From patchwork Fri Nov 3 12:12:21 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Joe Ramsay X-Patchwork-Id: 1858884 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (1024-bit key; unprotected) header.d=armh.onmicrosoft.com header.i=@armh.onmicrosoft.com header.a=rsa-sha256 header.s=selector2-armh-onmicrosoft-com header.b=q632viV3; dkim=pass (1024-bit key) header.d=armh.onmicrosoft.com header.i=@armh.onmicrosoft.com header.a=rsa-sha256 header.s=selector2-armh-onmicrosoft-com header.b=q632viV3; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=sourceware.org (client-ip=8.43.85.97; helo=server2.sourceware.org; envelope-from=libc-alpha-bounces+incoming=patchwork.ozlabs.org@sourceware.org; receiver=patchwork.ozlabs.org) Received: from server2.sourceware.org (server2.sourceware.org [8.43.85.97]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (secp384r1) server-digest SHA384) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SMKQv2Gq1z1yQ5 for ; Fri, 3 Nov 2023 23:13:43 +1100 (AEDT) Received: from server2.sourceware.org (localhost [IPv6:::1]) by sourceware.org (Postfix) with ESMTP id A579C3857C44 for ; Fri, 3 Nov 2023 12:13:40 +0000 (GMT) X-Original-To: libc-alpha@sourceware.org Delivered-To: libc-alpha@sourceware.org Received: from EUR05-AM6-obe.outbound.protection.outlook.com (mail-am6eur05on2060.outbound.protection.outlook.com [40.107.22.60]) by sourceware.org (Postfix) with ESMTPS id B3C293858CDB for ; Fri, 3 Nov 2023 12:12:52 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.2 sourceware.org B3C293858CDB Authentication-Results: sourceware.org; dmarc=pass (p=none dis=none) header.from=arm.com Authentication-Results: sourceware.org; spf=pass smtp.mailfrom=arm.com ARC-Filter: OpenARC Filter v1.0.0 sourceware.org B3C293858CDB Authentication-Results: server2.sourceware.org; arc=pass smtp.remote-ip=40.107.22.60 ARC-Seal: i=3; a=rsa-sha256; d=sourceware.org; s=key; t=1699013583; cv=pass; b=InSEFXyCA+43PIOIumVs8/IWOAEqklkt/bZvM7NW1re0IVPl6fcN4B4wqeUiMmQojhZiJl5GSW/LCMX1deqUSZej/6uNE5L4/+d49PkteCMM4Teb/Ww6XsoA94kTtFrZPH4E3vt2m12Owjz8IhJxWuBzmH5s3XnWGvU0K0nqx6Q= ARC-Message-Signature: i=3; a=rsa-sha256; d=sourceware.org; s=key; t=1699013583; c=relaxed/simple; bh=nABuAgbh+ONFnuUSzhfwY5ro7M+0tRcJZuozxNnzk8Y=; h=DKIM-Signature:DKIM-Signature:From:To:Subject:Date:Message-ID: MIME-Version; b=MAozRfIGaQckUq8iuGOU4H7xRHSC7dwvA+3ZRWXmA8mf7ZwLnNk6kibPoOe0f03ezeskxrv/t9gHdF6J3vjcDIXrsa6bkRL8OKl6mM4SZNeffXc1RGQoWO9/VI/UYmsqsANRwDSeikOHS+ge0M61cCjj0xv4qjXErABEaQGK/QY= ARC-Authentication-Results: i=3; server2.sourceware.org ARC-Seal: i=2; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=pass; b=TJdvy/ruAckrAkBnn7wKv0/ZZvXLmo5aZANS0hYdIsm71yMljiH02H3mYnxCx760vJ92J7C8blZt9MCM15pv+PZkxNmPpvUlsMctzo0y4/+4l7NPNdqF0Mvzj/1Ig+Qd9dDh5NCw+6gwiorlaa9zHDR+t/QFJxXPyPcvofE0/KV04FjRCo21hsr1NFMoeoB6cPdKroiT+pRyQy3redzxuXd4zGarDOKgDI9A4ZWVox2CegWMfHdiUcSVDsV3ElhD6Oxu8hCNHKU0bS/Cn8wlwGiD2FI+aZThNPpX4GZ9kLvg+yLEcAuegmZmNXKwWtp4DJWPbuhtjSR8QAywC/HhIQ== ARC-Message-Signature: i=2; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-AntiSpam-MessageData-ChunkCount:X-MS-Exchange-AntiSpam-MessageData-0:X-MS-Exchange-AntiSpam-MessageData-1; bh=NG34qTnEC4bRa4aNckHtvsEacuiHuQpwh7P8T+m7OCc=; b=EXhEQLsubWE+dqn+iR3m/r73cuhuL71u38lW0L8ui2yQ6QXqRVQNPDzXBosNmEPu+b1mfK2Ie1pU8fnUdtIrw6PKtMsl9Vd8Z8+nvJWJ5wUdT8lDST5lEBygnVVSZxvgNzsEmeaBrAcBrv7LtwaoZBbsr1P45iwdZ+gb/r4GKWh8dZFIArtzOw9F1RNmZfP5V/WN5IIu/GRFVSMSuCBHTM/NQL5YLDaLgpYq8MmXrOnWHvvO61gbCbEII1eRaG6I14mvqP3WwKO8Tmvb0xk0K9DPQFhR/N6yr//sbvK7dMUP9Lr2B2JpJdJnMcaQiMj9lFa4AzEd2GvoLeOGmL0S5Q== ARC-Authentication-Results: i=2; mx.microsoft.com 1; spf=pass (sender ip is 63.35.35.123) smtp.rcpttodomain=sourceware.org smtp.mailfrom=arm.com; dmarc=pass (p=none sp=none pct=100) action=none header.from=arm.com; dkim=pass (signature was verified) header.d=armh.onmicrosoft.com; arc=pass (0 oda=1 ltdi=1 spf=[1, 1, smtp.mailfrom=arm.com] dmarc=[1, 1, header.from=arm.com]) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=armh.onmicrosoft.com; s=selector2-armh-onmicrosoft-com; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=NG34qTnEC4bRa4aNckHtvsEacuiHuQpwh7P8T+m7OCc=; b=q632viV3a84DjKvzyAvpJn9i5OVjVw902tzXUFFXVrttHXhDndjYzJfZlUxwtuy6buW2v3w9gseVxTCd8xXTKLRK/weNj9ic28ivlQik08yd6sQQYQeWc8RrWH9BS8GO+tgIm++exr8rfhJjPeb2kDdOTeuAm4UfqXSv4svWT5k= Received: from AS9PR06CA0572.eurprd06.prod.outlook.com (2603:10a6:20b:486::7) by DB8PR08MB5434.eurprd08.prod.outlook.com (2603:10a6:10:116::20) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.6954.21; Fri, 3 Nov 2023 12:12:47 +0000 Received: from AMS0EPF000001B4.eurprd05.prod.outlook.com (2603:10a6:20b:486:cafe::3) by AS9PR06CA0572.outlook.office365.com (2603:10a6:20b:486::7) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.6954.22 via Frontend Transport; Fri, 3 Nov 2023 12:12:46 +0000 X-MS-Exchange-Authentication-Results: spf=pass (sender IP is 63.35.35.123) smtp.mailfrom=arm.com; dkim=pass (signature was verified) header.d=armh.onmicrosoft.com;dmarc=pass action=none header.from=arm.com; Received-SPF: Pass (protection.outlook.com: domain of arm.com designates 63.35.35.123 as permitted sender) receiver=protection.outlook.com; client-ip=63.35.35.123; helo=64aa7808-outbound-1.mta.getcheckrecipient.com; pr=C Received: from 64aa7808-outbound-1.mta.getcheckrecipient.com (63.35.35.123) by AMS0EPF000001B4.mail.protection.outlook.com (10.167.16.168) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.6954.19 via Frontend Transport; Fri, 3 Nov 2023 12:12:46 +0000 Received: ("Tessian outbound 7671e7ddc218:v228"); Fri, 03 Nov 2023 12:12:46 +0000 X-CheckRecipientChecked: true X-CR-MTA-CID: 175ec358b4dceb3b X-CR-MTA-TID: 64aa7808 Received: from ff83208177d0.1 by 64aa7808-outbound-1.mta.getcheckrecipient.com id 8B332F6C-3A00-4A55-98AC-B31FEB60A089.1; Fri, 03 Nov 2023 12:12:40 +0000 Received: from EUR04-DB3-obe.outbound.protection.outlook.com by 64aa7808-outbound-1.mta.getcheckrecipient.com with ESMTPS id ff83208177d0.1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384); Fri, 03 Nov 2023 12:12:40 +0000 ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=hgT+GAgpGe3KY93fP1xUs0sTTeRN8u3p0qnyFUqxzucA8o+22QzTptKgzuaZQGrQB5rpRHwJr5HG7ppdzdjqT9xtzJXRC/Ravr2BghwdRiFxSSaZJs6lPpbn1YmnoPie81fjJ1aBSGRVcjT6gHMBNyrsbUgwR6xANj65GMlIwFJ3u3zht1cATjNit5wN3G4ptViwbW+WRUailfICreJqSBpM4M0quryLUkST7LCFvpo9Hdy0CQ1FcX1PP7OoYcAHfQq/SHhCwx8Zo7B0qVpuRgGCSwzcZncr+Jcyb85zAfoqm1bl8FaUAhS6dpXapaRYYd8UQ7lYwsK1DV9T+qR0rQ== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-AntiSpam-MessageData-ChunkCount:X-MS-Exchange-AntiSpam-MessageData-0:X-MS-Exchange-AntiSpam-MessageData-1; bh=NG34qTnEC4bRa4aNckHtvsEacuiHuQpwh7P8T+m7OCc=; b=Ix6sV14iFZbr5NaSA8IBH1P+KDf2FcWOOzfftTffaN9Yuv5Ba731v5GvfuDnc8ptmPOBiDsbKx2outAjEudjWVeKa3vnP1SKSr5fHAgQKdKkaqcwHObgztJyTMT1J1fCDagTFgLLhQGY7hidvluVWopTHlcoi0G2foHd8rGyf3CcCNP79rdUQ/h6EcN9iRpn1+FjjubuHpjS8mkl5eSZMmdpBqt+DlJqBOgpst8mLzgHimG5n5UoBFbPSdxU5RGm/AVuPKP8BiNKgZftITzDPiV9uUqj+7ff9maMtftmiejoqN+5axecPLu6jRxw9yu/P39iPflh619SVRUUeGo1Xw== ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass (sender ip is 40.67.248.234) smtp.rcpttodomain=sourceware.org smtp.mailfrom=arm.com; dmarc=pass (p=none sp=none pct=100) action=none header.from=arm.com; dkim=none (message not signed); arc=none (0) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=armh.onmicrosoft.com; s=selector2-armh-onmicrosoft-com; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=NG34qTnEC4bRa4aNckHtvsEacuiHuQpwh7P8T+m7OCc=; b=q632viV3a84DjKvzyAvpJn9i5OVjVw902tzXUFFXVrttHXhDndjYzJfZlUxwtuy6buW2v3w9gseVxTCd8xXTKLRK/weNj9ic28ivlQik08yd6sQQYQeWc8RrWH9BS8GO+tgIm++exr8rfhJjPeb2kDdOTeuAm4UfqXSv4svWT5k= Received: from AS9PR05CA0178.eurprd05.prod.outlook.com (2603:10a6:20b:496::27) by PA4PR08MB5952.eurprd08.prod.outlook.com (2603:10a6:102:e9::10) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.6954.21; Fri, 3 Nov 2023 12:12:37 +0000 Received: from AM3PEPF00009B9F.eurprd04.prod.outlook.com (2603:10a6:20b:496:cafe::3c) by AS9PR05CA0178.outlook.office365.com (2603:10a6:20b:496::27) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.6954.21 via Frontend Transport; Fri, 3 Nov 2023 12:12:37 +0000 X-MS-Exchange-Authentication-Results: spf=pass (sender IP is 40.67.248.234) smtp.mailfrom=arm.com; dkim=none (message not signed) header.d=none;dmarc=pass action=none header.from=arm.com; Received-SPF: Pass (protection.outlook.com: domain of arm.com designates 40.67.248.234 as permitted sender) receiver=protection.outlook.com; client-ip=40.67.248.234; helo=nebula.arm.com; pr=C Received: from nebula.arm.com (40.67.248.234) by AM3PEPF00009B9F.mail.protection.outlook.com (10.167.16.24) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.20.6954.19 via Frontend Transport; Fri, 3 Nov 2023 12:12:37 +0000 Received: from AZ-NEU-EX02.Emea.Arm.com (10.251.26.5) by AZ-NEU-EX04.Arm.com (10.251.24.32) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.32; Fri, 3 Nov 2023 12:12:36 +0000 Received: from AZ-NEU-EX04.Arm.com (10.251.24.32) by AZ-NEU-EX02.Emea.Arm.com (10.251.26.5) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.32; Fri, 3 Nov 2023 12:12:36 +0000 Received: from vcn-man-apps.manchester.arm.com (10.32.108.22) by mail.arm.com (10.251.24.32) with Microsoft SMTP Server id 15.1.2507.32 via Frontend Transport; Fri, 3 Nov 2023 12:12:36 +0000 From: Joe Ramsay To: CC: Joe Ramsay Subject: [PATCH 3/6] aarch64: Add vector implementations of atan routines Date: Fri, 3 Nov 2023 12:12:21 +0000 Message-ID: <20231103121224.16835-3-Joe.Ramsay@arm.com> X-Mailer: git-send-email 2.27.0 In-Reply-To: <20231103121224.16835-1-Joe.Ramsay@arm.com> References: <20231103121224.16835-1-Joe.Ramsay@arm.com> MIME-Version: 1.0 X-EOPAttributedMessage: 1 X-MS-TrafficTypeDiagnostic: AM3PEPF00009B9F:EE_|PA4PR08MB5952:EE_|AMS0EPF000001B4:EE_|DB8PR08MB5434:EE_ X-MS-Office365-Filtering-Correlation-Id: b17e172a-c15d-40e3-57e1-08dbdc6630d4 x-checkrecipientrouted: true NoDisclaimer: true X-MS-Exchange-SenderADCheck: 1 X-MS-Exchange-AntiSpam-Relay: 0 X-Microsoft-Antispam-Untrusted: BCL:0; X-Microsoft-Antispam-Message-Info-Original: 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 X-Forefront-Antispam-Report-Untrusted: CIP:40.67.248.234; CTRY:IE; LANG:en; SCL:1; SRV:; IPV:NLI; SFV:NSPM; H:nebula.arm.com; PTR:InfoDomainNonexistent; CAT:NONE; SFS:(13230031)(4636009)(376002)(136003)(346002)(39860400002)(396003)(230922051799003)(82310400011)(64100799003)(1800799009)(186009)(451199024)(36840700001)(40470700004)(46966006)(82740400003)(83380400001)(36860700001)(47076005)(40480700001)(40460700003)(41300700001)(70206006)(70586007)(478600001)(6916009)(316002)(4326008)(8676002)(356005)(2906002)(6666004)(8936002)(30864003)(86362001)(7696005)(5660300002)(1076003)(36756003)(26005)(336012)(81166007)(426003)(2616005)(2004002)(36900700001); DIR:OUT; SFP:1101; X-MS-Exchange-Transport-CrossTenantHeadersStamped: PA4PR08MB5952 X-MS-Exchange-Transport-CrossTenantHeadersStripped: AMS0EPF000001B4.eurprd05.prod.outlook.com X-MS-PublicTrafficType: Email X-MS-Office365-Filtering-Correlation-Id-Prvs: 78a9ca07-cd6f-465e-6b19-08dbdc662b44 X-Microsoft-Antispam: BCL:0; X-Microsoft-Antispam-Message-Info: 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 X-Forefront-Antispam-Report: CIP:63.35.35.123; CTRY:IE; LANG:en; SCL:1; SRV:; IPV:CAL; SFV:NSPM; H:64aa7808-outbound-1.mta.getcheckrecipient.com; PTR:ec2-63-35-35-123.eu-west-1.compute.amazonaws.com; CAT:NONE; SFS:(13230031)(4636009)(346002)(39860400002)(136003)(396003)(376002)(230922051799003)(82310400011)(64100799003)(451199024)(186009)(1800799009)(46966006)(36840700001)(40470700004)(8676002)(6666004)(40480700001)(40460700003)(478600001)(2906002)(2616005)(1076003)(336012)(26005)(426003)(86362001)(82740400003)(81166007)(47076005)(36756003)(83380400001)(36860700001)(7696005)(5660300002)(6916009)(70206006)(70586007)(316002)(41300700001)(8936002)(4326008)(30864003)(2004002); DIR:OUT; SFP:1101; X-OriginatorOrg: arm.com X-MS-Exchange-CrossTenant-OriginalArrivalTime: 03 Nov 2023 12:12:46.6273 (UTC) X-MS-Exchange-CrossTenant-Network-Message-Id: b17e172a-c15d-40e3-57e1-08dbdc6630d4 X-MS-Exchange-CrossTenant-Id: f34e5979-57d9-4aaa-ad4d-b122a662184d X-MS-Exchange-CrossTenant-OriginalAttributedTenantConnectingIp: TenantId=f34e5979-57d9-4aaa-ad4d-b122a662184d; Ip=[63.35.35.123]; Helo=[64aa7808-outbound-1.mta.getcheckrecipient.com] X-MS-Exchange-CrossTenant-AuthSource: AMS0EPF000001B4.eurprd05.prod.outlook.com X-MS-Exchange-CrossTenant-AuthAs: Anonymous X-MS-Exchange-CrossTenant-FromEntityHeader: HybridOnPrem X-MS-Exchange-Transport-CrossTenantHeadersStamped: DB8PR08MB5434 X-Spam-Status: No, score=-12.4 required=5.0 tests=BAYES_00, DKIM_SIGNED, DKIM_VALID, FORGED_SPF_HELO, GIT_PATCH_0, KAM_DMARC_NONE, KAM_SHORT, RCVD_IN_DNSWL_NONE, RCVD_IN_MSPIKE_H2, SPF_HELO_PASS, SPF_NONE, TXREP, T_SCC_BODY_TEXT_LINE, UNPARSEABLE_RELAY autolearn=ham autolearn_force=no version=3.4.6 X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on server2.sourceware.org X-BeenThere: libc-alpha@sourceware.org X-Mailman-Version: 2.1.30 Precedence: list List-Id: Libc-alpha mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: libc-alpha-bounces+incoming=patchwork.ozlabs.org@sourceware.org --- Thanks, Joe sysdeps/aarch64/fpu/Makefile | 1 + sysdeps/aarch64/fpu/Versions | 4 + sysdeps/aarch64/fpu/atan_advsimd.c | 104 +++++++++++++++++ sysdeps/aarch64/fpu/atan_sve.c | 90 +++++++++++++++ sysdeps/aarch64/fpu/atanf_advsimd.c | 109 ++++++++++++++++++ sysdeps/aarch64/fpu/atanf_sve.c | 79 +++++++++++++ sysdeps/aarch64/fpu/bits/math-vector.h | 4 + .../fpu/test-double-advsimd-wrappers.c | 1 + .../aarch64/fpu/test-double-sve-wrappers.c | 1 + .../aarch64/fpu/test-float-advsimd-wrappers.c | 1 + sysdeps/aarch64/fpu/test-float-sve-wrappers.c | 1 + sysdeps/aarch64/libm-test-ulps | 8 ++ .../unix/sysv/linux/aarch64/libmvec.abilist | 4 + 13 files changed, 407 insertions(+) create mode 100644 sysdeps/aarch64/fpu/atan_advsimd.c create mode 100644 sysdeps/aarch64/fpu/atan_sve.c create mode 100644 sysdeps/aarch64/fpu/atanf_advsimd.c create mode 100644 sysdeps/aarch64/fpu/atanf_sve.c Reviewed-by: Szabolcs Nagy diff --git a/sysdeps/aarch64/fpu/Makefile b/sysdeps/aarch64/fpu/Makefile index 606fdd804f..5bd77a749d 100644 --- a/sysdeps/aarch64/fpu/Makefile +++ b/sysdeps/aarch64/fpu/Makefile @@ -1,5 +1,6 @@ libmvec-supported-funcs = acos \ asin \ + atan \ cos \ exp \ exp10 \ diff --git a/sysdeps/aarch64/fpu/Versions b/sysdeps/aarch64/fpu/Versions index 1037cd92bd..dfc3d2dad3 100644 --- a/sysdeps/aarch64/fpu/Versions +++ b/sysdeps/aarch64/fpu/Versions @@ -26,6 +26,10 @@ libmvec { _ZGVnN2v_asin; _ZGVsMxv_asinf; _ZGVsMxv_asin; + _ZGVnN4v_atanf; + _ZGVnN2v_atan; + _ZGVsMxv_atanf; + _ZGVsMxv_atan; _ZGVnN4v_exp10f; _ZGVnN2v_exp10; _ZGVsMxv_exp10f; diff --git a/sysdeps/aarch64/fpu/atan_advsimd.c b/sysdeps/aarch64/fpu/atan_advsimd.c new file mode 100644 index 0000000000..d52c07d8a0 --- /dev/null +++ b/sysdeps/aarch64/fpu/atan_advsimd.c @@ -0,0 +1,104 @@ +/* Double-precision AdvSIMD inverse tan + + Copyright (C) 2023 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Lesser General Public + License as published by the Free Software Foundation; either + version 2.1 of the License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with the GNU C Library; if not, see + . */ + +#include "v_math.h" +#include "poly_advsimd_f64.h" + +static const struct data +{ + float64x2_t pi_over_2; + float64x2_t poly[20]; +} data = { + /* Coefficients of polynomial P such that atan(x)~x+x*P(x^2) on + [2**-1022, 1.0]. */ + .poly = { V2 (-0x1.5555555555555p-2), V2 (0x1.99999999996c1p-3), + V2 (-0x1.2492492478f88p-3), V2 (0x1.c71c71bc3951cp-4), + V2 (-0x1.745d160a7e368p-4), V2 (0x1.3b139b6a88ba1p-4), + V2 (-0x1.11100ee084227p-4), V2 (0x1.e1d0f9696f63bp-5), + V2 (-0x1.aebfe7b418581p-5), V2 (0x1.842dbe9b0d916p-5), + V2 (-0x1.5d30140ae5e99p-5), V2 (0x1.338e31eb2fbbcp-5), + V2 (-0x1.00e6eece7de8p-5), V2 (0x1.860897b29e5efp-6), + V2 (-0x1.0051381722a59p-6), V2 (0x1.14e9dc19a4a4ep-7), + V2 (-0x1.d0062b42fe3bfp-9), V2 (0x1.17739e210171ap-10), + V2 (-0x1.ab24da7be7402p-13), V2 (0x1.358851160a528p-16), }, + .pi_over_2 = V2 (0x1.921fb54442d18p+0), +}; + +#define SignMask v_u64 (0x8000000000000000) +#define TinyBound 0x3e10000000000000 /* asuint64(0x1p-30). */ +#define BigBound 0x4340000000000000 /* asuint64(0x1p53). */ + +/* Fast implementation of vector atan. + Based on atan(x) ~ shift + z + z^3 * P(z^2) with reduction to [0,1] using + z=1/x and shift = pi/2. Maximum observed error is 2.27 ulps: + _ZGVnN2v_atan (0x1.0005af27c23e9p+0) got 0x1.9225645bdd7c1p-1 + want 0x1.9225645bdd7c3p-1. */ +float64x2_t VPCS_ATTR V_NAME_D1 (atan) (float64x2_t x) +{ + const struct data *d = ptr_barrier (&data); + + /* Small cases, infs and nans are supported