nengel@2: /*
nengel@2:  * Copyright (c) 2010 Mans Rullgard <mans@mansr.com>
nengel@2:  *
nengel@2:  * This file is part of FFmpeg.
nengel@2:  *
nengel@2:  * FFmpeg is free software; you can redistribute it and/or
nengel@2:  * modify it under the terms of the GNU Lesser General Public
nengel@2:  * License as published by the Free Software Foundation; either
nengel@2:  * version 2.1 of the License, or (at your option) any later version.
nengel@2:  *
nengel@2:  * FFmpeg is distributed in the hope that it will be useful,
nengel@2:  * but WITHOUT ANY WARRANTY; without even the implied warranty of
nengel@2:  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
nengel@2:  * Lesser General Public License for more details.
nengel@2:  *
nengel@2:  * You should have received a copy of the GNU Lesser General Public
nengel@2:  * License along with FFmpeg; if not, write to the Free Software
nengel@2:  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
nengel@2:  */
nengel@2: 
nengel@2: #include "asm.S"
nengel@2: 
nengel@2: function ff_dca_lfe_fir_neon, export=1
nengel@2:         push            {r4-r6,lr}
nengel@2: 
nengel@2:         add             r4,  r0,  r3,  lsl #2   @ out2
nengel@2:         add             r5,  r2,  #256*4-16     @ cf1
nengel@2:         sub             r1,  r1,  #12
nengel@2:         cmp             r3,  #32
nengel@2:         moveq           r6,  #256/32
nengel@2:         movne           r6,  #256/64
nengel@2: NOVFP   vldr            d0,  [sp, #16]          @ scale, bias
nengel@2:         mov             lr,  #-16
nengel@2: 1:
nengel@2:         vmov.f32        q2,  #0.0               @ v0
nengel@2:         vmov.f32        q3,  #0.0               @ v1
nengel@2:         mov             r12, r6
nengel@2: 2:
nengel@2:         vld1.32         {q8},     [r2,:128]!    @ cf0
nengel@2:         vld1.32         {q9},     [r5,:128], lr @ cf1
nengel@2:         vld1.32         {q1},     [r1], lr      @ in
nengel@2:         subs            r12, r12, #4
nengel@2:         vrev64.32       q10, q8
nengel@2:         vmla.f32        q3,  q1,  q9
nengel@2:         vmla.f32        d4,  d2,  d21
nengel@2:         vmla.f32        d5,  d3,  d20
nengel@2:         bne             2b
nengel@2: 
nengel@2:         add             r1,  r1,  r6,  lsl #2
nengel@2:         subs            r3,  r3,  #1
nengel@2:         vadd.f32        d4,  d4,  d5
nengel@2:         vadd.f32        d6,  d6,  d7
nengel@2:         vpadd.f32       d4,  d4,  d6
nengel@2:         vdup.32         d5,  d0[1]
nengel@2:         vmla.f32        d5,  d4,  d0[0]
nengel@2:         vst1.32         {d5[0]},  [r0,:32]!
nengel@2:         vst1.32         {d5[1]},  [r4,:32]!
nengel@2:         bne             1b
nengel@2: 
nengel@2:         pop             {r4-r6,pc}
nengel@2: endfunc