@/*****************************************************************************
@*
@* 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_inter_pred_chroma_copy_w16out_neon.s
@*
@* @brief
@* contains function definitions for inter prediction interpolation.
@* functions are coded using neon intrinsics and can be compiled using
@* rvct
@*
@* @author
@* yogeswaran rs
@*
@* @par list of functions:
@*
@*
@* @remarks
@* none
@*
@*******************************************************************************
@*/
@/**
@*******************************************************************************
@*
@* @brief
@* chroma interprediction filter for copy
@*
@* @par description:
@* copies the array of width 'wd' and height 'ht' from the location pointed
@* by 'src' to the location pointed by 'dst'
@*
@* @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] pi1_coeff
@* word8 pointer to the filter coefficients
@*
@* @param[in] ht
@* integer height of the array
@*
@* @param[in] wd
@* integer width of the array
@*
@* @returns
@*
@* @remarks
@* none
@*
@*******************************************************************************
@*/
@void ihevc_inter_pred_chroma_copy_w16out(uword8 *pu1_src,
@ word16 *pi2_dst,
@ word32 src_strd,
@ word32 dst_strd,
@ word8 *pi1_coeff,
@ word32 ht,
@ word32 wd)
@**************variables vs registers*****************************************
@r0 => *pu1_src
@r1 => *pi2_dst
@r2 => src_strd
@r3 => dst_strd
@r4 => *pi1_coeff
@r5 => ht
@r6 => wd
.text
.align 4
.globl ihevc_inter_pred_chroma_copy_w16out_a9q
.type ihevc_inter_pred_chroma_copy_w16out_a9q, %function
ihevc_inter_pred_chroma_copy_w16out_a9q:
stmfd sp!, {r4-r12, r14} @stack stores the values of the arguments
ldr r12,[sp,#48] @loads wd
lsl r12,r12,#1 @2*wd
ldr r7,[sp,#44] @loads ht
cmp r7,#0 @ht condition(ht == 0)
ble end_loops @loop
and r8,r7,#3 @check ht for mul of 2
sub r9,r7,r8 @check the rounded height value
and r11,r7,#6
cmp r11,#6
beq loop_ht_6
tst r12,#7 @conditional check for wd (multiples)
beq core_loop_wd_8
loop_ht_6:
sub r11,r12,#4
lsls r6,r3,#1
cmp r9,#0
beq outer_loop_wd_4_ht_2
outer_loop_wd_4:
subs r4,r12,#0 @wd conditional subtract
ble end_inner_loop_wd_4
inner_loop_wd_4:
vld1.8 {d0},[r0] @vld1_u8(pu1_src_tmp)
add r5,r0,r2 @pu1_src +src_strd
vmovl.u8 q0,d0 @vmovl_u8(vld1_u8(pu1_src_tmp)
add r10,r1,r6
subs r4,r4,#4 @wd - 4
vshl.i64 q0,q0,#6 @vshlq_n_s64(temp, 6)
vld1.8 {d22},[r5],r2 @vld1_u8(pu1_src_tmp)
add r0,r0,#4 @pu1_src += 4
vst1.64 {d0},[r1] @vst1q_lane_s64(pi2_dst_tmp, temp, 0)
add r1,r1,#8
