From patchwork Tue Jun 11 12:28:57 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Ramiro Polla X-Patchwork-Id: 49794 Delivered-To: ffmpegpatchwork2@gmail.com Received: by 2002:a59:613:0:b0:460:55fa:d5ed with SMTP id 19csp362887vqg; Tue, 11 Jun 2024 05:29:33 -0700 (PDT) X-Forwarded-Encrypted: i=2; AJvYcCXfN0e5kKHXf9CxrkTD2ijWGiyqwYyOwCdJC8SUQ8jlh4AL0ZpoSFih1H81koGBKDi8iuYY+ZoEp3TX2RojuT/u1Bp0F0Y74OQRDA== X-Google-Smtp-Source: AGHT+IG0T7/zfUFfh66xGie0sIknIw/Uv94DOItk5wHQzIDCCDZTVaUEXVClgEtCvnRjkt1BfaKg X-Received: by 2002:a17:906:d7a4:b0:a6f:1b59:e877 with SMTP id a640c23a62f3a-a6f1b59e8bamr498776566b.75.1718108972713; Tue, 11 Jun 2024 05:29:32 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1718108972; cv=none; d=google.com; s=arc-20160816; b=kjXrvAnkUJiiwRq19qYBcSYU5tMR2GIHRsQOLj9gJiCg2iLNN7hd/myhIKJj4Jx+9z bghRcshLLnVfhzR1nKOMYUELGMZfdZaFmo6Q60aYYCf/foJ/c3+UQm2oRmIn5H0XKNEu fJznQMxpqWySgNo7iFis3vOZDcUYbMYB86CA8IRmM7l4skOuI+FacDXsJTTx5YDxKKlj xzCOMtttWjxQRsTBjoxeKiD6fhD7QEIv0SxUx9H1k+VFpTmNK+EVhug1X4M85v29ZV25 oQ89DFRauA5QrY6qoPUaFBXSn7n689V5T/9g1794YT1V2yaoiB9g+r9M1cfvjwogjUF6 zv4Q== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=sender:errors-to:content-transfer-encoding:reply-to:list-subscribe :list-help:list-post:list-archive:list-unsubscribe:list-id :precedence:subject:mime-version:references:in-reply-to:message-id :date:to:from:dkim-signature:delivered-to; bh=XYhGunaVnoZ493NvB41yaodb9TkI5ttUvLOEYq1ebEM=; fh=YOA8vD9MJZuwZ71F/05pj6KdCjf6jQRmzLS+CATXUQk=; b=NLThkRABMOkL+mc8KbBKeVCopYjgGb6a2SKdgRs2xbUfO9UyeWzvXOv08Gz0sKnNN3 7nblgdCtkCoF4lRNPWGK8UNlJOd2TI3vGYC8JAskF385LptOPRoBu9sLNOERzAW4jM1p 9HHxMXa88yCxjwl26sloV/BqtXK0WD/aF6yds8PUGTzklrT+6PH5Y8zACjX2rPzBGW4z pl49SNudHGybmhLtLtqUfcxlBQ9yy2yOchKd9lVKsiIsdewe+FsZL+H/JBfYFwyF2ikP eaPpN8CKw85Fq++JvXq2wLOgGoAugCETIFcFx3V5tykTDQo8eBH9ISOBx16YOdg+PLb+ Ml6w==; dara=google.com ARC-Authentication-Results: i=1; mx.google.com; dkim=neutral (body hash did not verify) header.i=@gmail.com header.s=20230601 header.b=mgJFuHu3; 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 a640c23a62f3a-a6ef6ca28f0si373020366b.680.2024.06.11.05.29.32; Tue, 11 Jun 2024 05:29:32 -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=20230601 header.b=mgJFuHu3; 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 9475A68D845; Tue, 11 Jun 2024 15:29:12 +0300 (EEST) X-Original-To: ffmpeg-devel@ffmpeg.org Delivered-To: ffmpeg-devel@ffmpeg.org Received: from mail-wr1-f45.google.com (mail-wr1-f45.google.com [209.85.221.45]) by ffbox0-bg.mplayerhq.hu (Postfix) with ESMTPS id 1B92C68D586 for ; Tue, 11 Jun 2024 15:29:05 +0300 (EEST) Received: by mail-wr1-f45.google.com with