1
0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-12-23 12:43:46 +02:00
FFmpeg/libavcodec/arm/vp8dsp_armv6.S
Ronald S. Bultje a5dfeb612e VP8: armv6 optimizations.
From 52.503s (~40fps) to 27.973sec (~80fps) decoding of 480p sintel
trailer, i.e. a ~2x speedup overall, on a Nexus S.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2011-10-03 01:49:36 +02:00

2329 lines
100 KiB
ArmAsm

/**
* VP8 ARMv6 optimisations
*
* Copyright (c) 2011 The WebM project authors. All Rights Reserved.
* Copyright (c) 2010 Rob Clark <rob@ti.com>
* Copyright (c) 2011 Mans Rullgard <mans@mansr.com>
*
* This file is part of Libav.
*
* Libav 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.
*
* Libav 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 Libav; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*
* This code was partially ported from libvpx, which uses this license:
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*
* (Note that the "LICENSE", "AUTHORS" and "PATENTS" files can be
* found in the libvpx source tree.)
*/
#include "asm.S"
@ idct
@ void vp8_luma_dc_wht(DCTELEM block[4][4][16], DCTELEM dc[16])
function ff_vp8_luma_dc_wht_armv6, export=1
push {r4 - r10, lr}
@ load dc[] and zero memory
mov r12, #0
ldr r2, [r1] @ dc0[0,1]
ldr r3, [r1, #4] @ dc0[2,3]
ldr r4, [r1, #8] @ dc1[0,1]
ldr r5, [r1, #12] @ dc1[2,3]
ldr r6, [r1, #16] @ dc2[0,1]
ldr r7, [r1, #20] @ dc2[2,3]
ldr r8, [r1, #24] @ dc3[0,1]
ldr r9, [r1, #28] @ dc3[2,3]
str r12,[r1]
str r12,[r1, #4]
str r12,[r1, #8]
str r12,[r1, #12]
str r12,[r1, #16]
str r12,[r1, #20]
str r12,[r1, #24]
str r12,[r1, #28]
@ loop1
uadd16 r12, r2, r8 @ t0[0,1]
uadd16 r14, r3, r9 @ t0[2,3]
usub16 r2, r2, r8 @ t3[0,1]
usub16 r3, r3, r9 @ t3[2,3]
uadd16 r8, r4, r6 @ t1[0,1]
uadd16 r9, r5, r7 @ t1[2,3]
usub16 r4, r4, r6 @ t2[0,1]
usub16 r5, r5, r7 @ t2[2,3]
uadd16 r6, r12, r8 @ dc0[0,1]
uadd16 r7, r14, r9 @ dc0[2,3]
usub16 r12, r12, r8 @ dc2[0,1]
usub16 r14, r14, r9 @ dc2[2,3]
uadd16 r8, r2, r4 @ dc1[0,1]
uadd16 r9, r3, r5 @ dc1[2,3]
usub16 r2, r2, r4 @ dc3[0,1]
usub16 r3, r3, r5 @ dc3[2,3]
mov r1, #3
orr r1, r1, #0x30000 @ 3 | 3 (round)
@ "transpose"
pkhbt r4, r6, r8, lsl #16 @ dc{0,1}[0]
pkhtb r6, r8, r6, asr #16 @ dc{0,1}[1]
pkhbt r5, r12, r2, lsl #16 @ dc{2,3}[0]
pkhtb r12, r2, r12, asr #16 @ dc{2,3}[1]
pkhbt r8, r7, r9, lsl #16 @ dc{0,1}[2]
uadd16 r4, r4, r1
uadd16 r5, r5, r1
pkhtb r7, r9, r7, asr #16 @ dc{0,1}[3]
pkhbt r2, r14, r3, lsl #16 @ dc{2,3}[2]
pkhtb r14, r3, r14, asr #16 @ dc{2,3}[3]
@ loop2
uadd16 r9, r4, r7 @ t0[0,1]
uadd16 r3, r5, r14 @ t0[2,3]
usub16 r4, r4, r7 @ t3[0,1]
usub16 r5, r5, r14 @ t3[2,3]
uadd16 r7, r6, r8 @ t1[0,1]
uadd16 r14, r12, r2 @ t1[2,3]
usub16 r6, r6, r8 @ t2[0,1]
usub16 r12, r12, r2 @ t2[2,3]
uadd16 r8, r9, r7 @ block[0,1][0]
uadd16 r2, r3, r14 @ block[2,3][0]
usub16 r9, r9, r7 @ block[0,1][2]
usub16 r3, r3, r14 @ block[2,3][2]
uadd16 r7, r4, r6 @ block[0,1][1]
uadd16 r14, r5, r12 @ block[2,3][1]
usub16 r4, r4, r6 @ block[0,1][3]
usub16 r5, r5, r12 @ block[2,3][3]
@ store
mov r6, r8, asr #19 @ block[1][0]
mov r12, r7, asr #19 @ block[1][1]
mov r1, r9, asr #19 @ block[1][2]
mov r10, r4, asr #19 @ block[1][3]
sxth r8, r8
sxth r7, r7
sxth r9, r9
sxth r4, r4
asr r8, #3 @ block[0][0]
asr r7, #3 @ block[0][1]
asr r9, #3 @ block[0][2]
asr r4, #3 @ block[0][3]
strh r8, [r0], #32
strh r7, [r0], #32
strh r9, [r0], #32
strh r4, [r0], #32
strh r6, [r0], #32
strh r12,[r0], #32
strh r1, [r0], #32
strh r10,[r0], #32
mov r6, r2, asr #19 @ block[3][0]
mov r12, r14, asr #19 @ block[3][1]
mov r1, r3, asr #19 @ block[3][2]
mov r10, r5, asr #19 @ block[3][3]
sxth r2, r2
sxth r14, r14
sxth r3, r3
sxth r5, r5
asr r2, #3 @ block[2][0]
asr r14, #3 @ block[2][1]
asr r3, #3 @ block[2][2]
asr r5, #3 @ block[2][3]
strh r2, [r0], #32
strh r14,[r0], #32
strh r3, [r0], #32
strh r5, [r0], #32
strh r6, [r0], #32
strh r12,[r0], #32
strh r1, [r0], #32
strh r10,[r0], #32
pop {r4 - r10, pc}
endfunc
@ void vp8_luma_dc_wht_dc(DCTELEM block[4][4][16], DCTELEM dc[16])
function ff_vp8_luma_dc_wht_dc_armv6, export=1
ldrsh r2, [r1]
mov r3, #0
add r2, r2, #3
strh r3, [r1]
asr r2, r2, #3
.rept 16
strh r2, [r0], #32
.endr
bx lr
endfunc
@ void vp8_idct_add(uint8_t *dst, DCTELEM block[16], int stride)
function ff_vp8_idct_add_armv6, export=1
push {r4 - r11, lr}
sub sp, sp, #32
mov r3, #0x00004E00 @ cos
orr r3, r3, #0x0000007B @ cospi8sqrt2minus1 = 20091
mov r4, #0x00008A00 @ sin
orr r4, r4, #0x0000008C @ sinpi8sqrt2 = 35468
mov r5, #0x2 @ i=2
1:
ldr r6, [r1, #8] @ i5 | i4 = block1[1] | block1[0]
ldr r12,[r1, #24] @ i13 | i12 = block3[1] | block3[0]
ldr r14,[r1, #16] @ i9 | i8 = block2[1] | block2[0]
smulwt r9, r3, r6 @ (ip[5] * cospi8sqrt2minus1) >> 16
smulwb r7, r3, r6 @ (ip[4] * cospi8sqrt2minus1) >> 16
smulwt r10, r4, r6 @ (ip[5] * sinpi8sqrt2) >> 16
smulwb r8, r4, r6 @ (ip[4] * sinpi8sqrt2) >> 16
pkhbt r7, r7, r9, lsl #16 @ 5c | 4c
smulwt r11, r3, r12 @ (ip[13] * cospi8sqrt2minus1) >> 16
pkhbt r8, r8, r10, lsl #16 @ 5s | 4s = t2 first half
uadd16 r6, r6, r7 @ 5c+5 | 4c+4 = t3 first half
smulwt r7, r4, r12 @ (ip[13] * sinpi8sqrt2) >> 16
smulwb r9, r3, r12 @ (ip[12] * cospi8sqrt2minus1) >> 16
smulwb r10, r4, r12 @ (ip[12] * sinpi8sqrt2) >> 16
subs r5, r5, #1 @ i--
pkhbt r9, r9, r11, lsl #16 @ 13c | 12c
ldr r11,[r1] @ i1 | i0
pkhbt r10, r10, r7, lsl #16 @ 13s | 12s = t3 second half
uadd16 r7, r12, r9 @ 13c+13 | 12c+12 = t2 second half
usub16 r7, r8, r7 @ c = t2
uadd16 r6, r6, r10 @ d = t3
uadd16 r10, r11, r14 @ a = t0
usub16 r8, r11, r14 @ b = t1
uadd16 r9, r10, r6 @ a+d = tmp{0,1}[0]
usub16 r10, r10, r6 @ a-d = tmp{0,1}[3]
uadd16 r6, r8, r7 @ b+c = tmp{0,1}[1]
usub16 r7, r8, r7 @ b-c = tmp{0,1}[2]
mov r8, #0
str r6, [sp, #8] @ o5 | o4
str r7, [sp, #16] @ o9 | o8
str r10,[sp, #24] @ o13 | o12
str r9, [sp], #4 @ o1 | o0
str r8, [r1, #24]
str r8, [r1, #16]
str r8, [r1, #8]
str r8, [r1], #4
bne 1b
mov r5, #0x2 @ i=2
sub sp, sp, #8
2:
ldr r6, [sp, #8] @ i5 | i4 = tmp{0,1}[1]
ldr r14,[sp, #4] @ i3 | i2 = tmp{2,3}[0]
ldr r12,[sp, #12] @ i7 | i6 = tmp{2,3}[1]
ldr r1, [sp], #16 @ i1 | i0 = tmp{0,1}[0]
smulwt r9, r3, r6 @ (ip[5] * cospi8sqrt2minus1) >> 16
smulwt r7, r3, r1 @ (ip[1] * cospi8sqrt2minus1) >> 16
smulwt r10, r4, r6 @ (ip[5] * sinpi8sqrt2) >> 16
smulwt r8, r4, r1 @ (ip[1] * sinpi8sqrt2) >> 16
pkhbt r11, r1, r6, lsl #16 @ i4 | i0 = t0/t1 first half
pkhbt r7, r7, r9, lsl #16 @ 5c | 1c
pkhbt r8, r8, r10, lsl #16 @ 5s | 1s = temp1 = t2 first half
pkhtb r1, r6, r1, asr #16 @ i5 | i1
uadd16 r1, r7, r1 @ 5c+5 | 1c+1 = temp2 (d) = t3 first half
pkhbt r9, r14, r12, lsl #16 @ i6 | i2 = t0/t1 second half
uadd16 r10, r11, r9 @ a = t0
usub16 r9, r11, r9 @ b = t1
pkhtb r6, r12, r14, asr #16 @ i7 | i3
subs r5, r5, #0x1 @ i--
