2 * SIMD-optimized IDCT functions for HEVC decoding
3 * Copyright (c) Alexandra Hajkova
5 * This file is part of FFmpeg.
7 * FFmpeg is free software; you can redistribute it and/or
8 * modify it under the terms of the GNU Lesser General Public
9 * License as published by the Free Software Foundation; either
10 * version 2.1 of the License, or (at your option) any later version.
12 * FFmpeg is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15 * Lesser General Public License for more details.
17 * You should have received a copy of the GNU Lesser General Public
18 * License along with FFmpeg; if not, write to the Free Software
19 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
27 #include "libavutil/cpu.h"
28 #include "libavutil/ppc/cpu.h"
29 #include "libavutil/ppc/types_altivec.h"
30 #include "libavutil/ppc/util_altivec.h"
32 #include "libavcodec/hevcdsp.h"
35 static const vector int16_t trans4[4] = {
36 { 64, 64, 64, 64, 64, 64, 64, 64 },
37 { 83, 36, 83, 36, 83, 36, 83, 36 },
38 { 64, -64, 64, -64, 64, -64, 64, -64 },
39 { 36, -83, 36, -83, 36, -83, 36, -83 },
42 static const vec_u8 mask[2] = {
43 { 0x00, 0x01, 0x08, 0x09, 0x10, 0x11, 0x18, 0x19, 0x02, 0x03, 0x0A, 0x0B, 0x12, 0x13, 0x1A, 0x1B },
44 { 0x04, 0x05, 0x0C, 0x0D, 0x14, 0x15, 0x1C, 0x1D, 0x06, 0x07, 0x0E, 0x0F, 0x16, 0x17, 0x1E, 0x1F },
47 static void transform4x4(vector int16_t src_01, vector int16_t src_23,
48 vector int32_t res[4], const int shift, int16_t *coeffs)
50 vector int16_t src_02, src_13;
51 vector int32_t zero = vec_splat_s32(0);
52 vector int32_t e0, o0, e1, o1;
55 src_13 = vec_mergel(src_01, src_23);
56 src_02 = vec_mergeh(src_01, src_23);
58 e0 = vec_msums(src_02, trans4[0], zero);
59 o0 = vec_msums(src_13, trans4[1], zero);
60 e1 = vec_msums(src_02, trans4[2], zero);
61 o1 = vec_msums(src_13, trans4[3], zero);
63 add = vec_sl(vec_splat_s32(1), vec_splat_u32(shift - 1));
64 e0 = vec_add(e0, add);
65 e1 = vec_add(e1, add);
67 res[0] = vec_add(e0, o0);
68 res[1] = vec_add(e1, o1);
69 res[2] = vec_sub(e1, o1);
70 res[3] = vec_sub(e0, o0);
73 static void scale(vector int32_t res[4], vector int16_t res_packed[2], int shift)
76 vector unsigned int v_shift = vec_splat_u32(shift);
78 for (i = 0; i < 4; i++)
79 res[i] = vec_sra(res[i], v_shift);
82 res_packed[0] = vec_packs(res[0], res[1]);
83 res_packed[1] = vec_packs(res[2], res[3]);
86 #define FUNCDECL(a, depth) a ## _ ## depth ## _altivec
87 #define FUNC(a, b) FUNCDECL(a, b)
90 #include "hevcdsp_template.c"
94 #include "hevcdsp_template.c"
96 #endif /* HAVE_ALTIVEC */
98 av_cold void ff_hevc_dsp_init_ppc(HEVCDSPContext *c, const int bit_depth)
101 if (!PPC_ALTIVEC(av_get_cpu_flags()))
105 c->idct[0] = ff_hevc_idct_4x4_8_altivec;
107 c->idct[0] = ff_hevc_idct_4x4_10_altivec;
108 #endif /* HAVE_ALTIVEC */