From patchwork Sat Sep 3 21:45:54 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Paul B Mahol X-Patchwork-Id: 37641 Delivered-To: ffmpegpatchwork2@gmail.com Received: by 2002:a05:6a20:139a:b0:8f:1db5:eae2 with SMTP id w26csp1769172pzh; Sat, 3 Sep 2022 14:42:55 -0700 (PDT) X-Google-Smtp-Source: AA6agR7qnaJWKFw81DqOgRwzF3f3ggj5+XNTYEG9H44P9Sk6gE0Lf2e2fJL42JG1OJZU87whrqe6 X-Received: by 2002:a17:906:cc4d:b0:73d:f594:159b with SMTP id mm13-20020a170906cc4d00b0073df594159bmr27212306ejb.651.1662241375314; Sat, 03 Sep 2022 14:42:55 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1662241375; cv=none; d=google.com; s=arc-20160816; b=H5Xw46H9VxtM4rrJH3AYO0Bx7ThHqV9VTpIMU+vzHJhg6AOWP5IiZlNOTcnpiN7Wrf cnmY5lPNSpSJ0ZBkpt+qKCgV7yJpOb9gmq/wl408NWuouwoEFBn7+81NWdwropo7UZIv lD8hy1UTVtUSSbCVQprrCbcR+gPJ1Trxr0VWbrvDkUVBl1/K2o0bax62sDhB9IoqxvUp c0nK9CQeCYJIHP9e2zvR9JZI71SbyLSCn+sWwHe9eFbJ+UHxBNl/8JBHdO5lBuhuBVux 95PRgiDNVCPud9uS3P7DB5zORvkJvkhBysLteoAZEvWIVNp2trn12g16opU60sfMprnu ey4w== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=sender:errors-to:reply-to:list-subscribe:list-help:list-post :list-archive:list-unsubscribe:list-id:precedence:subject:to :message-id:date:from:mime-version:dkim-signature:delivered-to; bh=J5hchQeHEVh/ESFGXJQ0gqyUPAjsBczcWd4mr0ikZ68=; b=PFgpxdl1jR6QbECN1EtQRDZccHdwgRcJxkQ6RAyVe6tFKprLauj41/U9KtMvcnU+BQ rhd+fZYSz00qCEcVo8zkqE3M0MAQo3xMLSFqrh1bXOQXmarGSImAEODpGoPBOvsYTp6e wOTGNdZ31209VPPLsi2nDd3UrjUPuAFGlgJn8dHsrVLD5Wfn+OcXVVff/umhogtA1jsk aKBxYvEj/Rux+uMdz0fnlHH83Dfs2ucRjflGR55QULfc/82YomCWkt0hfcfqWTwcsxfN u76pIXsWlscjYEAzh0W7MgdGi+m0xC+tIRA/nZBjHtsJz/+kxireuX6Mbgcr5kjQbhJe bWwg== ARC-Authentication-Results: i=1; mx.google.com; dkim=neutral (body hash did not verify) header.i=@gmail.com header.s=20210112 header.b=JsKRp6lQ; 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=QUARANTINE 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 s17-20020a056402521100b00446e5ecf9bdsi5236084edd.37.2022.09.03.14.42.55; Sat, 03 Sep 2022 14:42:55 -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 header.s=20210112 header.b=JsKRp6lQ; 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=QUARANTINE 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 7B88468BA7A; Sun, 4 Sep 2022 00:42:52 +0300 (EEST) X-Original-To: ffmpeg-devel@ffmpeg.org Delivered-To: ffmpeg-devel@ffmpeg.org Received: from mail-yw1-f173.google.com (mail-yw1-f173.google.com [209.85.128.173]) by ffbox0-bg.mplayerhq.hu (Postfix) with ESMTPS id 4319E68B980 for ; Sun, 4 Sep 2022 00:42:46 +0300 (EEST) Received: by mail-yw1-f173.google.com with SMTP id 00721157ae682-3376851fe13so43463477b3.6 for ; Sat, 03 Sep 2022 14:42:46 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20210112; h=to:subject:message-id:date:from:mime-version:from:to:cc:subject :date; bh=osOcxiHb10+9Z4Drh3LbRsNchakNXzOVi73Ey69Z92E=; b=JsKRp6lQ6h//eg3YnRZtBiMBv4tVNt2P0XOIezJqjjXF0fQ0tJR9bkFuymMngkW8aN jFihq89ww6g6p9gx4V1IV373i3Lg6lGfGj/skgBWB2q6vy+yb2aTuTYDr3cCt+bHnSay De9uy82G1w+0WbaUTLdHytXCQrnkFSPAv3CWVFiV+FWAbWr+lZ60LcrzB5c6WtAPWW/x 291p2tqfgMqYK5IgRcw7k4REm/p/pkyFA9EMuzPUvPsxw/Thn8FCgj7NNfyOvkiP11Jc rqGIheoBF03xlqjqrldkrIsfX87kV+MBGIP5NVxmLvOAhqB20lQOblgKsCwVKs+X6/9y