1 /*
2 * Copyright (c) 2015 Manojkumar Bhosale (Manojkumar.Bhosale@imgtec.com)
3 *
4 * This file is part of FFmpeg.
5 *
6 * FFmpeg is free software; you can redistribute it and/or
7 * modify it under the terms of the GNU Lesser General Public
8 * License as published by the Free Software Foundation; either
9 * version 2.1 of the License, or (at your option) any later version.
10 *
11 * FFmpeg is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14 * Lesser General Public License for more details.
15 *
16 * You should have received a copy of the GNU Lesser General Public
17 * License along with FFmpeg; if not, write to the Free Software
18 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
19 */
20
21 #include "libavutil/mips/generic_macros_msa.h"
22 #include "idctdsp_mips.h"
23
put_pixels_clamped_msa(const int16_t * block,uint8_t * pixels,int32_t stride)24 static void put_pixels_clamped_msa(const int16_t *block, uint8_t *pixels,
25 int32_t stride)
26 {
27 uint64_t in0_d, in1_d, in2_d, in3_d, in4_d, in5_d, in6_d, in7_d;
28 v8i16 in0, in1, in2, in3, in4, in5, in6, in7;
29
30 LD_SH8(block, 8, in0, in1, in2, in3, in4, in5, in6, in7);
31 CLIP_SH8_0_255(in0, in1, in2, in3, in4, in5, in6, in7);
32 PCKEV_B4_SH(in0, in0, in1, in1, in2, in2, in3, in3, in0, in1, in2, in3);
33 PCKEV_B4_SH(in4, in4, in5, in5, in6, in6, in7, in7, in4, in5, in6, in7);
34
35 in0_d = __msa_copy_u_d((v2i64) in0, 0);
36 in1_d = __msa_copy_u_d((v2i64) in1, 0);
37 in2_d = __msa_copy_u_d((v2i64) in2, 0);
38 in3_d = __msa_copy_u_d((v2i64) in3, 0);
39 in4_d = __msa_copy_u_d((v2i64) in4, 0);
40 in5_d = __msa_copy_u_d((v2i64) in5, 0);
41 in6_d = __msa_copy_u_d((v2i64) in6, 0);
42 in7_d = __msa_copy_u_d((v2i64) in7, 0);
43 SD4(in0_d, in1_d, in2_d, in3_d, pixels, stride);
44 pixels += 4 * stride;
45 SD4(in4_d, in5_d, in6_d, in7_d, pixels, stride);
46 }
47
put_signed_pixels_clamped_msa(const int16_t * block,uint8_t * pixels,int32_t stride)48 static void put_signed_pixels_clamped_msa(const int16_t *block, uint8_t *pixels,
49 int32_t stride)
50 {
51 uint64_t in0_d, in1_d, in2_d, in3_d, in4_d, in5_d, in6_d, in7_d;
52 v8i16 in0, in1, in2, in3, in4, in5, in6, in7;
53
54 LD_SH8(block, 8, in0, in1, in2, in3, in4, in5, in6, in7);
55
56 in0 += 128;
57 in1 += 128;
58 in2 += 128;
59 in3 += 128;
60 in4 += 128;
61 in5 += 128;
62 in6 += 128;
63 in7 += 128;
64
65 CLIP_SH8_0_255(in0, in1, in2, in3, in4, in5, in6, in7);
66 PCKEV_B4_SH(in0, in0, in1, in1, in2, in2, in3, in3, in0, in1, in2, in3);
67 PCKEV_B4_SH(in4, in4, in5, in5, in6, in6, in7, in7, in4, in5, in6, in7);
68
69 in0_d = __msa_copy_u_d((v2i64) in0, 0);
70 in1_d = __msa_copy_u_d((v2i64) in1, 0);
71 in2_d = __msa_copy_u_d((v2i64) in2, 0);
72 in3_d = __msa_copy_u_d((v2i64) in3, 0);
73 in4_d = __msa_copy_u_d((v2i64) in4, 0);
74 in5_d = __msa_copy_u_d((v2i64) in5, 0);