vmovl.u8 q11,d22 @vmovl_u8(vld1_u8(pu1_src_tmp)
vld1.8 {d24},[r5],r2 @vld1_u8(pu1_src_tmp)
vshl.i64 q11,q11,#6 @vshlq_n_s64(temp, 6)
vmovl.u8 q12,d24 @vmovl_u8(vld1_u8(pu1_src_tmp)
vst1.64 {d22},[r10],r6 @vst1q_lane_s64(pi2_dst_tmp, temp, 0)
vshl.i64 q12,q12,#6 @vshlq_n_s64(temp, 6)
vld1.8 {d26},[r5],r2 @vld1_u8(pu1_src_tmp)
vst1.64 {d24},[r10],r6 @vst1q_lane_s64(pi2_dst_tmp, temp, 0)
vmovl.u8 q13,d26 @vmovl_u8(vld1_u8(pu1_src_tmp)
vshl.i64 q13,q13,#6 @vshlq_n_s64(temp, 6)
vst1.64 {d26},[r10],r6 @vst1q_lane_s64(pi2_dst_tmp, temp, 0)
bgt inner_loop_wd_4
end_inner_loop_wd_4:
subs r9,r9,#4 @ht - 4
sub r0,r5,r11
sub r1,r10,r11,lsl #1
bgt outer_loop_wd_4
cmp r8,#0
bgt outer_loop_wd_4_ht_2
end_loops:
ldmfd sp!,{r4-r12,r15} @reload the registers from sp
outer_loop_wd_4_ht_2:
subs r4,r12,#0 @wd conditional subtract
ble end_inner_loop_wd_4
inner_loop_wd_4_ht_2:
vld1.8 {d0},[r0] @vld1_u8(pu1_src_tmp)
add r5,r0,r2 @pu1_src +src_strd
vmovl.u8 q0,d0 @vmovl_u8(vld1_u8(pu1_src_tmp)
add r10,r1,r6
subs r4,r4,#4 @wd - 4
vshl.i64 q0,q0,#6 @vshlq_n_s64(temp, 6)
vld1.8 {d22},[r5],r2 @vld1_u8(pu1_src_tmp)
add r0,r0,#4 @pu1_src += 4
vst1.64 {d0},[r1] @vst1q_lane_s64(pi2_dst_tmp, temp, 0)
add r1,r1,#8
vmovl.u8 q11,d22 @vmovl_u8(vld1_u8(pu1_src_tmp)
vld1.8 {d24},[r5],r2 @vld1_u8(pu1_src_tmp)
vshl.i64 q11,q11,#6 @vshlq_n_s64(temp, 6)
vmovl.u8 q12,d24 @vmovl_u8(vld1_u8(pu1_src_tmp)
vst1.64 {d22},[r10],r6 @vst1q_lane_s64(pi2_dst_tmp, temp, 0)
bgt inner_loop_wd_4_ht_2
b end_loops
core_loop_wd_8:
@sub r11,r12,#8
lsls r5,r3,#1
rsb r11,r12,r3, lsl #2 @ r11 = (dst_strd * 4) - width
rsb r8,r12,r2,lsl #2 @r2->src_strd
mov r4,r12, lsr #3 @ divide by 8
mov r7,r9
mul r7, r4
sub r4,r12,#0 @wd conditional check
sub r7,r7,#4 @subtract one for epilog
cmp r9,#0
beq core_loop_wd_8_ht_2
prolog:
add r6,r0,r2 @pu1_src_tmp += src_strd
add r10,r1,r5
vld1.8 {d8},[r0]! @vld1_u8(pu1_src_tmp)
vld1.8 {d10},[r6],r2 @vld1_u8(pu1_src_tmp)
vld1.8 {d12},[r6],r2 @vld1_u8(pu1_src_tmp)
vld1.8 {d14},[r6],r2 @vld1_u8(pu1_src_tmp)
vmovl.u8 q8,d8 @vmovl_u8(vld1_u8(pu1_src_tmp))
vmovl.u8 q9,d10 @vmovl_u8(vld1_u8(pu1_src_tmp)
vmovl.u8 q10,d12 @vmovl_u8(vld1_u8(pu1_src_tmp)
vmovl.u8 q11,d14 @vmovl_u8(vld1_u8(pu1_src_tmp)
subs r4,r4,#8 @wd decrements by 8
vshl.i16 q0,q8,#6 @vshlq_n_s16(tmp, 6)
vshl.i16 q1,q9,#6 @vshlq_n_s16(tmp, 6)
vshl.i16 q2,q10,#6 @vshlq_n_s16(tmp, 6)
vshl.i16 q3,q11,#6 @vshlq_n_s16(tmp, 6)
addle r0,r0,r8
add r6,r0,r2 @pu1_src_tmp += src_strd
vld1.8 {d8},[r0]! @vld1_u8(pu1_src_tmp)
vld1.8 {d10},[r6],r2 @vld1_u8(pu1_src_tmp)
vld1.8 {d12},[r6],r2 @vld1_u8(pu1_src_tmp)
