xref: /external/libhevc/common/arm64/ihevc_intra_pred_luma_horz.s

///*****************************************************************************
//*
//* 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_intra_pred_luma_horz_neon.s
//*
//* @brief
//*  contains function definition for intra prediction  interpolation filters
//*
//*
//* @author
//*  parthiban v
//*
//* @par list of functions:
//*  - ihevc_intra_pred_luma_horz()
//*
//* @remarks
//*  none
//*
//*******************************************************************************
//*/
//
///**
//*******************************************************************************
//*
//* @brief
//*     intra prediction interpolation filter for horizontal luma variable.
//*
//* @par description:
//*      horizontal intraprediction(mode 10) with.extern  samples location
//*      pointed by 'pu1_ref' to the tu block  location pointed by 'pu1_dst'  refer
//*      to section 8.4.4.2.6 in the standard (special case)
//*
//* @param[in] pu1_src
//*  uword8 pointer to the source
//*
//* @param[out] pu1_dst
//*  uword8 pointer to the destination
//*
//* @param[in] src_strd
//*  integer source stride
//*
//* @param[in] dst_strd
//*  integer destination stride
//*
//* @param[in] nt
//*  integer transform block size
//*
//* @param[in] mode
//*  integer intraprediction mode
//*
//* @returns
//*
//* @remarks
//*  none
//*
//*******************************************************************************
//*/
//void ihevc_intra_pred_luma_horz(uword8 *pu1_ref,
//                                word32 src_strd,
//                                uword8 *pu1_dst,
//                                word32 dst_strd,
//                                word32 nt,
//                                word32 mode)
//**************variables vs registers*****************************************
//x0 => *pu1_ref
//x1 =>  src_strd
//x2 => *pu1_dst
//x3 =>  dst_strd

.text
.align 4
.include "ihevc_neon_macros.s"



.globl ihevc_intra_pred_luma_horz_av8

.type ihevc_intra_pred_luma_horz_av8, %function

ihevc_intra_pred_luma_horz_av8:

    // stmfd sp!, {x4-x12, x14}                //stack stores the values of the arguments

    stp         x19, x20,[sp,#-16]!

    //ldr          x5,[sp,#44]                        @loads mode

    lsl         x6,x4,#1                    //two_nt

    add         x12,x0,x6                   //*pu1_ref[two_nt]
    cmp         x4,#4                       //if nt == 4
    beq         core_loop_4

    cmp         x4,#8                       //if nt == 8
    beq         core_loop_8

    cmp         x4,#16                      //if nt == 16
    beq         core_loop_16
    sub         x12,x12,#16                 //move to 16th value pointer
    add         x9,x2,#16

core_loop_32:
    ld1         { v0.16b},[x12]             //load 16 values. d1[7] will have the 1st value.

    dup         v2.16b, v0.b[15]            //duplicate the i value.

    dup         v4.16b, v0.b[14]            //duplicate the ii value.
    dup         v6.16b, v0.b[13]            //duplicate the iii value.
    st1         { v2.16b},[x2],x3           //store in 1st row 0-16 columns
    st1         { v2.16b},[x9],x3           //store in 1st row 16-32 columns

    dup         v1.16b, v0.b[12]
    st1         { v4.16b},[x2],x3
    st1         { v4.16b},[x9],x3

    dup         v2.16b, v0.b[11]
    st1         { v6.16b},[x2],x3
    st1         { v6.16b},[x9],x3

    dup         v4.16b, v0.b[10]
    st1         { v1.16b},[x2],x3
    st1         { v1.16b},[x9],x3

    dup         v6.16b, v0.b[9]
    st1         { v2.16b},[x2],x3
    st1         { v2.16b},[x9],x3

    dup         v1.16b, v0.b[8]
    st1         { v4.16b},[x2],x3
    st1         { v4.16b},[x9],x3

    dup         v2.16b, v0.b[7]
    st1         { v6.16b},[x2],x3
    st1         { v6.16b},[x9],x3

    dup         v4.16b, v0.b[6]
    st1         { v1.16b},[x2],x3
    st1         { v1.16b},[x9],x3

    dup         v6.16b, v0.b[5]
    st1         { v2.16b},[x2],x3
    st1         { v2.16b},[x9],x3

    dup         v1.16b, v0.b[4]
    st1         { v4.16b},[x2],x3
    st1         { v4.16b},[x9],x3

    dup         v2.16b, v0.b[3]
    st1         { v6.16b},[x2],x3
    st1         { v6.16b},[x9],x3

    dup         v4.16b, v0.b[2]
    st1         { v1.16b},[x2],x3
    st1         { v1.16b},[x9],x3

    dup         v6.16b, v0.b[1]
    st1         { v2.16b},[x2],x3
    st1         { v2.16b},[x9],x3
    sub         x12,x12,#16                 //move to 16th value pointer

    dup         v1.16b, v0.b[0]
    st1         { v4.16b},[x2],x3
    st1         { v4.16b},[x9],x3

    subs        x4,x4,#16                   //decrement the loop count by 16
    st1         { v6.16b},[x2],x3
    st1         { v6.16b},[x9],x3

    st1         { v1.16b},[x2],x3
    st1         { v1.16b},[x9],x3
    bgt         core_loop_32
    // ldmfd sp!,{x4-x12,x15}                  //reload the registers from sp
    ldp         x19, x20,[sp],#16