smulwt r7, r3, r6 @ (ip[7] * cospi8sqrt2minus1) >> 16
smulwt r11, r4, r6 @ (ip[7] * sinpi8sqrt2) >> 16
smulwb r12, r3, r6 @ (ip[3] * cospi8sqrt2minus1) >> 16
smulwb r14, r4, r6 @ (ip[3] * sinpi8sqrt2) >> 16
pkhbt r7, r12, r7, lsl #16 @ 7c | 3c
pkhbt r11, r14, r11, lsl #16 @ 7s | 3s = temp1 (d) = t3 second half
mov r14, #0x4 @ set up 4's
orr r14, r14, #0x40000 @ 4|4
uadd16 r6, r7, r6 @ 7c+7 | 3c+3 = temp2 (c) = t2 second half
usub16 r12, r8, r6 @ c (o5 | o1) = t2
uadd16 r6, r11, r1 @ d (o7 | o3) = t3
uadd16 r10, r10, r14 @ t0 + 4
uadd16 r9, r9, r14 @ t1 + 4
uadd16 r7, r10, r6 @ a+d = dst{0,1}[0]
usub16 r6, r10, r6 @ a-d = dst{0,1}[3]
uadd16 r10, r9, r12 @ b+c = dst{0,1}[1]
usub16 r1, r9, r12 @ b-c = dst{0,1}[2]
mov r9, r6, asr #3 @ o[1][3]
mov r12, r1, asr #3 @ o[1][2]
pkhtb r8, r12, r7, asr #19 @ o[1][0,2]
pkhtb r11, r9, r10, asr #19 @ o[1][1,3]
ldr r12,[r0]
ldr r9, [r0, r2]
sxth r7, r7
sxth r6, r6
sxth r10, r10
sxth r1, r1
asr r7, #3 @ o[0][0]
asr r10, #3 @ o[0][1]
pkhbt r7, r7, r1, lsl #13 @ o[0][0,2]
pkhbt r10, r10, r6, lsl #13 @ o[0][1,3]
uxtab16 r7, r7, r12
uxtab16 r10, r10, r12, ror #8
uxtab16 r8, r8, r9
uxtab16 r11, r11, r9, ror #8
usat16 r7, #8, r7
usat16 r10, #8, r10
usat16 r8, #8, r8
usat16 r11, #8, r11
orr r7, r7, r10, lsl #8
orr r8, r8, r11, lsl #8
str r8, [r0, r2]
str_post r7, r0, r2, lsl #1
bne 2b
pop {r4 - r11, pc}
endfunc
@ void vp8_idct_dc_add(uint8_t *dst, DCTELEM block[16], int stride)
function ff_vp8_idct_dc_add_armv6, export=1
push {r4 - r5, lr}
ldrsh r3, [r1]
mov r4, #0
add r3, r3, #4
asr r3, #3
strh r4, [r1], #32
ldr r4, [r0, r2]
ldr_post r5, r0, r2, lsl #1
pkhbt r3, r3, r3, lsl #16
uxtab16 lr, r3, r5 @ a1+2 | a1+0
uxtab16 r5, r3, r5, ror #8 @ a1+3 | a1+1
uxtab16 r12, r3, r4
uxtab16 r4, r3, r4, ror #8
usat16 lr, #8, lr
usat16 r5, #8, r5
usat16 r12, #8, r12
usat16 r4, #8, r4
orr lr, lr, r5, lsl #8
orr r12, r12, r4, lsl #8
ldr r5, [r0]
ldr r4, [r0, r2]
sub r0, r0, r2, lsl #1
str r12,[r0, r2]
str_post lr, r0, r2, lsl #1
uxtab16 lr, r3, r5
uxtab16 r5, r3, r5, ror #8
uxtab16 r12, r3, r4
uxtab16 r4, r3, r4, ror #8
usat16 lr, #8, lr
usat16 r5, #8, r5
usat16 r12, #8, r12
usat16 r4, #8, r4
orr lr, lr, r5, lsl #8
orr r12, r12, r4, lsl #8
str r12,[r0, r2]
str_post lr, r0, r2, lsl #1
pop {r4 - r5, pc}
endfunc
@ void vp8_idct_dc_add4uv(uint8_t *dst, DCTELEM block[4][16], int stride)
function ff_vp8_idct_dc_add4uv_armv6, export=1
push {lr}
bl ff_vp8_idct_dc_add_armv6
sub r0, r0, r2, lsl #2
add r0, r0, #4
bl ff_vp8_idct_dc_add_armv6
sub r0, r0, #4
bl ff_vp8_idct_dc_add_armv6
sub r0, r0, r2, lsl #2
add r0, r0, #4
bl ff_vp8_idct_dc_add_armv6
pop {pc}
endfunc
@ void vp8_idct_dc_add4y(uint8_t *dst, DCTELEM block[4][16], int stride)
function ff_vp8_idct_dc_add4y_armv6, export=1
push {lr}
bl ff_vp8_idct_dc_add_armv6
sub r0, r0, r2, lsl #2
add r0, r0, #4
bl ff_vp8_idct_dc_add_armv6
sub r0, r0, r2, lsl #2
add r0, r0, #4
bl ff_vp8_idct_dc_add_armv6
sub r0, r0, r2, lsl #2
add r0, r0, #4
bl ff_vp8_idct_dc_add_armv6
pop {pc}
endfunc
@ loopfilter
@ void vp8_v_loop_filter16_simple(uint8_t *dst, int stride, int flim)
function ff_vp8_v_loop_filter16_simple_armv6, export=1
push {r4 - r11, lr}
ldr_dpren r3, r0, r1, lsl #1 @ p1
ldr_dpren r4, r0, r1 @ p0
ldr r5, [r0] @ q0
ldr r6, [r0, r1] @ q1
orr r2, r2, r2, lsl #16
mov r9, #4 @ count
mov lr, #0 @ need 0 in a couple places
orr r12, r2, r2, lsl #8 @ splat int -> byte
ldr r2, c0x80808080
1:
@ vp8_simple_filter_mask()
uqsub8 r7, r3, r6 @ p1 - q1
uqsub8 r8, r6, r3 @ q1 - p1
uqsub8 r10, r4, r5 @ p0 - q0
uqsub8 r11, r5, r4 @ q0 - p0
orr r8, r8, r7 @ abs(p1 - q1)
orr r10, r10, r11 @ abs(p0 - q0)
uqadd8 r10, r10, r10 @ abs(p0 - q0) * 2
uhadd8 r8, r8, lr @ abs(p1 - q2) >> 1
uqadd8 r10, r10, r8 @ abs(p0 - q0)*2 + abs(p1 - q1)/2
mvn r8, #0
usub8 r10, r12, r10 @ compare to flimit. usub8 sets GE flags
sel r10, r8, lr @ filter mask: F or 0
cmp r10, #0
beq 2f @ skip filtering if all masks are 0x00
@ vp8_simple_filter()
eor r3, r3, r2 @ p1 offset to convert to a signed value
eor r6, r6, r2 @ q1 offset to convert to a signed value
eor r4, r4, r2 @ p0 offset to convert to a signed value
eor r5, r5, r2 @ q0 offset to convert to a signed value
qsub8 r3, r3, r6 @ vp8_filter = p1 - q1
qsub8 r6, r5, r4 @ q0 - p0
qadd8 r3, r3, r6 @ += q0 - p0
ldr r7, c0x04040404
qadd8 r3, r3, r6 @ += q0 - p0
ldr r8, c0x03030303
qadd8 r3, r3, r6 @ vp8_filter = p1-q1 + 3*(q0-p0))
@STALL
and r3, r3, r10 @ vp8_filter &= mask
qadd8 r7, r3, r7 @ Filter1 = vp8_filter + 4
qadd8 r8, r3, r8 @ Filter2 = vp8_filter + 3
shadd8 r7, r7, lr
shadd8 r8, r8, lr
shadd8 r7, r7, lr
shadd8 r8, r8, lr
shadd8 r7, r7, lr @ Filter1 >>= 3
shadd8 r8, r8, lr @ Filter2 >>= 3
qsub8 r5, r5, r7 @ u = q0 - Filter1
qadd8 r4, r4, r8 @ u = p0 + Filter2
eor r5, r5, r2 @ *oq0 = u^0x80
eor r4, r4, r2 @ *op0 = u^0x80
T sub r7, r0, r1
str r5, [r0] @ store oq0 result
A str r4, [r0, -r1] @ store op0 result
T str r4, [r7]
2:
subs r9, r9, #1 @ counter--
add r0, r0, #4 @ next row
T itttt ne
A ldrne r3, [r0, -r1, lsl #1] @ p1
T subne r3, r0, r1, lsl #1
T ldrne r3, [r3] @ p1
A ldrne r4, [r0, -r1] @ p0
T subne r4, r0, r1
T ldrne r4, [r4] @ p0
T itt ne
ldrne r5, [r0] @ q0
ldrne r6, [r0, r1] @ q1
bne 1b
pop {r4 - r11, pc}
endfunc
c0x01010101: .long 0x01010101
c0x03030303: .long 0x03030303
c0x04040404: .long 0x04040404
c0x7F7F7F7F: .long 0x7F7F7F7F
c0x80808080: .long 0x80808080
@ void vp8_v_loop_filter16_inner(uint8_t *dst, int stride,
@ int fE, int fI, int hev_thresh)
@ and
@ void vp8_v_loop_filter8uv_inner(uint8_t *dstU, uint8_t *dstV, int stride,
@ int fE, int fI, int hev_thresh)
@ call:
@ void vp8_v_loop_filter_inner(uint8_t *dst, int stride,
@ int fE, int fI, int hev_thresh, int count)
function ff_vp8_v_loop_filter_inner_armv6, export=1
push {r4 - r11, lr}
sub r0, r0, r1, lsl #2 @ move r0 pointer down by 4 lines
ldr r5, [sp, #40] @ counter
ldr r6, [sp, #36] @ load thresh address
sub sp, sp, #16 @ create temp buffer
ldr r10,[r0, r1] @ p2
ldr_post r9, r0, r1, lsl #1 @ p3
ldr r12,[r0, r1] @ p0
ldr_post r11, r0, r1, lsl #1 @ p1
orr r2, r2, r2, lsl #16
orr r3, r3, r3, lsl #16
orr r6, r6, r6, lsl #16
orr r4, r2, r2, lsl #8 @ flimE splat int -> byte
orr r2, r3, r3, lsl #8 @ flimI splat int -> byte
orr r3, r6, r6, lsl #8 @ thresh splat int -> byte
1:
@ vp8_filter_mask() function
@ calculate breakout conditions
uqsub8 r6, r9, r10 @ p3 - p2
uqsub8 r7, r10, r9 @ p2 - p3
uqsub8 r8, r10, r11 @ p2 - p1
uqsub8 r10, r11, r10 @ p1 - p2
orr r6, r6, r7 @ abs (p3-p2)
orr r8, r8, r10 @ abs (p2-p1)
uqsub8 lr, r6, r2 @ compare to limit. lr: vp8_filter_mask
uqsub8 r8, r8, r2 @ compare to limit
uqsub8 r6, r11, r12 @ p1 - p0
orr lr, lr, r8
uqsub8 r7, r12, r11 @ p0 - p1
ldr r10,[r0, r1] @ q1
ldr_post r9, r0, r1, lsl #1 @ q0
orr r6, r6, r7 @ abs (p1-p0)
