1/*
2 * Loongson SIMD optimized pixblockdsp
3 *
4 * Copyright (c) 2015 Loongson Technology Corporation Limited
5 * Copyright (c) 2015 Zhou Xiaoyong <zhouxiaoyong@loongson.cn>
6 *
7 * This file is part of FFmpeg.
8 *
9 * FFmpeg is free software; you can redistribute it and/or
10 * modify it under the terms of the GNU Lesser General Public
11 * License as published by the Free Software Foundation; either
12 * version 2.1 of the License, or (at your option) any later version.
13 *
14 * FFmpeg is distributed in the hope that it will be useful,
15 * but WITHOUT ANY WARRANTY; without even the implied warranty of
16 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
17 * Lesser General Public License for more details.
18 *
19 * You should have received a copy of the GNU Lesser General Public
20 * License along with FFmpeg; if not, write to the Free Software
21 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
22 */
23
24#include "pixblockdsp_mips.h"
25#include "libavutil/mips/asmdefs.h"
26#include "libavutil/mips/mmiutils.h"
27
28void ff_get_pixels_8_mmi(int16_t *av_restrict block, const uint8_t *pixels,
29                         ptrdiff_t stride)
30{
31    double ftmp[7];
32    DECLARE_VAR_ALL64;
33    DECLARE_VAR_ADDRT;
34
35    __asm__ volatile (
36        "pxor       %[ftmp0],   %[ftmp0],       %[ftmp0]                \n\t"
37
38        MMI_LDC1(%[ftmp1], %[pixels], 0x00)
39        MMI_LDXC1(%[ftmp2], %[pixels], %[stride], 0x00)
40        "punpcklbh  %[ftmp3],   %[ftmp1],       %[ftmp0]                \n\t"
41        "punpckhbh  %[ftmp4],   %[ftmp1],       %[ftmp0]                \n\t"
42        "punpcklbh  %[ftmp5],   %[ftmp2],       %[ftmp0]                \n\t"
43        "punpckhbh  %[ftmp6],   %[ftmp2],       %[ftmp0]                \n\t"
44        MMI_SDC1(%[ftmp3], %[block], 0x00)
45        MMI_SDC1(%[ftmp4], %[block], 0x08)
46        MMI_SDC1(%[ftmp5], %[block], 0x10)
47        MMI_SDC1(%[ftmp6], %[block], 0x18)
48        PTR_ADDU   "%[pixels],  %[pixels],      %[stride_x2]            \n\t"
49
50        MMI_LDC1(%[ftmp1], %[pixels], 0x00)
51        MMI_LDXC1(%[ftmp2], %[pixels], %[stride], 0x00)
52        "punpcklbh  %[ftmp3],   %[ftmp1],       %[ftmp0]                \n\t"
53        "punpckhbh  %[ftmp4],   %[ftmp1],       %[ftmp0]                \n\t"
54        "punpcklbh  %[ftmp5],   %[ftmp2],       %[ftmp0]                \n\t"
55        "punpckhbh  %[ftmp6],   %[ftmp2],       %[ftmp0]                \n\t"
56        MMI_SDC1(%[ftmp3], %[block], 0x20)
57        MMI_SDC1(%[ftmp4], %[block], 0x28)
58        MMI_SDC1(%[ftmp5], %[block], 0x30)
59        MMI_SDC1(%[ftmp6], %[block], 0x38)
60        PTR_ADDU   "%[pixels],  %[pixels],      %[stride_x2]            \n\t"
61
62        MMI_LDC1(%[ftmp1], %[pixels], 0x00)
63        MMI_LDXC1(%[ftmp2], %[pixels], %[stride], 0x00)
64        "punpcklbh  %[ftmp3],   %[ftmp1],       %[ftmp0]                \n\t"
65        "punpckhbh  %[ftmp4],   %[ftmp1],       %[ftmp0]                \n\t"
66        "punpcklbh  %[ftmp5],   %[ftmp2],       %[ftmp0]                \n\t"
67        "punpckhbh  %[ftmp6],   %[ftmp2],       %[ftmp0]                \n\t"
68        MMI_SDC1(%[ftmp3], %[block], 0x40)
69        MMI_SDC1(%[ftmp4], %[block], 0x48)
70        MMI_SDC1(%[ftmp5], %[block], 0x50)
71        MMI_SDC1(%[ftmp6], %[block], 0x58)
