2013-07-13 01:12:58 +02:00
|
|
|
;
|
|
|
|
; Copyright (c) 2013 The WebM project authors. All Rights Reserved.
|
|
|
|
;
|
|
|
|
; 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.
|
|
|
|
;
|
|
|
|
|
|
|
|
|
|
|
|
; These functions are only valid when:
|
|
|
|
; x_step_q4 == 16
|
|
|
|
; w%4 == 0
|
|
|
|
; h%4 == 0
|
|
|
|
; taps == 8
|
|
|
|
; VP9_FILTER_WEIGHT == 128
|
|
|
|
; VP9_FILTER_SHIFT == 7
|
|
|
|
|
|
|
|
EXPORT |vp9_convolve8_avg_horiz_neon|
|
|
|
|
EXPORT |vp9_convolve8_avg_vert_neon|
|
|
|
|
IMPORT |vp9_convolve8_avg_horiz_c|
|
|
|
|
IMPORT |vp9_convolve8_avg_vert_c|
|
|
|
|
ARM
|
|
|
|
REQUIRE8
|
|
|
|
PRESERVE8
|
|
|
|
|
|
|
|
AREA ||.text||, CODE, READONLY, ALIGN=2
|
|
|
|
|
|
|
|
; Multiply and accumulate by q0
|
|
|
|
MACRO
|
|
|
|
MULTIPLY_BY_Q0 $dst, $src0, $src1, $src2, $src3, $src4, $src5, $src6, $src7
|
|
|
|
vmull.s16 $dst, $src0, d0[0]
|
|
|
|
vmlal.s16 $dst, $src1, d0[1]
|
|
|
|
vmlal.s16 $dst, $src2, d0[2]
|
|
|
|
vmlal.s16 $dst, $src3, d0[3]
|
|
|
|
vmlal.s16 $dst, $src4, d1[0]
|
|
|
|
vmlal.s16 $dst, $src5, d1[1]
|
|
|
|
vmlal.s16 $dst, $src6, d1[2]
|
|
|
|
vmlal.s16 $dst, $src7, d1[3]
|
|
|
|
MEND
|
|
|
|
|
|
|
|
; r0 const uint8_t *src
|
|
|
|
; r1 int src_stride
|
|
|
|
; r2 uint8_t *dst
|
|
|
|
; r3 int dst_stride
|
|
|
|
; sp[]const int16_t *filter_x
|
|
|
|
; sp[]int x_step_q4
|
|
|
|
; sp[]const int16_t *filter_y ; unused
|
|
|
|
; sp[]int y_step_q4 ; unused
|
|
|
|
; sp[]int w
|
|
|
|
; sp[]int h
|
|
|
|
|
|
|
|
|vp9_convolve8_avg_horiz_neon| PROC
|
2013-07-30 18:33:52 +02:00
|
|
|
ldr r12, [sp, #4] ; x_step_q4
|
|
|
|
cmp r12, #16
|
|
|
|
bne vp9_convolve8_avg_horiz_c
|
|
|
|
|
2013-07-13 01:12:58 +02:00
|
|
|
push {r4-r10, lr}
|
|
|
|
|
|
|
|
sub r0, r0, #3 ; adjust for taps
|
|
|
|
|
|
|
|
ldr r5, [sp, #32] ; filter_x
|
|
|
|
ldr r6, [sp, #48] ; w
|
|
|
|
ldr r7, [sp, #52] ; h
|
|
|
|
|
|
|
|
vld1.s16 {q0}, [r5] ; filter_x
|
|
|
|
|
2013-08-11 16:34:24 +02:00
|
|
|
sub r8, r1, r1, lsl #2 ; -src_stride * 3
|
|
|
|
add r8, r8, #4 ; -src_stride * 3 + 4
|
2013-07-13 01:12:58 +02:00
|
|
|
|
2013-08-11 16:34:24 +02:00
|
|
|
sub r4, r3, r3, lsl #2 ; -dst_stride * 3
|
|
|
|
add r4, r4, #4 ; -dst_stride * 3 + 4
|
2013-07-13 01:12:58 +02:00
|
|
|
|
2013-08-11 16:34:24 +02:00
|
|
|
rsb r9, r6, r1, lsl #2 ; reset src for outer loop
|
|
|
|
sub r9, r9, #7
|
2013-07-13 01:12:58 +02:00
|
|
|
rsb r12, r6, r3, lsl #2 ; reset dst for outer loop
|
|
|
|
|
|
|
|
mov r10, r6 ; w loop counter
|
|
|
|
|
2013-08-11 16:34:24 +02:00
|
|
|
loop_horiz_v
|
|
|
|
vld1.8 {d24}, [r0], r1
|
|
|
|
vld1.8 {d25}, [r0], r1
|
|
|
|
vld1.8 {d26}, [r0], r1
|
|
|
|
vld1.8 {d27}, [r0], r8
|
2013-07-13 01:12:58 +02:00
|
|
|
|
2013-07-30 19:11:06 +02:00
|
|
|
vtrn.16 q12, q13
|
|
|
|
vtrn.8 d24, d25
|
|
|
|
vtrn.8 d26, d27
|
|
|
|
|
2013-08-11 16:34:24 +02:00
|
|
|
pld [r0, r1, lsl #2]
|
|
|
|
|
2013-07-13 01:12:58 +02:00
|
|
|
vmovl.u8 q8, d24
|
|
|
|
vmovl.u8 q9, d25
|
|
|
|
vmovl.u8 q10, d26
|
|
|
|
vmovl.u8 q11, d27
|
2013-08-11 16:34:24 +02:00
|
|
|
|
|
|
|
; save a few instructions in the inner loop
|
|
|
|
vswp d17, d18
|
|
|
|
vmov d23, d21
|
|
|
|
|
|
|
|
add r0, r0, #3
|
|
|
|
|
|
|
|
loop_horiz
|
|
|
|
add r5, r0, #64
|
|
|
|
|
|
|
|
vld1.32 {d28[]}, [r0], r1
|
|
|
|
vld1.32 {d29[]}, [r0], r1
|
|
|
|
vld1.32 {d31[]}, [r0], r1
|
|
|
|
vld1.32 {d30[]}, [r0], r8
|
|
|
|
|
|
|
|
pld [r5]
|
|
|
|
|
|
|
|
vtrn.16 d28, d31
|
|
|
|
vtrn.16 d29, d30
|
|
|
|
vtrn.8 d28, d29
|
|
|
|
vtrn.8 d31, d30
|
|
|
|
|
|
|
|
pld [r5, r1]
|
|
|
|
|
|
|
|
; extract to s16
|
|
|
|
vtrn.32 q14, q15
|
2013-07-13 01:12:58 +02:00
|
|
|
vmovl.u8 q12, d28
|
2013-08-11 16:34:24 +02:00
|
|
|
vmovl.u8 q13, d29
|
|
|
|
|
|
|
|
pld [r5, r1, lsl #1]
|
2013-07-13 01:12:58 +02:00
|
|
|
|
|
|
|
; slightly out of order load to match the existing data
|
|
|
|
vld1.u32 {d6[0]}, [r2], r3
|
|
|
|
vld1.u32 {d7[0]}, [r2], r3
|
|
|
|
vld1.u32 {d6[1]}, [r2], r3
|
|
|
|
vld1.u32 {d7[1]}, [r2], r3
|
|
|
|
|
|
|
|
sub r2, r2, r3, lsl #2 ; reset for store
|
|
|
|
|
|
|
|
; src[] * filter_x
|
2013-08-11 16:34:24 +02:00
|
|
|
MULTIPLY_BY_Q0 q1, d16, d17, d20, d22, d18, d19, d23, d24
|
|
|
|
MULTIPLY_BY_Q0 q2, d17, d20, d22, d18, d19, d23, d24, d26
|
|
|
|
MULTIPLY_BY_Q0 q14, d20, d22, d18, d19, d23, d24, d26, d27
|
|
|
|
MULTIPLY_BY_Q0 q15, d22, d18, d19, d23, d24, d26, d27, d25
|
|
|
|
|
|
|
|
pld [r5, -r8]
|
2013-07-13 01:12:58 +02:00
|
|
|
|
|
|
|
; += 64 >> 7
|
|
|
|
vqrshrun.s32 d2, q1, #7
|
|
|
|
vqrshrun.s32 d3, q2, #7
|
|
|
|
vqrshrun.s32 d4, q14, #7
|
|
|
|
vqrshrun.s32 d5, q15, #7
|
|
|
|
|
|
|
|
; saturate
|
2013-07-30 19:08:17 +02:00
|
|
|
vqmovn.u16 d2, q1
|
|
|
|
vqmovn.u16 d3, q2
|
2013-07-13 01:12:58 +02:00
|
|
|
|
|
|
|
; transpose
|
|
|
|
vtrn.16 d2, d3
|
|
|
|
vtrn.32 d2, d3
|
|
|
|
vtrn.8 d2, d3
|
2013-08-11 16:34:24 +02:00
|
|
|
|
2013-07-13 01:12:58 +02:00
|
|
|
; average the new value and the dst value
|
2013-07-30 19:08:17 +02:00
|
|
|
vrhadd.u8 q1, q1, q3
|
2013-07-13 01:12:58 +02:00
|
|
|
|
|
|
|
vst1.u32 {d2[0]}, [r2], r3
|
|
|
|
vst1.u32 {d3[0]}, [r2], r3
|
|
|
|
vst1.u32 {d2[1]}, [r2], r3
|
|
|
|
vst1.u32 {d3[1]}, [r2], r4
|
|
|
|
|
2013-08-11 16:34:24 +02:00
|
|
|
vmov q8, q9
|
|
|
|
vmov d20, d23
|
|
|
|
vmov q11, q12
|
|
|
|
vmov q9, q13
|
|
|
|
|
2013-07-13 01:12:58 +02:00
|
|
|
subs r6, r6, #4 ; w -= 4
|
|
|
|
bgt loop_horiz
|
|
|
|
|
|
|
|
; outer loop
|
|
|
|
mov r6, r10 ; restore w counter
|
2013-08-11 16:34:24 +02:00
|
|
|
add r0, r0, r9 ; src += src_stride * 4 - w
|
2013-07-13 01:12:58 +02:00
|
|
|
add r2, r2, r12 ; dst += dst_stride * 4 - w
|
|
|
|
subs r7, r7, #4 ; h -= 4
|
2013-08-11 16:34:24 +02:00
|
|
|
bgt loop_horiz_v
|
2013-07-13 01:12:58 +02:00
|
|
|
|
|
|
|
pop {r4-r10, pc}
|
|
|
|
|
|
|
|
ENDP
|
|
|
|
|
|
|
|
|vp9_convolve8_avg_vert_neon| PROC
|
2013-07-30 18:33:52 +02:00
|
|
|
ldr r12, [sp, #12]
|
|
|
|
cmp r12, #16
|
|
|
|
bne vp9_convolve8_avg_vert_c
|
|
|
|
|
2013-08-12 16:37:48 +02:00
|
|
|
push {r4-r8, lr}
|
2013-07-13 01:12:58 +02:00
|
|
|
|
|
|
|
; adjust for taps
|
|
|
|
sub r0, r0, r1
|
|
|
|
sub r0, r0, r1, lsl #1
|
|
|
|
|
2013-08-12 16:37:48 +02:00
|
|
|
ldr r4, [sp, #32] ; filter_y
|
|
|
|
ldr r6, [sp, #40] ; w
|
|
|
|
ldr lr, [sp, #44] ; h
|
2013-07-13 01:12:58 +02:00
|
|
|
|
2013-08-12 16:37:48 +02:00
|
|
|
vld1.s16 {q0}, [r4] ; filter_y
|
2013-07-13 01:12:58 +02:00
|
|
|
|
2013-08-12 16:37:48 +02:00
|
|
|
lsl r1, r1, #1
|
|
|
|
lsl r3, r3, #1
|
2013-07-13 01:12:58 +02:00
|
|
|
|
2013-08-12 16:37:48 +02:00
|
|
|
loop_vert_h
|
|
|
|
mov r4, r0
|
|
|
|
add r7, r0, r1, asr #1
|
|
|
|
mov r5, r2
|
|
|
|
add r8, r2, r3, asr #1
|
|
|
|
mov r12, lr ; h loop counter
|
2013-07-13 01:12:58 +02:00
|
|
|
|
2013-08-12 16:37:48 +02:00
|
|
|
vld1.u32 {d16[0]}, [r4], r1
|
|
|
|
vld1.u32 {d16[1]}, [r7], r1
|
|
|
|
vld1.u32 {d18[0]}, [r4], r1
|
|
|
|
vld1.u32 {d18[1]}, [r7], r1
|
|
|
|
vld1.u32 {d20[0]}, [r4], r1
|
|
|
|
vld1.u32 {d20[1]}, [r7], r1
|
|
|
|
vld1.u32 {d22[0]}, [r4], r1
|
2013-07-13 01:12:58 +02:00
|
|
|
|
2013-08-12 16:37:48 +02:00
|
|
|
vmovl.u8 q8, d16
|
|
|
|
vmovl.u8 q9, d18
|
|
|
|
vmovl.u8 q10, d20
|
|
|
|
vmovl.u8 q11, d22
|
2013-07-13 01:12:58 +02:00
|
|
|
|
|
|
|
loop_vert
|
|
|
|
