From patchwork Sun Nov 19 11:59:04 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: =?utf-8?q?R=C3=A9mi_Denis-Courmont?= X-Patchwork-Id: 44719 Delivered-To: ffmpegpatchwork2@gmail.com Received: by 2002:a05:6a20:6a89:b0:181:818d:5e7f with SMTP id bi9csp995720pzb; Sun, 19 Nov 2023 03:59:14 -0800 (PST) X-Google-Smtp-Source: AGHT+IEeTp5vBM7aEwmHnx8k4aN5I3gGKdTYofjzAcJwQuCZb99Zn1QgHTSy7Fli6nfGi/ZEbTEO X-Received: by 2002:a17:906:7209:b0:9e0:5dab:a0f1 with SMTP id m9-20020a170906720900b009e05daba0f1mr3813933ejk.36.1700395154142; Sun, 19 Nov 2023 03:59:14 -0800 (PST) ARC-Seal: i=1; a=rsa-sha256; t=1700395154; cv=none; d=google.com; s=arc-20160816; b=aYw2LVy4eXBw2s6+/qSrAbowhsZARwtQeJ95Hkx4/llC/gvpBWtN/svh+elZmL7ik1 BW4f3GNK1A8KNKuZKbDXxYBZFKlRj7HaS+luafQvuOTmU8FA9+i42nJPsFWn2RVjqMhi oukBhmZtmcGiIA9ZPQBZT7Otvj3YjFVy2apgWcmcR1h0BVmEiLBkGu+nWHAXMFA7sI45 beozRfc1qzPtz45ekes77Ie8Kxcs3zGBctm0/ThEYG+rG/nxHgw8Ys1hmkR9dfN0I4v1 jdsHtoic3ivjQ0PpA4bpwG4TUbWxy7qpQAsz8s96zGpjYbGmB73F3kr24xPonApZkBc4 lLxg== 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:message-id:date:to:from :delivered-to; bh=hSwxZBXE4NTpdOJEaO2/92rgu/i+F8V/F9LM9Atxaqc=; fh=YOA8vD9MJZuwZ71F/05pj6KdCjf6jQRmzLS+CATXUQk=; b=RJNXDOeGDcol5FMqIIBHbfsB3rX0TJntzplBRgfaa0eLjjw5PEikVDVi5Qi79md9JR mZyHCyduE8cxPohaPjV3rUP6FbDiykPG1NaAR4JETnqliU5djj1C+mrmdnPJjDikR4Ja YFxWxjK51doe4kq6nRdO4lV/ad12au6pmpsh7g37Xn1Aec+k+xH3eeHJN8aZd5YYXwzF ogJacYcwFWtwP0KrGdl3M/0jx6FdxTl5p/PpUEkXPT41jkqSwQB57ZanWYykeKgUMRQz EeqpEQXljvYR8+iUf8H/72qKq3AhNAYo8BFRCvo8gyCfnHUyL7KaXef4wOvbtVcNjkNz peSg== ARC-Authentication-Results: i=1; mx.google.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 Return-Path: Received: from ffbox0-bg.mplayerhq.hu (ffbox0-bg.ffmpeg.org. [79.124.17.100]) by mx.google.com with ESMTP id mc17-20020a170906eb5100b009b2849aebcfsi3520169ejb.752.2023.11.19.03.59.13; Sun, 19 Nov 2023 03:59:14 -0800 (PST) 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; 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 Received: from [127.0.1.1] (localhost [127.0.0.1]) by ffbox0-bg.mplayerhq.hu (Postfix) with ESMTP id C819468CD67; Sun, 19 Nov 2023 13:59:11 +0200 (EET) X-Original-To: ffmpeg-devel@ffmpeg.org Delivered-To: ffmpeg-devel@ffmpeg.org Received: from ursule.remlab.net (vps-a2bccee9.vps.ovh.net [51.75.19.47]) by ffbox0-bg.mplayerhq.hu (Postfix) with ESMTP id 1E3AB68C6BA for ; Sun, 19 Nov 2023 13:59:05 +0200 (EET) Received: from basile.remlab.net (localhost [IPv6:::1]) by ursule.remlab.net (Postfix) with ESMTP id B4313C018B for ; Sun, 19 Nov 2023 13:59:04 +0200 (EET) From: =?utf-8?q?R=C3=A9mi_Denis-Courmont?