72        PTR_ADDU   "%[pixels],  %[pixels],      %[stride_x2]            \n\t"
73
74        MMI_LDC1(%[ftmp1], %[pixels], 0x00)
75        MMI_LDXC1(%[ftmp2], %[pixels], %[stride], 0x00)
76        "punpcklbh  %[ftmp3],   %[ftmp1],       %[ftmp0]                \n\t"
77        "punpckhbh  %[ftmp4],   %[ftmp1],       %[ftmp0]                \n\t"
78        "punpcklbh  %[ftmp5],   %[ftmp2],       %[ftmp0]                \n\t"
79        "punpckhbh  %[ftmp6],   %[ftmp2],       %[ftmp0]                \n\t"
80        MMI_SDC1(%[ftmp3], %[block], 0x60)
81        MMI_SDC1(%[ftmp4], %[block], 0x68)
82        MMI_SDC1(%[ftmp5], %[block], 0x70)
83        MMI_SDC1(%[ftmp6], %[block], 0x78)
84        : [ftmp0]"=&f"(ftmp[0]),            [ftmp1]"=&f"(ftmp[1]),
85          [ftmp2]"=&f"(ftmp[2]),            [ftmp3]"=&f"(ftmp[3]),
86          [ftmp4]"=&f"(ftmp[4]),            [ftmp5]"=&f"(ftmp[5]),
87          [ftmp6]"=&f"(ftmp[6]),
88          RESTRICT_ASM_ALL64
89          RESTRICT_ASM_ADDRT
90          [pixels]"+&r"(pixels)
91        : [block]"r"((mips_reg)block),      [stride]"r"((mips_reg)stride),
92          [stride_x2]"r"((mips_reg)(stride<<1))
93        : "memory"
94    );
95}
96
97void ff_diff_pixels_mmi(int16_t *av_restrict block, const uint8_t *src1,
98        const uint8_t *src2, ptrdiff_t stride)
99{
100    double ftmp[5];
101    mips_reg tmp[1];
102    DECLARE_VAR_ALL64;
103
104    __asm__ volatile (
105        "li         %[tmp0],    0x08                                    \n\t"
106        "pxor       %[ftmp4],   %[ftmp4],       %[ftmp4]                \n\t"
107        "1:                                                             \n\t"
108        MMI_LDC1(%[ftmp0], %[src1], 0x00)
109        "por        %[ftmp1],   %[ftmp0],       %[ftmp0]                \n\t"
110        MMI_LDC1(%[ftmp2], %[src2], 0x00)
111        "por        %[ftmp3],   %[ftmp2],       %[ftmp2]                \n\t"
112        "punpcklbh  %[ftmp0],   %[ftmp0],       %[ftmp4]                \n\t"
113        "punpckhbh  %[ftmp1],   %[ftmp1],       %[ftmp4]                \n\t"
114        "punpcklbh  %[ftmp2],   %[ftmp2],       %[ftmp4]                \n\t"
115        "punpckhbh  %[ftmp3],   %[ftmp3],       %[ftmp4]                \n\t"
116        "psubh      %[ftmp0],   %[ftmp0],       %[ftmp2]                \n\t"
117        "psubh      %[ftmp1],   %[ftmp1],       %[ftmp3]                \n\t"
118        MMI_SDC1(%[ftmp0], %[block], 0x00)
119        MMI_SDC1(%[ftmp1], %[block], 0x08)
120        PTR_ADDI   "%[tmp0],    %[tmp0], -0x01                          \n\t"
121        PTR_ADDIU  "%[block],   %[block], 0x10                          \n\t"
122        PTR_ADDU   "%[src1],    %[src1],        %[stride]               \n\t"
123        PTR_ADDU   "%[src2],    %[src2],        %[stride]               \n\t"
124        "bgtz       %[tmp0],    1b                                      \n\t"
125        : [ftmp0]"=&f"(ftmp[0]),            [ftmp1]"=&f"(ftmp[1]),
126          [ftmp2]"=&f"(ftmp[2]),            [ftmp3]"=&f"(ftmp[3]),
127          [ftmp4]"=&f"(ftmp[4]),
128          [tmp0]"=&r"(tmp[0]),
129          RESTRICT_ASM_ALL64
130          [block]"+&r"(block),              [src1]"+&r"(src1),
131          [src2]"+&r"(src2)
132        : [stride]"r"((mips_reg)stride)
133        : "memory"
134    );
135}
136