; always process a 4x4 block at a time
|
2013-08-12 16:37:48 +02:00
|
|
|
vld1.u32 {d24[0]}, [r7], r1
|
|
|
|
vld1.u32 {d26[0]}, [r4], r1
|
|
|
|
vld1.u32 {d26[1]}, [r7], r1
|
|
|
|
vld1.u32 {d24[1]}, [r4], r1
|
2013-07-13 01:12:58 +02:00
|
|
|
|
|
|
|
; extract to s16
|
|
|
|
vmovl.u8 q12, d24
|
|
|
|
vmovl.u8 q13, d26
|
|
|
|
|
2013-08-12 16:37:48 +02:00
|
|
|
vld1.u32 {d6[0]}, [r5], r3
|
|
|
|
vld1.u32 {d6[1]}, [r8], r3
|
|
|
|
vld1.u32 {d7[0]}, [r5], r3
|
|
|
|
vld1.u32 {d7[1]}, [r8], r3
|
2013-07-13 01:12:58 +02:00
|
|
|
|
2013-08-12 16:37:48 +02:00
|
|
|
pld [r7]
|
|
|
|
pld [r4]
|
2013-07-13 01:12:58 +02:00
|
|
|
|
|
|
|
; src[] * filter_y
|
2013-08-12 16:37:48 +02:00
|
|
|
MULTIPLY_BY_Q0 q1, d16, d17, d18, d19, d20, d21, d22, d24
|
|
|
|
|
|
|
|
pld [r7, r1]
|
|
|
|
pld [r4, r1]
|
|
|
|
|
|
|
|
MULTIPLY_BY_Q0 q2, d17, d18, d19, d20, d21, d22, d24, d26
|
|
|
|
|
|
|
|
pld [r5]
|
|
|
|
pld [r8]
|
|
|
|
|
|
|
|
MULTIPLY_BY_Q0 q14, d18, d19, d20, d21, d22, d24, d26, d27
|
|
|
|
|
|
|
|
pld [r5, r3]
|
|
|
|
pld [r8, r3]
|
|
|
|
|
|
|
|
MULTIPLY_BY_Q0 q15, d19, d20, d21, d22, d24, d26, d27, d25
|
2013-07-13 01:12:58 +02:00
|
|
|
|
|
|
|
; += 64 >> 7
|
|
|
|
vqrshrun.s32 d2, q1, #7
|
|
|
|
vqrshrun.s32 d3, q2, #7
|
|
|
|
vqrshrun.s32 d4, q14, #7
|
|
|
|
vqrshrun.s32 d5, q15, #7
|
|
|
|
|
|
|
|
; saturate
|
2013-07-30 19:08:17 +02:00
|
|
|
vqmovn.u16 d2, q1
|
|
|
|
vqmovn.u16 d3, q2
|
2013-07-13 01:12:58 +02:00
|
|
|
|
|
|
|
; average the new value and the dst value
|
2013-07-30 19:08:17 +02:00
|
|
|
vrhadd.u8 q1, q1, q3
|
2013-07-13 01:12:58 +02:00
|
|
|
|
2013-08-12 16:37:48 +02:00
|
|
|
sub r5, r5, r3, lsl #1 ; reset for store
|
|
|
|
sub r8, r8, r3, lsl #1
|
|
|
|
|
|
|
|
vst1.u32 {d2[0]}, [r5], r3
|
|
|
|
vst1.u32 {d2[1]}, [r8], r3
|
|
|
|
vst1.u32 {d3[0]}, [r5], r3
|
|
|
|
vst1.u32 {d3[1]}, [r8], r3
|
|
|
|
|
|
|
|
vmov q8, q10
|
|
|
|
vmov d18, d22
|
|
|
|
vmov d19, d24
|
|
|
|
vmov q10, q13
|
|
|
|
vmov d22, d25
|
2013-07-13 01:12:58 +02:00
|
|
|
|
2013-08-12 16:37:48 +02:00
|
|
|
subs r12, r12, #4 ; h -= 4
|
2013-07-13 01:12:58 +02:00
|
|
|
bgt loop_vert
|
|
|
|
|
|
|
|
; outer loop
|
2013-08-12 16:37:48 +02:00
|
|
|
add r0, r0, #4
|
|
|
|
add r2, r2, #4
|
|
|
|
subs r6, r6, #4 ; w -= 4
|
|
|
|
bgt loop_vert_h
|
2013-07-13 01:12:58 +02:00
|
|
|
|
2013-08-12 16:37:48 +02:00
|
|
|
pop {r4-r8, pc}
|
2013-07-13 01:12:58 +02:00
|
|
|
|
|
|
|
ENDP
|
|
|
|
END
|