pi2_tmp += trans_size;
            zero_cols = zero_cols >> 1;
        }

        pi2_tmp = pi2_tmp_orig;

        /* Inverse Transform 2nd stage */
        shift = IT_SHIFT_STAGE_2;
        add = 1 << (shift - 1);

        for(j = 0; j < trans_size; j++)
        {
            WORD32 itrans_out;

            /* Utilizing symmetry properties to the maximum to minimize the number of multiplications */
            o[0] = g_ai2_ihevc_trans_4[1][0] * pi2_tmp[trans_size]
                            + g_ai2_ihevc_trans_4[3][0]
                                            * pi2_tmp[3 * trans_size];
            o[1] = g_ai2_ihevc_trans_4[1][1] * pi2_tmp[trans_size]
                            + g_ai2_ihevc_trans_4[3][1]
                                            * pi2_tmp[3 * trans_size];
            e[0] = g_ai2_ihevc_trans_4[0][0] * pi2_tmp[0]
                            + g_ai2_ihevc_trans_4[2][0]
                                            * pi2_tmp[2 * trans_size];
            e[1] = g_ai2_ihevc_trans_4[0][1] * pi2_tmp[0]
                            + g_ai2_ihevc_trans_4[2][1]
                                            * pi2_tmp[2 * trans_size];

            itrans_out =
                            CLIP_S16(((e[0] + o[0] + add) >> shift));
            pu1_dst[0] = CLIP_U8((itrans_out + pu1_pred[0]));

            itrans_out =
                            CLIP_S16(((e[1] + o[1] + add) >> shift));
            pu1_dst[1] = CLIP_U8((itrans_out + pu1_pred[1]));

            itrans_out =
                            CLIP_S16(((e[1] - o[1] + add) >> shift));
            pu1_dst[2] = CLIP_U8((itrans_out + pu1_pred[2]));

            itrans_out =
                            CLIP_S16(((e[0] - o[0] + add) >> shift));
            pu1_dst[3] = CLIP_U8((itrans_out + pu1_pred[3]));

            pi2_tmp++;
            pu1_pred += pred_strd;
            pu1_dst += dst_strd;

        }
    }
}