From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id C9A9AC5478C for ; Wed, 28 Feb 2024 12:21:49 +0000 (UTC) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1rfIw5-0003EP-Eu; Wed, 28 Feb 2024 07:21:21 -0500 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1rfIw3-0003CM-72 for qemu-devel@nongnu.org; Wed, 28 Feb 2024 07:21:20 -0500 Received: from us-smtp-delivery-124.mimecast.com ([170.10.129.124]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1rfIw1-0003Ht-HK for qemu-devel@nongnu.org; Wed, 28 Feb 2024 07:21:18 -0500 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1709122876; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=Eiqxwv9OHlEZjddgbK8H+3zz1PiTsktvR2pES473PUo=; b=fSOwlNiYOnMIil5QcXzU67LdYbbD9XzPWSM7zg7Apsry6AIvxsmbeQ+shWnftcL34zmP2/ 8R9GA8ZI1RuDdYihb49oGDPBx1vlLjjzm+ORBFIx4L0N3yVSJ/83tJkxSlsMxF1nyOi4bL dsqHO64LsFsdywZKfN0kJ7IdcOQ5Iek= Received: from mail-wr1-f71.google.com (mail-wr1-f71.google.com [209.85.221.71]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-443--HuZtuAVO2CzUQcZ8fjQ1w-1; Wed, 28 Feb 2024 07:21:14 -0500 X-MC-Unique: -HuZtuAVO2CzUQcZ8fjQ1w-1 Received: by mail-wr1-f71.google.com with SMTP id ffacd0b85a97d-33d1d766f83so2170396f8f.3 for ; Wed, 28 Feb 2024 04:21:14 -0800 (PST) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1709122873; x=1709727673; h=content-transfer-encoding:to:subject:message-id:date:from :in-reply-to:references:mime-version:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=Eiqxwv9OHlEZjddgbK8H+3zz1PiTsktvR2pES473PUo=; b=kJkA+kX1MxuC2MFvwpfA5cmf8gAXC2crtDqhkv/x1hO8LIYAYht4ug3nYkscgYEg6K QZNMlyxmly2hDSAKAtisduKuMb1DDKHWlK6MxDxJHWCBDULBwOBhdX2NuHY471LfgJMd ucpmDMRJ6mfsaVsSU/wxbEu4YxinbMKUDUjw9J0HoVqBpanuGPqHdRRvrKrfJTT6JGtL CiQagmk6m8nvMB13BHlPYLQniwu8SUrZbeu7VLIGOqC8GgUtJt5PgXk9viV5OQTNfFv4 kXnQVuOWXP9E1lqLr/l44iYSSSmkvQCguMvuyJXixQLQanWFSLELCSTbLN4WGyxXQBKc Msaw== X-Gm-Message-State: AOJu0YwxqWVwcrLkrsxdZUSsKSKN0Pwkhb5Tw1QTqopsNGI40eJw9QUf t5XebQdCa1FcbCvVLQBgKXb7cUTTwQF0BucPRRVnjXsBH2A23tPywfsWHYbOZYYPl9D96TjRZuf p60pRBFDlH9GlUPNZzNSl0Bv20VJPntPOvo/TB67B6v7qVrkzdNSpHHO/EiDZPO2JhygUNDGv2f Z7iIdjcmj61VGvZbN/ZCD8avKw8TKo75e5O78= X-Received: by 2002:adf:efc5:0:b0:33d:746b:f377 with SMTP id i5-20020adfefc5000000b0033d746bf377mr8796489wrp.45.1709122873172; Wed, 28 Feb 2024 04:21:13 -0800 (PST) X-Google-Smtp-Source: AGHT+IHB0Z6Px0e3dwReWVYF/id4vvYXwcU7kNdJtAu1vzQX5nig6/0guAgPjwhAnHrN8YfEA5HiyGtmi82ewCyw7xo= X-Received: by 2002:adf:efc5:0:b0:33d:746b:f377 with SMTP id i5-20020adfefc5000000b0033d746bf377mr8796474wrp.45.1709122872811; Wed, 28 Feb 2024 04:21:12 -0800 (PST) MIME-Version: 1.0 References: <20240228110626.287178-1-pbonzini@redhat.com> In-Reply-To: <20240228110626.287178-1-pbonzini@redhat.com> From: Paolo Bonzini Date: Wed, 28 Feb 2024 13:21:01 +0100 Message-ID: Subject: Re: [PATCH] tcg/optimize: lower some ANDs to two shifts To: qemu-devel@nongnu.org Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Received-SPF: pass client-ip=170.10.129.124; envelope-from=pbonzini@redhat.com; helo=us-smtp-delivery-124.mimecast.com X-Spam_score_int: -21 X-Spam_score: -2.2 X-Spam_bar: -- X-Spam_report: (-2.2 