d461a47317
This moves the fields needed by asm near the top, before any structs or other members which complicate the offset calculation. Modifying other structs will no longer require updating the offsets, and the asm code is slightly simpler due to the smaller offsets. Signed-off-by: Mans Rullgard <mans@mansr.com>
104 lines
3.6 KiB
ArmAsm
104 lines
3.6 KiB
ArmAsm
/*
|
|
* Copyright (c) 2010 Mans Rullgard
|
|
*
|
|
* This file is part of FFmpeg.
|
|
*
|
|
* FFmpeg is free software; you can redistribute it and/or
|
|
* modify it under the terms of the GNU Lesser General Public
|
|
* License as published by the Free Software Foundation; either
|
|
* version 2.1 of the License, or (at your option) any later version.
|
|
*
|
|
* FFmpeg is distributed in the hope that it will be useful,
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
* Lesser General Public License for more details.
|
|
*
|
|
* You should have received a copy of the GNU Lesser General Public
|
|
* License along with FFmpeg; if not, write to the Free Software
|
|
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
|
|
*/
|
|
|
|
#include "asm.S"
|
|
#include "asm-offsets.h"
|
|
|
|
function ff_dct_unquantize_h263_inter_neon, export=1
|
|
add r12, r0, #BLOCK_LAST_INDEX
|
|
ldr r12, [r12, r2, lsl #2]
|
|
add r0, r0, #INTER_SCANTAB_RASTER_END
|
|
ldrb r12, [r0, r12]
|
|
sub r2, r3, #1
|
|
lsl r0, r3, #1
|
|
orr r2, r2, #1
|
|
add r3, r12, #1
|
|
endfunc
|
|
|
|
function ff_dct_unquantize_h263_neon, export=1
|
|
vdup.16 q15, r0 @ qmul
|
|
vdup.16 q14, r2 @ qadd
|
|
vneg.s16 q13, q14
|
|
cmp r3, #4
|
|
mov r0, r1
|
|
ble 2f
|
|
1:
|
|
vld1.16 {q0}, [r0,:128]!
|
|
vclt.s16 q3, q0, #0
|
|
vld1.16 {q8}, [r0,:128]!
|
|
vceq.s16 q1, q0, #0
|
|
vmul.s16 q2, q0, q15
|
|
vclt.s16 q11, q8, #0
|
|
vmul.s16 q10, q8, q15
|
|
vbsl q3, q13, q14
|
|
vbsl q11, q13, q14
|
|
vadd.s16 q2, q2, q3
|
|
vceq.s16 q9, q8, #0
|
|
vadd.s16 q10, q10, q11
|
|
vbif q0, q2, q1
|
|
vbif q8, q10, q9
|
|
subs r3, r3, #16
|
|
vst1.16 {q0}, [r1,:128]!
|
|
vst1.16 {q8}, [r1,:128]!
|
|
bxle lr
|
|
cmp r3, #8
|
|
bgt 1b
|
|
2:
|
|
vld1.16 {d0}, [r0,:64]
|
|
vclt.s16 d3, d0, #0
|
|
vceq.s16 d1, d0, #0
|
|
vmul.s16 d2, d0, d30
|
|
vbsl d3, d26, d28
|
|
vadd.s16 d2, d2, d3
|
|
vbif d0, d2, d1
|
|
vst1.16 {d0}, [r1,:64]
|
|
bx lr
|
|
endfunc
|
|
|
|
function ff_dct_unquantize_h263_intra_neon, export=1
|
|
push {r4-r6,lr}
|
|
add r12, r0, #BLOCK_LAST_INDEX
|
|
ldr r6, [r0, #AC_PRED]
|
|
add lr, r0, #INTER_SCANTAB_RASTER_END
|
|
cmp r6, #0
|
|
movne r12, #63
|
|
bne 1f
|
|
ldr r12, [r12, r2, lsl #2]
|
|
ldrb r12, [lr, r12]
|
|
1: ldr r5, [r0, #H263_AIC]
|
|
ldrsh r4, [r1]
|
|
cmp r5, #0
|
|
mov r5, r1
|
|
movne r2, #0
|
|
bne 2f
|
|
cmp r2, #4
|
|
addge r0, r0, #4
|
|
sub r2, r3, #1
|
|
ldr r6, [r0, #Y_DC_SCALE]
|
|
orr r2, r2, #1
|
|
smulbb r4, r4, r6
|
|
2: lsl r0, r3, #1
|
|
add r3, r12, #1
|
|
bl ff_dct_unquantize_h263_neon
|
|
vmov.16 d0[0], r4
|
|
vst1.16 {d0[0]}, [r5]
|
|
pop {r4-r6,pc}
|
|
endfunc
|