SMTP id ffacd0b85a97d-35f0aeff7a8so836642f8f.2 for ; Tue, 11 Jun 2024 05:29:05 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20230601; t=1718108944; x=1718713744; darn=ffmpeg.org; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:to:from:from:to:cc:subject:date:message-id :reply-to; bh=pcJfh1P+4g5UipbakKlBdtQVW3ZPakzg//XjaNi0CxU=; b=mgJFuHu3E8FbSJ7qBS8f/CHL8lG3fMB1p01iFXoyCshNmJb+VGw4dOFSxlbS+rlYsR 8jhptGq1hh2pECYC87bkZZr2H0+k3h4djRpCmWXlsWi46ZFmtgFK9nV2MTq4eUIwYThF UXdxvNL4HEV9o0k4+MtE3EVVmCiIqouLL35FtCzuZi34I37y87uGHVxqKK9qCgh0NY5X NIamQFY253IRRe+UTwClmUfdRQPXndZBPOJ5NcaqiDM/zI55YodtQzRaOn8w+v3MJtFT /f4ysymByNrhu/OKewShYf5//drcEHZAy+vrkCklyIP0JbT2mPVzIR0G0oc+RqqIfRpP IMUw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1718108944; x=1718713744; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=pcJfh1P+4g5UipbakKlBdtQVW3ZPakzg//XjaNi0CxU=; b=QhU4x66ZbeYo2CjYojY+cF6fW8k60zeduEULpHVG5vNwWQoxQCM7pXQmg1mpL+ybxJ /NkBWLRvwUrEMh9NsoMa48o/T4qHkdich8r7FpCBYjtITXc1Ulkxff0+Dk477IJm4Fz8 Lfam7ssp6ZP0ZH4cHyq4kwIVb138OKcDcmrSOU5QDbBhJR1IDxQLWG/z+va3id1nqz9t RvAkZ2kxOsGBYFuyQx/DlgCYbAxMYZ3eCkjEz8p/pMTHKJ47rw8xebRaEzNAPvd0Gvi3 lKiiwisXvAS5PrFyu0njiU6L2o8gkmn//GIxDPynscZC4CIYeOt5nHPSTMyOS1PLY2O0 YbFQ== X-Gm-Message-State: AOJu0YxOfURIZb/X1AWxDPKfQKVB2LNfLdY5/Px22+ukOc6aAmjFyw42 BvAOJs+8DHRWOxLAy7vUkgQKf45nere2ZflfaENlMhd7yurmR3E9C2Vwpw== X-Received: by 2002:a5d:6d87:0:b0:35f:154b:1a03 with SMTP id ffacd0b85a97d-35f154b1a7amr9833861f8f.61.1718108943811; Tue, 11 Jun 2024 05:29:03 -0700 (PDT) Received: from localhost.localdomain (237.35-178-91.adsl-dyn.isp.belgacom.be. [91.178.35.237]) by smtp.gmail.com with ESMTPSA id ffacd0b85a97d-35f0b876d80sm10592856f8f.109.2024.06.11.05.29.02 for (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Tue, 11 Jun 2024 05:29:02 -0700 (PDT) From: Ramiro Polla To: ffmpeg-devel@ffmpeg.org Date: Tue, 11 Jun 2024 14:28:57 +0200 Message-Id: <20240611122858.196658-3-ramiro.polla@gmail.com> X-Mailer: git-send-email 2.30.2 In-Reply-To: <20240611122858.196658-1-ramiro.polla@gmail.com> References: <20240611122858.196658-1-ramiro.polla@gmail.com> MIME-Version: 1.0 Subject: [FFmpeg-devel] [PATCH v2 3/4] swscale/x86: add avx2 {lum, chr}ConvertRange 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: S54Hn8cDFKjk chrRangeFromJpeg_8_c: 24.1 chrRangeFromJpeg_8_sse4: 16.1 chrRangeFromJpeg_8_avx2: 19.9 chrRangeFromJpeg_24_c: 72.6 chrRangeFromJpeg_24_sse4: 34.6 chrRangeFromJpeg_24_avx2: 30.9 chrRangeFromJpeg_128_c: 341.1 chrRangeFromJpeg_128_sse4: 160.9 chrRangeFromJpeg_128_avx2: 94.1 chrRangeFromJpeg_144_c: 381.9 chrRangeFromJpeg_144_sse4: 183.6 chrRangeFromJpeg_144_avx2: 108.9 