75 in6_d = __msa_copy_u_d((v2i64) in6, 0);
76 in7_d = __msa_copy_u_d((v2i64) in7, 0);
77 SD4(in0_d, in1_d, in2_d, in3_d, pixels, stride);
78 pixels += 4 * stride;
79 SD4(in4_d, in5_d, in6_d, in7_d, pixels, stride);
80 }
81
add_pixels_clamped_msa(const int16_t * block,uint8_t * pixels,int32_t stride)82 static void add_pixels_clamped_msa(const int16_t *block, uint8_t *pixels,
83 int32_t stride)
84 {
85 uint64_t in0_d, in1_d, in2_d, in3_d, in4_d, in5_d, in6_d, in7_d;
86 v8i16 in0, in1, in2, in3, in4, in5, in6, in7;
87 v16u8 pix_in0, pix_in1, pix_in2, pix_in3;
88 v16u8 pix_in4, pix_in5, pix_in6, pix_in7;
89 v8u16 pix0, pix1, pix2, pix3, pix4, pix5, pix6, pix7;
90 v8i16 zero = { 0 };
91
92 LD_SH8(block, 8, in0, in1, in2, in3, in4, in5, in6, in7);
93 LD_UB8(pixels, stride, pix_in0, pix_in1, pix_in2,
94 pix_in3, pix_in4, pix_in5, pix_in6, pix_in7);
95
96 ILVR_B4_UH(zero, pix_in0, zero, pix_in1, zero, pix_in2, zero, pix_in3,
97 pix0, pix1, pix2, pix3);
98 ILVR_B4_UH(zero, pix_in4, zero, pix_in5, zero, pix_in6, zero, pix_in7,
99 pix4, pix5, pix6, pix7);
100
101 in0 += (v8i16) pix0;
102 in1 += (v8i16) pix1;
103 in2 += (v8i16) pix2;
104 in3 += (v8i16) pix3;
105 in4 += (v8i16) pix4;
106 in5 += (v8i16) pix5;
107 in6 += (v8i16) pix6;
108 in7 += (v8i16) pix7;
109
110 CLIP_SH8_0_255(in0, in1, in2, in3, in4, in5, in6, in7);
111 PCKEV_B4_SH(in0, in0, in1, in1, in2, in2, in3, in3, in0, in1, in2, in3);
112 PCKEV_B4_SH(in4, in4, in5, in5, in6, in6, in7, in7, in4, in5, in6, in7);
113
114 in0_d = __msa_copy_u_d((v2i64) in0, 0);
115 in1_d = __msa_copy_u_d((v2i64) in1, 0);
116 in2_d = __msa_copy_u_d((v2i64) in2, 0);
117 in3_d = __msa_copy_u_d((v2i64) in3, 0);
118 in4_d = __msa_copy_u_d((v2i64) in4, 0);
119 in5_d = __msa_copy_u_d((v2i64) in5, 0);
120 in6_d = __msa_copy_u_d((v2i64) in6, 0);
121 in7_d = __msa_copy_u_d((v2i64) in7, 0);
122 SD4(in0_d, in1_d, in2_d, in3_d, pixels, stride);
123 pixels += 4 * stride;
124 SD4(in4_d, in5_d, in6_d, in7_d, pixels, stride);
125 }
126
ff_put_pixels_clamped_msa(const int16_t * block,uint8_t * av_restrict pixels,ptrdiff_t line_size)127 void ff_put_pixels_clamped_msa(const int16_t *block,
128 uint8_t *av_restrict pixels,
129 ptrdiff_t line_size)
130 {
131 put_pixels_clamped_msa(block, pixels, line_size);
132 }
133
ff_put_signed_pixels_clamped_msa(const int16_t * block,uint8_t * av_restrict pixels,ptrdiff_t line_size)134 void ff_put_signed_pixels_clamped_msa(const int16_t *block,
135 uint8_t *av_restrict pixels,
136 ptrdiff_t line_size)
137 {
138 put_signed_pixels_clamped_msa(block, pixels, line_size);
139 }
140
ff_add_pixels_clamped_msa(const int16_t * block,uint8_t * av_restrict pixels,ptrdiff_t line_size)141 void ff_add_pixels_clamped_msa(const int16_t *block,
142 uint8_t *av_restrict pixels,
143 ptrdiff_t line_size)
144 {
145 add_pixels_clamped_msa(block, pixels, line_size);
146 }
147