@/******************************************************************************
@ *
@ * Copyright (C) 2015 The Android Open Source Project
@ *
@ * 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.
@ *
@ *****************************************************************************
@ * Originally developed and contributed by Ittiam Systems Pvt. Ltd, Bangalore
@*/
@**
@******************************************************************************
@* @file
@*  ih264_inter_pred_luma_horz_qpel_vert_qpel_a9q.s
@*
@* @brief
@*  Contains function definitions for inter prediction  interpolation.
@*
@* @author
@*  Mohit
@*
@* @par List of Functions:
@*
@*  - ih264_inter_pred_luma_horz_qpel_vert_qpel_a9q()
@*
@* @remarks
@*  None
@*
@*******************************************************************************
@*

@* All the functions here are replicated from ih264_inter_pred_filters.c
@

@*******************************************************************************
@*
@* @brief
@*   This function implements two six tap filters. It
@*    applies the six tap filter in the horizontal direction on the
@*    predictor values, then applies the same filter in the
@*    vertical direction on the predictor values. It then averages these
@*    two outputs to obtain quarter pel values in horizontal and vertical direction.
@*    The six tap filtering operation is described in sec 8.4.2.2.1 titled
@*    "Luma sample interpolation process"
@*
@* @par Description:
@*    This function is called to obtain pixels lying at the following
@*    location (1/4,1/4) or (3/4,1/4) or (1/4,3/4) or (3/4,3/4).
@*    The function interpolates the predictors first in the horizontal direction
@*    and then in the vertical direction, and then averages these two
@*    values.
@*
@* @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] ht
@*  integer height of the array
@*
@* @param[in] wd
@*  integer width of the array
@*
@* @param[in] pu1_tmp: temporary buffer
@*
@* @param[in] dydx: x and y reference offset for qpel calculations
@*
@* @returns
@*
@* @remarks
@*  None
@*
@*******************************************************************************
@*;

@void ih264_inter_pred_luma_horz_qpel_vert_qpel(UWORD8 *pu1_src,
@                                UWORD8 *pu1_dst,
@                                WORD32 src_strd,,
@                                WORD32 dst_strd,
@                                WORD32 ht,
@                                WORD32 wd,
@                                UWORD8* pu1_tmp,
@                                UWORD32 dydx)

@**************Variables Vs Registers*****************************************
@   r0 => *pu1_src
@   r1 => *pu1_dst
@   r2 =>  src_strd
@   r3 =>  dst_strd
@   r4 =>  ht
@   r5 =>  wd
@   r6 =>  dydx

.text
.p2align 2

    .global ih264_inter_pred_luma_horz_qpel_vert_qpel_a9q

ih264_inter_pred_luma_horz_qpel_vert_qpel_a9q:

    stmfd         sp!, {r4-r12, r14}    @store register values to stack
    vstmdb        sp!, {d8-d15}         @push neon registers to stack
    ldr           r4, [sp, #104]        @ loads ht
    ldr           r5, [sp, #108]        @ loads wd
    ldr           r6, [sp, #116]        @dydx
    and           r7, r6, #3
    add           r7, r0, r7, lsr #1    @pu1_pred_vert = pu1_src + (x_offset>>1)

    and           r6, r6, #12           @Finds y-offset
    lsr           r6, r6, #3            @dydx>>3
    mul           r6, r2, r6
    add           r6, r0, r6            @pu1_pred_horz = pu1_src + (y_offset>>1)*src_strd
    sub           r7, r7, r2, lsl #1    @pu1_pred_vert-2*src_strd
    sub           r6, r6, #2            @pu1_pred_horz-2
    vmov.u8       d30, #20              @ Filter coeff 20
    vmov.u8       d31, #5               @ Filter coeff 5

    subs          r12, r5, #4           @if wd=4 branch to loop_4
    beq           loop_4
    subs          r12, r5, #8           @if wd=8 branch to loop_8
    beq           loop_8

