1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
|
/*
* Copyright (c) 2022 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.
*/
#include <arm_neon.h>
#include "./vpx_dsp_rtcd.h"
#include "./vpx_config.h"
#include "vpx/vpx_integer.h"
#include "vpx_dsp/arm/mem_neon.h"
#include "vpx_dsp/arm/sum_neon.h"
#include "vpx_ports/mem.h"
static INLINE uint32_t highbd_mse8_8xh_neon_dotprod(const uint16_t *src_ptr,
int src_stride,
const uint16_t *ref_ptr,
int ref_stride, int h) {
uint32x4_t sse_u32 = vdupq_n_u32(0);
int i = h / 2;
do {
uint16x8_t s0, s1, r0, r1;
uint8x16_t s, r, diff;
s0 = vld1q_u16(src_ptr);
src_ptr += src_stride;
s1 = vld1q_u16(src_ptr);
src_ptr += src_stride;
r0 = vld1q_u16(ref_ptr);
ref_ptr += ref_stride;
r1 = vld1q_u16(ref_ptr);
ref_ptr += ref_stride;
s = vcombine_u8(vmovn_u16(s0), vmovn_u16(s1));
r = vcombine_u8(vmovn_u16(r0), vmovn_u16(r1));
diff = vabdq_u8(s, r);
sse_u32 = vdotq_u32(sse_u32, diff, diff);
} while (--i != 0);
return horizontal_add_uint32x4(sse_u32);
}
static INLINE uint32_t highbd_mse8_16xh_neon_dotprod(const uint16_t *src_ptr,
int src_stride,
const uint16_t *ref_ptr,
int ref_stride, int h) {
uint32x4_t sse_u32 = vdupq_n_u32(0);
int i = h;
do {
uint16x8_t s0, s1, r0, r1;
uint8x16_t s, r, diff;
s0 = vld1q_u16(src_ptr);
s1 = vld1q_u16(src_ptr + 8);
r0 = vld1q_u16(ref_ptr);
r1 = vld1q_u16(ref_ptr + 8);
s = vcombine_u8(vmovn_u16(s0), vmovn_u16(s1));
r = vcombine_u8(vmovn_u16(r0), vmovn_u16(r1));
diff = vabdq_u8(s, r);
sse_u32 = vdotq_u32(sse_u32, diff, diff);
src_ptr += src_stride;
ref_ptr += ref_stride;
} while (--i != 0);
return horizontal_add_uint32x4(sse_u32);
}
#define HIGHBD_MSE_WXH_NEON_DOTPROD(w, h) \
uint32_t vpx_highbd_8_mse##w##x##h##_neon_dotprod( \
const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, \
int ref_stride, uint32_t *sse) { \
uint16_t *src = CONVERT_TO_SHORTPTR(src_ptr); \
uint16_t *ref = CONVERT_TO_SHORTPTR(ref_ptr); \
*sse = \
highbd_mse8_##w##xh_neon_dotprod(src, src_stride, ref, ref_stride, h); \
return *sse; \
}
HIGHBD_MSE_WXH_NEON_DOTPROD(16, 16)
HIGHBD_MSE_WXH_NEON_DOTPROD(16, 8)
HIGHBD_MSE_WXH_NEON_DOTPROD(8, 16)
HIGHBD_MSE_WXH_NEON_DOTPROD(8, 8)
#undef HIGHBD_MSE_WXH_NEON_DOTPROD
|