1 ;******************************************************************************
2 ;* AAC Spectral Band Replication decoding functions
3 ;* Copyright (C) 2012 Christophe Gisquet <christophe.gisquet@gmail.com>
5 ;* This file is part of Libav.
7 ;* Libav is free software; you can redistribute it and/or
8 ;* modify it under the terms of the GNU Lesser General Public
9 ;* License as published by the Free Software Foundation; either
10 ;* version 2.1 of the License, or (at your option) any later version.
12 ;* Libav is distributed in the hope that it will be useful,
13 ;* but WITHOUT ANY WARRANTY; without even the implied warranty of
14 ;* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15 ;* Lesser General Public License for more details.
17 ;* You should have received a copy of the GNU Lesser General Public
18 ;* License along with Libav; if not, write to the Free Software
19 ;* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
20 ;******************************************************************************
22 %include "libavutil/x86/x86util.asm"
25 ; mask equivalent for multiply by -1.0 1.0
26 ps_mask times 2 dd 1<<31, 0
31 cglobal sbr_sum_square, 2, 3, 6
57 ; len is a multiple of 2, thus there are at least 4 elements to process
79 cglobal sbr_hf_g_filt, 5, 6, 5
80 lea r1, [r1 + 8*r4] ; offset by ixh elements into X_high
88 movlps m0, [r2 + 4*r3 + 0]
89 movlps m1, [r2 + 4*r3 + 8]
90 movlps m2, [r1 + 0*STEP]
91 movlps m3, [r1 + 2*STEP]
92 movhps m2, [r1 + 1*STEP]
93 movhps m3, [r1 + 3*STEP]
98 movu [r0 + 8*r3 + 0], m0
99 movu [r0 + 8*r3 + 16], m1
103 and r5, 3 ; number of single element loops
105 .loop1: ; element 0 and 1 can be computed at the same time
119 ; static void sbr_hf_gen_c(float (*X_high)[2], const float (*X_low)[2],
120 ; const float alpha0[2], const float alpha1[2],
121 ; float bw, int start, int end)
123 cglobal sbr_hf_gen, 4,4,8, X_high, X_low, alpha0, alpha1, BW, S, E
126 %if ARCH_X86_64 == 0 || WIN64
132 mulps m2, bw ; (a1[0] a1[1])*bw
133 mulps m1, bw ; (a0[0] a0[1])*bw = (a2 a3)
134 mulps m2, bw ; (a1[0] a1[1])*bw*bw = (a0 a1)
140 %if ARCH_X86_64 == 0 || WIN64
141 ; start and end 6th and 7th args on stack
147 ; BW does not actually occupy a register, so shift by 1
151 sub start, end ; neg num of loops
152 lea X_highq, [X_highq + end*2*4]
153 lea X_lowq, [X_lowq + end*2*4 - 2*2*4]
154 shl start, 3 ; offset from num loops
156 mova m0, [X_lowq + start]
157 movlhps m1, m1 ; (a2 a3 a2 a3)
158 movlhps m2, m2 ; (a0 a1 a0 a1)
159 shufps m3, m3, q0101 ; (a3 a2 a3 a2)
160 shufps m4, m4, q0101 ; (a1 a0 a1 a0)
161 xorps m3, m7 ; (-a3 a2 -a3 a2)
162 xorps m4, m7 ; (-a1 a0 -a1 a0)
166 shufps m0, m0, q2200 ; {Xl[-2][0],",Xl[-1][0],"}
167 shufps m5, m5, q3311 ; {Xl[-2][1],",Xl[-1][1],"}
172 mova m0, [X_lowq + start + 2*2*4]
173 shufps m6, m0, q0022 ; {Xl[-1][0],",Xl[0][0],"}
174 shufps m7, m0, q1133 ; {Xl[-1][1],",Xl[1][1],"}
180 mova [X_highq + start], m5