1 /*
2 * Copyright © 2018, VideoLAN and dav1d authors
3 * Copyright © 2018, Two Orioles, LLC
4 * All rights reserved.
5 *
6 * Redistribution and use in source and binary forms, with or without
7 * modification, are permitted provided that the following conditions are met:
8 *
9 * 1. Redistributions of source code must retain the above copyright notice, this
10 * list of conditions and the following disclaimer.
11 *
12 * 2. Redistributions in binary form must reproduce the above copyright notice,
13 * this list of conditions and the following disclaimer in the documentation
14 * and/or other materials provided with the distribution.
15 *
16 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
17 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
18 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
19 * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
20 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
21 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
22 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
23 * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
24 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
25 * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
26 */
27
28 #include "src/cpu.h"
29 #include "src/looprestoration.h"
30
31 #include "common/intops.h"
32
33 #define decl_wiener_filter_fns(ext) \
34 decl_lr_filter_fn(BF(dav1d_wiener_filter7, ext)); \
35 decl_lr_filter_fn(BF(dav1d_wiener_filter5, ext))
36
37 #define decl_sgr_filter_fns(ext) \
38 decl_lr_filter_fn(BF(dav1d_sgr_filter_5x5, ext)); \
39 decl_lr_filter_fn(BF(dav1d_sgr_filter_3x3, ext)); \
40 decl_lr_filter_fn(BF(dav1d_sgr_filter_mix, ext))
41
42 decl_wiener_filter_fns(sse2);
43 decl_wiener_filter_fns(ssse3);
44 decl_wiener_filter_fns(avx2);
45 decl_wiener_filter_fns(avx512icl);
46 decl_sgr_filter_fns(ssse3);
47 decl_sgr_filter_fns(avx2);
48 decl_sgr_filter_fns(avx512icl);
49
loop_restoration_dsp_init_x86(Dav1dLoopRestorationDSPContext * const c,const int bpc)50 static ALWAYS_INLINE void loop_restoration_dsp_init_x86(Dav1dLoopRestorationDSPContext *const c, const int bpc) {
51 const unsigned flags = dav1d_get_cpu_flags();
52
53 if (!(flags & DAV1D_X86_CPU_FLAG_SSE2)) return;
54 #if BITDEPTH == 8
55 c->wiener[0] = BF(dav1d_wiener_filter7, sse2);
56 c->wiener[1] = BF(dav1d_wiener_filter5, sse2);
57 #endif
58
59 if (!(flags & DAV1D_X86_CPU_FLAG_SSSE3)) return;
60 c->wiener[0] = BF(dav1d_wiener_filter7, ssse3);
61 c->wiener[1] = BF(dav1d_wiener_filter5, ssse3);
62 if (BITDEPTH == 8 || bpc == 10) {
63 c->sgr[0] = BF(dav1d_sgr_filter_5x5, ssse3);
64 c->sgr[1] = BF(dav1d_sgr_filter_3x3, ssse3);
65 c->sgr[2] = BF(dav1d_sgr_filter_mix, ssse3);
66 }
67
68 #if ARCH_X86_64
69 if (!(flags & DAV1D_X86_CPU_FLAG_AVX2)) return;
70
71 c->wiener[0] = BF(dav1d_wiener_filter7, avx2);
72 c->wiener[1] = BF(dav1d_wiener_filter5, avx2);
73 if (BITDEPTH == 8 || bpc == 10) {
74 c->sgr[0] = BF(dav1d_sgr_filter_5x5, avx2);
75 c->sgr[1] = BF(dav1d_sgr_filter_3x3, avx2);
76 c->sgr[2] = BF(dav1d_sgr_filter_mix, avx2);
77 }
78
79 if (!(flags & DAV1D_X86_CPU_FLAG_AVX512ICL)) return;
80
81 c->wiener[0] = BF(dav1d_wiener_filter7, avx512icl);
82 #if BITDEPTH == 8
83 /* With VNNI we don't need a 5-tap version. */
84 c->wiener[1] = c->wiener[0];
85 #else
86 c->wiener[1] = BF(dav1d_wiener_filter5, avx512icl);
87 #endif
88 if (BITDEPTH == 8 || bpc == 10) {
89 c->sgr[0] = BF(dav1d_sgr_filter_5x5, avx512icl);
90 c->sgr[1] = BF(dav1d_sgr_filter_3x3, avx512icl);
91 c->sgr[2] = BF(dav1d_sgr_filter_mix, avx512icl);
92 }
93 #endif
94 }
95