uqsub8 r7, r6, r2 @ compare to limit
uqsub8 r8, r6, r3 @ compare to thresh -- save r8 for later
orr lr, lr, r7
uqsub8 r6, r11, r10 @ p1 - q1
uqsub8 r7, r10, r11 @ q1 - p1
uqsub8 r11, r12, r9 @ p0 - q0
uqsub8 r12, r9, r12 @ q0 - p0
orr r6, r6, r7 @ abs (p1-q1)
ldr r7, c0x7F7F7F7F
orr r12, r11, r12 @ abs (p0-q0)
ldr_post r11, r0, r1 @ q2
uqadd8 r12, r12, r12 @ abs (p0-q0) * 2
and r6, r7, r6, lsr #1 @ abs (p1-q1) / 2
uqsub8 r7, r9, r10 @ q0 - q1
uqadd8 r12, r12, r6 @ abs (p0-q0)*2 + abs (p1-q1)/2
uqsub8 r6, r10, r9 @ q1 - q0
uqsub8 r12, r12, r4 @ compare to flimit
uqsub8 r9, r11, r10 @ q2 - q1
orr lr, lr, r12
ldr_post r12, r0, r1 @ q3
uqsub8 r10, r10, r11 @ q1 - q2
orr r6, r7, r6 @ abs (q1-q0)
orr r10, r9, r10 @ abs (q2-q1)
uqsub8 r7, r6, r2 @ compare to limit
uqsub8 r10, r10, r2 @ compare to limit
uqsub8 r6, r6, r3 @ compare to thresh -- save r6 for later
orr lr, lr, r7
orr lr, lr, r10
uqsub8 r10, r12, r11 @ q3 - q2
uqsub8 r9, r11, r12 @ q2 - q3
mvn r11, #0 @ r11 == -1
orr r10, r10, r9 @ abs (q3-q2)
uqsub8 r10, r10, r2 @ compare to limit
mov r12, #0
orr lr, lr, r10
sub r0, r0, r1, lsl #2
usub8 lr, r12, lr @ use usub8 instead of ssub8
sel lr, r11, r12 @ filter mask: lr
cmp lr, #0
beq 2f @ skip filtering
sub r0, r0, r1, lsl #1 @ move r0 pointer down by 6 lines
@vp8_hevmask() function
@calculate high edge variance
orr r10, r6, r8 @ calculate vp8_hevmask
usub8 r10, r12, r10 @ use usub8 instead of ssub8
sel r6, r12, r11 @ obtain vp8_hevmask: r6
@vp8_filter() function
ldr r8, [r0, r1] @ p0
ldr_post r7, r0, r1, lsl #1 @ p1
ldr r12, c0x80808080
ldr r10,[r0, r1] @ q1
ldr_post r9, r0, r1, lsl #1 @ q0
eor r7, r7, r12 @ p1 offset to convert to a signed value
eor r8, r8, r12 @ p0 offset to convert to a signed value
eor r9, r9, r12 @ q0 offset to convert to a signed value
eor r10, r10, r12 @ q1 offset to convert to a signed value
str r9, [sp] @ store qs0 temporarily
str r8, [sp, #4] @ store ps0 temporarily
str r10,[sp, #8] @ store qs1 temporarily
str r7, [sp, #12] @ store ps1 temporarily
qsub8 r7, r7, r10 @ vp8_signed_char_clamp(ps1-qs1)
qsub8 r8, r9, r8 @ vp8_signed_char_clamp(vp8_filter + 3 * ( qs0 - ps0))
and r7, r7, r6 @ vp8_filter (r7) &= hev
qadd8 r7, r7, r8
ldr r9, c0x03030303 @ r9 = 3 --modified for vp8
qadd8 r7, r7, r8
ldr r10, c0x04040404
qadd8 r7, r7, r8
and r7, r7, lr @ vp8_filter &= mask@
qadd8 r8, r7, r9 @ Filter2 (r8) = vp8_signed_char_clamp(vp8_filter+3)
qadd8 r7, r7, r10 @ vp8_filter = vp8_signed_char_clamp(vp8_filter+4)
mov r9, #0
shadd8 r8, r8, r9 @ Filter2 >>= 3
shadd8 r7, r7, r9 @ vp8_filter >>= 3
shadd8 r8, r8, r9
shadd8 r7, r7, r9
shadd8 lr, r8, r9 @ lr: Filter2
shadd8 r7, r7, r9 @ r7: filter
@calculate output
ldr r8, [sp] @ load qs0
ldr r9, [sp, #4] @ load ps0
ldr r10, c0x01010101
qsub8 r8, r8, r7 @ u = vp8_signed_char_clamp(qs0 - vp8_filter)
qadd8 r9, r9, lr @ u = vp8_signed_char_clamp(ps0 + Filter2)
mov lr, #0
sadd8 r7, r7, r10 @ vp8_filter += 1
shadd8 r7, r7, lr @ vp8_filter >>= 1
ldr r11,[sp, #12] @ load ps1
ldr r10,[sp, #8] @ load qs1
bic r7, r7, r6 @ vp8_filter &= ~hev
sub r0, r0, r1, lsl #2
qadd8 r11, r11, r7 @ u = vp8_signed_char_clamp(ps1 + vp8_filter)
qsub8 r10, r10, r7 @ u = vp8_signed_char_clamp(qs1 - vp8_filter)
eor r11, r11, r12 @ *op1 = u^0x80
eor r9, r9, r12 @ *op0 = u^0x80
eor r8, r8, r12 @ *oq0 = u^0x80
eor r10, r10, r12 @ *oq1 = u^0x80
str r9, [r0, r1] @ store op0 result
str_post r11, r0, r1, lsl #1 @ store op1
str r10,[r0, r1] @ store oq1
str_post r8, r0, r1, lsl #1 @ store oq0 result
sub r0, r0, r1, lsl #1
2:
add r0, r0, #4
sub r0, r0, r1, lsl #2
subs r5, r5, #1
T ittt ne
ldrne r10,[r0, r1] @ p2
A ldrne r9, [r0], r1, lsl #1 @ p3
T ldrne r9, [r0] @ p3
T addne r0, r0, r1, lsl #1
T ittt ne
ldrne r12,[r0, r1] @ p0
A ldrne r11,[r0], r1, lsl #1 @ p1
T ldrne r11,[r0] @ p3
T addne r0, r0, r1, lsl #1
bne 1b
add sp, sp, #16
pop {r4 - r11, pc}
endfunc
@ void vp8_v_loop_filter16(uint8_t *dst, int stride,
@ int fE, int fI, int hev_thresh)
@ and
@ void vp8_v_loop_filter8uv(uint8_t *dstU, uint8_t *dstV, int stride,
@ int fE, int fI, int hev_thresh)
@ call:
@ void vp8_v_loop_filter(uint8_t *dst, int stride,
@ int fE, int fI, int hev_thresh, int count)
function ff_vp8_v_loop_filter_armv6, export=1
push {r4 - r11, lr}
sub r0, r0, r1, lsl #2 @ move r0 pointer down by 4 lines
ldr r5, [sp, #40] @ counter
ldr r6, [sp, #36] @ load thresh address
sub sp, sp, #16 @ create temp buffer
ldr r10,[r0, r1] @ p2
ldr_post r9, r0, r1, lsl #1 @ p3
ldr r12,[r0, r1] @ p0
ldr_post r11, r0, r1, lsl #1 @ p1
orr r2, r2, r2, lsl #16
orr r3, r3, r3, lsl #16
orr r6, r6, r6, lsl #16
orr r4, r2, r2, lsl #8 @ flimE splat int -> byte
orr r2, r3, r3, lsl #8 @ flimI splat int -> byte
orr r3, r6, r6, lsl #8 @ thresh splat int -> byte
1:
@ vp8_filter_mask() function
@ calculate breakout conditions
uqsub8 r6, r9, r10 @ p3 - p2
uqsub8 r7, r10, r9 @ p2 - p3
uqsub8 r8, r10, r11 @ p2 - p1
uqsub8 r10, r11, r10 @ p1 - p2
orr r6, r6, r7 @ abs (p3-p2)
orr r8, r8, r10 @ abs (p2-p1)
uqsub8 lr, r6, r2 @ compare to limit. lr: vp8_filter_mask
uqsub8 r8, r8, r2 @ compare to limit
uqsub8 r6, r11, r12 @ p1 - p0
orr lr, lr, r8
uqsub8 r7, r12, r11 @ p0 - p1
ldr r10,[r0, r1] @ q1
ldr_post r9, r0, r1, lsl #1 @ q0
orr r6, r6, r7 @ abs (p1-p0)
uqsub8 r7, r6, r2 @ compare to limit
uqsub8 r8, r6, r3 @ compare to thresh -- save r8 for later
orr lr, lr, r7
uqsub8 r6, r11, r10 @ p1 - q1
uqsub8 r7, r10, r11 @ q1 - p1
uqsub8 r11, r12, r9 @ p0 - q0
uqsub8 r12, r9, r12 @ q0 - p0
orr r6, r6, r7 @ abs (p1-q1)
ldr r7, c0x7F7F7F7F
orr r12, r11, r12 @ abs (p0-q0)
ldr_post r11, r0, r1 @ q2
uqadd8 r12, r12, r12 @ abs (p0-q0) * 2
and r6, r7, r6, lsr #1 @ abs (p1-q1) / 2
uqsub8 r7, r9, r10 @ q0 - q1
uqadd8 r12, r12, r6 @ abs (p0-q0)*2 + abs (p1-q1)/2
uqsub8 r6, r10, r9 @ q1 - q0
uqsub8 r12, r12, r4 @ compare to flimit
uqsub8 r9, r11, r10 @ q2 - q1
orr lr, lr, r12
ldr_post r12, r0, r1 @ q3
uqsub8 r10, r10, r11 @ q1 - q2
orr r6, r7, r6 @ abs (q1-q0)
orr r10, r9, r10 @ abs (q2-q1)
uqsub8 r7, r6, r2 @ compare to limit
uqsub8 r10, r10, r2 @ compare to limit
uqsub8 r6, r6, r3 @ compare to thresh -- save r6 for later
orr lr, lr, r7
orr lr, lr, r10
uqsub8 r10, r12, r11 @ q3 - q2
uqsub8 r9, r11, r12 @ q2 - q3
mvn r11, #0 @ r11 == -1
orr r10, r10, r9 @ abs (q3-q2)
uqsub8 r10, r10, r2 @ compare to limit
mov r12, #0
orr lr, lr, r10
usub8 lr, r12, lr @ use usub8 instead of ssub8
sel lr, r11, r12 @ filter mask: lr
cmp lr, #0
beq 2f @ skip filtering
@vp8_hevmask() function
@calculate high edge variance
sub r0, r0, r1, lsl #2 @ move r0 pointer down by 6 lines
sub r0, r0, r1, lsl #1
orr r10, r6, r8
usub8 r10, r12, r10
sel r6, r12, r11 @ hev mask: r6
@vp8_mbfilter() function
@p2, q2 are only needed at the end. Do not need to load them in now.