= To: ffmpeg-devel@ffmpeg.org Date: Sun, 19 Nov 2023 13:59:04 +0200 Message-ID: <20231119115904.11806-1-remi@remlab.net> X-Mailer: git-send-email 2.42.0 MIME-Version: 1.0 Subject: [FFmpeg-devel] [PATCH] lavu/fixed_dsp: R-V V fmul_window_scaled 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: RL5CG8reBEMM vector_fmul_window_scaled_fixed_c: 4393.7 vector_fmul_window_scaled_fixed_rvv_i64: 1642.7 --- libavutil/riscv/fixed_dsp_init.c | 7 ++++- libavutil/riscv/fixed_dsp_rvv.S | 48 ++++++++++++++++++++++++++++++++ 2 files changed, 54 insertions(+), 1 deletion(-) diff --git a/libavutil/riscv/fixed_dsp_init.c b/libavutil/riscv/fixed_dsp_init.c index 6469b45374..cd318af486 100644 --- a/libavutil/riscv/fixed_dsp_init.c +++ b/libavutil/riscv/fixed_dsp_init.c @@ -25,6 +25,9 @@ #include "libavutil/cpu.h" #include "libavutil/fixed_dsp.h" +void ff_vector_fmul_window_scaled_rvv(int16_t *dst, const int32_t *src0, + const int32_t *src1, const int32_t *win, + int len, uint8_t bits); void ff_vector_fmul_window_fixed_rvv(int32_t *dst, const int32_t *src0, const int32_t *src1, const int32_t *win, int len); @@ -43,8 +46,10 @@ av_cold void ff_fixed_dsp_init_riscv(AVFixedDSPContext *fdsp) int flags = av_get_cpu_flags(); if ((flags & AV_CPU_FLAG_RVV_I32) && (flags & AV_CPU_FLAG_RVB_ADDR)) { - if (flags & AV_CPU_FLAG_RVV_I64) + if (flags & AV_CPU_FLAG_RVV_I64) { + fdsp->vector_fmul_window_scaled = ff_vector_fmul_window_scaled_rvv; fdsp->vector_fmul_window = ff_vector_fmul_window_fixed_rvv; + } fdsp->vector_fmul = ff_vector_fmul_fixed_rvv; fdsp->vector_fmul_reverse = ff_vector_fmul_reverse_fixed_rvv; diff --git a/libavutil/riscv/fixed_dsp_rvv.S b/libavutil/riscv/fixed_dsp_rvv.S index 68de6d7e1b..6bac5813b8 100644 --- a/libavutil/riscv/fixed_dsp_rvv.S +++ b/libavutil/riscv/fixed_dsp_rvv.S @@ -20,6 +20,54 @@ #include "asm.S" +func ff_vector_fmul_window_scaled_rvv, zve64x + csrwi vxrm, 0 + vsetvli t0, zero, e16, m1, ta, ma + sh2add a2, a4, a2 + vid.v v0 + sh3add t3, a4, a3 + vadd.vi v0, v0, 1 + sh2add t0, a4, a0 +1: + vsetvli t2, a4, e16, m1, ta, ma + slli t4, t2, 2 + slli t1, t2, 1 + vrsub.vx v2, v0, t2 + sub t3, t3, t4 + vsetvli zero, zero, e32, m2, ta, ma + sub a2, a2, t4 + vle32.v v8, (t3) + sub t0, t0, t1 + vle32.v v4, (a2) + sub a4, a4, t2 + vrgatherei16.vv v28, v8, v2 + vle32.v v16, (a1) + add a1, a1, t4 + vrgatherei16.vv v20, v4, v2 + vle32.v v24, (a3) + add a3, a3, t4 + vwmul.vv v12, v16, v28 + vwmul.vv v8, v16, v24 + // vwnmsac.vv does _not_ exist so multiply & subtract separately + vwmul.vv v4, v20, v24 + vwmacc.vv v8, v20, v28 + vsetvli zero, zero, e64, m4, ta, ma + vsub.vv v12, v12, v4 + vsetvli zero, zero, e32, m2, ta, ma + vnclip.wi v16, v8, 31 + vnclip.wi v20, v12, 31 + vsetvli zero, zero, e16, m1, ta, ma + vnclip.wx v8, v16, a5 + vnclip.wx v12, v20, a5 + vrgatherei16.vv v16, v8, v2 + vse16.v v12, (a0) + add a0, a0, t1 + vse16.v v16, (t0) + bnez a4, 1b + + ret +endfunc + func ff_vector_fmul_window_fixed_rvv, zve64x csrwi vxrm, 0 vsetvli t0, zero, e16, m1, ta, ma