loop_16:
    vld1.32       {q0}, [r7], r2        @ Vector load from src[0_0]
    vld1.32       {q1}, [r7], r2        @ Vector load from src[1_0]
    vld1.32       {q2}, [r7], r2        @ Vector load from src[2_0]
    vld1.32       {q3}, [r7], r2        @ Vector load from src[3_0]
    vld1.32       {q4}, [r7], r2        @ Vector load from src[4_0]
    add           r11, r6, #8
    vld1.32       {q5}, [r7], r2        @ Vector load from src[5_0]
    vld1.32       {q9}, [r6], r2        @ horz row0, col 0
    vaddl.u8      q12, d0, d10
    vmlal.u8      q12, d4, d30
    vmlal.u8      q12, d6, d30
    vmlsl.u8      q12, d2, d31
    vmlsl.u8      q12, d8, d31
    vext.8        d23, d18, d19, #5
    vext.8        d20, d18, d19, #2
    vext.8        d21, d18, d19, #3
    vext.8        d22, d18, d19, #4
    vext.8        d19, d18, d19, #1
    vqrshrun.s16  d26, q12, #5
    vaddl.u8      q14, d18, d23
    vmlal.u8      q14, d20, d30
    vmlal.u8      q14, d21, d30
    vmlsl.u8      q14, d19, d31
    vmlsl.u8      q14, d22, d31
    vld1.32       {q9}, [r11], r2       @ horz row 0, col 1
    vaddl.u8      q12, d1, d11
    vmlal.u8      q12, d5, d30
    vmlal.u8      q12, d7, d30
    vmlsl.u8      q12, d3, d31
    vmlsl.u8      q12, d9, d31
    vqrshrun.s16  d28, q14, #5
    vext.8        d23, d18, d19, #5
    vext.8        d20, d18, d19, #2
    vext.8        d21, d18, d19, #3
    vext.8        d22, d18, d19, #4
    vext.8        d19, d18, d19, #1
    vqrshrun.s16  d27, q12, #5
    vld1.32       {q6}, [r7], r2        @ src[6_0]

    vaddl.u8      q12, d18, d23
    vmlal.u8      q12, d20, d30
    vmlal.u8      q12, d21, d30
    vmlsl.u8      q12, d19, d31
    vmlsl.u8      q12, d22, d31

    vaddl.u8      q8, d2, d12
    vmlal.u8      q8, d6, d30
    vmlal.u8      q8, d8, d30
    vmlsl.u8      q8, d4, d31
    vmlsl.u8      q8, d10, d31
    vqrshrun.s16  d29, q12, #5
    vld1.32       {q9}, [r6], r2        @ horz row 1, col 0

    vaddl.u8      q12, d3, d13
    vmlal.u8      q12, d7, d30
    vmlal.u8      q12, d9, d30
    vmlsl.u8      q12, d5, d31
    vmlsl.u8      q12, d11, d31
    vrhadd.u8     q14, q14, q13
    vqrshrun.s16  d26, q8, #5
    vext.8        d23, d18, d19, #5
    vext.8        d20, d18, d19, #2
    vext.8        d21, d18, d19, #3
    vext.8        d22, d18, d19, #4
    vst1.32       {q14}, [r1], r3       @ store row 0
    vext.8        d19, d18, d19, #1
    vqrshrun.s16  d27, q12, #5

    vaddl.u8      q14, d18, d23
    vmlal.u8      q14, d20, d30
    vmlal.u8      q14, d21, d30
    vmlsl.u8      q14, d19, d31
    vmlsl.u8      q14, d22, d31

    vld1.32       {q9}, [r11], r2       @ horz row 1, col 1

    vext.8        d23, d18, d19, #5
    vext.8        d20, d18, d19, #2
    vext.8        d21, d18, d19, #3
    vext.8        d22, d18, d19, #4
    vext.8        d19, d18, d19, #1

    vqrshrun.s16  d28, q14, #5
    vaddl.u8      q12, d18, d23
    vmlal.u8      q12, d20, d30
    vmlal.u8      q12, d21, d30
    vmlsl.u8      q12, d19, d31
    vmlsl.u8      q12, d22, d31

    vqrshrun.s16  d29, q12, #5
    vrhadd.u8     q14, q14, q13
    vst1.32       {q14}, [r1], r3       @ store row 1

    subs          r4, r4, #2            @ 2 rows processed, decrement by 2
    subne         r7, r7 , r2, lsl #2
    subne         r7, r7, r2
    beq           end_func              @ Branch if height==4