ldr r8, [r0, r1] @ p0
ldr_post r7, r0, r1, lsl #1 @ p1
ldr r12, c0x80808080
ldr_post r9, r0, r1 @ q0
ldr r10,[r0] @ q1
eor r7, r7, r12 @ ps1
eor r8, r8, r12 @ ps0
eor r9, r9, r12 @ qs0
eor r10, r10, r12 @ qs1
qsub8 r12, r9, r8 @ vp8_signed_char_clamp(vp8_filter + 3 * ( qs0 - ps0))
str r7, [sp, #12] @ store ps1 temporarily
qsub8 r7, r7, r10 @ vp8_signed_char_clamp(ps1-qs1)
str r10,[sp, #8] @ store qs1 temporarily
qadd8 r7, r7, r12
str r9, [sp] @ store qs0 temporarily
qadd8 r7, r7, r12
str r8, [sp, #4] @ store ps0 temporarily
qadd8 r7, r7, r12 @ vp8_filter: r7
ldr r10, c0x03030303 @ r10 = 3 --modified for vp8
ldr r9, c0x04040404
and r7, r7, lr @ vp8_filter &= mask (lr is free)
mov r12, r7 @ Filter2: r12
and r12, r12, r6 @ Filter2 &= hev
@save bottom 3 bits so that we round one side +4 and the other +3
qadd8 r8, r12, r9 @ Filter1 (r8) = vp8_signed_char_clamp(Filter2+4)
qadd8 r12, r12, r10 @ Filter2 (r12) = vp8_signed_char_clamp(Filter2+3)
mov r10, #0
shadd8 r8, r8, r10 @ Filter1 >>= 3
shadd8 r12, r12, r10 @ Filter2 >>= 3
shadd8 r8, r8, r10
shadd8 r12, r12, r10
shadd8 r8, r8, r10 @ r8: Filter1
shadd8 r12, r12, r10 @ r12: Filter2
ldr r9, [sp] @ load qs0
ldr r11,[sp, #4] @ load ps0
qsub8 r9, r9, r8 @ qs0 = vp8_signed_char_clamp(qs0 - Filter1)
qadd8 r11, r11, r12 @ ps0 = vp8_signed_char_clamp(ps0 + Filter2)
bic r12, r7, r6 @ vp8_filter &= ~hev ( r6 is free)
@roughly 3/7th difference across boundary
mov lr, #0x1b @ 27
mov r7, #0x3f @ 63
sxtb16 r6, r12
sxtb16 r10, r12, ror #8
smlabb r8, r6, lr, r7
smlatb r6, r6, lr, r7
smlabb r7, r10, lr, r7
smultb r10, r10, lr
ssat r8, #8, r8, asr #7
ssat r6, #8, r6, asr #7
add r10, r10, #63
ssat r7, #8, r7, asr #7
ssat r10, #8, r10, asr #7
ldr lr, c0x80808080
pkhbt r6, r8, r6, lsl #16
pkhbt r10, r7, r10, lsl #16
uxtb16 r6, r6
uxtb16 r10, r10
sub r0, r0, r1
orr r10, r6, r10, lsl #8 @ u = vp8_signed_char_clamp((63 + Filter2 * 27)>>7)
qsub8 r8, r9, r10 @ s = vp8_signed_char_clamp(qs0 - u)
qadd8 r10, r11, r10 @ s = vp8_signed_char_clamp(ps0 + u)
eor r8, r8, lr @ *oq0 = s^0x80
str r8, [r0] @ store *oq0
sub r0, r0, r1
eor r10, r10, lr @ *op0 = s^0x80
str r10,[r0] @ store *op0
@roughly 2/7th difference across boundary
mov lr, #0x12 @ 18
mov r7, #0x3f @ 63
sxtb16 r6, r12
sxtb16 r10, r12, ror #8
smlabb r8, r6, lr, r7
smlatb r6, r6, lr, r7
smlabb r9, r10, lr, r7
smlatb r10, r10, lr, r7
ssat r8, #8, r8, asr #7
ssat r6, #8, r6, asr #7
ssat r9, #8, r9, asr #7
ssat r10, #8, r10, asr #7
ldr lr, c0x80808080
pkhbt r6, r8, r6, lsl #16
pkhbt r10, r9, r10, lsl #16
ldr r9, [sp, #8] @ load qs1
ldr r11, [sp, #12] @ load ps1
uxtb16 r6, r6
uxtb16 r10, r10
sub r0, r0, r1
orr r10, r6, r10, lsl #8 @ u = vp8_signed_char_clamp((63 + Filter2 * 18)>>7)
qadd8 r11, r11, r10 @ s = vp8_signed_char_clamp(ps1 + u)
qsub8 r8, r9, r10 @ s = vp8_signed_char_clamp(qs1 - u)
eor r11, r11, lr @ *op1 = s^0x80
str_post r11, r0, r1 @ store *op1
eor r8, r8, lr @ *oq1 = s^0x80
add r0, r0, r1, lsl #1
mov r7, #0x3f @ 63
str_post r8, r0, r1 @ store *oq1
@roughly 1/7th difference across boundary
mov lr, #0x9 @ 9
ldr r9, [r0] @ load q2
sxtb16 r6, r12
sxtb16 r10, r12, ror #8
smlabb r8, r6, lr, r7
smlatb r6, r6, lr, r7
smlabb r12, r10, lr, r7
smlatb r10, r10, lr, r7
ssat r8, #8, r8, asr #7
ssat r6, #8, r6, asr #7
ssat r12, #8, r12, asr #7
ssat r10, #8, r10, asr #7
sub r0, r0, r1, lsl #2
pkhbt r6, r8, r6, lsl #16
pkhbt r10, r12, r10, lsl #16
sub r0, r0, r1
ldr lr, c0x80808080
ldr r11, [r0] @ load p2
uxtb16 r6, r6
uxtb16 r10, r10
eor r9, r9, lr
eor r11, r11, lr
orr r10, r6, r10, lsl #8 @ u = vp8_signed_char_clamp((63 + Filter2 * 9)>>7)
qadd8 r8, r11, r10 @ s = vp8_signed_char_clamp(ps2 + u)
qsub8 r10, r9, r10 @ s = vp8_signed_char_clamp(qs2 - u)
eor r8, r8, lr @ *op2 = s^0x80
str_post r8, r0, r1, lsl #2 @ store *op2
add r0, r0, r1
eor r10, r10, lr @ *oq2 = s^0x80
str_post r10, r0, r1, lsl #1 @ store *oq2
2:
add r0, r0, #4
sub r0, r0, r1, lsl #3
subs r5, r5, #1
T ittt ne
ldrne r10,[r0, r1] @ p2
A ldrne r9, [r0], r1, lsl #1 @ p3
T ldrne r9, [r0] @ p3
T addne r0, r0, r1, lsl #1
T ittt ne
ldrne r12,[r0, r1] @ p0
A ldrne r11,[r0], r1, lsl #1 @ p1
T ldrne r11,[r0] @ p3
T addne r0, r0, r1, lsl #1
bne 1b
add sp, sp, #16
pop {r4 - r11, pc}
endfunc
.macro TRANSPOSE_MATRIX i0, i1, i2, i3, o3, o2, o1, o0
@ input: $0, $1, $2, $3
@ output: $4, $5, $6, $7
@ i0: 03 02 01 00
@ i1: 13 12 11 10
@ i2: 23 22 21 20
@ i3: 33 32 31 30
@ o3 o2 o1 o0
uxtb16 \o1, \i1 @ xx 12 xx 10
uxtb16 \o0, \i0 @ xx 02 xx 00
uxtb16 \o3, \i3 @ xx 32 xx 30
uxtb16 \o2, \i2 @ xx 22 xx 20
orr \o1, \o0, \o1, lsl #8 @ 12 02 10 00
orr \o3, \o2, \o3, lsl #8 @ 32 22 30 20
uxtb16 \i1, \i1, ror #8 @ xx 13 xx 11
uxtb16 \i3, \i3, ror #8 @ xx 33 xx 31
uxtb16 \i0, \i0, ror #8 @ xx 03 xx 01
uxtb16 \i2, \i2, ror #8 @ xx 23 xx 21
orr \i0, \i0, \i1, lsl #8 @ 13 03 11 01
orr \i2, \i2, \i3, lsl #8 @ 33 23 31 21
pkhtb \o2, \o3, \o1, asr #16 @ 32 22 12 02 -- p1
pkhbt \o0, \o1, \o3, lsl #16 @ 30 20 10 00 -- p3
pkhtb \o3, \i2, \i0, asr #16 @ 33 23 13 03 -- p0
pkhbt \o1, \i0, \i2, lsl #16 @ 31 21 11 01 -- p2
.endm
@ void vp8_h_loop_filter16_simple(uint8_t *dst, int stride, int flim)
function ff_vp8_h_loop_filter16_simple_armv6, export=1
push {r4 - r11, lr}
orr r12, r2, r2, lsl #16
ldr r2, c0x80808080
orr r12, r12, r12, lsl #8
@ load soure data to r7, r8, r9, r10
sub r0, r0, #2
ldr r8, [r0, r1]
ldr_post r7, r0, r1, lsl #1
ldr r10,[r0, r1]
ldr_post r9, r0, r1, lsl #1
add r0, r0, #2
mov r11, #4 @ count (r11) for 4-in-parallel
1:
@transpose r7, r8, r9, r10 to r3, r4, r5, r6
TRANSPOSE_MATRIX r7, r8, r9, r10, r6, r5, r4, r3
@ vp8_simple_filter_mask() function
uqsub8 r7, r3, r6 @ p1 - q1
uqsub8 r8, r6, r3 @ q1 - p1
uqsub8 r9, r4, r5 @ p0 - q0
uqsub8 r10, r5, r4 @ q0 - p0
orr r7, r7, r8 @ abs(p1 - q1)
orr r9, r9, r10 @ abs(p0 - q0)
mov r8, #0
uqadd8 r9, r9, r9 @ abs(p0 - q0) * 2
uhadd8 r7, r7, r8 @ abs(p1 - q1) / 2
uqadd8 r7, r7, r9 @ abs(p0 - q0)*2 + abs(p1 - q1)/2
mvn r10, #0 @ r10 == -1
usub8 r7, r12, r7 @ compare to flimit
sel lr, r10, r8 @ filter mask
cmp lr, #0
beq 2f @ skip filtering
@vp8_simple_filter() function
eor r3, r3, r2 @ p1 offset to convert to a signed value
eor r6, r6, r2 @ q1 offset to convert to a signed value
eor r4, r4, r2 @ p0 offset to convert to a signed value
eor r5, r5, r2 @ q0 offset to convert to a signed value
qsub8 r3, r3, r6 @ vp8_filter = p1 - q1
qsub8 r6, r5, r4 @ q0 - p0
qadd8 r3, r3, r6 @ vp8_filter += q0 - p0
ldr r9, c0x03030303 @ r9 = 3
qadd8 r3, r3, r6 @ vp8_filter += q0 - p0
ldr r7, c0x04040404
qadd8 r3, r3, r6 @ vp8_filter = p1-q1 + 3*(q0-p0))
@STALL
and r3, r3, lr @ vp8_filter &= mask
qadd8 r9, r3, r9 @ Filter2 = vp8_filter + 3
qadd8 r3, r3, r7 @ Filter1 = vp8_filter + 4
shadd8 r9, r9, r8
shadd8 r3, r3, r8
shadd8 r9, r9, r8
shadd8 r3, r3, r8
shadd8 r9, r9, r8 @ Filter2 >>= 3
shadd8 r3, r3, r8 @ Filter1 >>= 3
@calculate output
sub r0, r0, r1, lsl #2
qadd8 r4, r4, r9 @ u = p0 + Filter2
qsub8 r5, r5, r3 @ u = q0 - Filter1
eor r4, r4, r2 @ *op0 = u^0x80
eor r5, r5, r2 @ *oq0 = u^0x80
strb r4, [r0, #-1] @ store the result
mov r4, r4, lsr #8
strb_post r5, r0, r1
mov r5, r5, lsr #8
strb r4, [r0, #-1]
mov r4, r4, lsr #8
strb_post r5, r0, r1
mov r5, r5, lsr #8
strb r4, [r0, #-1]
mov r4, r4, lsr #8
strb_post r5, r0, r1
mov r5, r5, lsr #8
strb r4, [r0, #-1]
strb_post r5, r0, r1
2:
subs r11, r11, #1
@ load soure data to r7, r8, r9, r10
sub r0, r0, #2
T ittt ne
ldrne r8, [r0, r1]
A ldrne r7, [r0], r1, lsl #1
T ldrne r7, [r0]
T addne r0, r0, r1, lsl #1
T ittt ne
ldrne r10,[r0, r1]
A ldrne r9, [r0], r1, lsl #1
T ldrne r9, [r0]
T addne r0, r0, r1, lsl #1
add r0, r0, #2
bne 1b
pop {r4 - r11, pc}
endfunc
@ void vp8_h_loop_filter16_inner(uint8_t *dst, int stride,
@ int fE, int fI, int hev_thresh)
@ and
@ void vp8_h_loop_filter8uv_inner(uint8_t *dstU, uint8_t *dstV, int stride,
@ int fE, int fI, int hev_thresh)
@ call:
@ void vp8_h_loop_filter_inner(uint8_t *dst, int stride,
@ int fE, int fI, int hev_thresh, int count)
function ff_vp8_h_loop_filter_inner_armv6, export=1
push {r4 - r11, lr}
sub r0, r0, #4 @ move r0 pointer down by 4
ldr r5, [sp, #40] @ counter
ldr r9, [sp, #36] @ load thresh address
sub sp, sp, #16 @ create temp buffer
ldr r7, [r0, r1] @ transpose will make it into p3-p0
ldr_post r6, r0, r1, lsl #1 @ load source data
ldr lr, [r0, r1]
ldr_post r8, r0, r1, lsl #1
orr r2, r2, r2, lsl #16
orr r3, r3, r3, lsl #16
orr r9, r9, r9, lsl #16
orr r4, r2, r2, lsl #8 @ flimE splat int -> byte
orr r2, r3, r3, lsl #8 @ flimI splat int -> byte
orr r3, r9, r9, lsl #8 @ thresh splat int -> byte
1:
@ vp8_filter_mask() function
@ calculate breakout conditions
@ transpose the source data for 4-in-parallel operation
TRANSPOSE_MATRIX r6, r7, r8, lr, r12, r11, r10, r9
uqsub8 r7, r9, r10 @ p3 - p2
uqsub8 r8, r10, r9 @ p2 - p3
