GitList

libavcodec/aarch64/h264pred_neon.S

f56d8d8d	/* * Copyright (c) 2009 Mans Rullgard <mans@mansr.com> *
5d5f8b29	* This file is part of FFmpeg.
f56d8d8d	*
5d5f8b29	* FFmpeg is free software; you can redistribute it and/or
f56d8d8d	* modify it under the terms of the GNU Lesser General Public * License as published by the Free Software Foundation; either * version 2.1 of the License, or (at your option) any later version. *
5d5f8b29	* FFmpeg is distributed in the hope that it will be useful,
f56d8d8d	* but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU * Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public
5d5f8b29	* License along with FFmpeg; if not, write to the Free Software
f56d8d8d	* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA / #include "libavutil/aarch64/asm.S" .macro ldcol.8 rd, rs, rt, n=8, hi=0 .if \n >= 8 \|\| \hi == 0 ld1 {\rd\().b}[0], [\rs], \rt ld1 {\rd\().b}[1], [\rs], \rt ld1 {\rd\().b}[2], [\rs], \rt ld1 {\rd\().b}[3], [\rs], \rt .endif .if \n >= 8 \|\| \hi == 1 ld1 {\rd\().b}[4], [\rs], \rt ld1 {\rd\().b}[5], [\rs], \rt ld1 {\rd\().b}[6], [\rs], \rt ld1 {\rd\().b}[7], [\rs], \rt .endif .if \n == 16 ld1 {\rd\().b}[8], [\rs], \rt ld1 {\rd\().b}[9], [\rs], \rt ld1 {\rd\().b}[10], [\rs], \rt ld1 {\rd\().b}[11], [\rs], \rt ld1 {\rd\().b}[12], [\rs], \rt ld1 {\rd\().b}[13], [\rs], \rt ld1 {\rd\().b}[14], [\rs], \rt ld1 {\rd\().b}[15], [\rs], \rt .endif .endm function ff_pred16x16_128_dc_neon, export=1 movi v0.16b, #128 b .L_pred16x16_dc_end endfunc function ff_pred16x16_top_dc_neon, export=1 sub x2, x0, x1 ld1 {v0.16b}, [x2] uaddlv h0, v0.16b rshrn v0.8b, v0.8h, #4 dup v0.16b, v0.b[0] b .L_pred16x16_dc_end endfunc function ff_pred16x16_left_dc_neon, export=1 sub x2, x0, #1 ldcol.8 v0, x2, x1, 16 uaddlv h0, v0.16b rshrn v0.8b, v0.8h, #4 dup v0.16b, v0.b[0] b .L_pred16x16_dc_end endfunc function ff_pred16x16_dc_neon, export=1 sub x2, x0, x1 sub x3, x0, #1 ld1 {v0.16b}, [x2] ldcol.8 v1, x3, x1, 16 uaddlv h0, v0.16b uaddlv h1, v1.16b add v0.4h, v0.4h, v1.4h rshrn v0.8b, v0.8h, #5 dup v0.16b, v0.b[0] .L_pred16x16_dc_end: mov w3, #8 6: st1 {v0.16b}, [x0], x1 st1 {v0.16b}, [x0], x1 subs w3, w3, #1 b.ne 6b ret endfunc function ff_pred16x16_hor_neon, export=1 sub x2, x0, #1 mov w3, #16 1: ld1r {v0.16b}, [x2], x1 st1 {v0.16b}, [x0], x1 subs w3, w3, #1 b.ne 1b ret endfunc function ff_pred16x16_vert_neon, export=1 sub x2, x0, x1 add x1, x1, x1 ld1 {v0.16b}, [x2], x1 mov w3, #8 1: st1 {v0.16b}, [x0], x1 st1 {v0.16b}, [x2], x1 subs w3, w3, #1 b.ne 1b ret endfunc function ff_pred16x16_plane_neon, export=1 sub x3, x0, x1 movrel x4, p16weight add x2, x3, #8 sub x3, x3, #1 ld1 {v0.8b}, [x3] ld1 {v2.8b}, [x2], x1 ldcol.8 v1, x3, x1 add x3, x3, x1 ldcol.8 v3, x3, x1 rev64 v0.8b, v0.8b rev64 v1.8b, v1.8b uaddl v7.8h, v2.8b, v3.8b usubl v2.8h, v2.8b, v0.8b usubl v3.8h, v3.8b, v1.8b ld1 {v0.8h}, [x4] mul v2.8h, v2.8h, v0.8h mul v3.8h, v3.8h, v0.8h addp v2.8h, v2.8h, v3.8h addp v2.8h, v2.8h, v2.8h addp v2.4h, v2.4h, v2.4h sshll v3.4s, v2.4h, #2 saddw v2.4s, v3.4s, v2.4h rshrn v4.4h, v2.4s, #6 trn2 v5.4h, v4.4h, v4.4h add v2.4h, v4.4h, v5.4h shl v3.4h, v2.4h, #3 ext v7.16b, v7.16b, v7.16b, #14 sub v3.4h, v3.4h, v2.4h // 7 (b + c) add v7.4h, v7.4h, v0.4h shl v2.4h, v7.4h, #4 sub v2.4h, v2.4h, v3.4h shl v3.4h, v4.4h, #4 ext v0.16b, v0.16b, v0.16b, #14 sub v6.4h, v5.4h, v3.4h mov v0.h[0], wzr mul v0.8h, v0.8h, v4.h[0] dup v1.8h, v2.h[0] dup v2.8h, v4.h[0] dup v3.8h, v6.h[0] shl v2.8h, v2.8h, #3 add v1.8h, v1.8h, v0.8h add v3.8h, v3.8h, v2.8h mov w3, #16 1: sqshrun v0.8b, v1.8h, #5 add v1.8h, v1.8h, v2.8h sqshrun2 v0.16b, v1.8h, #5 add v1.8h, v1.8h, v3.8h st1 {v0.16b}, [x0], x1 subs w3, w3, #1 b.ne 1b ret endfunc const p16weight, align=4 .short 1,2,3,4,5,6,7,8 endconst const p8weight, align=4 .short 1,2,3,4,1,2,3,4 endconst function ff_pred8x8_hor_neon, export=1 sub x2, x0, #1 mov w3, #8 1: ld1r {v0.8b}, [x2], x1 st1 {v0.8b}, [x0], x1 subs w3, w3, #1 b.ne 1b ret endfunc function ff_pred8x8_vert_neon, export=1 sub x2, x0, x1 lsl x1, x1, #1 ld1 {v0.8b}, [x2], x1 mov w3, #4 1: st1 {v0.8b}, [x0], x1 st1 {v0.8b}, [x2], x1 subs w3, w3, #1 b.ne 1b ret endfunc function ff_pred8x8_plane_neon, export=1 sub x3, x0, x1 movrel x4, p8weight movrel x5, p16weight add x2, x3, #4 sub x3, x3, #1 ld1 {v0.s}[0], [x3] ld1 {v2.s}[0], [x2], x1 ldcol.8 v0, x3, x1, 4, hi=1 add x3, x3, x1 ldcol.8 v3, x3, x1, 4 uaddl v7.8h, v2.8b, v3.8b rev32 v0.8b, v0.8b trn1 v2.2s, v2.2s, v3.2s usubl v2.8h, v2.8b, v0.8b ld1 {v6.8h}, [x4] mul v2.8h, v2.8h, v6.8h ld1 {v0.8h}, [x5] saddlp v2.4s, v2.8h addp v2.4s, v2.4s, v2.4s shl v3.4s, v2.4s, #4 add v2.4s, v3.4s, v2.4s rshrn v5.4h, v2.4s, #5 addp v2.4h, v5.4h, v5.4h shl v3.4h, v2.4h, #1 add v3.4h, v3.4h, v2.4h rev64 v7.4h, v7.4h add v7.4h, v7.4h, v0.4h shl v2.4h, v7.4h, #4 sub v2.4h, v2.4h, v3.4h ext v0.16b, v0.16b, v0.16b, #14 mov v0.h[0], wzr mul v0.8h, v0.8h, v5.h[0] dup v1.8h, v2.h[0] dup v2.8h, v5.h[1] add v1.8h, v1.8h, v0.8h mov w3, #8 1: sqshrun