XvZA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=to:subject:message-id:date:from:mime-version:x-gm-message-state :from:to:cc:subject:date; bh=osOcxiHb10+9Z4Drh3LbRsNchakNXzOVi73Ey69Z92E=; b=Jw/eFH03E/FoOuYKAMYJdgqqsX8vMInqJjF5LGf97Qeh/PjCT1Q52zL2Uz4D0zY3tz DLUwZMlziBGWTUpMp4YrNhgT6MqEZhJYX5Qx+Qb5M8kquEkbYqeJP0enHT0WUrUybFsW WbS39tASAbbpMxcKcqqBPTrhwU8fTBiwMDayPxyeKo3gsGf17HHP9WI2v6GRi72uPbIT MW6he1+g791L0csoj+zV0zhWWTB+8TzlpJ7EdzYLIvbQdPQzm6YFDuXcryEKqnNt+AQz c//MgRkXQgXTV/6n8byvZV7CkHOwDM/e/UMo7u99Km3mgHg+yjs3N54ZWiFlUbRQ2lEN B17g== X-Gm-Message-State: ACgBeo0YEFOh1LW+JGPbnovA27W4uawGS7hlZe4k50vdaPhQ3taAoLO8 govOt02EtQ7dlv0lFoSYBq7MgV6Kv0p6UjUy9dCuKLgN X-Received: by 2002:a81:b09:0:b0:345:30d:77b7 with SMTP id 9-20020a810b09000000b00345030d77b7mr5151952ywl.177.1662241364774; Sat, 03 Sep 2022 14:42:44 -0700 (PDT) MIME-Version: 1.0 From: Paul B Mahol Date: Sat, 3 Sep 2022 23:45:54 +0200 Message-ID: To: FFmpeg development discussions and patches X-Content-Filtered-By: Mailman/MimeDel 2.1.29 Subject: [FFmpeg-devel] [PATCH] avcodec: flac x86 asm fix X-BeenThere: ffmpeg-devel@ffmpeg.org X-Mailman-Version: 2.1.29 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 Errors-To: ffmpeg-devel-bounces@ffmpeg.org Sender: "ffmpeg-devel" X-TUID: k+1SjiWWjrf7 Patch attached. From 0f220489eb6402c6be3bc1d897c95fa9bc10431c Mon Sep 17 00:00:00 2001 From: Paul B Mahol Date: Sat, 3 Sep 2022 23:41:38 +0200 Subject: [PATCH] avcodec/x86/flacdsp: fix bug in decorrelation Fixes #9297 Signed-off-by: Paul B Mahol --- libavcodec/x86/flacdsp.asm | 23 +++++++++++++++----- libavcodec/x86/flacdsp_init.c | 41 ++++++++++++++++++++++------------- 2 files changed, 44 insertions(+), 20 deletions(-) diff --git a/libavcodec/x86/flacdsp.asm b/libavcodec/x86/flacdsp.asm index 7138611526..6d755f4972 100644 --- a/libavcodec/x86/flacdsp.asm +++ b/libavcodec/x86/flacdsp.asm @@ -23,6 +23,10 @@ %include "libavutil/x86/x86util.asm" +SECTION_RODATA + +vector: db 0,1,4,5,8,9,12,13,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,0,1,4,5,8,9,12,13, + SECTION .text %macro PMACSDQL 5 @@ -89,6 +93,9 @@ LPC_32 sse4 ;---------------------------------------------------------------------------------- %macro FLAC_DECORRELATE_16 3-4 cglobal flac_decorrelate_%1_16, 2, 4, 4, out, in0, in1, len +%ifidn %1, indep2 + VBROADCASTI128 m2, [vector] +%endif %if ARCH_X86_32 mov lend, lenm %endif @@ -112,11 +119,17 @@ align 16 %endif %ifnidn %1, indep2 p%4d m2, m0, m1 + packssdw m%2, m%2 + packssdw m%3, m%3 + punpcklwd m%2, m%3 + psllw m%2, m3 +%else + pslld m%2, m3 + pslld m%3, m3 + pshufb m%2, m%2, m2 + pshufb m%3, m%3, m2 + punpcklwd m%2, m%3 %endif - packssdw m%2, m%2 - packssdw m%3, m%3 - punpcklwd m%2, m%3 - psllw m%2, m3 mova [outq + lenq], m%2 add lenq, 16 jl .loop @@ -292,7 +305,7 @@ align 16 REP_RET %endmacro -INIT_XMM sse2 +INIT_XMM ssse3 FLAC_DECORRELATE_16 indep2, 0, 1 ; Reuse stereo 16bits macro FLAC_DECORRELATE_INDEP 32, 2, 3, d FLAC_DECORRELATE_INDEP 16, 4, 3, w diff --git a/libavcodec/x86/flacdsp_init.c b/libavcodec/x86/flacdsp_init.c index 2deaf3117f..48e3e7c55c 100644 --- a/libavcodec/x86/flacdsp_init.c +++ b/libavcodec/x86/flacdsp_init.c @@ -34,7 +34,9 @@ void ff_flac_decorrelate_ls_##fmt##_##opt(uint8_t **out, int32_t **in, int chann void ff_flac_decorrelate_rs_##fmt##_##opt(uint8_t **out, int32_t **in, int channels, \ int len, int shift); \ void ff_flac_decorrelate_ms_##fmt##_##opt(uint8_t **out, int32_t **in, int channels, \ - int len, int shift); \ + int len, int shift); + +#define DECORRELATE_IFUNCS(fmt, opt) \ void ff_flac_decorrelate_indep2_##fmt##_##opt(uint8_t **out, int32_t **in, int channels, \ int len, int shift); \ void ff_flac_decorrelate_indep4_##fmt##_##opt(uint8_t **out, int32_t **in, int channels, \ @@ -48,6 +50,10 @@ DECORRELATE_FUNCS(16, sse2); DECORRELATE_FUNCS(16, avx); DECORRELATE_FUNCS(32, sse2); DECORRELATE_FUNCS(32, avx); +DECORRELATE_IFUNCS(16, ssse3); +DECORRELATE_IFUNCS(16, avx); +DECORRELATE_IFUNCS(32, ssse3); +DECORRELATE_IFUNCS(32, avx); av_cold void ff_flacdsp_init_x86(FLACDSPContext *c, enum AVSampleFormat fmt, int channels) { @@ -55,30 +61,35 @@ av_cold void ff_flacdsp_init_x86(FLACDSPContext *c, enum AVSampleFormat fmt, int int cpu_flags = av_get_cpu_flags(); if (EXTERNAL_SSE2(cpu_flags)) { + if (fmt == AV_SAMPLE_FMT_S16) { + c->decorrelate[1] = ff_flac_decorrelate_ls_16_sse2; + c->decorrelate[2] = ff_flac_decorrelate_rs_16_sse2; + c->decorrelate[3] = ff_flac_decorrelate_ms_16_sse2; + } else if (fmt == AV_SAMPLE_FMT_S32) { + c->decorrelate[1] = ff_flac_decorrelate_ls_32_sse2; + c->decorrelate[2] = ff_flac_decorrelate_rs_32_sse2; + c->decorrelate[3] = ff_flac_decorrelate_ms_32_sse2; + } + } + if (EXTERNAL_SSSE3(cpu_flags)) { if (fmt == AV_SAMPLE_FMT_S16) { if (channels == 2) - c->decorrelate[0] = ff_flac_decorrelate_indep2_16_sse2; + c->decorrelate[0] = ff_flac_decorrelate_indep2_16_ssse3; else if (channels == 4) - c->decorrelate[0] = ff_flac_decorrelate_indep4_16_sse2; + c->decorrelate[0] = ff_flac_decorrelate_indep4_16_ssse3; else if (channels == 6) - c->decorrelate[0] = ff_flac_decorrelate_indep6_16_sse2; + c->decorrelate[0] = ff_flac_decorrelate_indep6_16_ssse3; else if (ARCH_X86_64 && channels == 8) - c->decorrelate[0] = ff_flac_decorrelate_indep8_16_sse2; - c->decorrelate[1] = ff_flac_decorrelate_ls_16_sse2; - c->decorrelate[2] = ff_flac_decorrelate_rs_16_sse2; - c->decorrelate[3] = ff_flac_decorrelate_ms_16_sse2; + c->decorrelate[0] = ff_flac_decorrelate_indep8_16_ssse3; } else if (fmt == AV_SAMPLE_FMT_S32) { if (channels == 2) - c->decorrelate[0] = ff_flac_decorrelate_indep2_32_sse2; + c->decorrelate[0] = ff_flac_decorrelate_indep2_32_ssse3; else if (channels == 4) - c->decorrelate[0] = ff_flac_decorrelate_indep4_32_sse2; + c->decorrelate[0] = ff_flac_decorrelate_indep4_32_ssse3; else if (channels == 6) - c->decorrelate[0] = ff_flac_decorrelate_indep6_32_sse2; + c->decorrelate[0] = ff_flac_decorrelate_indep6_32_ssse3; else if (ARCH_X86_64 && channels == 8) - c->decorrelate[0] = ff_flac_decorrelate_indep8_32_sse2; - c->decorrelate[1] = ff_flac_decorrelate_ls_32_sse2; - c->decorrelate[2] = ff_flac_decorrelate_rs_32_sse2; - c->decorrelate[3] = ff_flac_decorrelate_ms_32_sse2; + c->decorrelate[0] = ff_flac_decorrelate_indep8_32_ssse3; } } if (EXTERNAL_SSE4(cpu_flags)) { -- 2.37.2