1/* 2 * Copyright (C) 2013 Xiaolei Yu <dreifachstein@gmail.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 "config.h" 22#if HAVE_AS_DN_DIRECTIVE 23#include "rgb2yuv_neon_common.S" 24 25/* downsampled R16G16B16 x8 */ 26alias_qw r16x8, q7 27alias_qw g16x8, q8 28alias_qw b16x8, q9 29 30alias n16x16_o, q11 31alias n16x16_ol, q11_l 32alias n16x16_oh, q11_h 33 34alias y32x16_el, q12 35alias y32x16_eh, q13 36alias y32x16_ol, q14 37alias y32x16_oh, q15 38 39alias y16x16_e, q12 40alias y16x16_el, q12_l 41alias y16x16_eh, q12_h 42alias y16x16_o, q13 43alias y16x16_ol, q13_l 44alias y16x16_oh, q13_h 45 46 47alias y8x16, y16x16_e 48 49 50.macro init src 51 // load s32x3x3, narrow to s16x3x3 52 vld3.i32 {q13_l, q14_l, q15_l}, [\src]! 53 vld3.i32 {q13_h[0], q14_h[0], q15_h[0]}, [\src] 54 55 vmovn.i32 CO_R, q13 56 vmovn.i32 CO_G, q14 57 vmovn.i32 CO_B, q15 58 59 vmov.u8 BIAS_Y, #16 60 vmov.u8 BIAS_U, #128 61.endm 62 63 64.macro compute_y_16x1_step action, s8x16, coeff 65 vmov.u8 n16x16_o, #0 66 vtrn.u8 \s8x16, n16x16_o 67 68 \action y32x16_el, \s8x16\()_l, \coeff 69 \action y32x16_eh, \s8x16\()_h, \coeff 70 \action y32x16_ol, n16x16_ol, \coeff 71 \action y32x16_oh, n16x16_oh, \coeff 72.endm 73 74/* 75 * in: r8x16, g8x16, b8x16 76 * out: y8x16 77 * clobber: q11-q15, r8x16, g8x16, b8x16 78 */ 79.macro compute_y_16x1 80 compute_y_16x1_step vmull, r8x16, CO_RY 81 compute_y_16x1_step vmlal, g8x16, CO_GY 82 compute_y_16x1_step vmlal, b8x16, CO_BY 83 84 vrshrn.i32 y16x16_el, y32x16_el, #15 85 vrshrn.i32 y16x16_eh, y32x16_eh, #15 86 vrshrn.i32 y16x16_ol, y32x16_ol, #15 87 vrshrn.i32 y16x16_oh, y32x16_oh, #15 88 89 vtrn.8 y16x16_e, y16x16_o 90 vadd.u8 y8x16, y8x16, BIAS_Y 91.endm 92 93alias c32x8_l, q14 94alias c32x8_h, q15 95 96alias_qw c16x8, q13 97alias_qw c8x8x2, q10 98 99.macro compute_chroma_8x1_step action, s16x8, coeff 100 \action c32x8_l, \s16x8\()_l, \coeff 101 \action c32x8_h, \s16x8\()_h, \coeff 102.endm 103 104/* 105 * in: r16x8, g16x8, b16x8 106 * out: c8x8 107 * clobber: q14-q15 108 */ 109.macro compute_chroma_8x1 c, C 110 compute_chroma_8x1_step vmull, r16x8, CO_R\C 111 compute_chroma_8x1_step vmlal, g16x8, CO_G\C 112 compute_chroma_8x1_step vmlal, b16x8, CO_B\C 113 114 vrshrn.i32 c16x8_l, c32x8_l, #15 115 vrshrn.i32 c16x8_h, c32x8_h, #15 116 vmovn.i16 \c\()8x8, c16x8 117 vadd.u8 \c\()8x8, \c\()8x8, BIAS_\C 118.endm 119 120 121 loop_420sp rgbx, nv12, init, kernel_420_16x2, 32 122#endif 123