by our approximation technique, + but do not set fenv flags correctly. Only trigger special case if we need + fenv. */ + uint64x2_t ix = vreinterpretq_u64_f64 (x); + uint64x2_t sign = vandq_u64 (ix, SignMask); + +#if WANT_SIMD_EXCEPT + uint64x2_t ia12 = vandq_u64 (ix, v_u64 (0x7ff0000000000000)); + uint64x2_t special = vcgtq_u64 (vsubq_u64 (ia12, v_u64 (TinyBound)), + v_u64 (BigBound - TinyBound)); + /* If any lane is special, fall back to the scalar routine for all lanes. */ + if (__glibc_unlikely (v_any_u64 (special))) + return v_call_f64 (atan, x, v_f64 (0), v_u64 (-1)); +#endif + + /* Argument reduction: + y := arctan(x) for x < 1 + y := pi/2 + arctan(-1/x) for x > 1 + Hence, use z=-1/a if x>=1, otherwise z=a. */ + uint64x2_t red = vcagtq_f64 (x, v_f64 (1.0)); + /* Avoid dependency in abs(x) in division (and comparison). */ + float64x2_t z = vbslq_f64 (red, vdivq_f64 (v_f64 (1.0), x), x); + float64x2_t shift = vreinterpretq_f64_u64 ( + vandq_u64 (red, vreinterpretq_u64_f64 (d->pi_over_2))); + /* Use absolute value only when needed (odd powers of z). */ + float64x2_t az = vbslq_f64 ( + SignMask, vreinterpretq_f64_u64 (vandq_u64 (SignMask, red)), z); + + /* Calculate the polynomial approximation. + Use split Estrin scheme for P(z^2) with deg(P)=19. Use split instead of + full scheme to avoid underflow in x^16. + The order 19 polynomial P approximates + (atan(sqrt(x))-sqrt(x))/x^(3/2). */ + float64x2_t z2 = vmulq_f64 (z, z); + float64x2_t x2 = vmulq_f64 (z2, z2); + float64x2_t x4 = vmulq_f64 (x2, x2); + float64x2_t x8 = vmulq_f64 (x4, x4); + float64x2_t y + = vfmaq_f64 (v_estrin_7_f64 (z2, x2, x4, d->poly), + v_estrin_11_f64 (z2, x2, x4, x8, d->poly + 8), x8); + + /* Finalize. y = shift + z + z^3 * P(z^2). */ + y = vfmaq_f64 (az, y, vmulq_f64 (z2, az)); + y = vaddq_f64 (y, shift); + + /* y = atan(x) if x>0, -atan(-x) otherwise. */ + y = vreinterpretq_f64_u64 (veorq_u64 (vreinterpretq_u64_f64 (y), sign)); + return y; +} diff --git a/sysdeps/aarch64/fpu/atan_sve.c b/sysdeps/aarch64/fpu/atan_sve.c new file mode 100644 index 0000000000..35587ef212 --- /dev/null +++ b/sysdeps/aarch64/fpu/atan_sve.c @@ -0,0 +1,90 @@ +/* Double-precision SVE inverse tan + + Copyright (C) 2023 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Lesser General Public + License as published by the Free Software Foundation; either + version 2.1 of the License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with the GNU C Library; if not, see + . */ + +#include "sv_math.h" +#include "poly_sve_f64.h" + +static const struct data +{ + float64_t poly[20]; + float64_t pi_over_2; +} data = { + /* Coefficients of polynomial P such that atan(x)~x+x*P(x^2) on + [2**-1022, 1.0]. */ + .poly = { -0x1.5555555555555p-2, 0x1.99999999996c1p-3, -0x1.2492492478f88p-3, + 0x1.c71c71bc3951cp-4, -0x1.745d160a7e368p-4, 0x1.3b139b6a88ba1p-4, + -0x1.11100ee084227p-4, 0x1.e1d0f9696f63bp-5, -0x1.aebfe7b418581p-5, + 0x1.842dbe9b0d916p-5, -0x1.5d30140ae5e99p-5, 0x1.338e31eb2fbbcp-5, + -0x1.00e6eece7de8p-5, 0x1.860897b29e5efp-6, -0x1.0051381722a59p-6, + 0x1.14e9dc19a4a4ep-7, -0x1.d0062b42fe3bfp-9, 0x1.17739e210171ap-10, + -0x1.ab24da7be7402p-13, 0x1.358851160a528p-16, }, + .pi_over_2 = 0x1.921fb54442d18p+0, +}; + +/* Useful constants. */ +#define SignMask (0x8000000000000000) + +/* Fast implementation of SVE atan. + Based on atan(x) ~ shift + z + z^3 * P(z^2) with reduction to [0,1] using + z=1/x and shift = pi/2. Largest errors are close to 1. The maximum observed + error is 2.27 ulps: + _ZGVsMxv_atan (0x1.0005af27c23e9p+0) got 0x1.9225645bdd7c1p-1 + want 0x1.9225645bdd7c3p-1. */ +svfloat64_t SV_NAME_D1 (atan) (svfloat64_t x, const svbool_t pg) +{ + const struct data *d = ptr_barrier (&data); + + /* No need to trigger special case. Small cases, infs and nans + are supported by our approximation technique. */ + svuint64_t ix = svreinterpret_u64 (x); + svuint64_t sign = svand_x (pg, ix, SignMask); + + /* Argument reduction: + y := arctan(x) for x < 1 + y := pi/2 + arctan(-1/x) for x > 1 + Hence, use z=-1/a if x>=1, otherwise z=a. */ + svbool_t red = svacgt (pg, x, 1.0); + /* Avoid dependency in abs(x) in division (and comparison). */ + svfloat64_t z = svsel (red, svdivr_x (pg, x, 1.0), x); + /* Use absolute value only when needed (odd powers of z). */ + svfloat64_t az = svabs_x (pg, z); + az = svneg_m (az, red, az); + + /* Use split Estrin scheme for P(z^2) with deg(P)=19. */ + svfloat64_t z2 = svmul_x (pg, z, z); + svfloat64_t x2 = svmul_x (pg, z2, z2); + svfloat64_t x4 = svmul_x (pg, x2, x2); + svfloat64_t x8 = svmul_x (pg, x4, x4); + + svfloat64_t y + = svmla_x (pg, sv_estrin_7_f64_x (pg, z2, x2, x4, d->poly), + sv_estrin_11_f64_x (pg, z2, x2, x4, x8, d->poly + 8), x8); + + /* y = shift + z + z^3 * P(z^2). */ + svfloat64_t z3 = svmul_x (pg, z2, az); + y = svmla_x (pg, az, z3, y); + + /* Apply shift as indicated by `red` predicate. */ + y = svadd_m (red, y, d->pi_over_2); + + /* y = atan(x) if x>0, -atan(-x) otherwise. */ + y = svreinterpret_f64 (sveor_x (pg, svreinterpret_u64 (y), sign)); + + return y; +} diff --git a/sysdeps/aarch64/fpu/atanf_advsimd.c b/sysdeps/aarch64/fpu/atanf_advsimd.c new file mode 100644 index 0000000000..589b0e8c96 --- /dev/null +++ b/sysdeps/aarch64/fpu/atanf_advsimd.c @@ -0,0 +1,109 @@ +/* Single-precision AdvSIMD inverse tan + + Copyright (C) 2023 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Lesser General Public + License as published by the Free Software Foundation; either + version 2.1 of the License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with the GNU C Library; if not, see + . */ + +#include "v_math.h" +#include "poly_advsimd_f32.h" + +static const struct data +{ + float32x4_t poly[8]; + float32x4_t pi_over_2; +} data = { + /* Coefficients of polynomial P such that atan(x)~x+x*P(x^2) on + [2**-128, 1.0]. + Generated using fpminimax between FLT_MIN and 1. */ + .poly = { V4 (-0x1.55555p-2f), V4 (0x1.99935ep-3f), V4 (-0x1.24051ep-3f), + V4 (0x1.bd7368p-4f), V4 (-0x1.491f0ep-4f), V4 (0x1.93a2c0p-5f), + V4 (-0x1.4c3c60p-6f), V4 (0x1.01fd88p-8f) }, + .pi_over_2 = V4 (0x1.921fb6p+0f), +}; + +#define SignMask v_u32 (0x80000000) + +#define P(i) d->poly[i] + +#define TinyBound 0x30800000 /* asuint(0x1p-30). */ +#define BigBound 0x4e800000 /* asuint(0x1p30). */ + +#if WANT_SIMD_EXCEPT +static float32x4_t VPCS_ATTR NOINLINE +special_case (float32x4_t x, float32x4_t y, uint32x4_t special) +{ + return v_call_f32 (atanf, x, y, special); +} +#endif + +/* Fast implementation of vector atanf based on + atan(x) ~ shift + z + z^3 * P(z^2) with reduction to [0,1] + using z=-1/x and shift = pi/2. Maximum observed error is 2.9ulps: + _ZGVnN4v_atanf (0x1.0468f6p+0) got 0x1.967f06p-1 want 0x1.967fp-1. */ +float32x4_t VPCS_ATTR V_NAME_F1 (atan) (float32x4_t x) +{ + const struct data *d = ptr_barrier (&data); + + /* Small cases, infs and nans are supported by our approximation technique, + but do not set fenv flags correctly. Only trigger special case if we need + fenv. */ + uint32x4_t ix = vreinterpretq_u32_f32 (x); + uint32x4_t sign = vandq_u32 (ix, SignMask); + +#if WANT_SIMD_EXCEPT + uint32x4_t ia = vandq_u32 (ix, v_u32 (0x7ff00000)); + uint32x4_t special = vcgtq_u32 (vsubq_u32 (ia, v_u32 (TinyBound)), + v_u32 (BigBound - TinyBound)); + /* If any lane is special, fall back to the scalar routine for all lanes. */ + if (__glibc_unlikely (v_any_u32 (special))) + return special_case (x, x, v_u32 (-1)); +#endif + + /* Argument reduction: + y := arctan(x) for x < 1 + y := pi/2 + arctan(-1/x) for x > 1 + Hence, use z=-1/a if x>=1, otherwise z=a. */ + uint32x4_t red = vcagtq_f32 (x, v_f32 (1.0)); + /* Avoid dependency in abs(x) in division (and comparison). */ + float32x4_t z = vbslq_f32 (red, vdivq_f32 (v_f32 (1.0f), x), x); + float32x4_t shift = vreinterpretq_f32_u32 ( + vandq_u32 (red, vreinterpretq_u32_f32 (d->pi_over_2))); + /* Use absolute value only when needed (odd powers of z). */ + float32x4_t az = vbslq_f32 ( + SignMask, vreinterpretq_f32_u32 (vandq_u32 (SignMask, red)), z); + + /* Calculate the polynomial approximation. + Use 2-level Estrin scheme for P(z^2) with deg(P)=7. However, + a standard implementation using z8 creates spurious underflow + in the very last fma (when z^8 is small enough). + Therefore, we split the last fma into a mul and an fma. + Horner and single-level Estrin have higher errors that exceed + threshold. */ + float32x4_t z2 = vmulq_f32 (z, z); + float32x4_t z4 = vmulq_f32 (z2, z2); + + float32x4_t y = vfmaq_f32 ( + v_pairwise_poly_3_f32 (z2, z4, d->poly), z4, + vmulq_f32 (z4, v_pairwise_poly_3_f32 (z2, z4, d->poly + 4))); + + /* y = shift + z * P(z^2). */ + y = vaddq_f32 (vfmaq_f32 (az, y, vmulq_f32 (z2, az)), shift); + + /* y = atan(x) if x>0, -atan(-x) otherwise. */ + y = vreinterpretq_f32_u32 (veorq_u32 (vreinterpretq_u32_f32 (y), sign)); + + return y; +} diff --git a/sysdeps/aarch64/fpu/atanf_sve.c b/sysdeps/aarch64/fpu/atanf_sve.c new file mode 100644 index 0000000000..9453e7aa29 --- /dev/null +++ b/sysdeps/aarch64/fpu/atanf_sve.c @@ -0,0 +1,79 @@ +/* Single-precision SVE inverse tan + + Copyright (C) 2023 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Lesser General Public + License as published by the Free Software Foundation; either + version 2.1 of the License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with the GNU C Library; if not, see + . */ + +#include "sv_math.h" +#include "poly_sve_f32.h" + +static const struct data +{ + float32_t poly[8]; + float32_t pi_over_2; +} data = { + /* Coefficients of polynomial P such that atan(x)~x+x*P(x^2) on + [2**-128, 1.0]. */ + .poly = { -0x1.55555p-2f, 0x1.99935ep-3f, -0x1.24051ep-3f, 0x1.bd7368p-4f, + -0x1.491f0ep-4f, 0x1.93a2c0p-5f, -0x1.4c3c60p-6f, 0x1.01fd88p-8f }, + .pi_over_2 = 0x1.921fb6p+0f, +}; + +#define SignMask (0x80000000) + +/* Fast implementation of SVE atanf based on + atan(x) ~ shift + z + z^3 * P(z^2) with reduction to [0,1] using + z=-1/x and shift = pi/2. + Largest observed error is 2.9 ULP, close to +/-1.0: + _ZGVsMxv_atanf (0x1.0468f6p+0) got -0x1.967f06p-1 + want -0x1.967fp-1. */ +svfloat32_t SV_NAME_F1 (atan) (svfloat32_t x, const svbool_t pg) +{ + const struct data *d = ptr_barrier (&data); + + /* No need to trigger special case. Small cases, infs and nans + are supported by our approximation technique. */ + svuint32_t ix = svreinterpret_u32 (x); + svuint32_t sign = svand_x (pg, ix, SignMask); + + /* Argument reduction: + y := arctan(x) for x < 1 + y := pi/2 + arctan(-1/x) for x > 1 + Hence, use z=-1/a if x>=1, otherwise z=a. */ + svbool_t red = svacgt (pg, x, 1.0f); + /* Avoid dependency in abs(x) in division (and comparison). */ + svfloat32_t z = svsel (red, svdiv_x (pg, sv_f32 (1.0f), x), x); + /* Use absolute value only when needed (odd powers of z). */ + svfloat32_t az = svabs_x (pg, z); + az = svneg_m (az, red, az); + + /* Use split Estrin scheme for P(z^2) with deg(P)=7. */ + svfloat32_t z2 = svmul_x (pg, z, z); + svfloat32_t z4 = svmul_x (pg, z2, z2); + svfloat32_t z8 = svmul_x (pg, z4, z4); + + svfloat32_t y = sv_estrin_7_f32_x (pg, z2, z4, z8, d->poly); + + /* y = shift + z + z^3 * P(z^2). */ + svfloat32_t z3 = svmul_x (pg, z2, az); + y = svmla_x (pg, az, z3, y); + + /* Apply shift as indicated by 'red' predicate. */ + y = svadd_m (red, y, sv_f32 (d->pi_over_2)); + + /* y = atan(x) if x>0, -atan(-x) otherwise. */ + return svreinterpret_f32 (sveor_x (pg, svreinterpret_u32 (y), sign)); +} diff --git a/sysdeps/aarch64/fpu/bits/math-vector.h b/sysdeps/aarch64/fpu/bits/math-vector.h index f313993d70..37aa74fe50 100644 --- a/sysdeps/aarch64/fpu/bits/math-vector.h +++ b/sysdeps/aarch64/fpu/bits/math-vector.h @@ -51,6 +51,7 @@ typedef __SVBool_t __sv_bool_t; __vpcs __f32x4_t _ZGVnN4v_acosf (__f32x4_t); __vpcs __f32x4_t _ZGVnN4v_asinf (__f32x4_t); +__vpcs __f32x4_t _ZGVnN4v_atanf (__f32x4_t); __vpcs __f32x4_t _ZGVnN4v_cosf (__f32x4_t); __vpcs __f32x4_t _ZGVnN4v_expf (__f32x4_t); __vpcs __f32x4_t _ZGVnN4v_exp10f (__f32x4_t); @@ -63,6 +64,7 @@ __vpcs __f32x4_t _ZGVnN4v_tanf (__f32x4_t); __vpcs __f64x2_t _ZGVnN2v_acos (__f64x2_t); __vpcs __f64x2_t _ZGVnN2v_asin (__f64x2_t); +__vpcs __f64x2_t _ZGVnN2v_atan (__f64x2_t); __vpcs __f64x2_t _ZGVnN2v_cos (__f64x2_t); __vpcs __f64x2_t _ZGVnN2v_exp (__f64x2_t); __vpcs __f64x2_t _ZGVnN2v_exp10 (__f64x2_t); @@ -80,6 +82,7 @@ __vpcs __f64x2_t _ZGVnN2v_tan (__f64x2_t); __sv_f32_t _ZGVsMxv_acosf (__sv_f32_t, __sv_bool_t); __sv_f32_t _ZGVsMxv_asinf (__sv_f32_t, __sv_bool_t); +__sv_f32_t _ZGVsMxv_atanf (__sv_f32_t, __sv_bool_t); __sv_f32_t _ZGVsMxv_cosf (__sv_f32_t, __sv_bool_t); __sv_f32_t _ZGVsMxv_expf (__sv_f32_t, __sv_bool_t); __sv_f32_t _ZGVsMxv_exp10f (__sv_f32_t, __sv_bool_t); @@ -92,6 +95,7 @@ __sv_f32_t _ZGVsMxv_tanf (__sv_f32_t, __sv_bool_t); __sv_f64_t _ZGVsMxv_acos (__sv_f64_t, __sv_bool_t); __sv_f64_t _ZGVsMxv_asin (__sv_f64_t, __sv_bool_t); +__sv_f64_t _ZGVsMxv_atan (__sv_f64_t, __sv_bool_t); __sv_f64_t _ZGVsMxv_cos (__sv_f64_t, __sv_bool_t); __sv_f64_t _ZGVsMxv_exp (__sv_f64_t, __sv_bool_t); __sv_f64_t _ZGVsMxv_exp10 (__sv_f64_t, __sv_bool_t); diff --git a/sysdeps/aarch64/fpu/test-double-advsimd-wrappers.c b/sysdeps/aarch64/fpu/test-double-advsimd-wrappers.c index 5a0cbf743b..6954fe7435 100644 --- a/sysdeps/aarch64/fpu/test-double-advsimd-wrappers.c +++ b/sysdeps/aarch64/fpu/test-double-advsimd-wrappers.c @@ -25,6 +25,7 @@ VPCS_VECTOR_WRAPPER (acos_advsimd, _ZGVnN2v_acos) VPCS_VECTOR_WRAPPER (asin_advsimd, _ZGVnN2v_asin) +VPCS_VECTOR_WRAPPER (atan_advsimd, _ZGVnN2v_atan) VPCS_VECTOR_WRAPPER (cos_advsimd, _ZGVnN2v_cos) VPCS_VECTOR_WRAPPER (exp_advsimd, _ZGVnN2v_exp) VPCS_VECTOR_WRAPPER (exp10_advsimd, _ZGVnN2v_exp10) diff --git a/sysdeps/aarch64/fpu/test-double-sve-wrappers.c b/sysdeps/aarch64/fpu/test-double-sve-wrappers.c index bd89ff6133..1173d8f9ae 100644 --- a/sysdeps/aarch64/fpu/test-double-sve-wrappers.c +++ b/sysdeps/aarch64/fpu/test-double-sve-wrappers.c @@ -34,6 +34,7 @@ SVE_VECTOR_WRAPPER (acos_sve, _ZGVsMxv_acos) SVE_VECTOR_WRAPPER (asin_sve, _ZGVsMxv_asin) +SVE_VECTOR_WRAPPER (atan_sve, _ZGVsMxv_atan) SVE_VECTOR_WRAPPER (cos_sve, _ZGVsMxv_cos) SVE_VECTOR_WRAPPER (exp_sve, _ZGVsMxv_exp) SVE_VECTOR_WRAPPER (exp10_sve, _ZGVsMxv_exp10) diff --git a/sysdeps/aarch64/fpu/test-float-advsimd-wrappers.c b/sysdeps/aarch64/fpu/test-float-advsimd-wrappers.c index 3fafca7557..387efc30f8 100644 --- a/sysdeps/aarch64/fpu/test-float-advsimd-wrappers.c +++ b/sysdeps/aarch64/fpu/test-float-advsimd-wrappers.c @@ -25,6 +25,7 @@ VPCS_VECTOR_WRAPPER (acosf_advsimd, _ZGVnN4v_acosf) VPCS_VECTOR_WRAPPER (asinf_advsimd, _ZGVnN4v_asinf) +VPCS_VECTOR_WRAPPER (atanf_advsimd, _ZGVnN4v_atanf) VPCS_VECTOR_WRAPPER (cosf_advsimd, _ZGVnN4v_cosf) VPCS_VECTOR_WRAPPER (expf_advsimd, _ZGVnN4v_expf) VPCS_VECTOR_WRAPPER (exp10f_advsimd, _ZGVnN4v_exp10f) diff --git a/sysdeps/aarch64/fpu/test-float-sve-wrappers.c b/sysdeps/aarch64/fpu/test-float-sve-wrappers.c index b4ec9f777b..dddd4cb213 100644 --- a/sysdeps/aarch64/fpu/test-float-sve-wrappers.c +++ b/sysdeps/aarch64/fpu/test-float-sve-wrappers.c @@ -34,6 +34,7 @@ SVE_VECTOR_WRAPPER (acosf_sve, _ZGVsMxv_acosf) SVE_VECTOR_WRAPPER (asinf_sve, _ZGVsMxv_asinf) +SVE_VECTOR_WRAPPER (atanf_sve, _ZGVsMxv_atanf) SVE_VECTOR_WRAPPER (cosf_sve, _ZGVsMxv_cosf) SVE_VECTOR_WRAPPER (expf_sve, _ZGVsMxv_expf) SVE_VECTOR_WRAPPER (exp10f_sve, _ZGVsMxv_exp10f) diff --git a/sysdeps/aarch64/libm-test-ulps b/sysdeps/aarch64/libm-test-ulps index c2b6f21b9d..24a99e10da 100644 --- a/sysdeps/aarch64/libm-test-ulps +++ b/sysdeps/aarch64/libm-test-ulps @@ -121,11 +121,19 @@ double: 1 float: 1 ldouble: 2 +Function: "atan_advsimd": +double: 1 +float: 1 + Function: "atan_downward": double: 1 float: 2 ldouble: 2 +Function: "atan_sve": +double: 1 +float: 1 + Function: "atan_towardzero": double: 1 float: 1 diff --git a/sysdeps/unix/sysv/linux/aarch64/libmvec.abilist b/sysdeps/unix/sysv/linux/aarch64/libmvec.abilist index f79eaaf241..a2d1b8fb6d 100644 --- a/sysdeps/unix/sysv/linux/aarch64/libmvec.abilist +++ b/sysdeps/unix/sysv/linux/aarch64/libmvec.abilist @@ -16,6 +16,7 @@ GLIBC_2.38 _ZGVsMxv_sin F GLIBC_2.38 _ZGVsMxv_sinf F GLIBC_2.39 _ZGVnN2v_acos F GLIBC_2.39 _ZGVnN2v_asin F +GLIBC_2.39 _ZGVnN2v_atan F GLIBC_2.39 _ZGVnN2v_exp10 F GLIBC_2.39 _ZGVnN2v_exp2 F GLIBC_2.39 _ZGVnN2v_log10 F @@ -23,6 +24,7 @@ GLIBC_2.39 _ZGVnN2v_log2 F GLIBC_2.39 _ZGVnN2v_tan F GLIBC_2.39 _ZGVnN4v_acosf F GLIBC_2.39 _ZGVnN4v_asinf F +GLIBC_2.39 _ZGVnN4v_atanf F GLIBC_2.39 _ZGVnN4v_exp10f F GLIBC_2.39 _ZGVnN4v_exp2f F GLIBC_2.39 _ZGVnN4v_log10f F @@ -32,6 +34,8 @@ GLIBC_2.39 _ZGVsMxv_acos F GLIBC_2.39 _ZGVsMxv_acosf F GLIBC_2.39 _ZGVsMxv_asin F GLIBC_2.39 _ZGVsMxv_asinf F +GLIBC_2.39 _ZGVsMxv_atan F +GLIBC_2.39 _ZGVsMxv_atanf F GLIBC_2.39 _ZGVsMxv_exp10 F GLIBC_2.39 _ZGVsMxv_exp10f F GLIBC_2.39 _ZGVsMxv_exp2 F