From patchwork Fri Jun 2 11:59:00 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Matthieu Bouron X-Patchwork-Id: 3792 Delivered-To: ffmpegpatchwork@gmail.com Received: by 10.103.10.2 with SMTP id 2csp116669vsk; Fri, 2 Jun 2017 03:04:39 -0700 (PDT) X-Received: by 10.28.157.11 with SMTP id g11mr2160478wme.113.1496397879595; Fri, 02 Jun 2017 03:04:39 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1496397879; cv=none; d=google.com; s=arc-20160816; b=f26SRiuaqtnt4YJ4xeMW6mHYR3l2d0uXTH3f0EP7U5ri0NU2b1lkc4jHNnV3iqe44/ IId1B+HRyl86hA+gHobM6m6+EUEMpzLcmzmAtBIqrbOOrprtZ1ZEu+jIMR5sFcVGWG99 gBV+NnQ22zLlIlPcgXXjg9uK7YgDXpI+N9ZBKEhbV003SiOyz89LlX1bIcMSiI0YaWLa Y8RbreZIgI5jKt5XOpmjr8BTu6t9C+YZ9ulEnMjTRgfeFRR2cXLgoU0or5xNLOmQOvHW JanWd17rxnvDrnuOS7F9/nU1g9B8bPw2E12jA9IMgJ1hCFxddJQD6wjEEoq3q6oWhMDj kWLA== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=sender:errors-to:content-transfer-encoding:mime-version:cc:reply-to :list-subscribe:list-help:list-post:list-archive:list-unsubscribe :list-id:precedence:subject:message-id:date:to:from:dkim-signature :delivered-to:arc-authentication-results; bh=MdeGHJtH8ZnUcbGVQfYWezkE3lQbDz1TINUt/LzUIJw=; b=jGe1uX0sjzoqOjB5p1CYjsvIxsH8GV4ha3Xs5F0AD22yv5dQ1zAGZUWd7WX3g/aag6 QKuc0RmTm5M4bQzr83LRwnoJG5EsvpElBfGyIPBmxn+09insVstatPV3YfleKBHfh2qH iAskzIx1Fy9Wl28+HXCo/9aF3kHfjhPpCziypw/bBimA4vMHg7sA7+5QnSqvAymH1FMv VT+KVPGIA8LT22UiGtQMzH7X7KYCLCcWi16BRnJKIwRTgc6YtZxlOD5HsiyENNgemHPJ Ko9WNVVngwwC9Wp7OoiyPBD/4uW/U6G5o4n2+avCI47dpNruNBp1CohwhGztwnipVOh3 +VFA== ARC-Authentication-Results: i=1; mx.google.com; dkim=neutral (body hash did not verify) header.i=@gmail.com; spf=pass (google.com: domain of ffmpeg-devel-bounces@ffmpeg.org designates 79.124.17.100 as permitted sender) smtp.mailfrom=ffmpeg-devel-bounces@ffmpeg.org; dmarc=fail (p=NONE sp=NONE dis=NONE) header.from=gmail.com Return-Path: Received: from ffbox0-bg.mplayerhq.hu (ffbox0-bg.ffmpeg.org. [79.124.17.100]) by mx.google.com with ESMTP id k16si21333054wrc.113.2017.06.02.03.04.39; Fri, 02 Jun 2017 03:04:39 -0700 (PDT) Received-SPF: pass (google.com: domain of ffmpeg-devel-bounces@ffmpeg.org designates 79.124.17.100 as permitted sender) client-ip=79.124.17.100; Authentication-Results: mx.google.com; dkim=neutral (body hash did not verify) header.i=@gmail.com; spf=pass (google.com: domain of ffmpeg-devel-bounces@ffmpeg.org designates 79.124.17.100 as permitted sender) smtp.mailfrom=ffmpeg-devel-bounces@ffmpeg.org; dmarc=fail (p=NONE sp=NONE dis=NONE) header.from=gmail.com Received: from [127.0.1.1] (localhost [127.0.0.1]) by ffbox0-bg.mplayerhq.hu (Postfix) with ESMTP id 647FC689B54; Fri, 2 Jun 2017 13:04:31 +0300 (EEST) X-Original-To: ffmpeg-devel@ffmpeg.org Delivered-To: ffmpeg-devel@ffmpeg.org Received: from mail-wm0-f67.google.com (mail-wm0-f67.google.com [74.125.82.67]) by ffbox0-bg.mplayerhq.hu (Postfix) with ESMTPS id DE090689A3F for ; Fri, 2 Jun 2017 13:04:24 +0300 (EEST) Received: by mail-wm0-f67.google.com with SMTP id d127so17727588wmf.1 for ; Fri, 02 Jun 2017 03:04:30 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id; bh=eoSQF3bQcMKNfXbbajNdXHVXmr3deB4ePzNwFxgNQGM=; b=cYsMTDIxlF/wmcmJ5O8OTOQiBdObU0Idy0kMy4Vi/mowzkUIR1ira4i9FouOiqLH4n 8ha68nSrlcCDqCUgX3q6L2I0qt5wL5dvYhHAnFs03z1JJQ+sGpBHUXc5Q49FEzOjSpg1 VTxom0d0gGKkCdhdpU8l+1WcfPHvqzcKLKaklZOtXqCm3q6p5WogCPXx9qLmaCQbgRSy ERh/vzTN+ptaMnKVnagTuA0nMYEEhXy9QjM1cP5GB9Y8mb34+IKG1lwcs/p+zQHVIWkW zmFWVJxfHHDZKHDmGkEzE02vrQGp7SIEloQOSC2Z2iYc1gUSlTJuQZSiLnTTaDgIhURt looA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id; bh=eoSQF3bQcMKNfXbbajNdXHVXmr3deB4ePzNwFxgNQGM=; b=e+BL9cz1s+2Qat6uStfX27FDkz2N3bMYHFVaIM/O/7AK8KSBNm8/nFswSYyPTJbMd/ t/pJWPz9yZGDvZCWYeOKrLozTpr4B/fuU3YwwE0qX3e3gQBvVWQ0yWm/1xHQcfSTIdMz IjNLVeB5hZBM/hvbS0QZ4Yr2HLSSa/jEK2TqYpZLVvdm+u91LMt0o7CWGf0VQlJZGk7K uz+IHpGsO3R8opS0ehhC5OuLKttMFKo/LMOV9jFnoQPkBfheAKCghZaKik82z6Zx/aK0 9cfPgyMg29hY76XoSVFV8jZjnPXudRqNQGvEC0WsjWdn2xBJR3GjxuHEzqYHdzVz3QzI 5m7A== X-Gm-Message-State: AODbwcAdR9b7uUBVNo13HbxUdcHa/nzFldSZQr+Tvw7FHNeJpXFU7C1c AWVd2OjQwc74uuKw X-Received: by 10.28.21.13 with SMTP id 13mr2589873wmv.13.1496397518558; Fri, 02 Jun 2017 02:58:38 -0700 (PDT) Received: from boken.gopro.lcl (73.136-14-84.ripe.coltfrance.com. [84.14.136.73]) by smtp.gmail.com with ESMTPSA id a9sm41864843wrc.22.2017.06.02.02.58.37 (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Fri, 02 Jun 2017 02:58:37 -0700 (PDT) From: Matthieu Bouron To: ffmpeg-devel@ffmpeg.org Date: Fri, 2 Jun 2017 13:59:00 +0200 Message-Id: <20170602115900.4554-1-matthieu.bouron@gmail.com> X-Mailer: git-send-email 2.13.0 Subject: [FFmpeg-devel] [PATCH] lavc/aarch64: add sbrdsp neon implementation X-BeenThere: ffmpeg-devel@ffmpeg.org X-Mailman-Version: 2.1.20 Precedence: list List-Id: FFmpeg development discussions and patches List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Reply-To: FFmpeg development discussions and patches Cc: Matthieu Bouron MIME-Version: 1.0 Errors-To: ffmpeg-devel-bounces@ffmpeg.org Sender: "ffmpeg-devel" --- Hello, The following patch adds an aarch64 neon