From patchwork Sat Jul 27 19:39:34 2024 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: 50761 Delivered-To: ffmpegpatchwork2@gmail.com Received: by 2002:a05:612c:16f:b0:489:2eb3:e4c4 with SMTP id h47csp371749vqi; Sat, 27 Jul 2024 12:39:48 -0700 (PDT) X-Forwarded-Encrypted: i=2; AJvYcCWZ5YFkUvUZtbiIPYS9vQgcktMW+Uz1L5JbQuaiCbNddM0qxrLmuhg3+vGjazOMSOI+Onh/8CnyxSpQvvR1uHytl1xazjYjySAb1g== X-Google-Smtp-Source: AGHT+IGvjvQ5htfPjAXZ/fypKUg7IJamPJnBOGBmdmC9lBdunWsnYKUxNCY9kJbwi5oIWTEHYFwo X-Received: by 2002:a05:6402:1ece:b0:5a3:d140:1a46 with SMTP id 4fb4d7f45d1cf-5b021d2222bmr2879400a12.23.1722109188029; Sat, 27 Jul 2024 12:39:48 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1722109188; cv=none; d=google.com; s=arc-20160816; b=SEcgmetl2ptNcVr7s0smqjck68Pjy5gJN/Gfi8vsDAVJB3+wHGxvIoF/eZQvSqNBoB FgKmnf63v79Z1TpVWuMInMh3SgMjaHB2jv2pmk0nmQq57WfEnawxhv8uH45s89LKXOCA g+LNdibj1JXJpGMHmBvkUAVhMJiANOh0ZnOoJdfhkSGj0I+BFafoa2vTdQRwNSisoDzS YKoz/5AMH29w9ZLVcMpn7SJTK6zPPkfGQjfxGOn6cSEAivAW3TCAIE/W34c7mi6SBAQt +S2GALJGRoswrPvjqdrdwsKDLVusr7wx8WXksyE/fM8CrWErGMLJIFkDgon2vGPqcsqg NQQA== 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=jgQlnJ6R8/32j0o/EgpOtmNQXnbcrOeJJe+DJvm596U=; fh=YOA8vD9MJZuwZ71F/05pj6KdCjf6jQRmzLS+CATXUQk=; b=BL41UC9hOARXblopXMeiF3s8LJXQcbs+zLkcvN46004xtZv4gFJEC8P4i/Sw9vFjbv rwW+QRTqNqIr8iu5sA67fooRuBtfNy1Q6HUZpCV52i/g5+q4L86hidm7ztABbm5FHWls +iEjcpiXLE6LAnirPoNHYFHZKrJRihuNLKR/R5ZWGeGWt2w/YB7zqhP8OA62pIrovsef FNXk8aDd7tg39cGNAjwE7KLpUhS5THvc2p9AD17D3vDy0NYQioi2snlAdILdaetj6Vb6 /zd83ZblU24bR4daptuWEWpHqg4IPLvBR/S2dh0zZDy+kM2hMk7dyTKBqcFlJ61o17L8 lVKg==; dara=google.com 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 4fb4d7f45d1cf-5ac65f3bf41si4196115a12.539.2024.07.27.12.39.47; Sat, 27 Jul 2024 12:39:48 -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; 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 8DB3168D69E; Sat, 27 Jul 2024 22:39:43 +0300 (EEST) 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 A872168D4AD for ; Sat, 27 Jul 2024 22:39:36 +0300 (EEST) Received: from basile.remlab.net (localhost [IPv6:::1]) by ursule.remlab.net (Postfix) with ESMTP id 01464C013E for ; Sat, 27 Jul 2024 22:39:35 +0300 (EEST) From: =?utf-8?q?R=C3=A9mi_Denis-Courmont?