C908 X60 avg_8_2x2_c : 1.2 1.0 avg_8_2x2_rvv_i32 : 0.7 0.7 avg_8_2x4_c : 2.0 2.2 avg_8_2x4_rvv_i32 : 1.2 1.2 avg_8_2x8_c : 3.7 4.0 avg_8_2x8_rvv_i32 : 1.7 1.5 avg_8_2x16_c : 7.2 7.7 avg_8_2x16_rvv_i32 : 3.0 2.7 avg_8_2x32_c : 14.2 15.2 avg_8_2x32_rvv_i32 : 5.5 5.0 avg_8_2x64_c : 51.0 43.7 avg_8_2x64_rvv_i32 : 39.2 29.7 avg_8_2x128_c : 100.5 79.2 avg_8_2x128_rvv_i32 : 79.7 68.2 avg_8_4x2_c : 1.7 2.0 avg_8_4x2_rvv_i32 : 1.0 0.7 avg_8_4x4_c : 3.5 3.7 avg_8_4x4_rvv_i32 : 1.2 1.2 avg_8_4x8_c : 6.7 7.0 avg_8_4x8_rvv_i32 : 1.7 1.5 avg_8_4x16_c : 13.5 14.0 avg_8_4x16_rvv_i32 : 3.0 2.7 avg_8_4x32_c : 26.2 27.7 avg_8_4x32_rvv_i32 : 5.5 4.7 avg_8_4x64_c : 73.0 73.7 avg_8_4x64_rvv_i32 : 39.0 32.5 avg_8_4x128_c : 143.0 137.2 avg_8_4x128_rvv_i32 : 72.7 68.0 avg_8_8x2_c : 3.5 3.5 avg_8_8x2_rvv_i32 : 1.0 0.7 avg_8_8x4_c : 6.2 6.5 avg_8_8x4_rvv_i32 : 1.5 1.0 avg_8_8x8_c : 12.7 13.2 avg_8_8x8_rvv_i32 : 2.0 1.5 avg_8_8x16_c : 25.0 26.5 avg_8_8x16_rvv_i32 : 3.2 2.7 avg_8_8x32_c : 50.0 52.7 avg_8_8x32_rvv_i32 : 6.2 5.0 avg_8_8x64_c : 118.7 122.5 avg_8_8x64_rvv_i32 : 40.2 31.5 avg_8_8x128_c : 236.7 220.2 avg_8_8x128_rvv_i32 : 85.2 67.7 avg_8_16x2_c : 6.2 6.7 avg_8_16x2_rvv_i32 : 1.2 0.7 avg_8_16x4_c : 12.5 13.0 avg_8_16x4_rvv_i32 : 1.7 1.0 avg_8_16x8_c : 24.5 26.0 avg_8_16x8_rvv_i32 : 3.0 1.7 avg_8_16x16_c : 49.0 51.5 avg_8_16x16_rvv_i32 : 5.5 3.0 avg_8_16x32_c : 97.5 102.5 avg_8_16x32_rvv_i32 : 10.5 5.5 avg_8_16x64_c : 213.7 222.0 avg_8_16x64_rvv_i32 : 48.5 34.2 avg_8_16x128_c : 434.7 420.0 avg_8_16x128_rvv_i32 : 97.7 74.0 avg_8_32x2_c : 12.2 12.7 avg_8_32x2_rvv_i32 : 1.5 1.0 avg_8_32x4_c : 24.5 25.5 avg_8_32x4_rvv_i32 : 3.0 1.7 avg_8_32x8_c : 48.5 50.7 avg_8_32x8_rvv_i32 : 5.2 2.7 avg_8_32x16_c : 96.7 101.2 avg_8_32x16_rvv_i32 : 10.2 5.0 avg_8_32x32_c : 192.7 202.2 avg_8_32x32_rvv_i32 : 19.7 9.5 avg_8_32x64_c : 427.5 426.5 avg_8_32x64_rvv_i32 : 64.2 18.2 avg_8_32x128_c : 816.5 821.0 avg_8_32x128_rvv_i32 : 135.2 75.5 avg_8_64x2_c : 24.0 25.2 avg_8_64x2_rvv_i32 : 2.7 1.5 avg_8_64x4_c : 48.2 50.5 avg_8_64x4_rvv_i32 : 5.0 2.7 avg_8_64x8_c : 96.0 100.7 avg_8_64x8_rvv_i32 : 9.7 4.5 avg_8_64x16_c : 207.7 201.2 avg_8_64x16_rvv_i32 : 19.0 9.0 avg_8_64x32_c : 383.2 402.0 avg_8_64x32_rvv_i32 : 37.5 17.5 avg_8_64x64_c : 837.2 828.7 avg_8_64x64_rvv_i32 : 84.7 35.5 