From patchwork Fri Nov 18 05:45:00 2016 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Jerin Jacob X-Patchwork-Id: 17081 Return-Path: X-Original-To: patchwork@dpdk.org Delivered-To: patchwork@dpdk.org Received: from [92.243.14.124] (localhost [IPv6:::1]) by dpdk.org (Postfix) with ESMTP id BAA555698; Fri, 18 Nov 2016 06:46:22 +0100 (CET) Received: from NAM01-BY2-obe.outbound.protection.outlook.com (mail-by2nam01on0057.outbound.protection.outlook.com [104.47.34.57]) by dpdk.org (Postfix) with ESMTP id EF6295681 for ; Fri, 18 Nov 2016 06:45:41 +0100 (CET) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=CAVIUMNETWORKS.onmicrosoft.com; s=selector1-cavium-com; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version; bh=PIqiPTLcykDhS8DwGUlUW84ZZ8OtGlzmpcH5NvlVbqc=; b=DYTKFwYGlrH1ydDtt8lrZDEd0KiUAON8rG+fV4NDOLBz414DlbvPd/NuxY06mkZdajIKYsRwJcde6t480MnZIQAE3cOfbVzb1bR/J4tvjcpYSqe2OWDBgfLGcR6nvad1VN996AQ4E7r0dh0v5n/Huu/D7md/QPSv+EMsttXaVQ8= Authentication-Results: spf=none (sender IP is ) smtp.mailfrom=Jerin.Jacob@cavium.com; Received: from localhost.net (50.254.132.37) by BY1PR0701MB1724.namprd07.prod.outlook.com (10.162.111.143) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA384_P384) id 15.1.693.12; Fri, 18 Nov 2016 05:45:32 +0000 From: Jerin Jacob To: CC: , , , , Jerin Jacob Date: Fri, 18 Nov 2016 11:15:00 +0530 Message-ID: <1479447902-3700-3-git-send-email-jerin.jacob@caviumnetworks.com> X-Mailer: git-send-email 2.5.5 In-Reply-To: <1479447902-3700-1-git-send-email-jerin.jacob@caviumnetworks.com> References: <1479447902-3700-1-git-send-email-jerin.jacob@caviumnetworks.com> MIME-Version: 1.0 X-Originating-IP: [50.254.132.37] X-ClientProxiedBy: CY1PR1101CA0016.namprd11.prod.outlook.com (10.169.17.26) To BY1PR0701MB1724.namprd07.prod.outlook.com (10.162.111.143) X-Microsoft-Exchange-Diagnostics: 1; BY1PR0701MB1724; 2:/M5ihsUghm1lEpoe8W2+PiCmEKTJx79voEmEt1aZHC32H4HuYlELW6gDacU9LWlKRJbRd18fqrr6HGo+BPdFrhD7bh609apFmI8sP9V9oSO09z/T3Qw6sC8+Ydl2Vxp1vPXgGwPF5JfdhQYpTZJtza1dsaKDCo0t9usP9/swZoA=; 3:BuKCRR3oEKWRetnog/6Wm428Zk2ENXQltR+71aSYjJpTqfOhxA3LXcU064HnwrFbVSdVpidDhHNgHdMcPv0bqTu6XWHRJQKrL7cQNGvEVpqG1p6BiRnIks5AOKAr+/yfKnCQrfQVVSptadg3RAg24pxqE4RC6VK3A2tkoQL0WUU=; 25:rlYFi++db2gA29iRHVzq7w5mpvvdTNpk9oKZUEgPypR2Ggg1vwi6O+XmqNA8MixikR67BbgvLZcdJkIr+H16ssi2NcdzJLcm6bocMLt9SU6yek7pzB5dvGnr4tp6i3jThjUpJoeMdT7LwqK4esdA2I4aCAZHDLNdVZt+9dz6zxRvdVfzZ7Xrx2VaWJRMFFLRx41dhXkA7I3tTVFf6cDEUGx52tnAl8Twn1cwe38mmAnEHbVY8BfMOpY68VC7FjAr1uJnZuu8hcaZfQod9qb2GWp44BW+/hvEoVSko3gD7mlE4XGh/Pxw4koqbixol6biUsBKzJMOqmZD/dyLHS+6+Zgm/1hDFlys0oe0Z7IUF0XVProWWLgVzFIsa77FFn/8XGwrGvlY0A2ZNa0u6yMyLlt2kmktNkleCSLgzabyWGtziTvQs/EdZ6xSa+dNVy6XJyy+W4a3XR7w8qeVWSZ5tw== X-MS-Office365-Filtering-Correlation-Id: f899f9b1-6ea9-4805-1e3d-08d40f761cba X-Microsoft-Antispam: UriScan:; BCL:0; PCL:0; RULEID:(22001); SRVR:BY1PR0701MB1724; X-Microsoft-Exchange-Diagnostics: 1; BY1PR0701MB1724; 31:5AJwBFmC4SbY4jpXZWw7YOxWicRCMyzzik+AjOy+cCTcvP+ir7xoC+vQa9cfihLF18Q+SxChEk0Ezksqy8Ov2iI0Xbg31Yw0vArHFhrmOmjVL/ult8HN3bMZG3zzz+aeGo4E8cKU28/z/dieDVB2v9Ufk+iZ6n6cNyWjgtI/xAi2luAep3uuBx2TI0nTSwVMVg67eeKqHiVtTrD7ePPSpBsm4f+mNnV0hpdIwMh/MSr5hBC0t9DiFLqVkp9GPd1QyGzrbesExNyP0dxItBW3Cg==; 20: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 X-Microsoft-Antispam-PRVS: X-Exchange-Antispam-Report-Test: UriScan:(278428928389397); X-Exchange-Antispam-Report-CFA-Test: BCL:0; PCL:0; RULEID:(6060326)(6040281)(601004)(2401047)(8121501046)(5005006)(10201501046)(3002001)(6061324)(6041223); SRVR:BY1PR0701MB1724; BCL:0; PCL:0; RULEID:; SRVR:BY1PR0701MB1724; X-Microsoft-Exchange-Diagnostics: 1; BY1PR0701MB1724; 4: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 X-Forefront-PRVS: 01304918F3 X-Forefront-Antispam-Report: SFV:NSPM; SFS:(10009020)(4630300001)(6069001)(6009001)(7916002)(189002)(199003)(68736007)(66066001)(3846002)(7736002)(50226002)(69596002)(81166006)(47776003)(4326007)(4001430100002)(48376002)(6116002)(8676002)(5890100001)(81156014)(8666005)(42186005)(33026002)(105586002)(2351001)(92566002)(2906002)(7846002)(50466002)(50986999)(110136003)(5660300001)(76176999)(42882006)(33646002)(6916009)(189998001)(21086003)(107886002)(36756003)(305945005)(2950100002)(6666003)(101416001)(575784001)(97736004)(77096005)(5003940100001)(106356001)(76506005)(7059030)(579004)(559001); DIR:OUT; SFP:1101; SCL:1; SRVR:BY1PR0701MB1724; H:localhost.net; FPR:; SPF:None; PTR:InfoNoRecords; MX:1; A:1; LANG:en; Received-SPF: None (protection.outlook.com: cavium.com does not designate permitted sender hosts) X-Microsoft-Exchange-Diagnostics: =?us-ascii?Q?1; BY1PR0701MB1724; 23:f9b02Rqob3uW06NLhuZXcSyR4kx+MK/XqB8Pxoz?= =?us-ascii?Q?BzpgH4YXJI4EtSSScTL9fny4PnSrrydV5V8DmNtafA7w98NuDM35VKlAV8f+?= =?us-ascii?Q?tTNpGMGe7uscDaNzUEqk4yCrvcORgEYxy024WGfG+Enn2nROSwAiGcqIU3PJ?= =?us-ascii?Q?jB8oCJWkSM2oYClLAwm1mGacue4GSxPLfVLaHzWXBMTNzHlf+LxPtfrmmc6e?= =?us-ascii?Q?ZOKcmScycZtFQ61V7KZoCiTVbw3LpzFNltrnSLBDCh5eHxv4GA6cG2KcH6EQ?= =?us-ascii?Q?5nq0HcJVttc7KadcUVm/zPk2LaP6oQ7NHkY1P5qRb/Ni+LyYGKPuK1kbOFo5?= =?us-ascii?Q?IHxBWkUdUBaNvfI5h42Fik1HUuCEoA4AtpuQ0QsQbmuomK1VcMJze/u5MJ8l?= =?us-ascii?Q?0wFHzXs3xD0scfcwgWb0KL6/dM2VT1PDxDMyRUwS5bi18MBJBYWhsAqAn0lQ?= =?us-ascii?Q?Vl/O2cDc3lNNIfydxM0T+AvjM9PWwcx2a1N/zFDnS45yaSd/Gi42vOxX11WP?= =?us-ascii?Q?0eSSJ4ulfXID0HrpadVXm4I/BRi6Bkr+o98jdZcPMcvnBFZS9Cu55efw7s4p?