OpenJPH
Open-source implementation of JPEG2000 Part-15
ojph_colour_avx.cpp
Go to the documentation of this file.
1//***************************************************************************/
2// This software is released under the 2-Clause BSD license, included
3// below.
4//
5// Copyright (c) 2019, Aous Naman
6// Copyright (c) 2019, Kakadu Software Pty Ltd, Australia
7// Copyright (c) 2019, The University of New South Wales, Australia
8//
9// Redistribution and use in source and binary forms, with or without
10// modification, are permitted provided that the following conditions are
11// met:
12//
13// 1. Redistributions of source code must retain the above copyright
14// notice, this list of conditions and the following disclaimer.
15//
16// 2. Redistributions in binary form must reproduce the above copyright
17// notice, this list of conditions and the following disclaimer in the
18// documentation and/or other materials provided with the distribution.
19//
20// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS
21// IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
22// TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
23// PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
24// HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
25// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED
26// TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
27// PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
28// LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
29// NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
30// SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
31//***************************************************************************/
32// This file is part of the OpenJPH software implementation.
33// File: ojph_colour_avx.cpp
34// Author: Aous Naman
35// Date: 11 October 2019
36//***************************************************************************/
37
38#include <cmath>
39
40#include "ojph_defs.h"
41#include "ojph_arch.h"
42#include "ojph_colour.h"
43#include "ojph_colour_local.h"
44
45#include <immintrin.h>
46
47namespace ojph {
48 namespace local {
49
51 void avx_cnvrt_si32_to_float_shftd(const si32 *sp, float *dp, float mul,
52 ui32 width)
53 {
54 __m256 shift = _mm256_set1_ps(0.5f);
55 __m256 m = _mm256_set1_ps(mul);
56 for (int i = (width + 7) >> 3; i > 0; --i, sp+=8, dp+=8)
57 {
58 __m256i t = _mm256_loadu_si256((__m256i*)sp);
59 __m256 s = _mm256_cvtepi32_ps(t);
60 s = _mm256_mul_ps(s, m);
61 s = _mm256_sub_ps(s, shift);
62 _mm256_store_ps(dp, s);
63 }
64 }
65
67 void avx_cnvrt_si32_to_float(const si32 *sp, float *dp, float mul,
68 ui32 width)
69 {
70 __m256 m = _mm256_set1_ps(mul);
71 for (int i = (width + 7) >> 3; i > 0; --i, sp+=8, dp+=8)
72 {
73 __m256i t = _mm256_loadu_si256((__m256i*)sp);
74 __m256 s = _mm256_cvtepi32_ps(t);
75 s = _mm256_mul_ps(s, m);
76 _mm256_store_ps(dp, s);
77 }
78 }
79
81 void avx_cnvrt_float_to_si32_shftd(const float *sp, si32 *dp, float mul,
82 ui32 width)
83 {
84 __m256 shift = _mm256_set1_ps(0.5f);
85 __m256 m = _mm256_set1_ps(mul);
86 for (int i = (width + 7) >> 3; i > 0; --i, sp+=8, dp+=8)
87 {
88 __m256 t = _mm256_load_ps(sp);
89 __m256 s = _mm256_add_ps(t, shift);
90 s = _mm256_mul_ps(s, m);
91 s = _mm256_round_ps(s, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