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.102, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_NONE=-0.0001, RCVD_IN_MSPIKE_H2=-0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sorry, that was sent incorrectly. Paolo On Wed, Feb 28, 2024 at 12:06=E2=80=AFPM Paolo Bonzini wrote: > > Signed-off-by: Paolo Bonzini > --- > tcg/optimize.c | 60 +++++++++++++++++++++++++++++++++++++++++++------- > 1 file changed, 52 insertions(+), 8 deletions(-) > > diff --git a/tcg/optimize.c b/tcg/optimize.c > index 3995bc047db..8ea1f287788 100644 > --- a/tcg/optimize.c > +++ b/tcg/optimize.c > @@ -1281,6 +1281,43 @@ static bool fold_add2(OptContext *ctx, TCGOp *op) > return fold_addsub2(ctx, op, true); > } > > +static bool fold_and_to_shifts(OptContext *ctx, uint64_t c, TCGOp *op) > +{ > + TCGOpcode shl_opc =3D tcg->type =3D=3D TCG_TYPE_I32 ? INDEX_op_shl_i= 32 : INDEX_op_shl_i64; > + TCGOpcode shr_opc =3D tcg->type =3D=3D TCG_TYPE_I32 ? INDEX_op_shr_i= 32 : INDEX_op_shr_i64; > + > + TCGOpcode first, second; > + int count; > + TCGOp *op2; > + > + unsigned_c =3D tcg->type =3D=3D TCG_TYPE_I32 ? (uint32_t) c : c; > + if (is_power_of_2(-c) && > + !tcg_op_imm_match(op->opc, c)) { > + /* AND with 11...11000, shift right then left. */ > + count =3D ctz64(c); > + first =3D shr_opc; > + } else if (is_power_of_2(c + 1) && > + !tcg_op_imm_match(INDEX_op_and_i64, c)) { > + /* AND with 00...00111, shift left then right. */ > + int bits =3D tcg->type =3D=3D TCG_TYPE_I32 ? 32 : 64; > + count =3D bits - cto64(c); > + first =3D shl_opc; > + } else { > + return false; > + } > + > + > + op->opc =3D first; > + op->args[2] =3D arg_new_constant(ctx, count); > + > + second =3D shl_opc ^ shr_opc ^ first; > + op2 =3D tcg_op_insert_after(ctx->tcg, op, second, 3); > + op2->args[0] =3D op->args[0]; > + op2->args[1] =3D op->args[0]; > + op2->args[2] =3D arg_new_constant(ctx, count); > + return true; > +} > + > static bool fold_and(OptContext *ctx, TCGOp *op) > { > uint64_t z1, z2; > @@ -1294,6 +1331,18 @@ static bool fold_and(OptContext *ctx, TCGOp *op) > > z1 =3D arg_info(op->args[1])->z_mask; > z2 =3D arg_info(op->args[2])->z_mask; > + > + /* > + * Known-zeros does not imply known-ones. Therefore unless > + * arg2 is constant, we can't infer affected bits from it. > + */ > + if (arg_is_const(op->args[2])) { > + if (fold_and_to_shifts(ctx, z2, op)) { > + return true; > + } > + ctx->a_mask =3D z1 & ~z2; > + } > + > ctx->z_mask =3D z1 & z2; > > /* > @@ -1303,14 +1352,6 @@ static bool fold_and(OptContext *ctx, TCGOp *op) > ctx->s_mask =3D arg_info(op->args[1])->s_mask > & arg_info(op->args[2])->s_mask; > > - /* > - * Known-zeros does not imply known-ones. Therefore unless > - * arg2 is constant, we can't infer affected bits from it. > - */ > - if (arg_is_const(op->args[2])) { > - ctx->a_mask =3D z1 & ~z2; > - } > - > return fold_masks(ctx, op); > } > > @@ -1333,6 +1374,9 @@ static bool fold_andc(OptContext *ctx, TCGOp *op) > */ > if (arg_is_const(op->args[2])) { > uint64_t z2 =3D ~arg_info(op->args[2])->z_mask; > + if (fold_and_to_shifts(ctx, z2, op)) { > + return true; > + } > ctx->a_mask =3D z1 & ~z2; > z1 &=3D z2; > } > -- > 2.43.2