avg_8_64x128_c : 1640.7 1640.2 avg_8_64x128_rvv_i32 : 206.0 153.0 avg_8_128x2_c : 48.7 51.0 avg_8_128x2_rvv_i32 : 5.2 2.7 avg_8_128x4_c : 96.7 101.5 avg_8_128x4_rvv_i32 : 10.2 5.0 avg_8_128x8_c : 192.2 202.0 avg_8_128x8_rvv_i32 : 19.7 9.2 avg_8_128x16_c : 400.7 403.2 avg_8_128x16_rvv_i32 : 38.7 18.5 avg_8_128x32_c : 786.7 805.7 avg_8_128x32_rvv_i32 : 77.0 36.2 avg_8_128x64_c : 1615.5 1655.5 avg_8_128x64_rvv_i32 : 189.7 80.7 avg_8_128x128_c : 3182.0 3238.0 avg_8_128x128_rvv_i32 : 397.5 308.5 w_avg_8_2x2_c : 1.7 1.2 w_avg_8_2x2_rvv_i32 : 1.2 1.0 w_avg_8_2x4_c : 2.7 2.7 w_avg_8_2x4_rvv_i32 : 1.7 1.5 w_avg_8_2x8_c : 21.7 4.7 w_avg_8_2x8_rvv_i32 : 2.7 2.5 w_avg_8_2x16_c : 9.5 9.2 w_avg_8_2x16_rvv_i32 : 4.7 4.2 w_avg_8_2x32_c : 19.0 18.7 w_avg_8_2x32_rvv_i32 : 9.0 8.0 w_avg_8_2x64_c : 62.0 50.2 w_avg_8_2x64_rvv_i32 : 47.7 33.5 w_avg_8_2x128_c : 116.7 87.7 w_avg_8_2x128_rvv_i32 : 80.0 69.5 w_avg_8_4x2_c : 2.5 2.5 w_avg_8_4x2_rvv_i32 : 1.2 1.0 w_avg_8_4x4_c : 4.7 4.5 w_avg_8_4x4_rvv_i32 : 1.7 1.7 w_avg_8_4x8_c : 9.0 8.7 w_avg_8_4x8_rvv_i32 : 2.7 2.5 w_avg_8_4x16_c : 17.7 17.5 w_avg_8_4x16_rvv_i32 : 4.7 4.2 w_avg_8_4x32_c : 35.0 35.0 w_avg_8_4x32_rvv_i32 : 9.0 8.0 w_avg_8_4x64_c : 100.5 84.5 w_avg_8_4x64_rvv_i32 : 42.2 33.7 w_avg_8_4x128_c : 203.5 151.2 w_avg_8_4x128_rvv_i32 : 83.0 69.5 w_avg_8_8x2_c : 4.5 4.5 w_avg_8_8x2_rvv_i32 : 1.2 1.2 w_avg_8_8x4_c : 8.7 8.7 w_avg_8_8x4_rvv_i32 : 2.0 1.7 w_avg_8_8x8_c : 17.0 17.0 w_avg_8_8x8_rvv_i32 : 3.2 2.5 w_avg_8_8x16_c : 34.0 33.5 w_avg_8_8x16_rvv_i32 : 5.5 4.2 w_avg_8_8x32_c : 86.0 67.5 w_avg_8_8x32_rvv_i32 : 10.5 8.0 w_avg_8_8x64_c : 187.2 149.5 w_avg_8_8x64_rvv_i32 : 45.0 35.5 w_avg_8_8x128_c : 342.7 290.0 w_avg_8_8x128_rvv_i32 : 108.7 70.2 w_avg_8_16x2_c : 8.5 8.2 w_avg_8_16x2_rvv_i32 : 2.0 1.2 w_avg_8_16x4_c : 16.7 16.7 w_avg_8_16x4_rvv_i32 : 3.0 1.7 w_avg_8_16x8_c : 33.2 33.5 w_avg_8_16x8_rvv_i32 : 5.5 3.0 w_avg_8_16x16_c : 66.2 66.7 w_avg_8_16x16_rvv_i32 : 10.5 5.0 w_avg_8_16x32_c : 132.5 131.0 w_avg_8_16x32_rvv_i32 : 20.0 9.7 w_avg_8_16x64_c : 340.0 283.5 w_avg_8_16x64_rvv_i32 : 60.5 37.2 w_avg_8_16x128_c : 641.2 597.5 w_avg_8_16x128_rvv_i32 : 118.7 77.7 w_avg_8_32x2_c : 16.5 16.7 w_avg_8_32x2_rvv_i32 : 3.2 1.7 w_avg_8_32x4_c : 33.2 33.2 w_avg_8_32x4_rvv_i32 : 5.5 2.7 w_avg_8_32x8_c : 66.0 