implementation of the sbrdsp (tested on an Odroid-C2). It hasn't been benchmarked yet and it lacks the hf_apply_noise{0,1,2,3} functions (which will be added later). --- libavcodec/aarch64/Makefile | 2 + libavcodec/aarch64/sbrdsp_init_aarch64.c | 55 +++++++ libavcodec/aarch64/sbrdsp_neon.S | 258 +++++++++++++++++++++++++++++++ libavcodec/sbrdsp.h | 1 + libavcodec/sbrdsp_template.c | 2 + 5 files changed, 318 insertions(+) create mode 100644 libavcodec/aarch64/sbrdsp_init_aarch64.c create mode 100644 libavcodec/aarch64/sbrdsp_neon.S diff --git a/libavcodec/aarch64/Makefile b/libavcodec/aarch64/Makefile index 104bc67802..116c343e42 100644 --- a/libavcodec/aarch64/Makefile +++ b/libavcodec/aarch64/Makefile @@ -11,6 +11,7 @@ OBJS-$(CONFIG_NEON_CLOBBER_TEST) += aarch64/neontest.o OBJS-$(CONFIG_VIDEODSP) += aarch64/videodsp_init.o # decoders/encoders +OBJS-$(CONFIG_AAC_DECODER) += aarch64/sbrdsp_init_aarch64.o OBJS-$(CONFIG_DCA_DECODER) += aarch64/synth_filter_init.o OBJS-$(CONFIG_RV40_DECODER) += aarch64/rv40dsp_init_aarch64.o OBJS-$(CONFIG_VC1DSP) += aarch64/vc1dsp_init_aarch64.o @@ -27,6 +28,7 @@ ARMV8-OBJS-$(CONFIG_VIDEODSP) += aarch64/videodsp.o # NEON optimizations # subsystems +NEON-OBJS-$(CONFIG_AAC_DECODER) += aarch64/sbrdsp_neon.o NEON-OBJS-$(CONFIG_FFT) += aarch64/fft_neon.o NEON-OBJS-$(CONFIG_FMTCONVERT) += aarch64/fmtconvert_neon.o NEON-OBJS-$(CONFIG_H264CHROMA) += aarch64/h264cmc_neon.o diff --git a/libavcodec/aarch64/sbrdsp_init_aarch64.c b/libavcodec/aarch64/sbrdsp_init_aarch64.c new file mode 100644 index 0000000000..21aaf869fe --- /dev/null +++ b/libavcodec/aarch64/sbrdsp_init_aarch64.c @@ -0,0 +1,55 @@ +/* + * + * This file is part of FFmpeg. + * + * FFmpeg is free software; you can redistribute it and/or + * 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. + * + * FFmpeg is distributed in the hope that it will be useful, + * 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 + * License along with FFmpeg; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include "config.h" +#include "libavutil/aarch64/cpu.h" +#include "libavutil/attributes.h" +#include "libavcodec/sbrdsp.h" + +void ff_sbr_sum64x5_neon(float *z); +float ff_sbr_sum_square_neon(float (*x)[2], int n); +void ff_sbr_neg_odd_64_neon(float *x); +void ff_sbr_qmf_pre_shuffle_neon(float *z); +void ff_sbr_qmf_post_shuffle_neon(float W[32][2], const float *z); +void ff_sbr_qmf_deint_neg_neon(float *v, const float *src); +void ff_sbr_qmf_deint_bfly_neon(float *v, const float *src0, const float *src1); +void