= =?us-ascii?Q?lvebmUwpLePzzR7RYuScKkiR4y+ShA3DI+6dQL5EcbuYC3gVR2zY3vKJr8Ax?= =?us-ascii?Q?vxxEICWla5OAM3x1mzO044PgOqJE++h1j86xq0jaHJO0PBZaoygFlPQAmfBV?= =?us-ascii?Q?1h/AtiZTNyZlnfwqiOb7fT7FSYUV7LyIpyQvays13XAdhxRy6Zn5zWvp4hLm?= =?us-ascii?Q?GFENwXDRyc5Y6YiPT1A/Nn6DY8QM6L69Hjg2yeF4ZVdERyta82jP4OD1UxaV?= =?us-ascii?Q?4hUc1LA/GkBxjsbBLhD9MtVJpHWZN5bW8bO9BM0VajrH+KvuprqKDfScG8DN?= =?us-ascii?Q?v/ABqw2OOPSaGgIAEXig/GsAguQsfjdMwGElxzOu5ra5uRqSdIMrsHEsKyt4?= =?us-ascii?Q?/aEWKMzQHVaB8eFh1cmIYZrwex7tPHJHof6Wt6zJ5VTW6X0K7UoUqqJbDbCK?= =?us-ascii?Q?/HWArukPzS1L6/62PqQfl5Hdec0nMhmofx3RN7lSgEvHiInCSKBf7Bvu21A/?= =?us-ascii?Q?TqDT74eGLkSYQizfHd8Ne325t/yLmreXZcLU4uExKnRwplW2T9IpgCrUYCC0?= =?us-ascii?Q?Hfc+EXXMD10do+nlLfEPm0hNoQ7mbY6QiEs0G3vr17hJiBMkHz8mNCop51oj?= =?us-ascii?Q?fD4Vc6izmO4+pa7lRhX10lSLSt6WxguQH4deFoiBGaAzt3qi6MaQqHGUoyVS?= =?us-ascii?Q?WrfiCuUS6NukyHEea1BPvGNffEhdDw9a2hyrPXbbnVQI8Z/3N+9HSNNuRc3S?= =?us-ascii?Q?djm7SWtVqCC/ubIMELHzlUz4=3D?= X-Microsoft-Exchange-Diagnostics: 1; BY1PR0701MB1724; 6:fiA+AZhA6eQjcQd0o7fkI3AA6Fo+KYFT4ebD8dUZJ1tPlSWowIpacdXY+E9rY0BGF6+0D7DN7I+ZeUlg5dFewQJqOx7tH57QF0CjAaUJc2+upstamvOCX2OzkVV2ia1btfwFDoNMAchlkHXK46GPK5vp0qZYca/vBfwCvhKk5TUB780Dg7cVvMgDx7Ko7VTFUCkhtP9WZFTTUK85yxuMYQdiF0VoNee6FqgxYI9PWa/0dEOtQvqxJZlRwMetaV3X/NFkgw2XL3/wvb8m+g7nV0skKcNftb703HRKKBq4x19yYMVjB9QAsajpt/mLobs1JZpq8t3kuo/r/9Qww9mK0swnukp3VpJ47c40ZM9Y23o=; 5:Rhxibqvy6TXZ8JaaGCnbvSvJCwX7zg5kfpvRuDL2V3AMQRt7PFdZR+FD80+WlC/oOuuhLAsVSakfmcuvqx3Bf+jKAkFOm/9ny2pJlwaVmuUvNDFrXGVt1gHgJWiw83chNXeOlSk3aUm/0DcnQWmjng==; 24:oWL0eccdVdX/YeWnrk/j6HghHI9OWgn/TNXeWrvLdv8j3+02qG+LYAv2RnNPYI8RYOUnQIPct7R72xRr7yTQgbivcM8EXPpTtwf5TQj2DHM= SpamDiagnosticOutput: 1:99 SpamDiagnosticMetadata: NSPM X-Microsoft-Exchange-Diagnostics: 1; BY1PR0701MB1724; 7:ELR3kVW6oiXycALTdav11yJngbEOSBkBue+uyNWLj/2B2uDt4qvCF61D/AXNYV+mts05fy322j20aq8LgG/4DAWI6LxAFHPUMP4McMtfNrgacAKHtpKfzqHO07hXuRRxZoomPKRfa5x8LQnixNzUeggACwJnb6whvtQwCEIXru3Oh16UQvYOMzNxOC6ELrsSGZJiU59jMefZUuFjgJcfQUuY4jCJGrHGSEhRUtyp7RB3DEskmtz1Zx8Q1diJQCuTQcKDFUq6HwYzk2qtYgajhbflqecl/x/nPbP0Vts24j7qgkUSFgY8vgfceiuyCCOI818YRQ9kJZ+ymhKkYraMSHlKmKTdktF4COVL+9k1h9E= X-OriginatorOrg: caviumnetworks.com X-MS-Exchange-CrossTenant-OriginalArrivalTime: 18 Nov 2016 05:45:32.2624 (UTC) X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted X-MS-Exchange-Transport-CrossTenantHeadersStamped: BY1PR0701MB1724 Subject: [dpdk-dev] [PATCH 2/4] eventdev: implement the northbound APIs X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: patches and discussions about DPDK List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" This patch set defines the southbound driver interface and implements the common code required for northbound eventdev API interface. Signed-off-by: Jerin Jacob --- config/common_base | 6 + lib/Makefile | 1 + lib/librte_eal/common/include/rte_log.h | 1 + lib/librte_eventdev/Makefile | 57 ++ lib/librte_eventdev/rte_eventdev.c | 1211 ++++++++++++++++++++++++++ lib/librte_eventdev/rte_eventdev_pmd.h | 504 +++++++++++ lib/librte_eventdev/rte_eventdev_version.map | 39 + mk/rte.app.mk | 1 + 8 files changed, 1820 insertions(+) create mode 100644 lib/librte_eventdev/Makefile create mode 100644 lib/librte_eventdev/rte_eventdev.c create mode 100644 lib/librte_eventdev/rte_eventdev_pmd.h create mode 100644 lib/librte_eventdev/rte_eventdev_version.map diff --git a/config/common_base b/config/common_base index 4bff83a..7a8814e 100644 --- a/config/common_base +++ b/config/common_base @@ -411,6 +411,12 @@ CONFIG_RTE_LIBRTE_PMD_ZUC_DEBUG=n CONFIG_RTE_LIBRTE_PMD_NULL_CRYPTO=y # +# Compile generic event device library +# +CONFIG_RTE_LIBRTE_EVENTDEV=y +CONFIG_RTE_LIBRTE_EVENTDEV_DEBUG=n +CONFIG_RTE_EVENT_MAX_DEVS=16 +CONFIG_RTE_EVENT_MAX_QUEUES_PER_DEV=64 # Compile librte_ring # CONFIG_RTE_LIBRTE_RING=y diff --git a/lib/Makefile b/lib/Makefile index 990f23a..1a067bf 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -41,6 +41,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_CFGFILE) += librte_cfgfile DIRS-$(CONFIG_RTE_LIBRTE_CMDLINE) += librte_cmdline DIRS-$(CONFIG_RTE_LIBRTE_ETHER) += librte_ether DIRS-$(CONFIG_RTE_LIBRTE_CRYPTODEV) += librte_cryptodev +DIRS-$(CONFIG_RTE_LIBRTE_EVENTDEV) += librte_eventdev DIRS-$(CONFIG_RTE_LIBRTE_VHOST) += librte_vhost DIRS-$(CONFIG_RTE_LIBRTE_HASH) += librte_hash DIRS-$(CONFIG_RTE_LIBRTE_LPM) += librte_lpm diff --git a/lib/librte_eal/common/include/rte_log.h b/lib/librte_eal/common/include/rte_log.h index 29f7d19..9a07d92 100644 --- a/lib/librte_eal/common/include/rte_log.h +++ b/lib/librte_eal/common/include/rte_log.h @@ -79,6 +79,7 @@ extern struct rte_logs rte_logs; #define RTE_LOGTYPE_PIPELINE 0x00008000 /**< Log related to pipeline. */ #define RTE_LOGTYPE_MBUF 0x00010000 /**< Log related to mbuf. */ #define RTE_LOGTYPE_CRYPTODEV 0x00020000 /**< Log related to cryptodev. */ +#define RTE_LOGTYPE_EVENTDEV 0x00040000 /**< Log related to eventdev. */ /* these log types can be used in an application */ #define RTE_LOGTYPE_USER1 0x01000000 /**< User-defined log type 1. */ diff --git a/lib/librte_eventdev/Makefile b/lib/librte_eventdev/Makefile new file mode 100644 index 0000000..dac0663 --- /dev/null +++ b/lib/librte_eventdev/Makefile @@ -0,0 +1,57 @@ +# BSD LICENSE +# +# Copyright(c) 2016 