62.5 w_avg_8_32x8_rvv_i32 : 10.5 5.0 w_avg_8_32x16_c : 131.5 132.0 w_avg_8_32x16_rvv_i32 : 20.2 9.5 w_avg_8_32x32_c : 261.7 272.0 w_avg_8_32x32_rvv_i32 : 39.7 18.0 w_avg_8_32x64_c : 575.2 545.5 w_avg_8_32x64_rvv_i32 : 105.5 58.7 w_avg_8_32x128_c : 1154.2 1088.0 w_avg_8_32x128_rvv_i32 : 207.0 98.0 w_avg_8_64x2_c : 33.0 33.0 w_avg_8_64x2_rvv_i32 : 6.2 2.7 w_avg_8_64x4_c : 65.5 66.0 w_avg_8_64x4_rvv_i32 : 11.5 5.0 w_avg_8_64x8_c : 131.2 132.5 w_avg_8_64x8_rvv_i32 : 22.5 9.5 w_avg_8_64x16_c : 268.2 262.5 w_avg_8_64x16_rvv_i32 : 44.2 18.0 w_avg_8_64x32_c : 561.5 528.7 w_avg_8_64x32_rvv_i32 : 88.0 35.2 w_avg_8_64x64_c : 1136.2 1124.0 w_avg_8_64x64_rvv_i32 : 222.0 82.2 w_avg_8_64x128_c : 2345.0 2312.7 w_avg_8_64x128_rvv_i32 : 423.0 190.5 w_avg_8_128x2_c : 65.7 66.5 w_avg_8_128x2_rvv_i32 : 11.2 5.5 w_avg_8_128x4_c : 131.2 132.2 w_avg_8_128x4_rvv_i32 : 22.0 10.2 w_avg_8_128x8_c : 263.5 312.0 w_avg_8_128x8_rvv_i32 : 43.2 19.7 w_avg_8_128x16_c : 528.7 526.2 w_avg_8_128x16_rvv_i32 : 85.5 39.5 w_avg_8_128x32_c : 1067.7 1062.7 w_avg_8_128x32_rvv_i32 : 171.7 78.2 w_avg_8_128x64_c : 2234.7 2168.7 w_avg_8_128x64_rvv_i32 : 400.0 159.0 w_avg_8_128x128_c : 4752.5 4295.0 w_avg_8_128x128_rvv_i32 : 757.7 365.5
Unified · Split
| @@ -0,0 +1,2 @@ | ||
| 1 | +OBJS-$(CONFIG_VVC_DECODER) += riscv/vvc/vvcdsp_init.o | |
| 2 | +RVV-OBJS-$(CONFIG_VVC_DECODER) += riscv/vvc/vvc_mc_rvv.o | |
| @@ -0,0 +1,287 @@ | ||
| 1 | +/* | |
| 2 | + * Copyright (c) 2024 Institue of Software Chinese Academy of Sciences (ISCAS). | |
| 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/riscv/asm.S" | |
| 22 | + | |
| 23 | +.macro vsetvlstatic8 w, vlen | |
| 24 | + .if \w == 2 && \vlen == 128 | |
| 25 | + vsetivli zero, \w, e8, mf8, ta, ma | |
| 26 | + .elseif \w == 4 && \vlen == 128 | |
| 27 | + vsetivli zero, \w, e8, mf4, ta, ma | |
| 28 | + .elseif \w == 8 && \vlen == 128 | |
| 29 | + vsetivli zero, \w, e8, mf2, ta, ma | |
| 30 | + .elseif \w == 16 && \vlen == 128 | |
| 31 | + vsetivli zero, \w, e8, m1, ta, ma | |
| 32 | + .elseif \w == 32 && \vlen == 128 | |
| 33 | + li t0, \w | |
| 34 | + vsetvli zero, t0, e8, m2, ta, ma | |
| 35 | + .elseif \w <= 4 && \vlen == 256 | |
| 36 | + vsetivli zero, \w, e8, mf8, ta, ma | |
| 37 | + .elseif \w == 8 && \vlen == 256 | |