= To: ffmpeg-devel@ffmpeg.org Date: Sat, 27 Jul 2024 22:39:34 +0300 Message-ID: <20240727193935.85724-1-remi@remlab.net> X-Mailer: git-send-email 2.45.2 MIME-Version: 1.0 Subject: [FFmpeg-devel] [PATCH 1/2] lavc/vc1dsp: unify R-V V DC bypass functions 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: f8iiVMyfDc2P --- libavcodec/riscv/vc1dsp_rvv.S | 126 ++++++++++------------------------ 1 file changed, 35 insertions(+), 91 deletions(-) diff --git a/libavcodec/riscv/vc1dsp_rvv.S b/libavcodec/riscv/vc1dsp_rvv.S index 5189d5e855..548ef9d3bf 100644 --- a/libavcodec/riscv/vc1dsp_rvv.S +++ b/libavcodec/riscv/vc1dsp_rvv.S @@ -21,101 +21,45 @@ #include "libavutil/riscv/asm.S" -func ff_vc1_inv_trans_8x8_dc_rvv, zve64x, zba +.macro inv_trans_dc rows, cols, w, mat_lmul, row_lmul +func ff_vc1_inv_trans_\cols\()x\rows\()_dc_rvv, zve64x, zba lpad 0 - lh t2, (a2) - vsetivli zero, 8, e8, mf2, ta, ma - vlse64.v v0, (a0), a1 - sh1add t2, t2, t2 - addi t2, t2, 1 - srai t2, t2, 1 - sh1add t2, t2, t2 - addi t2, t2, 16 - srai t2, t2, 5 - li t0, 8*8 - vsetvli zero, t0, e16, m8, ta, ma - vzext.vf2 v8, v0 - vadd.vx v8, v8, t2 - vmax.vx v8, v8, zero - vsetvli zero, zero, e8, m4, ta, ma - vnclipu.wi v0, v8, 0 - vsetivli zero, 8, e8, mf2, ta, ma - vsse64.v v0, (a0), a1 - ret -endfunc - -func ff_vc1_inv_trans_4x8_dc_rvv, zve32x, zba - lpad 0 - lh t2, (a2) - vsetivli zero, 8, e8, mf2, ta, ma - vlse32.v v0, (a0), a1 - slli t1, t2, 4 - add t2, t2, t1 - addi t2, t2, 4 - srai t2, t2, 3 - sh1add t2, t2, t2 - slli t2, t2, 2 - addi t2, t2, 64 - srai t2, t2, 7 - li t0, 4*8 - vsetvli zero, t0, e16, m4, ta, ma - vzext.vf2 v4, v0 - vadd.vx v4, v4, t2 - vmax.vx v4, v4, zero - vsetvli zero, zero, e8, m2, ta, ma - vnclipu.wi v0, v4, 0 - vsetivli zero, 8, e8, mf2, ta, ma - vsse32.v v0, (a0), a1 - ret -endfunc - -func ff_vc1_inv_trans_8x4_dc_rvv, zve64x, zba - lpad 0 - lh t2, (a2) - vsetivli zero, 4, e8, mf4, ta, ma - vlse64.v v0, (a0), a1 - sh1add t2, t2, t2 - addi t2, t2, 1 - srai t2, t2, 1 - slli t1, t2, 4 - add t2, t2, t1 - addi t2, t2, 64 - srai t2, t2, 7 - li t0, 8*4 - vsetvli zero, t0, e16, m4, ta, ma - vzext.vf2 v4, v0 - vadd.vx v4, v4, t2 - vmax.vx v4, v4, zero - vsetvli zero, zero, e8, m2, ta, ma - vnclipu.wi v0, v4, 0 - vsetivli zero, 4, e8, mf4, ta, ma - vsse64.v v0, (a0), a1 + lh t2, (a2) + li a4, 22 - (5 * \cols) / 4 + mul t2, t2, a4 + vsetivli zero, \rows, e8, m\row_lmul, ta, ma + vlse\w\().v v0, (a0), a1 + addi t2, t2, 4 + li a5, 22 - (5 * \rows) / 4 + srai t2, t2, 3 + mul t2, t2, a5 +.if \cols * \rows >= 32 + li t0, \cols * \rows +.endif + addi t2, t2, 64 + srai t2, t2, 7 +.if \rows * \cols == 64 + vsetvli zero, t0, e16, m8, ta, ma +.elseif \rows * \cols == 32 + vsetvli zero, t0, e16, m4, ta, ma +.else + vsetivli zero, \rows * \cols, e16, m2, ta, ma +.endif + vzext.vf2 v8, v0 + vadd.vx v8, v8, t2 + vmax.vx v8, v8, zero + vsetvli zero, zero, e8, m\mat_lmul, ta, ma + vnclipu.wi