Cavium networks. All rights reserved. +# +# Redistribution and use in source and binary forms, with or without +# modification, are permitted provided that the following conditions +# are met: +# +# * Redistributions of source code must retain the above copyright +# notice, this list of conditions and the following disclaimer. +# * Redistributions in binary form must reproduce the above copyright +# notice, this list of conditions and the following disclaimer in +# the documentation and/or other materials provided with the +# distribution. +# * Neither the name of Cavium networks nor the names of its +# contributors may be used to endorse or promote products derived +# from this software without specific prior written permission. +# +# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + +include $(RTE_SDK)/mk/rte.vars.mk + +# library name +LIB = librte_eventdev.a + +# library version +LIBABIVER := 1 + +# build flags +CFLAGS += -O3 +CFLAGS += $(WERROR_FLAGS) + +# library source files +SRCS-y += rte_eventdev.c + +# export include files +SYMLINK-y-include += rte_eventdev.h +SYMLINK-y-include += rte_eventdev_pmd.h + +# versioning export map +EXPORT_MAP := rte_eventdev_version.map + +# library dependencies +DEPDIRS-y += lib/librte_eal +DEPDIRS-y += lib/librte_mbuf + +include $(RTE_SDK)/mk/rte.lib.mk diff --git a/lib/librte_eventdev/rte_eventdev.c b/lib/librte_eventdev/rte_eventdev.c new file mode 100644 index 0000000..17ce5c3 --- /dev/null +++ b/lib/librte_eventdev/rte_eventdev.c @@ -0,0 +1,1211 @@ +/* + * BSD LICENSE + * + * Copyright(c) 2016 Cavium networks. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Cavium networks nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include "rte_eventdev.h" +#include "rte_eventdev_pmd.h" + +struct rte_eventdev rte_event_devices[RTE_EVENT_MAX_DEVS]; + +struct rte_eventdev *rte_eventdevs = &rte_event_devices[0]; + +static struct rte_eventdev_global eventdev_globals = { + .nb_devs = 0 +}; + +struct rte_eventdev_global *rte_eventdev_globals = &eventdev_globals; + +/* Event dev north bound API implementation */ + +uint8_t +rte_event_dev_count(void) +{ + return rte_eventdev_globals->nb_devs; +} + +int +rte_event_dev_get_dev_id(const char *name) +{ + int i; + + if (!name) + return -EINVAL; + + for (i = 0; i < rte_eventdev_globals->nb_devs; i++) + if ((strcmp(rte_event_devices[i].data->name, name) + == 0) && + (rte_event_devices[i].attached == + RTE_EVENTDEV_ATTACHED)) + return i; + return -ENODEV; +} + +int +rte_event_dev_socket_id(uint8_t dev_id) +{ + struct rte_eventdev *dev; + + RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL); + dev = &rte_eventdevs[dev_id]; + + return dev->data->socket_id; +} + +int +rte_event_dev_info_get(uint8_t dev_id, struct rte_event_dev_info *dev_info) +{ + struct rte_eventdev *dev; + + RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL); + dev = &rte_eventdevs[dev_id]; + + if (dev_info == NULL) + return -EINVAL; + + memset(dev_info, 0, sizeof(struct rte_event_dev_info)); + + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->dev_infos_get, -ENOTSUP); + (*dev->dev_ops->dev_infos_get)(dev, dev_info); + + dev_info->pci_dev = dev->pci_dev; + if (dev->driver) + dev_info->driver_name = dev->driver->pci_drv.driver.name; + return 0; +} + +static inline int +rte_event_dev_queue_config(struct rte_eventdev *dev, uint8_t nb_queues) +{ + uint8_t old_nb_queues = dev->data->nb_queues; + void **queues; + uint8_t *queues_prio; + unsigned int i; + + EDEV_LOG_DEBUG("Setup %d queues on device %u", nb_queues, + dev->data->dev_id); + + /* First time configuration */ + if (dev->data->queues == NULL && nb_queues != 0) { + dev->data->queues = rte_zmalloc_socket("eventdev->data->queues", + sizeof(dev->data->queues[0]) * nb_queues, + RTE_CACHE_LINE_SIZE, dev->data->socket_id); + if (dev->data->queues == NULL) { + dev->data->nb_queues = 0; + EDEV_LOG_ERR("failed to get memory for queue meta data," + "nb_queues %u", nb_queues); + return -(ENOMEM); + } + /* Allocate memory to store queue priority */ + dev->data->queues_prio = rte_zmalloc_socket( + "eventdev->data->queues_prio", + sizeof(dev->data->queues_prio[0]) * nb_queues, + RTE_CACHE_LINE_SIZE, dev->data->socket_id); + if (dev->data->queues_prio == NULL) { + dev->data->nb_queues = 0; + EDEV_LOG_ERR("failed to get memory for queue priority," + "nb_queues %u", nb_queues); + return -(ENOMEM); + } + + } else if (dev->data->queues != NULL && nb_queues != 0) {/* re-config */ + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->queue_release, -ENOTSUP); + + queues = dev->data->queues; + for (i = nb_queues; i < old_nb_queues; i++) + (*dev->dev_ops->queue_release)(queues[i]); + + queues = rte_realloc(queues, sizeof(queues[0]) * nb_queues, + RTE_CACHE_LINE_SIZE); + if (queues == NULL) { + EDEV_LOG_ERR("failed to realloc queue meta data," + " nb_queues %u", nb_queues); + return -(ENOMEM); + } + dev->data->queues = queues; + + /* Re allocate memory to store queue priority */ + queues_prio = dev->data->queues_prio; + queues_prio = rte_realloc(queues_prio, + sizeof(queues_prio[0]) * nb_queues, + RTE_CACHE_LINE_SIZE); + if (queues_prio == NULL) { + EDEV_LOG_ERR("failed to realloc queue priority," + " nb_queues %u", nb_queues); + return -(ENOMEM); + } + dev->data->queues_prio = queues_prio; + + if (nb_queues > old_nb_queues) { + uint8_t new_qs = nb_queues - old_nb_queues; + + memset(queues + old_nb_queues, 0, + sizeof(queues[0]) * new_qs); + memset(queues_prio + old_nb_queues, 0, + sizeof(queues_prio[0]) * new_qs); + } + } else if (dev->data->queues != NULL && nb_queues == 0) { + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->queue_release, -ENOTSUP); + + queues = dev->data->queues; + for (i = nb_queues; i < old_nb_queues; i++) + (*dev->dev_ops->queue_release)(queues[i]); + } + + dev->data->nb_queues = nb_queues; + return 0; +} + +static inline int +rte_event_dev_port_config(struct rte_eventdev *dev, uint8_t nb_ports) +{ + uint8_t old_nb_ports = dev->data->nb_ports; + void **ports; + uint16_t *links_map; + uint8_t *ports_dequeue_depth; + uint8_t *ports_enqueue_depth; + unsigned int i; + + EDEV_LOG_DEBUG("Setup %d ports on device %u", nb_ports, + dev->data->dev_id); + + /* First time configuration */ + if (dev->data->ports == NULL && nb_ports != 0) { + dev->data->ports = rte_zmalloc_socket("eventdev->data->ports", + sizeof(dev->data->ports[0]) * nb_ports, + RTE_CACHE_LINE_SIZE, dev->data->socket_id); + if (dev->data->ports == NULL) { + dev->data->nb_ports = 0; + EDEV_LOG_ERR("failed to get memory for port meta data," + "nb_ports %u", nb_ports); + return -(ENOMEM); + } + + /* Allocate memory to store ports dequeue depth */ + dev->data->ports_dequeue_depth = + rte_zmalloc_socket("eventdev->ports_dequeue_depth", + sizeof(dev->data->ports_dequeue_depth[0]) * nb_ports, + RTE_CACHE_LINE_SIZE, dev->data->socket_id); + if (dev->data->ports_dequeue_depth == NULL) { + dev->data->nb_ports = 0; + EDEV_LOG_ERR("failed to get memory for port deq meta," + "nb_ports %u", nb_ports); + return -(ENOMEM); + } + + /* Allocate memory to store ports enqueue depth */ + dev->data->ports_enqueue_depth = + rte_zmalloc_socket("eventdev->ports_enqueue_depth", + sizeof(dev->data->ports_enqueue_depth[0]) * nb_ports, + RTE_CACHE_LINE_SIZE, dev->data->socket_id); + if (dev->data->ports_enqueue_depth == NULL) { + dev->data->nb_ports = 0; + EDEV_LOG_ERR("failed to get memory for port enq meta," + "nb_ports %u", nb_ports); + return -(ENOMEM); + } + + /* Allocate memory to store queue to port link connection */ + dev->data->links_map = + rte_zmalloc_socket("eventdev->links_map", + sizeof(dev->data->links_map[0]) * nb_ports * + RTE_EVENT_MAX_QUEUES_PER_DEV, + RTE_CACHE_LINE_SIZE, dev->data->socket_id); + if (dev->data->links_map == NULL) { + dev->data->nb_ports = 0; + EDEV_LOG_ERR("failed to get memory for port_map area," + "nb_ports %u", nb_ports); + return -(ENOMEM); + } + } else if (dev->data->ports != NULL && nb_ports != 0) {/* re-config */ + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->port_release, -ENOTSUP); + + ports = dev->data->ports; + ports_dequeue_depth = dev->data->ports_dequeue_depth; + ports_enqueue_depth = dev->data->ports_enqueue_depth; + links_map = dev->data->links_map; + + for (i = nb_ports; i < old_nb_ports; i++) + (*dev->dev_ops->port_release)(ports[i]); + + /* Realloc memory for ports */ + ports = rte_realloc(ports, sizeof(ports[0]) * nb_ports, + RTE_CACHE_LINE_SIZE); + if (ports == NULL) { + EDEV_LOG_ERR("failed to realloc port meta data," + " nb_ports %u", nb_ports); + return -(ENOMEM); + } + + /* Realloc memory for ports_dequeue_depth */ + ports_dequeue_depth = rte_realloc(ports_dequeue_depth, + sizeof(ports_dequeue_depth[0]) * nb_ports, + RTE_CACHE_LINE_SIZE); + if (ports_dequeue_depth == NULL) { + EDEV_LOG_ERR("failed to realloc port deqeue meta data," + " nb_ports %u", nb_ports); + return -(ENOMEM); + } + + /* Realloc memory for ports_enqueue_depth */ + ports_enqueue_depth = rte_realloc(ports_enqueue_depth, + sizeof(ports_enqueue_depth[0]) * nb_ports, + RTE_CACHE_LINE_SIZE); + if (ports_enqueue_depth == NULL) { + EDEV_LOG_ERR("failed to realloc port enqueue meta data," + " nb_ports %u", nb_ports); + return -(ENOMEM); + } + + /* Realloc memory to store queue to port link connection */ + links_map = rte_realloc(links_map, + sizeof(dev->data->links_map[0]) * nb_ports * + RTE_EVENT_MAX_QUEUES_PER_DEV, + RTE_CACHE_LINE_SIZE); + if (dev->data->links_map == NULL) { + dev->data->nb_ports = 0; + EDEV_LOG_ERR("failed to realloc mem for port_map area," + "nb_ports %u", nb_ports); + return -(ENOMEM); + } + + if (nb_ports > old_nb_ports) { + uint8_t new_ps = nb_ports - old_nb_ports; + + memset(ports + old_nb_ports, 0, + sizeof(ports[0]) * new_ps); + memset(ports_dequeue_depth + old_nb_ports, 0, + sizeof(ports_dequeue_depth[0]) * new_ps); + memset(ports_enqueue_depth + old_nb_ports, 0, + sizeof(ports_enqueue_depth[0]) * new_ps); + memset(links_map + + (old_nb_ports * RTE_EVENT_MAX_QUEUES_PER_DEV), + 0, sizeof(ports_enqueue_depth[0]) * new_ps); + } + + dev->data->ports = ports; + dev->data->ports_dequeue_depth = ports_dequeue_depth; + dev->data->ports_enqueue_depth = ports_enqueue_depth; + dev->data->links_map = links_map; + } else if (dev->data->ports != NULL && nb_ports == 0) { + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->port_release, -ENOTSUP); + + ports = dev->data->ports; + for (i = nb_ports; i < old_nb_ports; i++) + (*dev->dev_ops->port_release)(ports[i]); + } + + dev->data->nb_ports = nb_ports; + return 0; +} + +int +rte_event_dev_configure(uint8_t