| 38 | + vsetivli zero, \w, e8, mf4, ta, ma | |
| 39 | + .elseif \w == 16 && \vlen == 256 | |
| 40 | + vsetivli zero, \w, e8, mf2, ta, ma | |
| 41 | + .elseif \w == 32 && \vlen == 256 | |
| 42 | + li t0, \w | |
| 43 | + vsetvli zero, t0, e8, m1, ta, ma | |
| 44 | + .elseif \w == 64 && \vlen == 256 | |
| 45 | + li t0, \w | |
| 46 | + vsetvli zero, t0, e8, m2, ta, ma | |
| 47 | + .else | |
| 48 | + li t0, \w | |
| 49 | + vsetvli zero, t0, e8, m4, ta, ma | |
| 50 | + .endif | |
| 51 | +.endm | |
| 52 | + | |
| 53 | +.macro vsetvlstatic16 w, vlen | |
| 54 | + .if \w == 2 && \vlen == 128 | |
| 55 | + vsetivli zero, \w, e16, mf4, ta, ma | |
| 56 | + .elseif \w == 4 && \vlen == 128 | |
| 57 | + vsetivli zero, \w, e16, mf2, ta, ma | |
| 58 | + .elseif \w == 8 && \vlen == 128 | |
| 59 | + vsetivli zero, \w, e16, m1, ta, ma | |
| 60 | + .elseif \w == 16 && \vlen == 128 | |
| 61 | + vsetivli zero, \w, e16, m2, ta, ma | |
| 62 | + .elseif \w == 32 && \vlen == 128 | |
| 63 | + li t0, \w | |
| 64 | + vsetvli zero, t0, e16, m4, ta, ma | |
| 65 | + .elseif \w <= 4 && \vlen == 256 | |
| 66 | + vsetivli zero, \w, e16, mf4, ta, ma | |
| 67 | + .elseif \w == 8 && \vlen == 256 | |
| 68 | + vsetivli zero, \w, e16, mf2, ta, ma | |
| 69 | + .elseif \w == 16 && \vlen == 256 | |
| 70 | + vsetivli zero, \w, e16, m1, ta, ma | |
| 71 | + .elseif \w == 32 && \vlen == 256 | |
| 72 | + li t0, \w | |
| 73 | + vsetvli zero, t0, e16, m2, ta, ma | |
| 74 | + .elseif \w == 64 && \vlen == 256 | |
| 75 | + li t0, \w | |
| 76 | + vsetvli zero, t0, e16, m4, ta, ma | |
| 77 | + .else | |
| 78 | + li t0, \w | |
| 79 | + vsetvli zero, t0, e16, m8, ta, ma | |
| 80 | + .endif | |
| 81 | +.endm | |
| 82 | + | |
| 83 | +.macro vsetvlstatic32 w, vlen | |
| 84 | + .if \w == 2 | |
| 85 | + vsetivli zero, \w, e32, mf2, ta, ma | |
| 86 | + .elseif \w == 4 && \vlen == 128 | |
| 87 | + vsetivli zero, \w, e32, m1, ta, ma | |
| 88 | + .elseif \w == 8 && \vlen == 128 | |
| 89 | + vsetivli zero, \w, e32, m2, ta, ma | |
| 90 | + .elseif \w == 16 && \vlen == 128 | |
| 91 | + vsetivli zero, \w, e32, m4, ta, ma | |
| 92 | + .elseif \w == 4 && \vlen == 256 | |
| 93 | + vsetivli zero, \w, e32, mf2, ta, ma | |
| 94 | + .elseif \w == 8 && \vlen == 256 | |
| 95 | + vsetivli zero, \w, e32, m1, ta, ma | |