92 _mm256_storeu_si256((__m256i*)dp, _mm256_cvtps_epi32(s));
93 }
94 }
95
97 void avx_cnvrt_float_to_si32(const float *sp, si32 *dp, float mul,
98 ui32 width)
99 {
100 __m256 m = _mm256_set1_ps(mul);
101 for (int i = (width + 7) >> 3; i > 0; --i, sp+=8, dp+=8)
102 {
103 __m256 t = _mm256_load_ps(sp);
104 __m256 s = _mm256_mul_ps(t, m);
105 s = _mm256_round_ps(s, _MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC);
106 _mm256_storeu_si256((__m256i*)dp, _mm256_cvtps_epi32(s));
107 }
108 }
109
111 void avx_ict_forward(const float *r, const float *g, const float *b,
112 float *y, float *cb, float *cr, ui32 repeat)
113 {
114 __m256 alpha_rf = _mm256_set1_ps(CT_CNST::ALPHA_RF);
115 __m256 alpha_gf = _mm256_set1_ps(CT_CNST::ALPHA_GF);
116 __m256 alpha_bf = _mm256_set1_ps(CT_CNST::ALPHA_BF);
117 __m256 beta_cbf = _mm256_set1_ps(CT_CNST::BETA_CbF);
118 __m256 beta_crf = _mm256_set1_ps(CT_CNST::BETA_CrF);
119 for (int i = (repeat + 7) >> 3; i > 0; --i)
120 {
121 __m256 mr = _mm256_load_ps(r);
122 __m256 mb = _mm256_load_ps(b);
123 __m256 my = _mm256_mul_ps(alpha_rf, mr);
124 my = _mm256_add_ps(my, _mm256_mul_ps(alpha_gf, _mm256_load_ps(g)));
125 my = _mm256_add_ps(my, _mm256_mul_ps(alpha_bf, mb));
126 _mm256_store_ps(y, my);
127 _mm256_store_ps(cb, _mm256_mul_ps(beta_cbf, _mm256_sub_ps(mb, my)));
128 _mm256_store_ps(cr, _mm256_mul_ps(beta_crf, _mm256_sub_ps(mr, my)));
129
130 r += 8; g += 8; b += 8;
131 y += 8; cb += 8; cr += 8;
132 }
133 }
134
136 void avx_ict_backward(const float *y, const float *cb, const float *cr,
137 float *r, float *g, float *b, ui32 repeat)
138 {
139 __m256 gamma_cr2g = _mm256_set1_ps(CT_CNST::GAMMA_CR2G);
140 __m256 gamma_cb2g = _mm256_set1_ps(CT_CNST::GAMMA_CB2G);
141 __m256 gamma_cr2r = _mm256_set1_ps(CT_CNST::GAMMA_CR2R);
142 __m256 gamma_cb2b = _mm256_set1_ps(CT_CNST::GAMMA_CB2B);
143 for (int i = (repeat + 7) >> 3; i > 0; --i)
144 {
145 __m256 my = _mm256_load_ps(y);
146 __m256 mcr = _mm256_load_ps(cr);
147 __m256 mcb = _mm256_load_ps(cb);
148 __m256 mg = _mm256_sub_ps(my, _mm256_mul_ps(gamma_cr2g, mcr));
149 _mm256_store_ps(g, _mm256_sub_ps(mg, _mm256_mul_ps(gamma_cb2g, mcb)));
150 _mm256_store_ps(r, _mm256_add_ps(my, _mm256_mul_ps(gamma_cr2r, mcr)));
151 _mm256_store_ps(b, _mm256_add_ps(my, _mm256_mul_ps(gamma_cb2b, mcb)));
152
153 y += 8; cb += 8; cr += 8;
154 r += 8; g += 8; b += 8;
155 }
156 }
157
158 }
159}
void avx_cnvrt_float_to_si32(const float *sp, si32 *dp, float mul, ui32 width)
void avx_ict_forward(const float *r, const float *g, const float *b, float *y, float *cb, float *cr, ui32 repeat)
void avx_cnvrt_si32_to_float_shftd(const si32 *sp, float *dp, float mul, ui32 width)
void avx_cnvrt_si32_to_float(const si32 *sp, float *dp, float mul, ui32 width)
void avx_cnvrt_float_to_si32_shftd(const float *sp, si32 *dp, float mul, ui32 width)
void avx_ict_backward(const float *y, const float *cb, const float *cr, float *r, float *g, float *b, ui32 repeat)
int32_t si32
Definition: ojph_defs.h:55
uint32_t ui32
Definition: ojph_defs.h:54
static const float GAMMA_CR2R
static const float BETA_CbF
static const float GAMMA_CB2B
static const float ALPHA_RF
static const float GAMMA_CB2G
static const float GAMMA_CR2G
static const float ALPHA_BF
static const float BETA_CrF
static const float ALPHA_GF