    b             loop_16               @ looping if height = 8 or 16


loop_8:
    vld1.32       d0, [r7], r2          @ Vector load from src[0_0]
    vld1.32       d1, [r7], r2          @ Vector load from src[1_0]
    vld1.32       d2, [r7], r2          @ Vector load from src[2_0]
    vld1.32       d3, [r7], r2          @ Vector load from src[3_0]
    vld1.32       d4, [r7], r2          @ Vector load from src[4_0]
    vld1.32       d5, [r7], r2          @ Vector load from src[5_0]
    vaddl.u8      q5, d0, d5
    vmlal.u8      q5, d2, d30
    vmlal.u8      q5, d3, d30
    vmlsl.u8      q5, d1, d31
    vmlsl.u8      q5, d4, d31
    vld1.32       {q6}, [r6], r2        @horz row 0
    vext.8        d17, d12, d13, #5
    vext.8        d14, d12, d13, #2
    vext.8        d15, d12, d13, #3
    vext.8        d16, d12, d13, #4
    vext.8        d13, d12, d13, #1
    vqrshrun.s16  d26, q5, #5
    vld1.32       d6, [r7], r2          @ src[6_0]
    vaddl.u8      q5, d12, d17
    vmlal.u8      q5, d14, d30
    vmlal.u8      q5, d15, d30
    vmlsl.u8      q5, d13, d31
    vmlsl.u8      q5, d16, d31
    vld1.32       {q6}, [r6], r2        @ horz row 1
    vaddl.u8      q9, d1, d6
    vmlal.u8      q9, d3, d30
    vmlal.u8      q9, d4, d30
    vmlsl.u8      q9, d2, d31
    vmlsl.u8      q9, d5, d31
    vqrshrun.s16  d28, q5, #5
    vext.8        d17, d12, d13, #5
    vext.8        d14, d12, d13, #2
    vext.8        d15, d12, d13, #3
    vext.8        d16, d12, d13, #4
    vext.8        d13, d12, d13, #1
    vqrshrun.s16  d27, q9, #5
    vaddl.u8      q5, d12, d17
    vmlal.u8      q5, d14, d30
    vmlal.u8      q5, d15, d30
    vmlsl.u8      q5, d13, d31
    vmlsl.u8      q5, d16, d31
    vqrshrun.s16  d29, q5, #5
    vrhadd.u8     q13, q13, q14
    vst1.32       d26, [r1], r3
    vst1.32       d27, [r1], r3

    subs          r4, r4, #2            @ 2 rows processed, decrement by 2
    subne         r7, r7 , r2, lsl #2
    subne         r7, r7, r2
    beq           end_func              @ Branch if height==4
    b             loop_8                @looping if height == 8 or 16

loop_4:
    vld1.32       d0[0], [r7], r2       @ Vector load from src[0_0]
    vld1.32       d1[0], [r7], r2       @ Vector load from src[1_0]
    vld1.32       d2[0], [r7], r2       @ Vector load from src[2_0]
    vld1.32       d3[0], [r7], r2       @ Vector load from src[3_0]
    vld1.32       d4[0], [r7], r2       @ Vector load from src[4_0]
    vld1.32       d5[0], [r7], r2       @ Vector load from src[5_0]
    vaddl.u8      q5, d0, d5
    vmlal.u8      q5, d2, d30
    vmlal.u8      q5, d3, d30
    vmlsl.u8      q5, d1, d31
    vmlsl.u8      q5, d4, d31
    vld1.32       {q6}, [r6], r2        @load for horz filter row 0
    vext.8        d17, d12, d13, #5
    vext.8        d14, d12, d13, #2
    vext.8        d15, d12, d13, #3
    vext.8        d16, d12, d13, #4
    vext.8        d13, d12, d13, #1
    vqrshrun.s16  d26, q5, #5
    vld1.32       d6[0], [r7], r2       @ Vector load from src[6_0]
    vaddl.u8      q5, d12, d17
    vmlal.u8      q5, d14, d30
    vmlal.u8      q5, d15, d30
    vmlsl.u8      q5, d13, d31
    vmlsl.u8      q5, d16, d31
    vld1.32       {q6}, [r6], r2        @horz row 1
    vaddl.u8      q9, d1, d6
    vmlal.u8      q9, d3, d30
    vmlal.u8      q9, d4, d30
    vmlsl.u8      q9, d2, d31
    vmlsl.u8      q9, d5, d31
    vqrshrun.s16  d28, q5, #5
    vext.8        d17, d12, d13, #5
    vext.8        d14, d12, d13, #2
    vext.8        d15, d12, d13, #3
    vext.8        d16, d12, d13, #4
    vext.8        d13, d12, d13, #1
    vqrshrun.s16  d27, q9, #5
    vaddl.u8      q5, d12, d17
    vmlal.u8      q5, d14, d30
    vmlal.u8      q5, d15, d30
    vmlsl.u8      q5, d13, d31
    vmlsl.u8      q5, d16, d31
    vqrshrun.s16  d29, q5, #5
    vrhadd.u8     q13, q13, q14
    vst1.32       d26[0], [r1], r3
    vst1.32       d27[0], [r1], r3

    subs          r4, r4, #2            @ 2 rows processed, decrement by 2
    subne         r7, r7 , r2, lsl #2
    subne         r7, r7, r2
    beq           end_func              @ Branch if height==4
    b             loop_4                @ Loop if height==8
end_func:
    vldmia        sp!, {d8-d15}         @ Restore neon registers that were saved
    ldmfd         sp!, {r4-r12, pc}     @Restoring registers from stack