| 96 | + .elseif \w == 16 && \vlen == 256 | |
| 97 | + vsetivli zero, \w, e32, m2, ta, ma | |
| 98 | + .elseif \w == 32 && \vlen == 256 | |
| 99 | + li t0, \w | |
| 100 | + vsetvli zero, t0, e32, m4, ta, ma | |
| 101 | + .else | |
| 102 | + li t0, \w | |
| 103 | + vsetvli zero, t0, e32, m8, ta, ma | |
| 104 | + .endif | |
| 105 | +.endm | |
| 106 | + | |
| 107 | +.macro avg w, vlen, id | |
| 108 | +\id\w\vlen: | |
| 109 | +.if \w < 128 | |
| 110 | + vsetvlstatic16 \w, \vlen | |
| 111 | + addi t0, a2, 128*2 | |
| 112 | + addi t1, a3, 128*2 | |
| 113 | + add t2, a0, a1 | |
| 114 | + vle16.v v0, (a2) | |
| 115 | + vle16.v v8, (a3) | |
| 116 | + addi a5, a5, -2 | |
| 117 | + vle16.v v16, (t0) | |
| 118 | + vle16.v v24, (t1) | |
| 119 | + vadd.vv v8, v8, v0 | |
| 120 | + vadd.vv v24, v24, v16 | |
| 121 | + vmax.vx v8, v8, zero | |
| 122 | + vmax.vx v24, v24, zero | |
| 123 | + vsetvlstatic8 \w, \vlen | |
| 124 | + addi a2, a2, 128*4 | |
| 125 | + vnclipu.wi v8, v8, 7 | |
| 126 | + vnclipu.wi v24, v24, 7 | |
| 127 | + addi a3, a3, 128*4 | |
| 128 | + vse8.v v8, (a0) | |
| 129 | + vse8.v v24, (t2) | |
| 130 | + sh1add a0, a1, a0 | |
| 131 | +.else | |
| 132 | + addi a5, a5, -1 | |
| 133 | + mv t1, a0 | |
| 134 | + mv t2, a2 | |
| 135 | + mv t3, a3 | |
| 136 | + mv t4, a4 | |
| 137 | +1: | |
| 138 | + vsetvli t0, a4, e16, m8, ta, ma | |
| 139 | + sub a4, a4, t0 | |
| 140 | + vle16.v v0, (a2) | |
| 141 | + vle16.v v8, (a3) | |
| 142 | + vadd.vv v8, v8, v0 | |
| 143 | + vmax.vx v8, v8, zero | |
| 144 | + vsetvli zero, zero, e8, m4, ta, ma | |
| 145 | + vnclipu.wi v8, v8, 7 | |
| 146 | + vse8.v v8, (a0) | |
| 147 | + sh1add a2, t0, a2 | |
| 148 | + sh1add a3, t0, a3 | |
| 149 | + add a0, a0, t0 | |
| 150 | + bnez a4, 1b | |
| 151 | + add a0, t1, a1 | |
| 152 | + addi a2, t2, 128*2 | |
| 153 | + addi a3, t3, 128*2 | |
| 154 | + mv a4, t4 | |
| 155 | +.endif | |
| 156 | + bnez a5, \id\w\vlen\()b | |
| 157 | + ret | |
| 158 | +.endm | |
| 159 | + | |
| 160 | + | |
| 161 | +.macro AVG_JMP_TABLE id, vlen | |
| 162 | +const jmp_table_\id\vlen | |
| 163 | + .4byte \id\()2\vlen\()f - jmp_table_\id\vlen | |
| 164 | + .4byte \id\()4\vlen\()f - jmp_table_\id\vlen | |
| 165 | + .4byte \id\()8\vlen\()f - jmp_table_\id\vlen | |
| 166 | + .4byte \id\()16\vlen\()f - jmp_table_\id\vlen | |