uqsub8 r9, r10, r11 @ p2 - p1
uqsub8 r10, r11, r10 @ p1 - p2
orr r7, r7, r8 @ abs (p3-p2)
orr r10, r9, r10 @ abs (p2-p1)
uqsub8 lr, r7, r2 @ compare to limit. lr: vp8_filter_mask
uqsub8 r10, r10, r2 @ compare to limit
sub r0, r0, r1, lsl #2 @ move r0 pointer down by 4 lines
orr lr, lr, r10
uqsub8 r6, r11, r12 @ p1 - p0
uqsub8 r7, r12, r11 @ p0 - p1
add r0, r0, #4 @ move r0 pointer up by 4
orr r6, r6, r7 @ abs (p1-p0)
str r11,[sp, #12] @ save p1
uqsub8 r10, r6, r2 @ compare to limit
uqsub8 r11, r6, r3 @ compare to thresh
orr lr, lr, r10
@ transpose uses 8 regs(r6 - r12 and lr). Need to save reg value now
@ transpose the source data for 4-in-parallel operation
str r11,[sp] @ push r11 to stack
ldr r7, [r0, r1]
ldr_post r6, r0, r1, lsl #1 @ load source data
str r12,[sp, #4] @ save current reg before load q0 - q3 data
str lr, [sp, #8]
ldr lr, [r0, r1]
ldr_post r8, r0, r1, lsl #1
TRANSPOSE_MATRIX r6, r7, r8, lr, r12, r11, r10, r9
ldr lr, [sp, #8] @ load back (f)limit accumulator
uqsub8 r6, r12, r11 @ q3 - q2
uqsub8 r7, r11, r12 @ q2 - q3
uqsub8 r12, r11, r10 @ q2 - q1
uqsub8 r11, r10, r11 @ q1 - q2
orr r6, r6, r7 @ abs (q3-q2)
orr r7, r12, r11 @ abs (q2-q1)
uqsub8 r6, r6, r2 @ compare to limit
uqsub8 r7, r7, r2 @ compare to limit
ldr r11,[sp, #4] @ load back p0
ldr r12,[sp, #12] @ load back p1
orr lr, lr, r6
orr lr, lr, r7
uqsub8 r6, r11, r9 @ p0 - q0
uqsub8 r7, r9, r11 @ q0 - p0
uqsub8 r8, r12, r10 @ p1 - q1
uqsub8 r11, r10, r12 @ q1 - p1
orr r6, r6, r7 @ abs (p0-q0)
ldr r7, c0x7F7F7F7F
orr r8, r8, r11 @ abs (p1-q1)
uqadd8 r6, r6, r6 @ abs (p0-q0) * 2
and r8, r7, r8, lsr #1 @ abs (p1-q1) / 2
uqsub8 r11, r10, r9 @ q1 - q0
uqadd8 r6, r8, r6 @ abs (p0-q0)*2 + abs (p1-q1)/2
uqsub8 r12, r9, r10 @ q0 - q1
uqsub8 r6, r6, r4 @ compare to flimit
orr r9, r11, r12 @ abs (q1-q0)
uqsub8 r8, r9, r2 @ compare to limit
uqsub8 r10, r9, r3 @ compare to thresh
orr lr, lr, r6
orr lr, lr, r8
mvn r11, #0 @ r11 == -1
mov r12, #0
usub8 lr, r12, lr
ldr r9, [sp] @ load the compared result
sel lr, r11, r12 @ filter mask: lr
cmp lr, #0
beq 2f @ skip filtering
@vp8_hevmask() function
@calculate high edge variance
sub r0, r0, r1, lsl #2 @ move r0 pointer down by 4 lines
orr r9, r9, r10
ldrh r7, [r0, #-2]
ldrh_post r8, r0, r1
usub8 r9, r12, r9
sel r6, r12, r11 @ hev mask: r6
@vp8_filter() function
@ load soure data to r6, r11, r12, lr
ldrh r9, [r0, #-2]
ldrh_post r10, r0, r1
pkhbt r12, r7, r8, lsl #16
ldrh r7, [r0, #-2]
ldrh_post r8, r0, r1
pkhbt r11, r9, r10, lsl #16
ldrh r9, [r0, #-2]
ldrh_post r10, r0, r1
@ Transpose needs 8 regs(r6 - r12, and lr). Save r6 and lr first
str r6, [sp]
str lr, [sp, #4]
pkhbt r6, r7, r8, lsl #16
pkhbt lr, r9, r10, lsl #16
@transpose r12, r11, r6, lr to r7, r8, r9, r10
TRANSPOSE_MATRIX r12, r11, r6, lr, r10, r9, r8, r7
@load back hev_mask r6 and filter_mask lr
ldr r12, c0x80808080
ldr r6, [sp]
ldr lr, [sp, #4]
eor r7, r7, r12 @ p1 offset to convert to a signed value
eor r8, r8, r12 @ p0 offset to convert to a signed value
eor r9, r9, r12 @ q0 offset to convert to a signed value
eor r10, r10, r12 @ q1 offset to convert to a signed value
str r9, [sp] @ store qs0 temporarily
str r8, [sp, #4] @ store ps0 temporarily
str r10,[sp, #8] @ store qs1 temporarily
str r7, [sp, #12] @ store ps1 temporarily
qsub8 r7, r7, r10 @ vp8_signed_char_clamp(ps1-qs1)
qsub8 r8, r9, r8 @ vp8_signed_char_clamp(vp8_filter + 3 * ( qs0 - ps0))
and r7, r7, r6 @ vp8_filter (r7) &= hev (r7 : filter)
qadd8 r7, r7, r8
ldr r9, c0x03030303 @ r9 = 3 --modified for vp8
qadd8 r7, r7, r8
ldr r10, c0x04040404
qadd8 r7, r7, r8
and r7, r7, lr @ vp8_filter &= mask
qadd8 r8, r7, r9 @ Filter2 (r8) = vp8_signed_char_clamp(vp8_filter+3)
qadd8 r7, r7, r10 @ vp8_filter = vp8_signed_char_clamp(vp8_filter+4)
mov r9, #0
shadd8 r8, r8, r9 @ Filter2 >>= 3
shadd8 r7, r7, r9 @ vp8_filter >>= 3
shadd8 r8, r8, r9
shadd8 r7, r7, r9
shadd8 lr, r8, r9 @ lr: filter2
shadd8 r7, r7, r9 @ r7: filter
@calculate output
ldr r8, [sp] @ load qs0
ldr r9, [sp, #4] @ load ps0
ldr r10, c0x01010101
qsub8 r8, r8, r7 @ u = vp8_signed_char_clamp(qs0 - vp8_filter)
qadd8 r9, r9, lr @ u = vp8_signed_char_clamp(ps0 + Filter2)
eor r8, r8, r12
eor r9, r9, r12
mov lr, #0
sadd8 r7, r7, r10
shadd8 r7, r7, lr
ldr r10,[sp, #8] @ load qs1
ldr r11,[sp, #12] @ load ps1
bic r7, r7, r6 @ r7: vp8_filter
qsub8 r10, r10, r7 @ u = vp8_signed_char_clamp(qs1 - vp8_filter)
qadd8 r11, r11, r7 @ u = vp8_signed_char_clamp(ps1 + vp8_filter)
eor r10, r10, r12
eor r11, r11, r12
sub r0, r0, r1, lsl #2
@we can use TRANSPOSE_MATRIX macro to transpose output - input: q1, q0, p0, p1
TRANSPOSE_MATRIX r11, r9, r8, r10, lr, r12, r7, r6
strh r6, [r0, #-2] @ store the result
mov r6, r6, lsr #16
strh_post r6, r0, r1
strh r7, [r0, #-2]
mov r7, r7, lsr #16
strh_post r7, r0, r1
strh r12, [r0, #-2]
mov r12, r12, lsr #16
strh_post r12, r0, r1
strh lr, [r0, #-2]
mov lr, lr, lsr #16
strh_post lr, r0, r1
2:
sub r0, r0, #4
subs r5, r5, #1
T ittt ne
ldrne r7, [r0, r1]
A ldrne r6, [r0], r1, lsl #1 @ load source data
T ldrne r6, [r0] @ load source data
T addne r0, r0, r1, lsl #1
T ittt ne
ldrne lr, [r0, r1]
A ldrne r8, [r0], r1, lsl #1
T ldrne r8, [r0]
T addne r0, r0, r1, lsl #1
bne 1b
add sp, sp, #16
pop {r4 - r11, pc}
endfunc
@ void vp8_h_loop_filter16(uint8_t *dst, int stride,
@ int fE, int fI, int hev_thresh)
@ and
@ void vp8_h_loop_filter8uv(uint8_t *dstU, uint8_t *dstV, int stride,
@ int fE, int fI, int hev_thresh)
@ call:
@ void vp8_h_loop_filter(uint8_t *dst, int stride,
@ int fE, int fI, int hev_thresh, int count)
function ff_vp8_h_loop_filter_armv6, export=1
push {r4 - r11, lr}
sub r0, r0, #4 @ move r0 pointer down by 4
ldr r5, [sp, #40] @ counter
ldr r9, [sp, #36] @ load thresh address
sub sp, sp, #16 @ create temp buffer
ldr r7, [r0, r1] @ transpose will make it into p3-p0
ldr_post r6, r0, r1, lsl #1 @ load source data
ldr lr, [r0, r1]
ldr_post r8, r0, r1, lsl #1
orr r2, r2, r2, lsl #16
orr r3, r3, r3, lsl #16
orr r9, r9, r9, lsl #16
orr r4, r2, r2, lsl #8 @ flimE splat int -> byte
orr r2, r3, r3, lsl #8 @ flimI splat int -> byte
orr r3, r9, r9, lsl #8 @ thresh splat int -> byte
1:
@ vp8_filter_mask() function
@ calculate breakout conditions
@ transpose the source data for 4-in-parallel operation
TRANSPOSE_MATRIX r6, r7, r8, lr, r12, r11, r10, r9
uqsub8 r7, r9, r10 @ p3 - p2
uqsub8 r8, r10, r9 @ p2 - p3
uqsub8 r9, r10, r11 @ p2 - p1
uqsub8 r10, r11, r10 @ p1 - p2
orr r7, r7, r8 @ abs (p3-p2)
orr r10, r9, r10 @ abs (p2-p1)
uqsub8 lr, r7, r2 @ compare to limit. lr: vp8_filter_mask
uqsub8 r10, r10, r2 @ compare to limit
sub r0, r0, r1, lsl #2 @ move r0 pointer down by 4 lines
orr lr, lr, r10
uqsub8 r6, r11, r12 @ p1 - p0
uqsub8 r7, r12, r11 @ p0 - p1
add r0, r0, #4 @ move r0 pointer up by 4
orr r6, r6, r7 @ abs (p1-p0)
str r11,[sp, #12] @ save p1
uqsub8 r10, r6, r2 @ compare to limit
uqsub8 r11, r6, r3 @ compare to thresh
orr lr, lr, r10
@ transpose uses 8 regs(r6 - r12 and lr). Need to save reg value now
@ transpose the source data for 4-in-parallel operation
str r11,[sp] @ push r11 to stack
ldr r7, [r0, r1]
ldr_post r6, r0, r1, lsl #1 @ load source data
str r12,[sp, #4] @ save current reg before load q0 - q3 data
str lr, [sp, #8]
ldr lr, [r0, r1]
ldr_post r8, r0, r1, lsl #1
TRANSPOSE_MATRIX r6, r7, r8, lr, r12, r11, r10, r9
ldr lr, [sp, #8] @ load back (f)limit accumulator
uqsub8 r6, r12, r11 @ q3 - q2
uqsub8 r7, r11, r12 @ q2 - q3
uqsub8 r12, r11, r10 @ q2 - q1
uqsub8 r11, r10, r11 @ q1 - q2
orr r6, r6, r7 @ abs (q3-q2)
orr r7, r12, r11 @ abs (q2-q1)
uqsub8 r6, r6, r2 @ compare to limit
uqsub8 r7, r7, r2 @ compare to limit
ldr r11,[sp, #4] @ load back p0
ldr r12,[sp, #12] @ load back p1
orr lr, lr, r6
orr lr, lr, r7
uqsub8 r6, r11, r9 @ p0 - q0
uqsub8 r7, r9, r11 @ q0 - p0
uqsub8 r8, r12, r10 @ p1 - q1
uqsub8 r11, r10, r12 @ q1 - p1
orr r6, r6, r7 @ abs (p0-q0)
ldr r7, c0x7F7F7F7F
orr r8, r8, r11 @ abs (p1-q1)
uqadd8 r6, r6, r6 @ abs (p0-q0) * 2
and r8, r7, r8, lsr #1 @ abs (p1-q1) / 2
uqsub8 r11, r10, r9 @ q1 - q0
uqadd8 r6, r8, r6 @ abs (p0-q0)*2 + abs (p1-q1)/2
uqsub8 r12, r9, r10 @ q0 - q1
uqsub8 r6, r6, r4 @ compare to flimit
orr r9, r11, r12 @ abs (q1-q0)
uqsub8 r8, r9, r2 @ compare to limit
uqsub8 r10, r9, r3 @ compare to thresh
orr lr, lr, r6
orr lr, lr, r8
mvn r11, #0 @ r11 == -1
mov r12, #0
usub8 lr, r12, lr
ldr r9, [sp] @ load the compared result
sel lr, r11, r12 @ filter mask: lr
cmp lr, #0
beq 2f @ skip filtering
@vp8_hevmask() function
@calculate high edge variance
sub r0, r0, r1, lsl #2 @ move r0 pointer down by 4 lines
orr r9, r9, r10
ldrh r7, [r0, #-2]
ldrh_post r8, r0, r1
usub8 r9, r12, r9
sel r6, r12, r11 @ hev mask: r6
@ vp8_mbfilter() function
@ p2, q2 are only needed at the end. do not need to load them in now.
@ Transpose needs 8 regs(r6 - r12, and lr). Save r6 and lr first
@ load soure data to r6, r11, r12, lr
ldrh r9, [r0, #-2]
ldrh_post r10, r0, r1
pkhbt r12, r7, r8, lsl #16
ldrh r7, [r0, #-2]
ldrh_post r8, r0, r1
pkhbt r11, r9, r10, lsl #16
ldrh r9, [r0, #-2]
ldrh_post r10, r0, r1
str r6, [sp] @ save r6
str lr, [sp, #4] @ save lr
pkhbt r6, r7, r8, lsl #16
pkhbt lr, r9, r10, lsl #16
@transpose r12, r11, r6, lr to p1, p0, q0, q1
TRANSPOSE_MATRIX r12, r11, r6, lr, r10, r9, r8, r7
@load back hev_mask r6 and filter_mask lr
ldr r12, c0x80808080
ldr r6, [sp]
ldr lr, [sp, #4]
eor r7, r7, r12 @ ps1
eor r8, r8, r12 @ ps0
eor r9, r9, r12 @ qs0
eor r10, r10, r12 @ qs1
qsub8 r12, r9, r8 @ vp8_signed_char_clamp(vp8_filter + 3 * ( qs0 - ps0))
str r7, [sp, #12] @ store ps1 temporarily
qsub8 r7, r7, r10 @ vp8_signed_char_clamp(ps1-qs1)
str r10,[sp, #8] @ store qs1 temporarily
qadd8 r7, r7, r12
str r9, [sp] @ store qs0 temporarily
qadd8 r7, r7, r12
str r8, [sp, #4] @ store ps0 temporarily
qadd8 r7, r7, r12 @ vp8_filter: r7
ldr r10, c0x03030303 @ r10 = 3 --modified for vp8
ldr r9, c0x04040404
and r7, r7, lr @ vp8_filter &= mask (lr is free)
mov r12, r7 @ Filter2: r12
and r12, r12, r6 @ Filter2 &= hev
@save bottom 3 bits so that we round one side +4 and the other +3
qadd8 r8, r12, r9 @ Filter1 (r8) = vp8_signed_char_clamp(Filter2+4)
qadd8 r12, r12, r10 @ Filter2 (r12) = vp8_signed_char_clamp(Filter2+3)
mov r10, #0
shadd8 r8, r8, r10 @ Filter1 >>= 3
shadd8 r12, r12, r10 @ Filter2 >>= 3
shadd8 r8, r8, r10
shadd8 r12, r12, r10
shadd8 r8, r8, r10 @ r8: Filter1
shadd8 r12, r12, r10 @ r12: Filter2
ldr r9, [sp] @ load qs0
ldr r11,[sp, #4] @ load ps0
qsub8 r9, r9, r8 @ qs0 = vp8_signed_char_clamp(qs0 - Filter1)
qadd8 r11, r11, r12 @ ps0 = vp8_signed_char_clamp(ps0 + Filter2)
bic r12, r7, r6 @vp8_filter &= ~hev ( r6 is free)
@roughly 3/7th difference across boundary
mov lr, #0x1b @ 27
mov r7, #0x3f @ 63
sxtb16 r6, r12
sxtb16 r10, r12, ror #8
smlabb r8, r6, lr, r7
smlatb r6, r6, lr, r7
smlabb r7, r10, lr, r7
smultb r10, r10, lr
ssat r8, #8, r8, asr #7
ssat r6, #8, r6, asr #7
add r10, r10, #63
ssat r7, #8, r7, asr #7
ssat r10, #8, r10, asr #7
ldr lr, c0x80808080
pkhbt r6, r8, r6, lsl #16
pkhbt r10, r7, r10, lsl #16
uxtb16 r6, r6
uxtb16 r10, r10
sub r0, r0, r1, lsl #2 @ move r0 pointer down by 4 lines
orr r10, r6, r10, lsl #8 @ u = vp8_signed_char_clamp((63 + Filter2 * 27)>>7)
qsub8 r8, r9, r10 @ s = vp8_signed_char_clamp(qs0 - u)
qadd8 r10, r11, r10 @ s = vp8_signed_char_clamp(ps0 + u)
eor r8, r8, lr @ *oq0 = s^0x80
eor r10, r10, lr @ *op0 = s^0x80
strb r10,[r0, #-1] @ store op0 result
strb_post r8, r0, r1 @ store oq0 result
mov r10, r10, lsr #8
mov r8, r8, lsr #8
strb r10,[r0, #-1]
strb_post r8, r0, r1
mov r10, r10, lsr #8
mov r8, r8, lsr #8
strb r10,[r0, #-1]
strb_post r8, r0, r1
mov r10, r10, lsr #8
mov r8, r8, lsr #8
strb r10,[r0, #-1]
strb_post r8, r0, r1
@roughly 2/7th difference across boundary
mov lr, #0x12 @ 18
mov r7, #0x3f @ 63
sxtb16 r6, r12
sxtb16 r10, r12, ror #8
smlabb r8, r6, lr, r7
smlatb r6, r6, lr, r7
smlabb r9, r10, lr, r7
smlatb r10, r10, lr, r7
ssat r8, #8, r8, asr #7
ssat r6, #8, r6, asr #7
ssat r9, #8, r9, asr #7
ssat r10, #8, r10, asr #7
sub r0, r0, r1, lsl #2 @ move r0 pointer down by 4 lines
pkhbt r6, r8, r6, lsl #16
pkhbt r10, r9, r10, lsl #16
ldr r9, [sp, #8] @ load qs1
ldr r11,[sp, #12] @ load ps1
ldr lr, c0x80808080
uxtb16 r6, r6
uxtb16 r10, r10
add r0, r0, #2
orr r10, r6, r10, lsl #8 @ u = vp8_signed_char_clamp((63 + Filter2 * 18)>>7)
qsub8 r8, r9, r10 @ s = vp8_signed_char_clamp(qs1 - u)
qadd8 r10, r11, r10 @ s = vp8_signed_char_clamp(ps1 + u)
eor r8, r8, lr @ *oq1 = s^0x80
eor r10, r10, lr @ *op1 = s^0x80
ldrb r11,[r0, #-5] @ load p2 for 1/7th difference across boundary
strb r10,[r0, #-4] @ store op1
strb r8, [r0, #-1] @ store oq1
ldrb_post r9, r0, r1 @ load q2 for 1/7th difference across boundary
mov r10, r10, lsr #8
mov r8, r8, lsr #8
ldrb r6, [r0, #-5]
strb r10,[r0, #-4]
strb r8, [r0, #-1]
ldrb_post r7, r0, r1
mov r10, r10, lsr #8
mov r8, r8, lsr #8
orr r11, r11, r6, lsl #8
orr r9, r9, r7, lsl #8
ldrb r6, [r0, #-5]
strb r10,[r0, #-4]
strb r8, [r0, #-1]
ldrb_post r7, r0, r1
mov r10, r10, lsr #8
mov r8, r8, lsr #8
orr r11, r11, r6, lsl #16
orr r9, r9, r7, lsl #16
ldrb r6, [r0, #-5]
strb r10,[r0, #-4]
strb r8, [r0, #-1]
ldrb_post r7, r0, r1
orr r11, r11, r6, lsl #24
orr r9, r9, r7, lsl #24
@roughly 1/7th difference across boundary
eor r9, r9, lr
eor r11, r11, lr
mov lr, #0x9 @ 9
mov r7, #0x3f @ 63
sxtb16 r6, r12
sxtb16 r10, r12, ror #8
smlabb r8, r6, lr, r7
smlatb r6, r6, lr, r7
smlabb r12, r10, lr, r7
smlatb r10, r10, lr, r7
ssat r8, #8, r8, asr #7
ssat r6, #8, r6, asr #7
ssat r12, #8, r12, asr #7
ssat r10, #8, r10, asr #7
sub r0, r0, r1, lsl #2
pkhbt r6, r8, r6, lsl #16
pkhbt r10, r12, r10, lsl #16
uxtb16 r6, r6
uxtb16 r10, r10
ldr lr, c0x80808080
orr r10, r6, r10, lsl #8 @ u = vp8_signed_char_clamp((63 + Filter2 * 9)>>7)
qadd8 r8, r11, r10 @ s = vp8_signed_char_clamp(ps2 + u)
qsub8 r10, r9, r10 @ s = vp8_signed_char_clamp(qs2 - u)
eor r8, r8, lr @ *op2 = s^0x80
eor r10, r10, lr @ *oq2 = s^0x80
strb r8, [r0, #-5] @ store *op2
strb_post r10, r0, r1 @ store *oq2
mov r8, r8, lsr #8
mov r10, r10, lsr #8
strb r8, [r0, #-5]
strb_post r10, r0, r1
mov r8, r8, lsr #8
mov r10, r10, lsr #8
strb r8, [r0, #-5]
strb_post r10, r0, r1
mov r8, r8, lsr #8
mov r10, r10, lsr #8
strb r8, [r0, #-5]
strb_post r10, r0, r1
@adjust r0 pointer for next loop
sub r0, r0, #2
2:
sub r0, r0, #4
subs r5, r5, #1
T ittt ne
ldrne r7, [r0, r1]
A ldrne r6, [r0], r1, lsl #1 @ load source data
T ldrne r6, [r0]
T addne r0, r0, r1, lsl #1
T ittt ne
ldrne lr, [r0, r1]
A ldrne r8, [r0], r1, lsl #1
T ldrne r8, [r0]
T addne r0, r0, r1, lsl #1
bne 1b
add sp, sp, #16
pop {r4 - r11, pc}
endfunc
@ MC
@ void put_vp8_pixels16(uint8_t *dst, int dststride, uint8_t *src,
@ int srcstride, int h, int mx, int my)
function ff_put_vp8_pixels16_armv6, export=1
push {r4 - r11}
ldr r12,[sp, #32] @ h
1:
subs r12, r12, #2
ldr r5, [r2, #4]
ldr r6, [r2, #8]
ldr r7, [r2, #12]
ldr_post r4, r2, r3
ldr r9, [r2, #4]
ldr r10,[r2, #8]
ldr r11,[r2, #12]
ldr_post r8, r2, r3
strd r6, r7, [r0, #8]
strd_post r4, r5, r0, r1
strd r10, r11,[r0, #8]
strd_post r8, r9, r0, r1
bgt 1b
pop {r4 - r11}
bx lr
endfunc
@ void put_vp8_pixels8(uint8_t *dst, int dststride, uint8_t *src,
@ int srcstride, int h, int mx, int my)
function ff_put_vp8_pixels8_armv6, export=1
push {r4 - r11}
ldr r12,[sp, #32] @ h
1:
subs r12, r12, #4
ldr r5, [r2, #4]
ldr_post r4, r2, r3
ldr r7, [r2, #4]
ldr_post r6, r2, r3
ldr r9, [r2, #4]
ldr_post r8, r2, r3
ldr r11,[r2, #4]
ldr_post r10, r2, r3
strd_post r4, r5, r0, r1
strd_post r6, r7, r0, r1
strd_post r8, r9, r0, r1
strd_post r10, r11, r0, r1
bgt 1b
pop {r4 - r11}
bx lr
endfunc
@ void put_vp8_pixels4(uint8_t *dst, int dststride, uint8_t *src,
@ int srcstride, int h, int mx, int my)
function ff_put_vp8_pixels4_armv6, export=1
ldr r12, [sp, #0] @ h
push {r4 - r6, lr}
1:
subs r12, r12, #4
ldr r5, [r2, r3]
ldr_post r4, r2, r3, lsl #1
ldr lr, [r2, r3]
ldr_post r6, r2, r3, lsl #1
str r5, [r0, r1]
str_post r4, r0, r1, lsl #1
str lr, [r0, r1]
str_post r6, r0, r1, lsl #1
bgt 1b
pop {r4 - r6, pc}
endfunc
@ note: worst case sum of all 6-tap filter values * 255 is 0x7f80 so 16 bit
@ arithmatic can be used to apply filters
const sixtap_filters_13245600, align=4
.short 2, 108, -11, 36, -8, 1, 0, 0
.short 3, 77, -16, 77, -16, 3, 0, 0
.short 1, 36, -8, 108, -11, 2, 0, 0
endconst
const fourtap_filters_1324, align=4
.short -6, 12, 123, -1
.short -9, 50, 93, -6
.short -6, 93, 50, -9
.short -1, 123, 12, -6
endconst
@ void put_vp8_epel_h6(uint8_t *dst, int dststride, uint8_t *src,
@ int srcstride, int w, int h, int mx)
function ff_put_vp8_epel_h6_armv6, export=1
push {r4 - r11, lr}
sub r2, r2, #2
movrel lr, sixtap_filters_13245600 - 16
ldr r12,[sp, #44] @ vp8_filter index
ldr r4, [sp, #36] @ width
add lr, lr, r12, lsl #3
sub r3, r3, r4 @ src_stride - block_width
sub r1, r1, r4 @ dst_stride - block_width
lsr r4, #2
str r4, [sp, #36] @ "4-in-parallel" loop counter @40
str r3, [sp, #44] @ src_stride - block_width @48
push {r1} @ dst_stride - block_width @0
@ height @44
ldr r1, [lr], #4 @ coefficients
ldr r3, [lr], #4
ldr lr, [lr]
1:
@ 3 loads, 10 shuffles and then mul/acc/add/shr
@ o0: i0/i1/i2/i3/i4/i5 -> i0/i2 (ld1) | i1/i3 (ld1) | i4/i5 (ld2)
@ o1: i1/i2/i3/i4/i5/i6 -> i1/i3 (ld1) | i2/i4 (ld2) | i5/i6 (ld2/3)
@ o2: i2/i3/i4/i5/i6/i7 -> i2/i4 (ld2) | i3/i5 (ld2) | i6/i7 (ld3)
@ o3: i3/i4/i5/i6/i7/i8 -> i3/i5 (ld2) | i4/i6 (ld2/3) | i7/i8 (ld3)
ldr r7, [r2, #5] @ ld3 -> src[5-8]
ldr r6, [r2, #2] @ ld2 -> src[2-5]
ldr r5, [r2], #4 @ ld1 -> src[0-3]
pkhtb r7, r7, r7, asr #8 @ src[8,7,7,6]
uxtb16 r9, r6, ror #8 @ src[5] | src[3]
uxtb16 r6, r6 @ src[4] | src[2]
uxtb16 r8, r5, ror #8 @ src[3] | src[1]
uxtb16 r11, r7, ror #8 @ src[8] | src[7]
uxtb16 r7, r7 @ src[7] | src[6]
pkhtb r10, r9, r6, asr #16 @ src[5] | src[4]
uxtb16 r5, r5 @ src[2] | src[0]
smuad r11, r11, lr @ filter[3][2] -> r11
subs r4, r4, #1
pkhbt r12, r10, r7, lsl #16 @ src[6] | src[4]
smuad r7, r7, lr @ filter[2][2] -> r7
smuad r5, r5, r1 @ filter[0][0] -> r5
smlad r11, r9, r1, r11 @ filter[3][0] -> r11
smlad r7, r9, r3, r7 @ filter[2][1] -> r7
smuad r9, r8, r1 @ filter[1][0] -> r9
smlad r5, r8, r3, r5 @ filter[0][1] -> r5
pkhtb r8, r12, r10, asr #16 @ src[6] | src[5]
smlad r11, r12, r3, r11 @ filter[3][1] -> r11
smlad r9, r6, r3, r9 @ filter[1][1] -> r9
smlad r5, r10, lr, r5 @ filter[0][2] -> r5
smlad r7, r6, r1, r7 @ filter[2][0] -> r7
smlad r9, r8, lr, r9 @ filter[1][2] -> r9
add r5, r5, #0x40 @ round_shift_and_clamp[0]
add r9, r9, #0x40 @ round_shift_and_clamp[1]
add r7, r7, #0x40 @ round_shift_and_clamp[2]
add r11, r11, #0x40 @ round_shift_and_clamp[3]
usat r5, #8, r5, asr #7
usat r9, #8, r9, asr #7
usat r7, #8, r7, asr #7
usat r11, #8, r11, asr #7
strb r5, [r0], #1 @ store res[0]
strb r9, [r0], #1 @ store res[1]
strb r7, [r0], #1 @ store res[2]
strb r11,[r0], #1 @ store res[3]
bne 1b
ldr r12,[sp, #44] @ height = outer-loop counter
subs r12, r12, #1
T itttt ne
ldrne r4, [sp, #40] @ 4-in-parallel loop counter
ldrne r5, [sp, #48]
ldrne r6, [sp]
strne r12,[sp, #44]
add r2, r2, r5 @ move to next input/output lines
add r0, r0, r6
bne 1b
add sp, sp, #4 @ restore stack after push{r1} above
pop {r4 - r11, pc}
endfunc
@ void put_vp8_epel_v6(uint8_t *dst, int dststride, uint8_t *src,
@ int srcstride, int w, int h, int my)
function ff_put_vp8_epel_v6_armv6, export=1
push {r4 - r11, lr}
movrel lr, sixtap_filters_13245600 - 16
ldr r12,[sp, #44] @ vp8_filter index
ldr r4, [sp, #36] @ width
add lr, lr, r12, lsl #3
sub r1, r1, r4 @ dst_stride - block_width
lsr r4, #2
str r4, [sp, #36] @ "4-in-parallel" loop counter @40
str r3, [sp, #44] @ src_stride - block_width @48
push {r1} @ dst_stride - block_width @0
@ height @44
1:
add r1, r3, r3, lsl #1 @ stride * 3
ldr_dpren r5, r2, r3 @ src[0,1,2,3 + stride * 1]
ldr r6, [r2, r3] @ src[0,1,2,3 + stride * 3]
ldr r7, [r2, r3, lsl #1] @ src[0,1,2,3 + stride * 4]
ldr r8, [r2, r1] @ src[0,1,2,3 + stride * 5]
@ byte -> word and "transpose"
uxtb16 r9, r5, ror #8 @ src[3 + stride*1] | src[1 + stride*1]
uxtb16 r10, r6, ror #8 @ src[3 + stride*3] | src[1 + stride*3]
uxtb16 r11, r7, ror #8 @ src[3 + stride*4] | src[1 + stride*4]
uxtb16 r12, r8, ror #8 @ src[3 + stride*5] | src[1 + stride*5]
uxtb16 r5, r5 @ src[2 + stride*1] | src[0 + stride*1]
uxtb16 r6, r6 @ src[2 + stride*3] | src[0 + stride*3]
uxtb16 r7, r7 @ src[2 + stride*4] | src[0 + stride*4]
uxtb16 r8, r8 @ src[2 + stride*5] | src[0 + stride*5]
pkhbt r1, r9, r10, lsl #16 @ src[1 + stride*3] | src[1 + stride*1]
pkhtb r9, r10, r9, asr #16 @ src[3 + stride*3] | src[3 + stride*1]
pkhbt r10, r11, r12, lsl #16 @ src[1 + stride*5] | src[1 + stride*4]
pkhtb r11, r12, r11, asr #16 @ src[3 + stride*5] | src[3 + stride*4]
pkhbt r12, r5, r6, lsl #16 @ src[0 + stride*3] | src[0 + stride*1]
pkhtb r5, r6, r5, asr #16 @ src[2 + stride*3] | src[2 + stride*1]
pkhbt r6, r7, r8, lsl #16 @ src[0 + stride*5] | src[0 + stride*4]
pkhtb r7, r8, r7, asr #16 @ src[2 + stride*5] | src[2 + stride*4]
ldr r8, [lr, #4] @ stall - if only I had more registers...
smuad r12, r12, r8 @ filter[0][1]
smuad r1, r1, r8 @ filter[1][1]
smuad r5, r5, r8 @ filter[2][1]
smuad r9, r9, r8 @ filter[3][1]
ldr r8, [lr, #8] @ stall - if only I had more registers...
smlad r12, r6, r8, r12 @ filter[0][2]
smlad r1, r10, r8, r1 @ filter[1][2]
ldr_dpren r6, r2, r3, lsl #1 @ src[0,1,2,3 + stride * 0]
ldr r10,[r2], #4 @ src[0,1,2,3 + stride * 2]
smlad r5, r7, r8, r5 @ filter[2][2]
smlad r9, r11, r8, r9 @ filter[3][2]
uxtb16 r7, r6, ror #8 @ src[3 + stride*0] | src[1 + stride*0]
uxtb16 r11, r10, ror #8 @ src[3 + stride*2] | src[1 + stride*2]
uxtb16 r6, r6 @ src[2 + stride*0] | src[0 + stride*0]
uxtb16 r10, r10 @ src[2 + stride*2] | src[0 + stride*2]
pkhbt r8, r7, r11, lsl #16 @ src[1 + stride*2] | src[1 + stride*0]
pkhtb r7, r11, r7, asr #16 @ src[3 + stride*2] | src[3 + stride*0]
pkhbt r11, r6, r10, lsl #16 @ src[0 + stride*2] | src[0 + stride*0]
pkhtb r6, r10, r6, asr #16 @ src[2 + stride*2] | src[2 + stride*0]
ldr r10,[lr] @ stall - if only I had more registers...
