xref: /aosp_15_r20/external/libaom/av1/encoder/arm/highbd_rdopt_neon.c (revision 77c1e3ccc04c968bd2bc212e87364f250e820521)
1 /*
2  * Copyright (c) 2023, Alliance for Open Media. All rights reserved.
3  *
4  * This source code is subject to the terms of the BSD 2 Clause License and
5  * the Alliance for Open Media Patent License 1.0. If the BSD 2 Clause License
6  * was not distributed with this source code in the LICENSE file, you can
7  * obtain it at www.aomedia.org/license/software. If the Alliance for Open
8  * Media Patent License 1.0 was not distributed with this source code in the
9  * PATENTS file, you can obtain it at www.aomedia.org/license/patent.
10  */
11 
12 #include <assert.h>
13 #include <arm_neon.h>
14 
15 #include "config/av1_rtcd.h"
16 #include "aom_dsp/arm/sum_neon.h"
17 
av1_highbd_block_error_neon(const tran_low_t * coeff,const tran_low_t * dqcoeff,intptr_t block_size,int64_t * ssz,int bd)18 int64_t av1_highbd_block_error_neon(const tran_low_t *coeff,
19                                     const tran_low_t *dqcoeff,
20                                     intptr_t block_size, int64_t *ssz, int bd) {
21   uint64x2_t err_u64 = vdupq_n_u64(0);
22   int64x2_t ssz_s64 = vdupq_n_s64(0);
23 
24   const int shift = 2 * (bd - 8);
25   const int rounding = shift > 0 ? 1 << (shift - 1) : 0;
26 
27   assert(block_size >= 16);
28   assert((block_size % 16) == 0);
29 
30   do {
31     const int32x4_t c = vld1q_s32(coeff);
32     const int32x4_t d = vld1q_s32(dqcoeff);
33 
34     const uint32x4_t diff = vreinterpretq_u32_s32(vabdq_s32(c, d));
35 
36     err_u64 = vmlal_u32(err_u64, vget_low_u32(diff), vget_low_u32(diff));
37     err_u64 = vmlal_u32(err_u64, vget_high_u32(diff), vget_high_u32(diff));
38 
39     ssz_s64 = vmlal_s32(ssz_s64, vget_low_s32(c), vget_low_s32(c));
40     ssz_s64 = vmlal_s32(ssz_s64, vget_high_s32(c), vget_high_s32(c));
41 
42     coeff += 4;
43     dqcoeff += 4;
44     block_size -= 4;
45   } while (block_size != 0);
46 
47   *ssz = (horizontal_add_s64x2(ssz_s64) + rounding) >> shift;
48   return ((int64_t)horizontal_add_u64x2(err_u64) + rounding) >> shift;
49 }
50