    ret
    b           end_func

core_loop_16:
    ldrb        w14,[x12],#1                //pu1_ref[two_nt]
    sxtw        x14,w14
    ld1         { v30.8b},[x12],#8          //pu1_ref[two_nt + 1 + col]
    ld1         { v31.8b},[x12]             //pu1_ref[two_nt + 1 + col]
    sub         x12,x12,#8

    dup         v28.8b,w14
    sub         x12,x12,#17
    ld1         { v0.16b},[x12]
    dup         v26.8b, v0.b[15]
    uxtl        v26.8h, v26.8b

    dup         v2.16b, v0.b[14]
    usubl       v24.8h, v30.8b, v28.8b

    dup         v4.16b, v0.b[13]
    sshr        v24.8h, v24.8h,#1

    dup         v6.16b, v0.b[12]
    sqadd       v22.8h,  v26.8h ,  v24.8h

    dup         v1.16b, v0.b[11]
    sqxtun      v22.8b, v22.8h

    st1         {v22.8b},[x2],#8

    dup         v18.16b, v0.b[10]
    usubl       v24.8h, v31.8b, v28.8b

    dup         v19.16b, v0.b[9]
    sshr        v24.8h, v24.8h,#1

    dup         v20.16b, v0.b[8]
    sqadd       v22.8h,  v26.8h ,  v24.8h

    dup         v16.16b, v0.b[7]
    sqxtun      v22.8b, v22.8h

    st1         {v22.8b},[x2],x3
    sub         x2,x2,#8

    st1         { v2.16b},[x2],x3

    st1         { v4.16b},[x2],x3
    st1         { v6.16b},[x2],x3
    st1         { v1.16b},[x2],x3

    dup         v2.16b, v0.b[6]
    st1         { v18.16b},[x2],x3

    dup         v4.16b, v0.b[5]
    st1         { v19.16b},[x2],x3

    dup         v6.16b, v0.b[4]
    st1         { v20.16b},[x2],x3

    dup         v1.16b, v0.b[3]
    st1         { v16.16b},[x2],x3

    dup         v18.16b, v0.b[2]
    st1         { v2.16b},[x2],x3

    dup         v19.16b, v0.b[1]
    st1         { v4.16b},[x2],x3

    dup         v20.16b, v0.b[0]
    st1         { v6.16b},[x2],x3

    st1         { v1.16b},[x2],x3
    st1         { v18.16b},[x2],x3
    st1         { v19.16b},[x2],x3
    st1         { v20.16b},[x2],x3

    // ldmfd sp!,{x4-x12,x15}                  //reload the registers from sp
    ldp         x19, x20,[sp],#16

    ret
    b           end_func


core_loop_8:
    ldrb        w14,[x12]                   //pu1_ref[two_nt]
    sxtw        x14,w14
    add         x12,x12,#1                  //pu1_ref[two_nt + 1]
    ld1         {v30.8b},[x12]              //pu1_ref[two_nt + 1 + col]

    sub         x12,x12,#9
    ld1         {v0.8b},[x12]
    dup         v26.8b, v0.b[7]
    dup         v28.8b,w14

    dup         v3.8b, v0.b[6]
    uxtl        v26.8h, v26.8b

    dup         v4.8b, v0.b[5]
    usubl       v24.8h, v30.8b, v28.8b

    dup         v5.8b, v0.b[4]
    sshr        v24.8h, v24.8h,#1

    dup         v6.8b, v0.b[3]
    sqadd       v22.8h,  v26.8h ,  v24.8h

    dup         v7.8b, v0.b[2]
    sqxtun      v22.8b, v22.8h

    st1         {v22.8b},[x2],x3
    st1         {v3.8b},[x2],x3

    dup         v1.8b, v0.b[1]
    st1         {v4.8b},[x2],x3
    st1         {v5.8b},[x2],x3

    dup         v17.8b, v0.b[0]
    st1         {v6.8b},[x2],x3
    st1         {v7.8b},[x2],x3

    st1         {v1.8b},[x2],x3
    st1         {v17.8b},[x2],x3
    // ldmfd sp!,{x4-x12,x15}                  //reload the registers from sp
    ldp         x19, x20,[sp],#16

    ret
    b           end_func


core_loop_4:
    ldrb        w14,[x12]                   //pu1_ref[two_nt]
    sxtw        x14,w14
    add         x12,x12,#1                  //pu1_ref[two_nt + 1]
    ld1         {v30.8b},[x12]              //pu1_ref[two_nt + 1 + col]

    sub         x12,x12,#5
    ld1         {v0.8b},[x12]
    dup         v28.8b,w14
    dup         v26.8b, v0.b[3]
    uxtl        v26.8h, v26.8b

    dup         v3.8b, v0.b[2]
    usubl       v24.8h, v30.8b, v28.8b

    dup         v4.8b, v0.b[1]
    sshr        v24.8h, v24.8h,#1

    dup         v5.8b, v0.b[0]
    sqadd       v22.8h,  v26.8h ,  v24.8h

    sqxtun      v22.8b, v22.8h

    st1         {v22.s}[0],[x2],x3
    st1         {v3.s}[0],[x2],x3
    st1         {v4.s}[0],[x2],x3
    st1         {v5.s}[0],[x2],x3

    // ldmfd sp!,{x4-x12,x15}                  //reload the registers from sp
    ldp         x19, x20,[sp],#16

    ret
end_func: