From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from ffbox0-bg.mplayerhq.hu (ffbox0-bg.ffmpeg.org [79.124.17.100]) by master.gitmailbox.com (Postfix) with ESMTP id 7773E4B00B for ; Sun, 26 May 2024 01:51:47 +0000 (UTC) Received: from [127.0.1.1] (localhost [127.0.0.1]) by ffbox0-bg.mplayerhq.hu (Postfix) with ESMTP id E237268D4E1; Sun, 26 May 2024 04:51:44 +0300 (EEST) Received: from mail-pf1-f180.google.com (mail-pf1-f180.google.com [209.85.210.180]) by ffbox0-bg.mplayerhq.hu (Postfix) with ESMTPS id EED1C68CFFE for ; Sun, 26 May 2024 04:51:38 +0300 (EEST) Received: by mail-pf1-f180.google.com with SMTP id d2e1a72fcca58-6f6bddf57f6so5441423b3a.0 for ; Sat, 25 May 2024 18:51:38 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20230601; t=1716688296; x=1717293096; darn=ffmpeg.org; h=content-transfer-encoding:in-reply-to:from:content-language :references:to:subject:user-agent:mime-version:date:message-id:from :to:cc:subject:date:message-id:reply-to; bh=XBa4peLz+cxtODBl9WoIGU9fh1FePG3E8gJf48qiiHk=; b=Uf0yxhOnM/qfzuGGi1J+Sm4f0tn1SOiZknZQwgZuQnaBepPEtXHtBbqJHt3OP5UUpI +/1GbeQLWjP6ekUHviMEFrEfnxkYMQncIS0WMVWNHkEQi0mMAncujzuqLn+VLTOAXeXJ JFcX9wrxKXLwgtZFxVNST9ls28Uo+4BA4/SQx4v+Z5i56WpXN/u4W/cjVSUohtGtwWWj WycszWftyvZiZ4AlzXaFzShdnZ+4naHDiDVYY8GpEmt9K+396BZ5UTQ/wIlhTFfmguX6 dK3RhzQ5GHkbqffwDEs0d4UmqjrfiwhvGCI4kf4gYuNZf2w+PkfDjOtGn4mjzInj6fmy UR4g== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1716688296; x=1717293096; h=content-transfer-encoding:in-reply-to:from:content-language :references:to:subject:user-agent:mime-version:date:message-id :x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=XBa4peLz+cxtODBl9WoIGU9fh1FePG3E8gJf48qiiHk=; b=lXTsHoxmUxDOA4SZ4NFqcXG8NQB+yDOVqOM++j3SSCxLCdnSeWxjlIqY7GNRsgiXBu hMwf8ScBruulchMJmp3omIxggvl7feWKvePEqysmQlKyaG6+liRSe7YjBlP1d/1KT71O ND2YZGfaXObsVfjFBFfv2QtYCjgEA+FanYgdqd+y1ehsTMVELXum0NPptf6I18qzuMCY hoOxU/yRidgZ24w+W06ZYQnwBDky4ZSrFxuDMqRoBi8bVENTfagNGYuajFOcgm/R+MhD Je62bCDMy3CFUu3jQ0jrP0pv2T/KW8Pqn3dZS3qc+ErLgo2eElTGcbp63CW7MqLL0LWS 5BBw== X-Gm-Message-State: AOJu0Yz1m8dnozcghET7YiFWev0OkCVxvn1ZpKgzPsbLixm6dQcyF3Pv wltxLcPMTMyaE7sJwe3Bn71Hj08lIeBRPRAKiYuSEYbQYjhcZi/jtKWipQ== X-Google-Smtp-Source: AGHT+IFfL/P3GdkYAoLAvwf71PZJ1nNjpmWU9cwRDQbrcVJfaQVecubzJGH2Mg7SfC39IvON4FhzTw== X-Received: by 2002:a05:6300:808d:b0:1af:d810:9d0e with SMTP id adf61e73a8af0-1b212dfe5fdmr5998756637.49.1716688295757; Sat, 25 May 2024 18:51:35 -0700 (PDT) Received: from [192.168.0.10] ([190.194.167.233]) by smtp.gmail.com with ESMTPSA id 98e67ed59e1d1-2bf5f9b28c2sm3415302a91.51.2024.05.25.18.51.34 for (version=TLS1_3 cipher=TLS_AES_128_GCM_SHA256 bits=128/128); Sat, 25 May 2024 18:51:35 -0700 (PDT) Message-ID: <54b8aabb-ab4c-46f5-88b2-fb0a0bf16f7f@gmail.com> Date: Sat, 25 May 2024 22:51:41 -0300 MIME-Version: 1.0 User-Agent: Mozilla Thunderbird To: ffmpeg-devel@ffmpeg.org References: <20240525205731.2578146-1-dev@lynne.ee> <20240526014207.2697057-1-dev@lynne.ee> Content-Language: en-US From: James Almer In-Reply-To: <20240526014207.2697057-1-dev@lynne.ee> Subject: Re: [FFmpeg-devel] [PATCH v2] lpc: rewrite lpc_compute_autocorr in external asm 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 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset="us-ascii"; Format="flowed" Errors-To: ffmpeg-devel-bounces@ffmpeg.org Sender: "ffmpeg-devel" Archived-At: List-Archive: List-Post: On 5/25/2024 10:42 PM, Lynne via ffmpeg-devel wrote: > The inline asm function had issues running under checkasm. > So I came to finish what I started, and wrote the last part > of LPC computation in assembly. > --- > libavcodec/x86/lpc.asm | 91 +++++++++++++++++++++++++++++++++++++++ > libavcodec/x86/lpc_init.c | 87 ++++--------------------------------- > 2 files changed, 100 insertions(+), 78 deletions(-) > > diff --git a/libavcodec/x86/lpc.asm b/libavcodec/x86/lpc.asm > index a585c17ef5..9c359ae480 100644 > --- a/libavcodec/x86/lpc.asm > +++ b/libavcodec/x86/lpc.asm > @@ -261,3 +261,94 @@ APPLY_WELCH_FN > INIT_YMM avx2 > APPLY_WELCH_FN > %endif > + > +%macro COMPUTE_AUTOCORR_FN 0 > +cglobal lpc_compute_autocorr, 4, 7, 3, data, len, lag, autoc, lag_p, data_l, len_p > + shl lagd, 3 > + shl lenq, 3 > + xor lag_pq, lag_pq > + > +.lag_l: > + movaps m2, [one_tab] Super nit: movapd > + > + mov len_pq, lag_pq > + > + lea data_lq, [lag_pq + mmsize - 8] > + neg data_lq ; -j - mmsize > + add data_lq, dataq ; data[-j - mmsize] > +.len_l: > + > +%if mmsize == 32 > + vbroadcastsd m0, [dataq + len_pq] > + vpermpd m1, [data_lq + len_pq], q0123 > +%else > + movupd m1, [data_lq + len_pq] ; data[i - j] > + movsd xm0, [dataq + len_pq] ; data[i] > + shufpd m1, m1, m1, 01b > +%endif > + > + shufpd m0, m0, m0, 1100b This is not needed for mmsize == 32. The broadcast set every qword to the value movsd loaded. > + > + ; fmadd actually hurts performance in this case due to > + ; the earlier loads + shuffles > + mulpd m0, m1 > + addpd m2, m0 ; sum += data[i]*data[i-j] > + > + add len_pq, 8 > + cmp len_pq, lenq > + jl .len_l > + > + movupd [autocq + lag_pq], m2 ; autoc[j] = sum > + add lag_pq, mmsize > + cmp lag_pq, lagq > + jl .lag_l > + > + ; The tail computation is guaranteed never to happen > + ; as long as we're doing multiples of 4, rather than 2. > +%if mmsize != 32 > + jg .end > + ; If lag_p == lag fallthrough > + > +.tail: > + movaps m2, [one_tab] > + > + mov len_pq, lag_pq > + sub len_pq, mmsize > + > + lea data_lq, [lag_pq] > + neg data_lq ; -j > + add data_lq, dataq ; data[-j] > + > +.tail_l: > + movupd m0, [dataq + len_pq] > + movupd m1, [data_lq + len_pq] > + > + mulpd m0, m1 > + addpd m2, m0 ; sum += data[i]*data[i-j] > + > + add len_pq, mmsize > + cmp len_pq, lenq > + jl .tail_l > + > + shufpd m1, m2, m2, 01b > + addpd m2, m1 > + > + ; Leave this here just in case its ever needed > +%if mmsize == 32 > + vperm2f128 m1, m2, m2, 0x01 > + addpd xm2, xm1 > + movupd [autocq + lag_pq], xm2 > +%else > + movhpd [autocq + lag_pq], xm2 > +%endif > + > +.end: > +%endif > + > + RET > +%endmacro > + > +INIT_XMM sse2 > +COMPUTE_AUTOCORR_FN > +INIT_YMM avx vpermpd is avx2, so it needs to be that. > +COMPUTE_AUTOCORR_FN > diff --git a/libavcodec/x86/lpc_init.c b/libavcodec/x86/lpc_init.c > index