chrRangeFromJpeg_256_c: 646.1 chrRangeFromJpeg_256_sse4: 320.4 chrRangeFromJpeg_256_avx2: 190.6 chrRangeFromJpeg_512_c: 1255.9 chrRangeFromJpeg_512_sse4: 654.1 chrRangeFromJpeg_512_avx2: 392.4 chrRangeToJpeg_8_c: 36.9 chrRangeToJpeg_8_sse4: 13.9 chrRangeToJpeg_8_avx2: 20.6 chrRangeToJpeg_24_c: 113.4 chrRangeToJpeg_24_sse4: 29.6 chrRangeToJpeg_24_avx2: 28.9 chrRangeToJpeg_128_c: 632.1 chrRangeToJpeg_128_sse4: 162.4 chrRangeToJpeg_128_avx2: 94.6 chrRangeToJpeg_144_c: 709.9 chrRangeToJpeg_144_sse4: 183.9 chrRangeToJpeg_144_avx2: 108.1 chrRangeToJpeg_256_c: 2672.9 chrRangeToJpeg_256_sse4: 334.4 chrRangeToJpeg_256_avx2: 190.6 chrRangeToJpeg_512_c: 2500.9 chrRangeToJpeg_512_sse4: 654.1 chrRangeToJpeg_512_avx2: 379.6 lumRangeFromJpeg_8_c: 10.9 lumRangeFromJpeg_8_sse4: 12.4 lumRangeFromJpeg_8_avx2: 17.6 lumRangeFromJpeg_24_c: 38.4 lumRangeFromJpeg_24_sse4: 16.9 lumRangeFromJpeg_24_avx2: 20.6 lumRangeFromJpeg_128_c: 233.6 lumRangeFromJpeg_128_sse4: 79.9 lumRangeFromJpeg_128_avx2: 51.6 lumRangeFromJpeg_144_c: 263.9 lumRangeFromJpeg_144_sse4: 90.1 lumRangeFromJpeg_144_avx2: 57.6 lumRangeFromJpeg_256_c: 436.9 lumRangeFromJpeg_256_sse4: 162.1 lumRangeFromJpeg_256_avx2: 100.6 lumRangeFromJpeg_512_c: 878.4 lumRangeFromJpeg_512_sse4: 335.1 lumRangeFromJpeg_512_avx2: 199.4 lumRangeToJpeg_8_c: 19.1 lumRangeToJpeg_8_sse4: 11.6 lumRangeToJpeg_8_avx2: 17.6 lumRangeToJpeg_24_c: 56.9 lumRangeToJpeg_24_sse4: 17.6 lumRangeToJpeg_24_avx2: 21.4 lumRangeToJpeg_128_c: 335.9 lumRangeToJpeg_128_sse4: 79.1 lumRangeToJpeg_128_avx2: 48.9 lumRangeToJpeg_144_c: 372.9 lumRangeToJpeg_144_sse4: 91.6 lumRangeToJpeg_144_avx2: 55.4 lumRangeToJpeg_256_c: 651.9 lumRangeToJpeg_256_sse4: 163.6 lumRangeToJpeg_256_avx2: 99.1 lumRangeToJpeg_512_c: 1289.9 lumRangeToJpeg_512_sse4: 333.6 lumRangeToJpeg_512_avx2: 211.1 --- libswscale/x86/range_convert.asm | 46 ++++++++++++++++++++++++++------ libswscale/x86/swscale.c | 5 +++- 2 files changed, 42 insertions(+), 9 deletions(-) diff --git a/libswscale/x86/range_convert.asm b/libswscale/x86/range_convert.asm index 13983a386b..54c2f64769 100644 --- a/libswscale/x86/range_convert.asm +++ b/libswscale/x86/range_convert.asm @@ -22,20 +22,20 @@ SECTION_RODATA -chr_to_mult: times 4 dd 4663 -chr_to_offset: times 4 dd -9289992 +chr_to_mult: times 8 dd 4663 +chr_to_offset: times 8 dd -9289992 %define chr_to_shift 12 -chr_from_mult: times 4 dd 1799 -chr_from_offset: times 4 dd 4081085 +chr_from_mult: times 8 dd 1799 +chr_from_offset: times 8 dd 4081085 %define chr_from_shift 11 -lum_to_mult: times 4 dd 19077 -lum_to_offset: times 4 dd -39057361 +lum_to_mult: times 8 dd 19077 +lum_to_offset: times 8 dd -39057361 %define lum_to_shift 14 -lum_from_mult: times 4 dd 14071 -lum_from_offset: times 4 dd 33561947 +lum_from_mult: times 8 dd 14071 +lum_from_offset: times 8 dd 33561947 %define lum_from_shift 14 SECTION .text @@ -66,10 +66,19 @@ cglobal %1, 2, 3, 3, dst, width, x paddd m1, m5 psrad m0, %4 psrad m1, %4 +%if mmsize == 16 packssdw m0, m0 packssdw m1, m1 movq [dstq+xq*2], m0 movq [dstq+xq*2+mmsize/2], m1 +%else + vextracti128 xm7, ym0, 1 + packssdw xm0, xm7 + vextracti128 xm7, ym1, 1 + packssdw xm1, xm7 + movdqu [dstq+xq*2], xm0 + movdqu [dstq+xq*2+mmsize/2], xm1 +%endif add xq, mmsize / 2 cmp xd, widthd jl .loop @@ -107,6 +116,7 @@ cglobal %1, 3, 4, 4, dstU, dstV, width, x psrad m1, %4 psrad m2, %4 psrad m3, %4 +%if mmsize == 16 packssdw m0, m0 packssdw m1, m1 packssdw m2, m2 @@ -115,6 +125,20 @@ cglobal %1, 3, 4, 4, dstU, dstV, width, x movq [dstUq+xq*2+mmsize/2], m1 movq [dstVq+xq*2], m2 movq [dstVq+xq*2+mmsize/2], m3 +%else + vextracti128 xm7, ym0, 1 + packssdw xm0, xm7 + vextracti128 xm7, ym1, 1 + packssdw xm1, xm7 + vextracti128 xm7, ym2, 1 + packssdw xm2, xm7 + vextracti128 xm7, ym3, 1 + packssdw xm3, xm7 + movdqu [dstUq+xq*2], xm0 + movdqu [dstUq+xq*2+mmsize/2], xm1 + movdqu [dstVq+xq*2], xm2 + movdqu [dstVq+xq*2+mmsize/2], xm3 +%endif add xq, mmsize / 2 cmp xd, widthd jl .loop @@ -127,4 +151,10 @@ LUMCONVERTRANGE lumRangeToJpeg, lum_to_mult, lum_to_offset, lum_to_shift CHRCONVERTRANGE chrRangeToJpeg, chr_to_mult, chr_to_offset, chr_to_shift LUMCONVERTRANGE lumRangeFromJpeg, lum_from_mult, lum_from_offset, lum_from_shift CHRCONVERTRANGE chrRangeFromJpeg, chr_from_mult, chr_from_offset, chr_from_shift + +INIT_YMM avx2 +LUMCONVERTRANGE lumRangeToJpeg, lum_to_mult, lum_to_offset, lum_to_shift +CHRCONVERTRANGE chrRangeToJpeg, chr_to_mult, chr_to_offset, chr_to_shift +LUMCONVERTRANGE lumRangeFromJpeg, lum_from_mult, lum_from_offset, lum_from_shift +CHRCONVERTRANGE chrRangeFromJpeg, chr_from_mult, chr_from_offset, chr_from_shift %endif diff --git a/libswscale/x86/swscale.c b/libswscale/x86/swscale.c index 8f477b7b72..704e5f9c85 100644 --- a/libswscale/x86/swscale.c +++ b/libswscale/x86/swscale.c @@ -473,12 +473,15 @@ void ff_lumRangeToJpeg_ ##opt(int16_t *dst, int width); \ void ff_chrRangeToJpeg_ ##opt(int16_t *dstU, int16_t *dstV, int width); \ RANGE_CONVERT_FUNCS_DECL(sse4); +RANGE_CONVERT_FUNCS_DECL(avx2); av_cold void ff_sws_init_range_convert_x86(SwsContext *c) { if (c->srcRange != c->dstRange && !isAnyRGB(c->dstFormat)) { int cpu_flags = av_get_cpu_flags(); - if (EXTERNAL_SSE4(cpu_flags)) { + if (EXTERNAL_AVX2_FAST(cpu_flags)) { + RANGE_CONVERT_FUNCS(avx2); + } else if (EXTERNAL_SSE4(cpu_flags)) { RANGE_CONVERT_FUNCS(sse4); } }