ff_sbr_hf_g_filt_neon(float (*Y)[2], const float (*X_high)[40][2], + const float *g_filt, int m_max, intptr_t ixh); +void ff_sbr_hf_gen_neon(float (*X_high)[2], const float (*X_low)[2], + const float alpha0[2], const float alpha1[2], + float bw, int start, int end); +void ff_sbr_autocorrelate_neon(const float x[40][2], float phi[3][2][2]); + +av_cold void ff_sbrdsp_init_aarch64(SBRDSPContext *s) +{ + int cpu_flags = av_get_cpu_flags(); + + if (have_neon(cpu_flags)) { + s->sum64x5 = ff_sbr_sum64x5_neon; + s->sum_square = ff_sbr_sum_square_neon; + s->neg_odd_64 = ff_sbr_neg_odd_64_neon; + s->qmf_pre_shuffle = ff_sbr_qmf_pre_shuffle_neon; + s->qmf_post_shuffle = ff_sbr_qmf_post_shuffle_neon; + s->qmf_deint_neg = ff_sbr_qmf_deint_neg_neon; + s->qmf_deint_bfly = ff_sbr_qmf_deint_bfly_neon; + s->hf_g_filt = ff_sbr_hf_g_filt_neon; + s->hf_gen = ff_sbr_hf_gen_neon; + s->autocorrelate = ff_sbr_autocorrelate_neon; + } +} diff --git a/libavcodec/aarch64/sbrdsp_neon.S b/libavcodec/aarch64/sbrdsp_neon.S new file mode 100644 index 0000000000..8f638e93c7 --- /dev/null +++ b/libavcodec/aarch64/sbrdsp_neon.S @@ -0,0 +1,258 @@ +/* + * Copyright (c) 2017 Matthieu Bouron + * + * This file is part of FFmpeg. + * + * FFmpeg is free software; you can redistribute it and/or + * 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. + * + * FFmpeg is distributed in the hope that it will be useful, + * 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 + * License along with FFmpeg; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include "libavutil/aarch64/asm.S" + +function ff_sbr_sum64x5_neon, export=1 + add x1, x0, #64*4 + add x2, x0, #128*4 + add x3, x0, #192*4 + add x4, x0, #256*4 + mov x5, #64 +1: ld1 {v0.4S}, [x0] + ld1 {v1.4S}, [x1], #16 + fadd v0.4S, v0.4S, v1.4S + ld1 {v2.4S}, [x2], #16 + fadd v0.4S, v0.4S, v2.4S + ld1 {v3.4S}, [x3], #16 + fadd v0.4S, v0.4S, v3.4S + ld1 {v4.4S}, [x4], #16 + fadd v0.4S, v0.4S, v4.4S + st1 {v0.4S}, [x0], #16 + subs x5, x5, #4 + bgt 1b + ret +endfunc + +function ff_sbr_sum_square_neon, export=1 + movi v0.4S, #0 +1: ld1 {v1.4S}, [x0], #16 + fmla v0.4S, v1.4S, v1.4S + subs w1, w1, #2 + bgt 1b + faddp v0.4S, v0.4S, v0.4S + faddp v0.4S, v0.4S, v0.4S + mov w0, v0.S[0] + ret +endfunc + +function ff_sbr_neg_odd_64_neon, export=1 + mov x1, x0 + movi v5.4S, #1<<7, lsl #24 + ld2 {v0.4S, v1.4S}, [x0], #32 + eor v1.16B, v1.16B, v5.16B + ld2 {v2.4S, v3.4S}, [x0], #32 +.rept 3 + st2 {v0.4S, v1.4S}, [x1], #32 + eor v3.16B, v3.16B, v5.16B + ld2 {v0.4S, v1.4S}, [x0], #32 + st2 {v2.4S, v3.4S}, [x1], #32 + eor v1.16B, v1.16B, v5.16B + ld2 {v2.4S, v3.4S}, [x0], #32 +.endr + eor v3.16B, v3.16B, v5.16B + st2 {v0.4S, v1.4S}, [x1], #32 + st2 {v2.4S, v3.4S}, [x1], #32 + ret +endfunc + +function ff_sbr_qmf_pre_shuffle_neon, export=1 + add x1, x0, #60*4 + add x2, x0, #64*4 + mov x3, #-16 + mov x4, #28 + mov x5, #-4 + movi v6.4S, #1<<7, lsl #24 + ld1 {v0.2S}, [x0], #8 + st1 {v0.2S}, [x2], #8 +1: ld1 {v1.4S}, [x1], x3 + ld1 {v2.4S}, [x0], #16 + eor v1.16B, v1.16B, v6.16B + rev64 v1.4S, v1.4S + ext v1.16B, v1.16B, v1.16B, #8 + st2 {v1.4S, v2.4S}, [x2], #32 + subs x4, x4, #4 + bgt 1b + add x1, x1, #8 + ld1 {v1.2S}, [x1], x5 + ld1 {v2.2S}, [x0], #8 + ld1 {v1.S}[3], [x1] + ld1 {v2.S}[2], [x0] + eor v1.16B, v1.16B, v6.16B + rev64 v1.4S, v1.4S + st2 {v1.2S, v2.2S}, [x2], #16 + st1 {v1.S}[2], [x2], #4 + st1 {v2.S}[2], [x2] + ret +endfunc + +function ff_sbr_qmf_post_shuffle_neon, export=1 + add x2, x1, #60*4 + mov x3, #-16 + mov x4, #32 + movi v6.4S, #1<<7, lsl #24 +1: ld1 {v0.4S}, [x2], x3 + ld1 {v1.4S}, [x1], #16 + eor v0.16B, v0.16B, v6.16B + rev64 v0.4S, v0.4S + ext v0.16B, v0.16B, v0.16B, #8 + st2 {v0.4S, v1.4S}, [x0], #32 + subs x4, x4, #4 + bgt 1b + ret +endfunc + +function ff_sbr_qmf_deint_neg_neon, export=1 + add x1, x1, #56*4 + add x2, x0, #60*4 + mov x3, #-32 + mov x4, #32 + movi v2.4S, #1<<7, lsl #24 +1: ld2 {v0.4S, v1.4S}, [x1], x3 + eor v0.16B, v0.16B, v2.16B + rev64 v1.4S, v1.4S + ext v1.16B, v1.16B, v1.16B, #8 + st1 {v0.4S}, [x2] + st1 {v1.4S}, [x0], #16 + sub x2, x2, #16 + subs x4, x4, #4 + bgt 1b + ret +endfunc + +function ff_sbr_qmf_deint_bfly_neon, export=1 + add x2, x2, #60*4 + add x3, x0, #124*4 + mov x4, #64 + mov x5, #-16 +1: ld1 {v0.4S}, [x1], #16 + ld1 {v1.4S}, [x2], x5 + rev64 v2.4S, v0.4S + ext v2.16B, v2.16B, v2.16B, #8 + rev64 v3.4S, v1.4S + ext v3.16B, v3.16B, v3.16B, #8 + fadd v1.4S, v1.4S, v2.4S + fsub v0.4S, v0.4S, v3.4S + st1 {v0.4S}, [x0], #16 + st1 {v1.4S}, [x3], x5 + subs x4, x4, #4 + bgt 1b + ret +endfunc + +const coeff, align=4 + .float 1.0, -1.0, 1.0, -1.0 +endconst + +function ff_sbr_hf_gen_neon, export=1 + sxtw x4, w4 + sxtw x5, w5 + movrel x6, coeff + ld1 {v7.4S}, [x6] + dup v1.4S, v0.S[0] + mov v2.8B, v1.8B + mov v2.S[2], v7.S[0] + mov v2.S[3], v7.S[0] + fmul v1.4S, v1.4S, v2.4S + ld1 {v0.D}[0], [x3] + ld1 {v0.D}[1], [x2] + fmul v0.4S, v0.4S, v1.4S + fmul v1.4S, v0.4S, v7.4S + rev64 v0.4S, v0.4S + sub x7, x5, x4 + add x0, x0, x4, lsl #3 + add x1, x1, x4, lsl #3 + sub x1, x1, #16 +1: ld1 {v2.4S}, [x1], #16 + ld1 {v3.2S}, [x1] + fmul v4.4S, v2.4S, v1.4S + fmul v5.4S, v2.4S, v0.4S + faddp v4.4S, v4.4S, v4.4S + faddp v5.4S, v5.4S, v5.4S + faddp v4.4S, v4.4S, v4.4S + faddp v5.4S, v5.4S, v5.4S + mov v4.S[1], v5.S[0] + fadd v4.2S, v4.2S, v3.2S + st1 {v4.2S}, [x0], #8 + sub x1, x1, #8 + subs x7, x7, #1 + bgt 1b + ret +endfunc + +function ff_sbr_hf_g_filt_neon, export=1 + sxtw x3, w3 + sxtw x4, w4 + mov x5, #40*2*4 + add x1, x1, x4, lsl #3 +1: ld1 {v0.2S}, [x1], x5 + ld1 {v1.S}[0], [x2], #4 + fmul v2.4S, v0.4S, v1.S[0] + st1 {v2.2S}, [x0], #8 + subs x3, x3, #1 + bgt 1b + ret +endfunc + +function ff_sbr_autocorrelate_neon, export=1 + mov x2, #38 + movrel x3, coeff + ld1 {v0.4S}, [x3] + movi v1.4S, #0 + movi v2.4S, #0 + movi v3.4S, #0 + ld1 {v4.2S}, [x0], #8 + ld1 {v5.2S}, [x0], #8 + fmul v16.2S, v4.2S, v4.2S + fmul v17.2S, v5.2S, v4.S[0] + fmul v18.2S, v5.2S, v4.S[1] +1: ld1 {v5.D}[1], [x0], #8 + fmla v1.2S, v4.2S, v4.2S + fmla v2.4S, v5.4S, v4.S[0] + fmla v3.4S, v5.4S, v4.S[1] + mov v4.D[0], v5.D[0] + mov v5.D[0], v5.D[1] + subs x2, x2, #1 + bgt 1b + fmul v19.2S, v4.2S, v4.2S + fmul v20.2S, v5.2S, v4.S[0] + fmul v21.2S, v5.2S, v4.S[1] + fadd v22.4S, v2.4S, v20.4S + fsub v22.4S, v22.4S, v17.4S + fadd v23.4S, v3.4S, v21.4S + fsub v23.4S, v23.4S, v18.4S + rev64 v23.4S, v23.4S + fmul v23.4S, v23.4S, v0.4S + fadd v22.4S, v22.4S, v23.4S + st1 {v22.4S}, [x1], #16 + fadd v23.2S, v1.2S, v19.2S + fsub v23.2S, v23.2S, v16.2S + faddp v23.2S, v23.2S, v23.2S + st1 {v23.S}[0], [x1] + add x1, x1, #8 + rev64 v3.2S, v3.2S + fmul v3.2S, v3.2S, v0.2S + fadd v2.2S, v2.2S, v3.2S + st1 {v2.2S}, [x1] + add x1, x1, #16 + faddp v1.2S, v1.2S, v1.2S + st1 {v1.S}[0], [x1] + ret +endfunc diff --git a/libavcodec/sbrdsp.h b/libavcodec/sbrdsp.h index 66852de618..e6fd76d8ed 100644 --- a/libavcodec/sbrdsp.h +++ b/libavcodec/sbrdsp.h @@ -48,6 +48,7 @@ extern const INTFLOAT AAC_RENAME(ff_sbr_noise_table)[][2]; void AAC_RENAME(ff_sbrdsp_init)(SBRDSPContext *s); void ff_sbrdsp_init_arm(SBRDSPContext *s); +void ff_sbrdsp_init_aarch64(SBRDSPContext *s); void ff_sbrdsp_init_x86(SBRDSPContext *s); void ff_sbrdsp_init_mips(SBRDSPContext *s); diff --git a/libavcodec/sbrdsp_template.c b/libavcodec/sbrdsp_template.c index 897a3bbffb..37a3365b97 100644 --- a/libavcodec/sbrdsp_template.c +++ b/libavcodec/sbrdsp_template.c @@ -94,6 +94,8 @@ av_cold void AAC_RENAME(ff_sbrdsp_init)(SBRDSPContext *s) #if !USE_FIXED if (ARCH_ARM) ff_sbrdsp_init_arm(s); + if (ARCH_AARCH64) + ff_sbrdsp_init_aarch64(s); if (ARCH_X86) ff_sbrdsp_init_x86(s); if (ARCH_MIPS)