f2fca53799..bb174be53e 100644 > --- a/libavcodec/x86/lpc_init.c > +++ b/libavcodec/x86/lpc_init.c > @@ -28,89 +28,20 @@ void ff_lpc_apply_welch_window_sse2(const int32_t *data, ptrdiff_t len, > double *w_data); > void ff_lpc_apply_welch_window_avx2(const int32_t *data, ptrdiff_t len, > double *w_data); > - > -DECLARE_ASM_CONST(16, double, pd_1)[2] = { 1.0, 1.0 }; > - > -#if HAVE_SSE2_INLINE > - > -static void lpc_compute_autocorr_sse2(const double *data, ptrdiff_t len, int lag, > - double *autoc) > -{ > - int j; > - > - if((x86_reg)data & 15) > - data++; > - > - for(j=0; j - x86_reg i = -len*sizeof(double); > - if(j == lag-2) { > - __asm__ volatile( > - "movsd "MANGLE(pd_1)", %%xmm0 \n\t" > - "movsd "MANGLE(pd_1)", %%xmm1 \n\t" > - "movsd "MANGLE(pd_1)", %%xmm2 \n\t" > - "1: \n\t" > - "movapd (%2,%0), %%xmm3 \n\t" > - "movupd -8(%3,%0), %%xmm4 \n\t" > - "movapd (%3,%0), %%xmm5 \n\t" > - "mulpd %%xmm3, %%xmm4 \n\t" > - "mulpd %%xmm3, %%xmm5 \n\t" > - "mulpd -16(%3,%0), %%xmm3 \n\t" > - "addpd %%xmm4, %%xmm1 \n\t" > - "addpd %%xmm5, %%xmm0 \n\t" > - "addpd %%xmm3, %%xmm2 \n\t" > - "add $16, %0 \n\t" > - "jl 1b \n\t" > - "movhlps %%xmm0, %%xmm3 \n\t" > - "movhlps %%xmm1, %%xmm4 \n\t" > - "movhlps %%xmm2, %%xmm5 \n\t" > - "addsd %%xmm3, %%xmm0 \n\t" > - "addsd %%xmm4, %%xmm1 \n\t" > - "addsd %%xmm5, %%xmm2 \n\t" > - "movsd %%xmm0, (%1) \n\t" > - "movsd %%xmm1, 8(%1) \n\t" > - "movsd %%xmm2, 16(%1) \n\t" > - :"+&r"(i) > - :"r"(autoc+j), "r"(data+len), "r"(data+len-j) > - NAMED_CONSTRAINTS_ARRAY_ADD(pd_1) > - :"memory" > - ); > - } else { > - __asm__ volatile( > - "movsd "MANGLE(pd_1)", %%xmm0 \n\t" > - "movsd "MANGLE(pd_1)", %%xmm1 \n\t" > - "1: \n\t" > - "movapd (%3,%0), %%xmm3 \n\t" > - "movupd -8(%4,%0), %%xmm4 \n\t" > - "mulpd %%xmm3, %%xmm4 \n\t" > - "mulpd (%4,%0), %%xmm3 \n\t" > - "addpd %%xmm4, %%xmm1 \n\t" > - "addpd %%xmm3, %%xmm0 \n\t" > - "add $16, %0 \n\t" > - "jl 1b \n\t" > - "movhlps %%xmm0, %%xmm3 \n\t" > - "movhlps %%xmm1, %%xmm4 \n\t" > - "addsd %%xmm3, %%xmm0 \n\t" > - "addsd %%xmm4, %%xmm1 \n\t" > - "movsd %%xmm0, %1 \n\t" > - "movsd %%xmm1, %2 \n\t" > - :"+&r"(i), "=m"(autoc[j]), "=m"(autoc[j+1]) > - :"r"(data+len), "r"(data+len-j) > - NAMED_CONSTRAINTS_ARRAY_ADD(pd_1) > - ); > - } > - } > -} > - > -#endif /* HAVE_SSE2_INLINE */ > +void ff_lpc_compute_autocorr_sse2(const double *data, ptrdiff_t len, int lag, > + double *autoc); > +void ff_lpc_compute_autocorr_avx(const double *data, ptrdiff_t len, int lag, > + double *autoc); > > av_cold void ff_lpc_init_x86(LPCContext *c) > { > int cpu_flags = av_get_cpu_flags(); > > -#if HAVE_SSE2_INLINE > - if (INLINE_SSE2_SLOW(cpu_flags)) > - c->lpc_compute_autocorr = lpc_compute_autocorr_sse2; > -#endif > + if (EXTERNAL_SSE2(cpu_flags)) > + c->lpc_compute_autocorr = ff_lpc_compute_autocorr_sse2; Place this with ff_lpc_apply_welch_window_sse2 below. > + > + if (EXTERNAL_AVX_FAST(cpu_flags)) > + c->lpc_compute_autocorr = ff_lpc_compute_autocorr_avx; > > if (EXTERNAL_SSE2(cpu_flags)) > c->lpc_apply_welch_window = ff_lpc_apply_welch_window_sse2; _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".