v0, v8, 0 + vsetivli zero, \rows, e8, m\row_lmul, ta, ma + vsse\w\().v v0, (a0), a1 ret endfunc +.endm -func ff_vc1_inv_trans_4x4_dc_rvv, zve32x - lpad 0 - lh t2, (a2) - vsetivli zero, 4, e8, mf4, ta, ma - vlse32.v v0, (a0), a1 - slli t1, t2, 4 - add t2, t2, t1 - addi t2, t2, 4 - srai t2, t2, 3 - slli t1, t2, 4 - add t2, t2, t1 - addi t2, t2, 64 - srai t2, t2, 7 - vsetivli zero, 4*4, e16, m2, ta, ma - vzext.vf2 v2, v0 - vadd.vx v2, v2, t2 - vmax.vx v2, v2, zero - vsetvli zero, zero, e8, m1, ta, ma - vnclipu.wi v0, v2, 0 - vsetivli zero, 4, e8, mf4, ta, ma - vsse32.v v0, (a0), a1 - ret -endfunc +inv_trans_dc 8, 8, 64, 4, f2 +inv_trans_dc 4, 8, 64, 2, f4 +inv_trans_dc 8, 4, 32, 2, f2 +inv_trans_dc 4, 4, 32, 1, f4 .variant_cc ff_vc1_inv_trans_8_rvv func ff_vc1_inv_trans_8_rvv, zve32x From patchwork Sat Jul 27 19:39:35 2024 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: 50762 Delivered-To: ffmpegpatchwork2@gmail.com Received: by 2002:a05:612c:16f:b0:489:2eb3:e4c4 with SMTP id h47csp371801vqi; Sat, 27 Jul 2024 12:40:01 -0700 (PDT) X-Forwarded-Encrypted: i=2; AJvYcCWvSfA2j8zYxPtjPZ54khLOAgWbZAmsAJb4oCN7e2b+467SNtJbTShf8vv+Dq6sQ3Eqwq+diAPoGyWgz0VqJp3iisZypNnJkc+4HA== X-Google-Smtp-Source: AGHT+IEsBzpDdaslg38HQWaatnJkhazmm4amHJsvCKjfOJMeBRYeRVvTsz6q0j2MEzaCHxSiMZ7y X-Received: by 2002:a2e:b178:0:b0:2ef:2593:334d with SMTP id 38308e7fff4ca-2f12ee579aemr19162181fa.47.1722109200798; Sat, 27 Jul 2024 12:40:00 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1722109200; cv=none; d=google.com; s=arc-20160816; b=cGcZOJNb+5gIIgWmOfH/CFXWCQOCJvyFKxZCIAEuXlXwL43cbPhW+iiKZXkBY8JFQx ddCpsjebGsrIojMsALT5Ns1BWf798ClnQZGyYQKbRbP0uxS+BDYgkaiAPe8zNT25uAWs NqNAywKOqza1rQG6bhv4t+dbcd9ozKMwAxA3+JtOrIkGkp4vqEIl36w/8Nv7FaIrZHGE Csj0D6y4fE2LZS/oOGtpH/BXN2HS0+XI+o56iekpziwGCHwzd8vW97Cr/25tw1UgCJF5 dKf8XBPaEIUOQoqh4APGUaMraTi+mY3k6rKETcPN7T4VA/j1nC3Vl3ANOI9eCDS6ukKN yrig== 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:delivered-to; bh=pjgWn7eCdJYhihxDLAzwdF0r4p+rVTlhgBzc2ZZ7r/E=; fh=YOA8vD9MJZuwZ71F/05pj6KdCjf6jQRmzLS+CATXUQk=; b=YTufgIr1TscEYk0qltD/3M8IOuVv5qyLpNFL1IbI829izfhPxOZsKdJp+eWt6JImuP pFmy6OahnAl12dfaELCXuPxV6dX6JHtbkInYxfS6VEds1wa4w83064fGUlG3de7o+Mx7 JNwhoSP5TCJPkJFHMX5u0mY0pIvSnFXKD88Ju/dRuaqVskGNIIT3uWm5uMeoxhUp5Nmk Fx7X1QFN8DLZv3gHBwjzOTlR1xIAH3pIC8nRMwoxOjuBmJRBI3b1AzeEtXIz0ye3nnjx LlsBVbhv1u4XPgm1gyFcG6m/0WuyqnOwrF/ZloIlrQHk28xlGkxzK4rJAAkrkB4SLK05 vduQ==; dara=google.com 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 38308e7fff4ca-2f03d13eadesi16244511fa.205.2024.07.27.12.40.00; Sat, 27 Jul 2024 12:40:00 -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; 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 9F06768CC49; Sat, 27 Jul 2024 22:39:44 +0300 (EEST) 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 B70FC68D506 for ; Sat, 27 Jul 2024 22:39:36 +0300 (EEST) Received: from basile.remlab.net (localhost [IPv6:::1]) by ursule.remlab.net (Postfix) with ESMTP id 35405C01A0 for ; Sat, 27 Jul 2024 22:39:36 +0300 (EEST) From: =?utf-8?q?R=C3=A9mi_Denis-Courmont?= To: ffmpeg-devel@ffmpeg.org Date: Sat, 27 Jul 2024 22:39:35 +0300 Message-ID: <20240727193935.85724-2-remi@remlab.net> X-Mailer: git-send-email 2.45.2 In-Reply-To: <20240727193935.85724-1-remi@remlab.net> References: <20240727193935.85724-1-remi@remlab.net> MIME-Version: 1.0 Subject: [FFmpeg-devel] [PATCH 2/2] lavc/vc1dsp: use saturating arithmetic for RVV inv_trans_dc 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: MdabgIPSm7gt T-Head C908 (cycles): vc1dsp.vc1_inv_trans_4x4_dc_c: 113.7 vc1dsp.vc1_inv_trans_4x4_dc_rvv_i32: 46.5 (before) vc1dsp.vc1_inv_trans_4x4_dc_rvv_i32: 45.5 (after) vc1dsp.vc1_inv_trans_4x8_dc_c: 230.7 vc1dsp.vc1_inv_trans_4x8_dc_rvv_i32: 65.7 (before) vc1dsp.vc1_inv_trans_4x8_dc_rvv_i32: 52.5 (after) vc1dsp.vc1_inv_trans_8x4_dc_c: 246.7 vc1dsp.vc1_inv_trans_8x4_dc_rvv_i64: 56.7 (before) vc1dsp.vc1_inv_trans_8x4_dc_rvv_i64: 45.5 (after) vc1dsp.vc1_inv_trans_8x8_dc_c: 419.7 vc1dsp.vc1_inv_trans_8x8_dc_rvv_i64: 81.2 (before) vc1dsp.vc1_inv_trans_8x8_dc_rvv_i64: 53.5 (after) --- libavcodec/riscv/vc1dsp_rvv.S | 22 ++++++++++++---------- 1 file changed, 12 insertions(+), 10 deletions(-) diff --git a/libavcodec/riscv/vc1dsp_rvv.S b/libavcodec/riscv/vc1dsp_rvv.S index 548ef9d3bf..f9b59688ae 100644 --- a/libavcodec/riscv/vc1dsp_rvv.S +++ b/libavcodec/riscv/vc1dsp_rvv.S @@ -38,18 +38,20 @@ func ff_vc1_inv_trans_\cols\()x\rows\()_dc_rvv, zve64x, zba .endif addi t2, t2, 64 srai t2, t2, 7 -.if \rows * \cols == 64 - vsetvli zero, t0, e16, m8, ta, ma -.elseif \rows * \cols == 32 - vsetvli zero, t0, e16, m4, ta, ma +.if \rows * \cols >= 32 + vsetvli zero, t0, e8, m\mat_lmul, ta, ma .else - vsetivli zero, \rows * \cols, e16, m2, ta, ma + vsetivli zero, \rows * \cols, e8, m\mat_lmul, ta, ma .endif - vzext.vf2 v8, v0 - vadd.vx v8, v8, t2 - vmax.vx v8, v8, zero - vsetvli zero, zero, e8, m\mat_lmul, ta, ma - vnclipu.wi v0, v8, 0 + bgez t2, 1f + + neg t2, t2 + vssubu.vx v0, v0, t2 + vsetivli zero, \rows, e8, m\row_lmul, ta, ma + vsse\w\().v v0, (a0), a1 + ret +1: + vsaddu.vx v0, v0, t2 vsetivli zero, \rows, e8, m\row_lmul, ta, ma vsse\w\().v v0, (a0), a1 ret