| 167 | + .4byte \id\()32\vlen\()f - jmp_table_\id\vlen | |
| 168 | + .4byte \id\()64\vlen\()f - jmp_table_\id\vlen | |
| 169 | + .4byte \id\()128\vlen\()f - jmp_table_\id\vlen | |
| 170 | +endconst | |
| 171 | +.endm | |
| 172 | + | |
| 173 | +.macro AVG_J vlen, id | |
| 174 | + clz t1, a4 | |
| 175 | + neg t1, t1 | |
| 176 | + lla t5, jmp_table_\id\vlen | |
| 177 | + sh2add t1, t1, t5 | |
| 178 | + lw t1, ((__riscv_xlen-2)<<2)(t1) | |
| 179 | + add t1, t1, t5 | |
| 180 | + jr t1 | |
| 181 | +.endm | |
| 182 | + | |
| 183 | +.macro func_avg vlen | |
| 184 | +func ff_vvc_avg_8_rvv_\vlen\(), zve32x, zbb, zba | |
| 185 | + lpad 0 | |
| 186 | + AVG_JMP_TABLE 1, \vlen | |
| 187 | + csrwi vxrm, 0 | |
| 188 | + AVG_J \vlen, 1 | |
| 189 | + .irp w,2,4,8,16,32,64,128 | |
| 190 | + avg \w, \vlen, 1 | |
| 191 | + .endr | |
| 192 | +endfunc | |
| 193 | +.endm | |
| 194 | + | |
| 195 | +func_avg 128 | |
| 196 | +func_avg 256 | |
| 197 | + | |
| 198 | +#if (__riscv_xlen == 64) | |
| 199 | +.macro w_avg w, vlen, id | |
| 200 | +\id\w\vlen: | |
| 201 | +.if \w <= 32 || (\w == 64 && \vlen == 256) | |
| 202 | + vsetvlstatic16 \w, \vlen | |
| 203 | + addi t0, a2, 128*2 | |
| 204 | + addi t1, a3, 128*2 | |
| 205 | + vle16.v v0, (a2) | |
| 206 | + vle16.v v4, (a3) | |
| 207 | + addi a5, a5, -2 | |
| 208 | + vle16.v v8, (t0) | |
| 209 | + vle16.v v12, (t1) | |
| 210 | + vwmul.vx v16, v0, a7 | |
| 211 | + vwmul.vx v24, v8, a7 | |
| 212 | + vwmacc.vx v16, t3, v4 | |
| 213 | + vwmacc.vx v24, t3, v12 | |
| 214 | + vsetvlstatic32 \w, \vlen | |
| 215 | + add t2, a0, a1 | |
| 216 | + vadd.vx v16, v16, t4 | |
| 217 | + vadd.vx v24, v24, t4 | |
| 218 | + vsetvlstatic16 \w, \vlen | |
| 219 | + vnsrl.wx v16, v16, t6 | |
| 220 | + vnsrl.wx v24, v24, t6 | |
| 221 | + vmax.vx v16, v16, zero | |
| 222 | + vmax.vx v24, v24, zero | |
| 223 | + vsetvlstatic8 \w, \vlen | |
| 224 | + addi a2, a2, 128*4 | |
| 225 | + vnclipu.wi v16, v16, 0 | |
| 226 | + vnclipu.wi v24, v24, 0 | |
| 227 | + vse8.v v16, (a0) | |
| 228 | + addi a3, a3, 128*4 | |
| 229 | + vse8.v v24, (t2) | |
| 230 | + sh1add a0, a1, a0 | |
| 231 | +.else | |
| 232 | + addi a5, a5, -1 | |
| 233 | + mv t1, a0 | |
| 234 | + mv t2, a2 | |
| 235 | + mv t5, a3 | |
| 236 | + mv a6, a4 | |
| 237 | +1: | |
| 238 | + vsetvli t0, a4, e16, m4, ta, ma | |
| 239 | + sub a4, a4, t0 | |
| 240 | + vle16.v v0, (a2) | |
| 241 | + vle16.v v4, (a3) | |
| 242 | + vwmul.vx v16, v0, a7 | |
| 243 | + vwmacc.vx v16, t3, v4 | |
| 244 | + vsetvli zero, zero, e32, m8, ta, ma | |
| 245 | + vadd.vx v16, v16, t4 | |
| 246 | + vsetvli zero, zero, e16, m4, ta, ma | |
| 247 | + vnsrl.wx v16, v16, t6 | |
| 248 | + vmax.vx v16, v16, zero | |
| 249 | + vsetvli zero, zero, e8, m2, ta, ma | |
| 250 | + vnclipu.wi v16, v16, 0 | |
| 251 | + vse8.v v16, (a0) | |
| 252 | + sh1add a2, t0, a2 | |
| 253 | + sh1add a3, t0, a3 | |
| 254 | + add a0, a0, t0 | |
| 255 | + bnez a4, 1b | |
| 256 | + add a0, t1, a1 | |
| 257 | + addi a2, t2, 128*2 | |
| 258 | + addi a3, t5, 128*2 | |
| 259 | + mv a4, a6 | |
| 260 | +.endif | |
| 261 | + bnez a5, \id\w\vlen\()b | |
| 262 | + ret | |
| 263 | +.endm | |
| 264 | + | |
| 265 | +.macro func_w_avg vlen | |
| 266 | +func ff_vvc_w_avg_8_rvv_\vlen\(), zve32x, zbb, zba | |
| 267 | + lpad 0 | |
| 268 | + AVG_JMP_TABLE 2, \vlen | |
| 269 | + csrwi vxrm, 0 | |
| 270 | + addi t6, a6, 7 | |
| 271 | + ld t3, (sp) | |
| 272 | + ld t4, 8(sp) | |
| 273 | + ld t5, 16(sp) | |
| 274 | + addi t4, t4, 1 // o0 + o1 + 1 | |
| 275 | + add t4, t4, t5 | |
| 276 | + addi t5, t6, -1 // shift - 1 | |
| 277 | + sll t4, t4, t5 | |
| 278 | + AVG_J \vlen, 2 | |
| 279 | + .irp w,2,4,8,16,32,64,128 | |
| 280 | + w_avg \w, \vlen, 2 | |
| 281 | + .endr | |
| 282 | +endfunc | |
| 283 | +.endm | |
| 284 | + | |
| 285 | +func_w_avg 128 | |
| 286 | +func_w_avg 256 | |
| 287 | +#endif | |
| @@ -0,0 +1,72 @@ | ||
| 1 | +/* | |
| 2 | + * Copyright (c) 2024 Institue of Software Chinese Academy of Sciences (ISCAS). | |
| 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 "config.h" | |
| 22 | + | |
| 23 | +#include "libavutil/attributes.h" | |
| 24 | +#include "libavutil/cpu.h" | |
| 25 | +#include "libavutil/riscv/cpu.h" | |
| 26 | +#include "libavcodec/vvc/dsp.h" | |
| 27 | + | |
| 28 | +#define bf(fn, bd, opt) fn##_##bd##_##opt | |
| 29 | + | |
| 30 | +#define AVG_PROTOTYPES(bd, opt) \ | |
| 31 | +void bf(ff_vvc_avg, bd, opt)(uint8_t *dst, ptrdiff_t dst_stride, \ | |
| 32 | + const int16_t *src0, const int16_t *src1, int width, int height); \ | |