subs r4, r4, #1 @ counter--
smlad r12, r11, r10, r12 @ filter[0][0]
smlad r1, r8, r10, r1 @ filter[1][0]
smlad r5, r6, r10, r5 @ filter[2][0]
smlad r9, r7, r10, r9 @ filter[3][0]
add r12, r12, #0x40 @ round_shift_and_clamp[0]
add r1, r1, #0x40 @ round_shift_and_clamp[1]
add r5, r5, #0x40 @ round_shift_and_clamp[2]
add r9, r9, #0x40 @ round_shift_and_clamp[3]
usat r12, #8, r12, asr #7
usat r1, #8, r1, asr #7
usat r5, #8, r5, asr #7
usat r9, #8, r9, asr #7
strb r12,[r0], #1 @ store res[0]
strb r1, [r0], #1 @ store res[1]
strb r5, [r0], #1 @ store res[2]
strb r9, [r0], #1 @ store res[3]
bne 1b
ldr r12,[sp, #44] @ height = outer-loop counter
subs r12, r12, #1
T itttt ne
ldrne r4, [sp, #40] @ 4-in-parallel loop counter
ldrne r6, [sp, #0]
subne r2, r2, r4, lsl #2
strne r12,[sp, #44]
add r0, r0, r6
add r2, r2, r3 @ move to next input/output lines
bne 1b
add sp, sp, #4 @ restore stack after push{r1} above
pop {r4 - r11, pc}
endfunc
@ void put_vp8_epel_h4(uint8_t *dst, int dststride, uint8_t *src,
@ int srcstride, int w, int h, int mx)
function ff_put_vp8_epel_h4_armv6, export=1
push {r4 - r11, lr}
subs r2, r2, #1
movrel lr, fourtap_filters_1324 - 4
ldr r4, [sp, #36] @ width
ldr r12,[sp, #44] @ vp8_filter index
add lr, lr, r12, lsl #2
sub r3, r3, r4 @ src_stride - block_width
sub r1, r1, r4 @ dst_stride - block_width
ldr r5, [lr]
ldr r6, [lr, #4]
asr r4, #2
ldr lr, [sp, #40] @ height = outer-loop counter
str r4, [sp, #36] @ "4-in-parallel" inner loop counter
1:
@ 3 loads, 5 uxtb16s and then mul/acc/add/shr
@ o0: i0/i1/i2/i3 -> i0/i2(ld1) + i1/i3(ld1)
@ o1: i1/i2/i3/i4 -> i1/i3(ld1) + i2/i4(ld2)
@ o2: i2/i3/i4/i5 -> i2/i4(ld2) + i3/i5(ld2)
@ o3: i3/i4/i5/i6 -> i3/i5(ld2) + i4/i6(ld3)
ldr r9, [r2, #3] @ load source data
ldr r8, [r2, #2]
ldr r7, [r2], #4
uxtb16 r9, r9, ror #8 @ src[6] | src[4]
uxtb16 r10, r8, ror #8 @ src[5] | src[3]
uxtb16 r8, r8 @ src[4] | src[2]
uxtb16 r11, r7, ror #8 @ src[3] | src[1]
uxtb16 r7, r7 @ src[2] | src[0]
smuad r9, r9, r6 @ filter[3][1] -> r9
smuad r12, r10, r6 @ filter[2][1] -> r12
smuad r7, r7, r5 @ filter[0][0] -> r7
smlad r9, r10, r5, r9 @ filter[3][0] -> r9
smuad r10, r11, r5 @ filter[1][0] -> r10
smlad r12, r8, r5, r12 @ filter[2][0] -> r12
smlad r7, r11, r6, r7 @ filter[0][1] -> r7
smlad r10, r8, r6, r10 @ filter[1][1] -> r10
subs r4, r4, #1 @ counter--
add r7, r7, #0x40 @ round_shift_and_clamp[0]
add r10, r10, #0x40 @ round_shift_and_clamp[1]
add r12, r12, #0x40 @ round_shift_and_clamp[2]
add r9, r9, #0x40 @ round_shift_and_clamp[3]
usat r7, #8, r7, asr #7
usat r10, #8, r10, asr #7
usat r12, #8, r12, asr #7
usat r9, #8, r9, asr #7
strb r7, [r0], #1 @ store res[0]
strb r10,[r0], #1 @ store res[1]
strb r12,[r0], #1 @ store res[2]
strb r9, [r0], #1 @ store res[3]
bne 1b
subs lr, lr, #1
T it ne
ldrne r4, [sp, #36] @ 4-in-parallel loop counter
add r2, r2, r3 @ move to next input/output lines
add r0, r0, r1
bne 1b
pop {r4 - r11, pc}
endfunc
@ void put_vp8_epel_v4(uint8_t *dst, int dststride, uint8_t *src,
@ int srcstride, int w, int h, int my)
function ff_put_vp8_epel_v4_armv6, export=1
push {r4 - r11, lr}
movrel lr, fourtap_filters_1324 - 4
ldr r12,[sp, #44] @ vp8_filter index
ldr r4, [sp, #36] @ width
add lr, lr, r12, lsl #2
sub r1, r1, r4 @ dst_stride - block_width
asr r4, #2
ldr r5, [lr]
ldr r6, [lr, #4]
str r4, [sp, #36] @ "4-in-parallel" loop counter @40
str r3, [sp, #44] @ src_stride @48
push {r1} @ dst_stride - block_width @36
@ height @44
1:
ldr lr, [r2, r3, lsl #1] @ load source pixels
ldr r12,[r2, r3]
ldr_dpren r7, r2, r3
ldr r11,[r2], #4
@ byte -> word and "transpose"
uxtb16 r8, lr, ror #8 @ src[3 + stride*3] | src[1 + stride*3]
uxtb16 r9, r12, ror #8 @ src[3 + stride*2] | src[1 + stride*2]
uxtb16 r3, r7, ror #8 @ src[3 + stride*0] | src[1 + stride*0]
uxtb16 r1, r11, ror #8 @ src[3 + stride*1] | src[1 + stride*1]
uxtb16 lr, lr @ src[2 + stride*3] | src[0 + stride*3]
uxtb16 r12, r12 @ src[2 + stride*2] | src[0 + stride*2]
uxtb16 r7, r7 @ src[2 + stride*0] | src[0 + stride*0]
uxtb16 r11, r11 @ src[2 + stride*1] | src[0 + stride*1]
pkhbt r10, r1, r8, lsl #16 @ src[1 + stride*3] | src[1 + stride*1]
pkhtb r1, r8, r1, asr #16 @ src[3 + stride*3] | src[3 + stride*1]
pkhbt r8, r3, r9, lsl #16 @ src[1 + stride*2] | src[1 + stride*0]
pkhtb r3, r9, r3, asr #16 @ src[3 + stride*2] | src[3 + stride*0]
pkhbt r9, r11, lr, lsl #16 @ src[0 + stride*3] | src[0 + stride*1]
pkhtb r11, lr, r11, asr #16 @ src[2 + stride*3] | src[2 + stride*1]
pkhbt lr, r7, r12, lsl #16 @ src[0 + stride*2] | src[0 + stride*0]
pkhtb r7, r12, r7, asr #16 @ src[2 + stride*2] | src[2 + stride*0]
smuad r9, r9, r6 @ filter[0][1]
smuad r10, r10, r6 @ filter[1][1]
smuad r11, r11, r6 @ filter[2][1]
smuad r1, r1, r6 @ filter[3][1]
smlad r9, lr, r5, r9 @ filter[0][0]
smlad r10, r8, r5, r10 @ filter[1][0]
smlad r11, r7, r5, r11 @ filter[2][0]
smlad r1, r3, r5, r1 @ filter[3][0]
subs r4, r4, #1 @ counter--
ldr r3, [sp, #48] @ FIXME prevent clobber of r3 above?
add r9, r9, #0x40 @ round_shift_and_clamp[0]
add r10, r10, #0x40 @ round_shift_and_clamp[1]
add r11, r11, #0x40 @ round_shift_and_clamp[2]
add r1, r1, #0x40 @ round_shift_and_clamp[3]
usat r9, #8, r9, asr #7
usat r10, #8, r10, asr #7
usat r11, #8, r11, asr #7
usat r1, #8, r1, asr #7
strb r9, [r0], #1 @ store result
strb r10,[r0], #1
strb r11,[r0], #1
strb r1, [r0], #1
bne 1b
ldr r12,[sp, #44] @ height = outer-loop counter
subs r12, r12, #1
T ittt ne
ldrne r4, [sp, #40] @ 4-in-parallel loop counter
ldrne r9, [sp, #0]
strne r12,[sp, #44]
sub r2, r2, r4, lsl #2
add r0, r0, r9
add r2, r2, r3 @ move to next input/output lines
bne 1b
add sp, sp, #4 @ restore stack after push{r1} above
pop {r4 - r11, pc}
endfunc
@ void put_vp8_bilin_h(uint8_t *dst, int dststride, uint8_t *src,
@ int srcstride, int w, int h, int mx)
function ff_put_vp8_bilin_h_armv6, export=1
push {r4 - r9, lr}
ldr r8, [sp, #36] @ vp8_filter index
ldr r12,[sp, #32] @ height = outer-loop counter
ldr r4, [sp, #28] @ width
lsl r5, r8, #16 @ mx << 16
sub r3, r3, r4 @ src_stride - block_width
sub r1, r1, r4 @ dst_stride - block_width
asr r4, #2
sub r5, r5, r8 @ (mx << 16) | (-mx)
str r4, [sp, #28] @ "4-in-parallel" loop counter
add r5, r5, #8 @ (8 - mx) | (mx << 16) = filter coefficients
1:
ldrb r6, [r2], #1 @ load source data
ldrb r7, [r2], #1
ldrb r8, [r2], #1
ldrb r9, [r2], #1
ldrb lr, [r2]
pkhbt r6, r6, r7, lsl #16 @ src[1] | src[0]
pkhbt r7, r7, r8, lsl #16 @ src[2] | src[1]
pkhbt r8, r8, r9, lsl #16 @ src[3] | src[2]
pkhbt r9, r9, lr, lsl #16 @ src[4] | src[3]
smuad r6, r6, r5 @ apply the filter
smuad r7, r7, r5
smuad r8, r8, r5
smuad r9, r9, r5
subs r4, r4, #1 @ counter--
add r6, r6, #0x4 @ round_shift_and_clamp
add r7, r7, #0x4
add r8, r8, #0x4
add r9, r9, #0x4
asr r6, #3
asr r7, #3
pkhbt r6, r6, r8, lsl #13
pkhbt r7, r7, r9, lsl #13
orr r6, r6, r7, lsl #8
str r6, [r0], #4 @ store result
bne 1b
ldr r4, [sp, #28] @ 4-in-parallel loop counter
subs r12, r12, #1
add r2, r2, r3 @ move to next input/output lines
add r0, r0, r1
bne 1b
pop {r4 - r9, pc}
endfunc
@ void put_vp8_bilin_v(uint8_t *dst, int dststride, uint8_t *src,
@ int srcstride, int w, int h, int my)
function ff_put_vp8_bilin_v_armv6, export=1
push {r4 - r11, lr}
ldr r11,[sp, #44] @ vp8_filter index
ldr r4, [sp, #36] @ width
mov r5, r11, lsl #16 @ mx << 16
ldr r12,[sp, #40] @ height = outer-loop counter
sub r1, r1, r4
sub r5, r5, r11 @ (mx << 16) | (-mx)
asr r4, #2
add r5, r5, #8 @ (8 - mx) | (mx << 16) = filter coefficients
str r4, [sp, #36] @ "4-in-parallel" loop counter
1:
ldrb r10,[r2, r3] @ load the data
ldrb r6, [r2], #1
ldrb r11,[r2, r3]
ldrb r7, [r2], #1
ldrb lr, [r2, r3]
ldrb r8, [r2], #1
ldrb r9, [r2, r3]
pkhbt r6, r6, r10, lsl #16
ldrb r10,[r2], #1
pkhbt r7, r7, r11, lsl #16
pkhbt r8, r8, lr, lsl #16
pkhbt r9, r10, r9, lsl #16
smuad r6, r6, r5 @ apply the filter
smuad r7, r7, r5
smuad r8, r8, r5
smuad r9, r9, r5
subs r4, r4, #1 @ counter--
add r6, r6, #0x4 @ round_shift_and_clamp
add r7, r7, #0x4
add r8, r8, #0x4
add r9, r9, #0x4
asr r6, #3
asr r7, #3
pkhbt r6, r6, r8, lsl #13
pkhbt r7, r7, r9, lsl #13
orr r6, r6, r7, lsl #8
str r6, [r0], #4 @ store result
bne 1b
ldr r4, [sp, #36] @ 4-in-parallel loop counter
subs r12, r12, #1
add r2, r2, r3 @ move to next input/output lines
add r0, r0, r1
sub r2, r2, r4, lsl #2
bne 1b
pop {r4 - r11, pc}
endfunc