vld1.8 {d14},[r6],r2 @vld1_u8(pu1_src_tmp)
vst1.16 {d0,d1},[r1]! @vst1q_s16(pi2_dst_tmp, tmp)
addle r1,r1,r11,lsl #1
suble r4,r12,#0 @wd conditional check
subs r7,r7,#4 @ht - 4
blt epilog_end @jumps to epilog_end
beq epilog @jumps to epilog
outer_loop_wd_8:
vst1.16 {d2,d3},[r10],r5 @vst1q_s16(pi2_dst_tmp, tmp)
vmovl.u8 q8,d8 @vmovl_u8(vld1_u8(pu1_src_tmp))
vst1.16 {d4,d5},[r10],r5 @vst1q_s16(pi2_dst_tmp, tmp)
vmovl.u8 q9,d10 @vmovl_u8(vld1_u8(pu1_src_tmp)
vst1.16 {d6,d7},[r10],r5 @vst1q_s16(pi2_dst_tmp, tmp)
vmovl.u8 q10,d12 @vmovl_u8(vld1_u8(pu1_src_tmp)
vmovl.u8 q11,d14 @vmovl_u8(vld1_u8(pu1_src_tmp)
subs r4,r4,#8 @wd decrements by 8
addle r0,r0,r8
add r6,r0,r2 @pu1_src_tmp += src_strd
vld1.8 {d8},[r0]! @vld1_u8(pu1_src_tmp)
vshl.i16 q0,q8,#6 @vshlq_n_s16(tmp, 6)
vld1.8 {d10},[r6],r2 @vld1_u8(pu1_src_tmp)
vshl.i16 q1,q9,#6 @vshlq_n_s16(tmp, 6)
vld1.8 {d12},[r6],r2 @vld1_u8(pu1_src_tmp)
vshl.i16 q2,q10,#6 @vshlq_n_s16(tmp, 6)
vld1.8 {d14},[r6],r2 @vld1_u8(pu1_src_tmp)
add r10,r1,r5
vshl.i16 q3,q11,#6 @vshlq_n_s16(tmp, 6)
vst1.16 {d0,d1},[r1]! @vst1q_s16(pi2_dst_tmp, tmp)
addle r1,r1,r11,lsl #1
suble r4,r12,#0 @wd conditional check
subs r7,r7,#4 @ht - 4
bgt outer_loop_wd_8
epilog:
vst1.16 {d2,d3},[r10],r5 @vst1q_s16(pi2_dst_tmp, tmp)
vmovl.u8 q8,d8 @vmovl_u8(vld1_u8(pu1_src_tmp))
vst1.16 {d4,d5},[r10],r5 @vst1q_s16(pi2_dst_tmp, tmp)
vmovl.u8 q9,d10 @vmovl_u8(vld1_u8(pu1_src_tmp)
vst1.16 {d6,d7},[r10],r5 @vst1q_s16(pi2_dst_tmp, tmp)
vmovl.u8 q10,d12 @vmovl_u8(vld1_u8(pu1_src_tmp)
vmovl.u8 q11,d14 @vmovl_u8(vld1_u8(pu1_src_tmp)
@add r6,r0,r2 @pu1_src_tmp += src_strd
vshl.i16 q0,q8,#6 @vshlq_n_s16(tmp, 6)
vshl.i16 q1,q9,#6 @vshlq_n_s16(tmp, 6)
vshl.i16 q2,q10,#6 @vshlq_n_s16(tmp, 6)
add r10,r1,r5
vshl.i16 q3,q11,#6 @vshlq_n_s16(tmp, 6)
vst1.16 {d0,d1},[r1]! @vst1q_s16(pi2_dst_tmp, tmp)
epilog_end:
vst1.16 {d2,d3},[r10],r5 @vst1q_s16(pi2_dst_tmp, tmp)
vst1.16 {d4,d5},[r10],r5 @vst1q_s16(pi2_dst_tmp, tmp)
vst1.16 {d6,d7},[r10],r5 @vst1q_s16(pi2_dst_tmp, tmp)
b end_loops
core_loop_wd_8_ht_2:
add r6,r0,r2 @pu1_src_tmp += src_strd
add r10,r1,r5
vld1.8 {d8},[r0]! @vld1_u8(pu1_src_tmp)
vld1.8 {d10},[r6],r2 @vld1_u8(pu1_src_tmp)
vmovl.u8 q8,d8 @vmovl_u8(vld1_u8(pu1_src_tmp))
vmovl.u8 q9,d10 @vmovl_u8(vld1_u8(pu1_src_tmp)
subs r12,r12,#8 @wd decrements by 8
vshl.i16 q0,q8,#6 @vshlq_n_s16(tmp, 6)
vshl.i16 q1,q9,#6 @vshlq_n_s16(tmp, 6)
vst1.16 {d0,d1},[r1]! @vst1q_s16(pi2_dst_tmp, tmp)
vst1.16 {d2,d3},[r10],r5 @vst1q_s16(pi2_dst_tmp, tmp)
bgt core_loop_wd_8_ht_2
ldmfd sp!,{r4-r12,r15} @reload the registers from sp