| 33 | +void bf(ff_vvc_w_avg, bd, opt)(uint8_t *dst, ptrdiff_t dst_stride, \ | |
| 34 | + const int16_t *src0, const int16_t *src1, int width, int height, \ | |
| 35 | + int denom, int w0, int w1, int o0, int o1); | |
| 36 | + | |
| 37 | +AVG_PROTOTYPES(8, rvv_128) | |
| 38 | +AVG_PROTOTYPES(8, rvv_256) | |
| 39 | + | |
| 40 | +void ff_vvc_dsp_init_riscv(VVCDSPContext *const c, const int bd) | |
| 41 | +{ | |
| 42 | +#if HAVE_RVV | |
| 43 | + const int flags = av_get_cpu_flags(); | |
| 44 | + int vlenb = ff_get_rv_vlenb(); | |
| 45 | + | |
| 46 | + if ((flags & AV_CPU_FLAG_RVV_I32) && (flags & AV_CPU_FLAG_RVB) && | |
| 47 | + vlenb >= 32) { | |
| 48 | + switch (bd) { | |
| 49 | + case 8: | |
| 50 | + c->inter.avg = ff_vvc_avg_8_rvv_256; | |
| 51 | +# if (__riscv_xlen == 64) | |
| 52 | + c->inter.w_avg = ff_vvc_w_avg_8_rvv_256; | |
| 53 | +# endif | |
| 54 | + break; | |
| 55 | + default: | |
| 56 | + break; | |
| 57 | + } | |
| 58 | + } else if ((flags & AV_CPU_FLAG_RVV_I32) && (flags & AV_CPU_FLAG_RVB) && | |
| 59 | + vlenb >= 16) { | |
| 60 | + switch (bd) { | |
| 61 | + case 8: | |
| 62 | + c->inter.avg = ff_vvc_avg_8_rvv_128; | |
| 63 | +# if (__riscv_xlen == 64) | |
| 64 | + c->inter.w_avg = ff_vvc_w_avg_8_rvv_128; | |
| 65 | +# endif | |
| 66 | + break; | |
| 67 | + default: | |
| 68 | + break; | |
| 69 | + } | |
| 70 | + } | |
| 71 | +#endif | |
| 72 | +} | |
| @@ -102,6 +102,8 @@ void ff_vvc_dsp_init(VVCDSPContext *vvcdsp, int bit_depth) · 101 unmodified lines | ||
| 102 | 102 | |
| 103 | 103 | #if ARCH_AARCH64 |
| 104 | 104 | ff_vvc_dsp_init_aarch64(vvcdsp, bit_depth); |
| 105 | +#elif ARCH_RISCV | |
| 106 | + ff_vvc_dsp_init_riscv(vvcdsp, bit_depth); | |
| 105 | 107 | #elif ARCH_X86 |
| 106 | 108 | ff_vvc_dsp_init_x86(vvcdsp, bit_depth); |
| 107 | 109 | #endif |
| @@ -179,6 +179,7 @@ typedef struct VVCDSPContext { · 178 unmodified lines | ||
| 179 | 179 | void ff_vvc_dsp_init(VVCDSPContext *hpc, int bit_depth); |
| 180 | 180 | |
| 181 | 181 | void ff_vvc_dsp_init_aarch64(VVCDSPContext *hpc, const int bit_depth); |
| 182 | +void ff_vvc_dsp_init_riscv(VVCDSPContext *hpc, const int bit_depth); | |
| 182 | 183 | void ff_vvc_dsp_init_x86(VVCDSPContext *hpc, const int bit_depth); |
| 183 | 184 | |
| 184 | 185 | #endif /* AVCODEC_VVC_DSP_H */ |