v0.8b, v1.8h, #5 add v1.8h, v1.8h, v2.8h st1 {v0.8b}, [x0], x1 subs w3, w3, #1 b.ne 1b ret endfunc function ff_pred8x8_128_dc_neon, export=1 movi v0.8b, #128 movi v1.8b, #128 b .L_pred8x8_dc_end endfunc function ff_pred8x8_top_dc_neon, export=1 sub x2, x0, x1 ld1 {v0.8b}, [x2] uaddlp v0.4h, v0.8b addp v0.4h, v0.4h, v0.4h zip1 v0.8h, v0.8h, v0.8h rshrn v2.8b, v0.8h, #2 zip1 v0.8b, v2.8b, v2.8b zip1 v1.8b, v2.8b, v2.8b b .L_pred8x8_dc_end endfunc function ff_pred8x8_left_dc_neon, export=1 sub x2, x0, #1 ldcol.8 v0, x2, x1 uaddlp v0.4h, v0.8b addp v0.4h, v0.4h, v0.4h rshrn v2.8b, v0.8h, #2 dup v1.8b, v2.b[1] dup v0.8b, v2.b[0] b .L_pred8x8_dc_end endfunc function ff_pred8x8_dc_neon, export=1 sub x2, x0, x1 sub x3, x0, #1 ld1 {v0.8b}, [x2] ldcol.8 v1, x3, x1 uaddlp v0.4h, v0.8b uaddlp v1.4h, v1.8b trn1 v2.2s, v0.2s, v1.2s trn2 v3.2s, v0.2s, v1.2s addp v4.4h, v2.4h, v3.4h addp v5.4h, v4.4h, v4.4h rshrn v6.8b, v5.8h, #3 rshrn v7.8b, v4.8h, #2 dup v0.8b, v6.b[0] dup v2.8b, v7.b[2] dup v1.8b, v7.b[3] dup v3.8b, v6.b[1] zip1 v0.2s, v0.2s, v2.2s zip1 v1.2s, v1.2s, v3.2s .L_pred8x8_dc_end: mov w3, #4 add x2, x0, x1, lsl #2 6: st1 {v0.8b}, [x0], x1 st1 {v1.8b}, [x2], x1 subs w3, w3, #1 b.ne 6b ret endfunc function ff_pred8x8_l0t_dc_neon, export=1 sub x2, x0, x1 sub x3, x0, #1 ld1 {v0.8b}, [x2] ldcol.8 v1, x3, x1, 4 zip1 v0.4s, v0.4s, v1.4s uaddlp v0.8h, v0.16b addp v0.8h, v0.8h, v0.8h addp v1.4h, v0.4h, v0.4h rshrn v2.8b, v0.8h, #2 rshrn v3.8b, v1.8h, #3 dup v4.8b, v3.b[0] dup v6.8b, v2.b[2] dup v5.8b, v2.b[0] zip1 v0.2s, v4.2s, v6.2s zip1 v1.2s, v5.2s, v6.2s b .L_pred8x8_dc_end endfunc function ff_pred8x8_l00_dc_neon, export=1 sub x2, x0, #1 ldcol.8 v0, x2, x1, 4 uaddlp v0.4h, v0.8b addp v0.4h, v0.4h, v0.4h rshrn v0.8b, v0.8h, #2 movi v1.8b, #128 dup v0.8b, v0.b[0] b .L_pred8x8_dc_end endfunc function ff_pred8x8_0lt_dc_neon, export=1 add x3, x0, x1, lsl #2 sub x2, x0, x1 sub x3, x3, #1 ld1 {v0.8b}, [x2] ldcol.8 v1, x3, x1, 4, hi=1 zip1 v0.4s, v0.4s, v1.4s uaddlp v0.8h, v0.16b addp v0.8h, v0.8h, v0.8h addp v1.4h, v0.4h, v0.4h rshrn v2.8b, v0.8h, #2 rshrn v3.8b, v1.8h, #3 dup v4.8b, v2.b[0] dup v5.8b, v2.b[3] dup v6.8b, v2.b[2] dup v7.8b, v3.b[1] zip1 v0.2s, v4.2s, v6.2s zip1 v1.2s, v5.2s, v7.2s b .L_pred8x8_dc_end endfunc function ff_pred8x8_0l0_dc_neon, export=1 add x2, x0, x1, lsl #2 sub x2, x2, #1 ldcol.8 v1, x2, x1, 4 uaddlp v2.4h, v1.8b addp v2.4h, v2.4h, v2.4h rshrn v1.8b, v2.8h, #2 movi v0.8b, #128 dup v1.8b, v1.b[0] b .L_pred8x8_dc_end endfunc