dev_id, struct rte_event_dev_config *dev_conf) +{ + struct rte_eventdev *dev; + struct rte_event_dev_info info; + int diag; + + RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL); + dev = &rte_eventdevs[dev_id]; + + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->dev_infos_get, -ENOTSUP); + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->dev_configure, -ENOTSUP); + + if (dev->data->dev_started) { + EDEV_LOG_ERR( + "device %d must be stopped to allow configuration", dev_id); + return -EBUSY; + } + + if (dev_conf == NULL) + return -EINVAL; + + (*dev->dev_ops->dev_infos_get)(dev, &info); + + /* Check dequeue_wait_ns value is in limit */ + if (!dev_conf->event_dev_cfg & RTE_EVENT_DEV_CFG_PER_DEQUEUE_WAIT) { + if (dev_conf->dequeue_wait_ns < info.min_dequeue_wait_ns || + dev_conf->dequeue_wait_ns > info.max_dequeue_wait_ns) { + EDEV_LOG_ERR("dev%d invalid dequeue_wait_ns=%d" + " min_dequeue_wait_ns=%d max_dequeue_wait_ns=%d", + dev_id, dev_conf->dequeue_wait_ns, + info.min_dequeue_wait_ns, + info.max_dequeue_wait_ns); + return -EINVAL; + } + } + + /* Check nb_events_limit is in limit */ + if (dev_conf->nb_events_limit > info.max_num_events) { + EDEV_LOG_ERR("dev%d nb_events_limit=%d > max_num_events=%d", + dev_id, dev_conf->nb_events_limit, info.max_num_events); + return -EINVAL; + } + + /* Check nb_event_queues is in limit */ + if (!dev_conf->nb_event_queues) { + EDEV_LOG_ERR("dev%d nb_event_queues cannot be zero", dev_id); + return -EINVAL; + } + if (dev_conf->nb_event_queues > info.max_event_queues) { + EDEV_LOG_ERR("dev%d nb_event_queues=%d > max_event_queues=%d", + dev_id, dev_conf->nb_event_queues, info.max_event_queues); + return -EINVAL; + } + + /* Check nb_event_ports is in limit */ + if (!dev_conf->nb_event_ports) { + EDEV_LOG_ERR("dev%d nb_event_ports cannot be zero", dev_id); + return -EINVAL; + } + if (dev_conf->nb_event_ports > info.max_event_ports) { + EDEV_LOG_ERR("dev%d nb_event_ports=%d > max_event_ports= %d", + dev_id, dev_conf->nb_event_ports, info.max_event_ports); + return -EINVAL; + } + + /* Check nb_event_queue_flows is in limit */ + if (!dev_conf->nb_event_queue_flows) { + EDEV_LOG_ERR("dev%d nb_flows cannot be zero", dev_id); + return -EINVAL; + } + if (dev_conf->nb_event_queue_flows > info.max_event_queue_flows) { + EDEV_LOG_ERR("dev%d nb_flows=%x > max_flows=%x", + dev_id, dev_conf->nb_event_queue_flows, + info.max_event_queue_flows); + return -EINVAL; + } + + /* Check nb_event_port_dequeue_depth is in limit */ + if (!dev_conf->nb_event_port_dequeue_depth) { + EDEV_LOG_ERR("dev%d nb_dequeue_depth cannot be zero", dev_id); + return -EINVAL; + } + if (dev_conf->nb_event_port_dequeue_depth > + info.max_event_port_dequeue_depth) { + EDEV_LOG_ERR("dev%d nb_dequeue_depth=%d > max_dequeue_depth=%d", + dev_id, dev_conf->nb_event_port_dequeue_depth, + info.max_event_port_dequeue_depth); + return -EINVAL; + } + + /* Check nb_event_port_enqueue_depth is in limit */ + if (!dev_conf->nb_event_port_enqueue_depth) { + EDEV_LOG_ERR("dev%d nb_enqueue_depth cannot be zero", dev_id); + return -EINVAL; + } + if (dev_conf->nb_event_port_enqueue_depth > + info.max_event_port_enqueue_depth) { + EDEV_LOG_ERR("dev%d nb_enqueue_depth=%d > max_enqueue_depth=%d", + dev_id, dev_conf->nb_event_port_enqueue_depth, + info.max_event_port_enqueue_depth); + return -EINVAL; + } + + /* Copy the dev_conf parameter into the dev structure */ + memcpy(&dev->data->dev_conf, dev_conf, sizeof(dev->data->dev_conf)); + + /* Setup new number of queues and reconfigure device. */ + diag = rte_event_dev_queue_config(dev, dev_conf->nb_event_queues); + if (diag != 0) { + EDEV_LOG_ERR("dev%d rte_event_dev_queue_config = %d", + dev_id, diag); + return diag; + } + + /* Setup new number of ports and reconfigure device. */ + diag = rte_event_dev_port_config(dev, dev_conf->nb_event_ports); + if (diag != 0) { + rte_event_dev_queue_config(dev, 0); + EDEV_LOG_ERR("dev%d rte_event_dev_port_config = %d", + dev_id, diag); + return diag; + } + + /* Configure the device */ + diag = (*dev->dev_ops->dev_configure)(dev); + if (diag != 0) { + EDEV_LOG_ERR("dev%d dev_configure = %d", dev_id, diag); + rte_event_dev_queue_config(dev, 0); + rte_event_dev_port_config(dev, 0); + } + + dev->data->event_dev_cap = info.event_dev_cap; + return diag; +} + +static inline int +is_valid_queue(struct rte_eventdev *dev, uint8_t queue_id) +{ + if (queue_id < dev->data->nb_queues && queue_id < + RTE_EVENT_MAX_QUEUES_PER_DEV) + return 1; + else + return 0; +} + +int +rte_event_queue_default_conf_get(uint8_t dev_id, uint8_t queue_id, + struct rte_event_queue_conf *queue_conf) +{ + struct rte_eventdev *dev; + + RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL); + dev = &rte_eventdevs[dev_id]; + + if (queue_conf == NULL) + return -EINVAL; + + if (!is_valid_queue(dev, queue_id)) { + EDEV_LOG_ERR("Invalid queue_id=%" PRIu8, queue_id); + return -EINVAL; + } + + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->queue_def_conf, -ENOTSUP); + memset(queue_conf, 0, sizeof(struct rte_event_queue_conf)); + (*dev->dev_ops->queue_def_conf)(dev, queue_id, queue_conf); + return 0; +} + +static inline int +is_valid_atomic_queue_conf(struct rte_event_queue_conf *queue_conf) +{ + if (queue_conf && ( + ((queue_conf->event_queue_cfg & RTE_EVENT_QUEUE_CFG_TYPE_MASK) + == RTE_EVENT_QUEUE_CFG_ALL_TYPES) || + ((queue_conf->event_queue_cfg & RTE_EVENT_QUEUE_CFG_TYPE_MASK) + == RTE_EVENT_QUEUE_CFG_ATOMIC_ONLY) + )) + return 1; + else + return 0; +} + +int +rte_event_queue_setup(uint8_t dev_id, uint8_t queue_id, + struct rte_event_queue_conf *queue_conf) +{ + struct rte_eventdev *dev; + struct rte_event_queue_conf def_conf; + + RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL); + dev = &rte_eventdevs[dev_id]; + + if (!is_valid_queue(dev, queue_id)) { + EDEV_LOG_ERR("Invalid queue_id=%" PRIu8, queue_id); + return -EINVAL; + } + + /* Check nb_atomic_flows limit */ + if (is_valid_atomic_queue_conf(queue_conf)) { + if (queue_conf->nb_atomic_flows == 0 || + queue_conf->nb_atomic_flows > + dev->data->dev_conf.nb_event_queue_flows) { + EDEV_LOG_ERR( + "dev%d queue%d Invalid nb_atomic_flows=%d max_flows=%d", + dev_id, queue_id, queue_conf->nb_atomic_flows, + dev->data->dev_conf.nb_event_queue_flows); + return -EINVAL; + } + } + + /* Check nb_atomic_order_sequences limit */ + if (is_valid_atomic_queue_conf(queue_conf)) { + if (queue_conf->nb_atomic_order_sequences == 0 || + queue_conf->nb_atomic_order_sequences > + dev->data->dev_conf.nb_event_queue_flows) { + EDEV_LOG_ERR( + "dev%d queue%d Invalid nb_atomic_order_seq=%d max_flows=%d", + dev_id, queue_id, queue_conf->nb_atomic_order_sequences, + dev->data->dev_conf.nb_event_queue_flows); + return -EINVAL; + } + } + + if (dev->data->dev_started) { + EDEV_LOG_ERR( + "device %d must be stopped to allow queue setup", dev_id); + return -EBUSY; + } + + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->queue_setup, -ENOTSUP); + + if (queue_conf == NULL) { + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->queue_def_conf, + -ENOTSUP); + (*dev->dev_ops->queue_def_conf)(dev, queue_id, &def_conf); + def_conf.event_queue_cfg = RTE_EVENT_QUEUE_CFG_DEFAULT; + queue_conf = &def_conf; + } + + dev->data->queues_prio[queue_id] = queue_conf->priority; + return (*dev->dev_ops->queue_setup)(dev, queue_id, queue_conf); +} + +uint8_t +rte_event_queue_count(uint8_t dev_id) +{ + struct rte_eventdev *dev; + + dev = &rte_eventdevs[dev_id]; + return dev->data->nb_queues; +} + +uint8_t +rte_event_queue_priority(uint8_t dev_id, uint8_t queue_id) +{ + struct rte_eventdev *dev; + + dev = &rte_eventdevs[dev_id]; + if (dev->data->event_dev_cap & RTE_EVENT_DEV_CAP_QUEUE_QOS) + return dev->data->queues_prio[queue_id]; + else + return RTE_EVENT_QUEUE_PRIORITY_NORMAL; +} + +static inline int +is_valid_port(struct rte_eventdev *dev, uint8_t port_id) +{ + if (port_id < dev->data->nb_ports) + return 1; + else + return 0; +} + +int +rte_event_port_default_conf_get(uint8_t dev_id, uint8_t port_id, + struct rte_event_port_conf *port_conf) +{ + struct rte_eventdev *dev; + + RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL); + dev = &rte_eventdevs[dev_id]; + + if (port_conf == NULL) + return -EINVAL; + + if (!is_valid_port(dev, port_id)) { + EDEV_LOG_ERR("Invalid port_id=%" PRIu8, port_id); + return -EINVAL; + } + + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->port_def_conf, -ENOTSUP); + memset(port_conf, 0, sizeof(struct rte_event_port_conf)); + (*dev->dev_ops->port_def_conf)(dev, port_id, port_conf); + return 0; +} + +int +rte_event_port_setup(uint8_t dev_id, uint8_t port_id, + struct rte_event_port_conf *port_conf) +{ + struct rte_eventdev *dev; + struct rte_event_port_conf def_conf; + int diag; + + RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL); + dev = &rte_eventdevs[dev_id]; + + if (!is_valid_port(dev, port_id)) { + EDEV_LOG_ERR("Invalid port_id=%" PRIu8, port_id); + return -EINVAL; + } + + /* Check new_event_threshold limit */ + if ((port_conf && !port_conf->new_event_threshold) || + (port_conf && port_conf->new_event_threshold > + dev->data->dev_conf.nb_events_limit)) { + EDEV_LOG_ERR( + "dev%d port%d Invalid event_threshold=%d nb_events_limit=%d", + dev_id, port_id, port_conf->new_event_threshold, + dev->data->dev_conf.nb_events_limit); + return -EINVAL; + } + + /* Check dequeue_depth limit */ + if ((port_conf && !port_conf->dequeue_depth) || + (port_conf && port_conf->dequeue_depth > + dev->data->dev_conf.nb_event_port_dequeue_depth)) { + EDEV_LOG_ERR( + "dev%d port%d Invalid dequeue depth=%d max_dequeue_depth=%d", + dev_id, port_id, port_conf->dequeue_depth, + dev->data->dev_conf.nb_event_port_dequeue_depth); + return -EINVAL; + } + + /* Check enqueue_depth limit */ + if ((port_conf && !port_conf->enqueue_depth) || + (port_conf && port_conf->enqueue_depth > + dev->data->dev_conf.nb_event_port_enqueue_depth)) { + EDEV_LOG_ERR( + "dev%d port%d Invalid enqueue depth=%d max_enqueue_depth=%d", + dev_id, port_id, port_conf->enqueue_depth, + dev->data->dev_conf.nb_event_port_enqueue_depth); + return -EINVAL; + } + + if (dev->data->dev_started) { + EDEV_LOG_ERR( + "device %d must be stopped to allow port setup", dev_id); + return -EBUSY; + } + + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->port_setup, -ENOTSUP); + + if (port_conf == NULL) { + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->port_def_conf, + -ENOTSUP); + (*dev->dev_ops->port_def_conf)(dev, port_id, &def_conf); + port_conf = &def_conf; + } + + dev->data->ports_dequeue_depth[port_id] = + port_conf->dequeue_depth; + dev->data->ports_enqueue_depth[port_id] = + port_conf->enqueue_depth; + + diag = (*dev->dev_ops->port_setup)(dev, port_id, port_conf); + + /* Unlink all the queues from this port(default state after setup) */ + if (!diag) + diag = rte_event_port_unlink(dev_id, port_id, NULL, 0); + + if (diag < 0) + return diag; + + return 0; +} + +uint8_t +rte_event_port_dequeue_depth(uint8_t dev_id, uint8_t port_id) +{ + struct rte_eventdev *dev; + + dev = &rte_eventdevs[dev_id]; + return dev->data->ports_dequeue_depth[port_id]; +} + +uint8_t +rte_event_port_enqueue_depth(uint8_t dev_id, uint8_t port_id) +{ + struct rte_eventdev *dev; + + dev = &rte_eventdevs[dev_id]; + return dev->data->ports_enqueue_depth[port_id]; +} + +uint8_t +rte_event_port_count(uint8_t dev_id) +{ + struct rte_eventdev *dev; + + dev = &rte_eventdevs[dev_id]; + return dev->data->nb_ports; +} + +int +rte_event_port_link(uint8_t dev_id, uint8_t port_id, + struct rte_event_queue_link link[], uint16_t nb_links) +{ + struct rte_eventdev *dev; + struct rte_event_queue_link all_queues[RTE_EVENT_MAX_QUEUES_PER_DEV]; + uint16_t *links_map; + int i, diag; + + RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL); + dev = &rte_eventdevs[dev_id]; + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->port_link, -ENOTSUP); + + if (!is_valid_port(dev, port_id)) { + EDEV_LOG_ERR("Invalid port_id=%" PRIu8, port_id); + return -EINVAL; + } + + if (link == NULL) { + for (i = 0; i < dev->data->nb_queues; i++) { + all_queues[i].queue_id = i; + all_queues[i].priority = + RTE_EVENT_QUEUE_SERVICE_PRIORITY_NORMAL; + } + link = all_queues; + nb_links = dev->data->nb_queues; + } + + for (i = 0; i < nb_links; i++) + if (link[i].queue_id >= RTE_EVENT_MAX_QUEUES_PER_DEV) + return -EINVAL; + + diag = (*dev->dev_ops->port_link)(dev->data->ports[port_id], link, + nb_links); + if (diag < 0) + return diag; + + links_map = dev->data->links_map; + /* Point links_map to this port specific area */ + links_map += (port_id * RTE_EVENT_MAX_QUEUES_PER_DEV); + for (i = 0; i < diag; i++) + links_map[link[i].queue_id] = (uint8_t)link[i].priority; + + return diag; +} + +#define EVENT_QUEUE_SERVICE_PRIORITY_INVALID (0xdead) + +int +rte_event_port_unlink(uint8_t dev_id, uint8_t port_id, + uint8_t queues[], uint16_t nb_unlinks) +{ + struct rte_eventdev *dev; + uint8_t all_queues[RTE_EVENT_MAX_QUEUES_PER_DEV]; + int i, diag; + uint16_t *links_map; + + RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL); + dev = &rte_eventdevs[dev_id]; + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->port_unlink, -ENOTSUP); + + if (!is_valid_port(dev, port_id)) { + EDEV_LOG_ERR("Invalid port_id=%" PRIu8, port_id); + return -EINVAL; + } + + if (queues == NULL) { + for (i = 0; i < dev->data->nb_queues; i++) + all_queues[i] = i; + queues = all_queues; + nb_unlinks = dev->data->nb_queues; + } + + for (i = 0; i < nb_unlinks; i++) + if (queues[i] >= RTE_EVENT_MAX_QUEUES_PER_DEV) + return -EINVAL; + + diag = (*dev->dev_ops->port_unlink)(dev->data->ports[port_id], queues, + nb_unlinks); + + if (diag < 0) + return diag; + + links_map = dev->data->links_map; + /* Point links_map to this port specific area */ + links_map += (port_id * RTE_EVENT_MAX_QUEUES_PER_DEV); + for (i = 0; i < diag; i++) + links_map[queues[i]] = EVENT_QUEUE_SERVICE_PRIORITY_INVALID; + + return diag; +} + +int +rte_event_port_links_get(uint8_t dev_id, uint8_t port_id, + struct rte_event_queue_link link[]) +{ + struct rte_eventdev *dev; + uint16_t *links_map; + int i, count = 0; + + RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL); + dev = &rte_eventdevs[dev_id]; + if (!is_valid_port(dev, port_id)) { + EDEV_LOG_ERR("Invalid port_id=%" PRIu8, port_id); + return -EINVAL; + } + + links_map = dev->data->links_map; + /* Point links_map to this port specific area */ + links_map += (port_id * RTE_EVENT_MAX_QUEUES_PER_DEV); + for (i = 0; i < RTE_EVENT_MAX_QUEUES_PER_DEV; i++) { + if (links_map[i] != EVENT_QUEUE_SERVICE_PRIORITY_INVALID) { + link[count].queue_id = i; + link[count].priority = (uint8_t)links_map[i]; + ++count; + } + } + return count; +} + +int +rte_event_dequeue_wait_time(uint8_t dev_id, uint64_t ns, uint64_t *wait_ticks) +{ + struct rte_eventdev *dev; + + RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL); + dev = &rte_eventdevs[dev_id]; + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->wait_time, -ENOTSUP); + + if (wait_ticks == NULL) + return -EINVAL; + + (*dev->dev_ops->wait_time)(dev, ns, wait_ticks); + return 0; +} + +int +rte_event_dev_dump(uint8_t dev_id, FILE *f) +{ + struct rte_eventdev *dev; + + RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL); + dev = &rte_eventdevs[dev_id]; + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->dump, -ENOTSUP); + + (*dev->dev_ops->dump)(dev, f); + return 0; + +} + +int +rte_event_dev_start(uint8_t dev_id) +{ + struct rte_eventdev *dev; + int diag; + + EDEV_LOG_DEBUG("Start dev_id=%" PRIu8, dev_id); + + RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL); + dev = &rte_eventdevs[dev_id]; + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->dev_start, -ENOTSUP); + + if (dev->data->dev_started != 0) { + EDEV_LOG_ERR("Device with dev_id=%" PRIu8 " already started", + dev_id); + return 0; + } + + diag = (*dev->dev_ops->dev_start)(dev); + if (diag == 0) + dev->data->dev_started = 1; + else + return diag; + + return 0; +} + +void +rte_event_dev_stop(uint8_t dev_id) +{ + struct rte_eventdev *dev; + + EDEV_LOG_DEBUG("Stop dev_id=%" PRIu8, dev_id); + + RTE_EVENTDEV_VALID_DEVID_OR_RET(dev_id); + dev = &rte_eventdevs[dev_id]; + RTE_FUNC_PTR_OR_RET(*dev->dev_ops->dev_stop); + + if (dev->data->dev_started == 0) { + EDEV_LOG_ERR("Device with dev_id=%" PRIu8 " already stopped", + dev_id); + return; + } + + dev->data->dev_started = 0; + (*dev->dev_ops->dev_stop)(dev); +} + +int +rte_event_dev_close(uint8_t dev_id) +{ + struct rte_eventdev *dev; + + RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL); + dev = &rte_eventdevs[dev_id]; + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->dev_close, -ENOTSUP); + + /* Device must be stopped before it can be closed */ + if (dev->data->dev_started == 1) { + EDEV_LOG_ERR("Device %u must be stopped before closing", + dev_id); + return -EBUSY; + } + + return (*dev->dev_ops->dev_close)(dev); +} + +static inline int +rte_eventdev_data_alloc(uint8_t dev_id, struct rte_eventdev_data **data, + int socket_id) +{ + char mz_name[RTE_EVENTDEV_NAME_MAX_LEN]; + const struct rte_memzone *mz; + int n; + + /* Generate memzone name */ + n = snprintf(mz_name, sizeof(mz_name), "rte_eventdev_data_%u", dev_id); + if (n >= (int)sizeof(mz_name)) + return -EINVAL; + + if (rte_eal_process_type() == RTE_PROC_PRIMARY) { + mz = rte_memzone_reserve(mz_name, + sizeof(struct rte_eventdev_data), + socket_id, 0); + } else + mz = rte_memzone_lookup(mz_name); + + if (mz == NULL) + return -ENOMEM; + + *data = mz->addr; + if (rte_eal_process_type() == RTE_PROC_PRIMARY) + memset(*data, 0, sizeof(struct rte_eventdev_data)); + + return 0; +} + +static uint8_t +rte_eventdev_find_free_device_index(void) +{ + uint8_t dev_id; + + for (dev_id = 0; dev_id < RTE_EVENT_MAX_DEVS; dev_id++) { + if (rte_eventdevs[dev_id].attached == + RTE_EVENTDEV_DETACHED) + return dev_id; + } + return RTE_EVENT_MAX_DEVS; +} + +struct rte_eventdev * +rte_eventdev_pmd_allocate(const char *name, int socket_id) +{ + struct rte_eventdev *eventdev; + uint8_t dev_id; + + if (rte_eventdev_pmd_get_named_dev(name) != NULL) { + EDEV_LOG_ERR("Event device with name %s already " + "allocated!", name); + return NULL; + } + + dev_id = rte_eventdev_find_free_device_index(); + if (dev_id == RTE_EVENT_MAX_DEVS) { + EDEV_LOG_ERR("Reached maximum number of event devices"); + return NULL; + } + + eventdev = &rte_eventdevs[dev_id]; + + if (eventdev->data == NULL) { + struct rte_eventdev_data *eventdev_data = NULL; + + int retval = rte_eventdev_data_alloc(dev_id, &eventdev_data, + socket_id); + + if (retval < 0 || eventdev_data == NULL) + return NULL; + + eventdev->data = eventdev_data; + + snprintf(eventdev->data->name, RTE_EVENTDEV_NAME_MAX_LEN, + "%s", name); + + eventdev->data->dev_id = dev_id; + eventdev->data->socket_id = socket_id; + eventdev->data->dev_started = 0; + + eventdev->attached = RTE_EVENTDEV_ATTACHED; + + eventdev_globals.nb_devs++; + } + + return eventdev; +} + +int +rte_eventdev_pmd_release(struct rte_eventdev *eventdev) +{ + int ret; + + if (eventdev == NULL) + return -EINVAL; + + ret = rte_event_dev_close(eventdev->data->dev_id); + if (ret < 0) + return ret; + + eventdev->attached = RTE_EVENTDEV_DETACHED; + eventdev_globals.nb_devs--; + eventdev->data = NULL; + + return 0; +} + +struct rte_eventdev * +rte_eventdev_pmd_vdev_init(const char *name, size_t dev_private_size, + int socket_id) +{ + struct rte_eventdev *eventdev; + + /* Allocate device structure */ + eventdev = rte_eventdev_pmd_allocate(name, socket_id); + if (eventdev == NULL) + return NULL; + + /* Allocate private device structure */ + if (rte_eal_process_type() == RTE_PROC_PRIMARY) { + eventdev->data->dev_private = + rte_zmalloc_socket("eventdev device private", + dev_private_size, + RTE_CACHE_LINE_SIZE, + socket_id); + + if (eventdev->data->dev_private == NULL) + rte_panic("Cannot allocate memzone for private device" + " data"); + } + + return eventdev; +} + +int +rte_eventdev_pmd_pci_probe(struct rte_pci_driver *pci_drv, + struct rte_pci_device *pci_dev) +{ + struct rte_eventdev_driver *eventdrv; + struct rte_eventdev *eventdev; + + char eventdev_name[RTE_EVENTDEV_NAME_MAX_LEN]; + + int retval; + + eventdrv = (struct rte_eventdev_driver *)pci_drv; + if (eventdrv == NULL) + return -ENODEV; + + rte_eal_pci_device_name(&pci_dev->addr, eventdev_name, + sizeof(eventdev_name)); + + eventdev = rte_eventdev_pmd_allocate(eventdev_name, + pci_dev->device.numa_node); + if (eventdev == NULL) + return -ENOMEM; + + if (rte_eal_process_type() == RTE_PROC_PRIMARY) { + eventdev->data->dev_private = + rte_zmalloc_socket( + "eventdev private structure", + eventdrv->dev_private_size, + RTE_CACHE_LINE_SIZE, + rte_socket_id()); + + if (eventdev->data->dev_private == NULL) + rte_panic("Cannot allocate memzone for private " + "device data"); + } + + eventdev->pci_dev = pci_dev; + eventdev->driver = eventdrv; + + /* Invoke PMD device initialization function */ + retval = (*eventdrv->eventdev_init)(eventdev); + if (retval == 0) + return 0; + + EDEV_LOG_ERR("driver %s: event_dev_init(vendor_id=0x%x device_id=0x%x)" + " failed", pci_drv->driver.name, + (unsigned int) pci_dev->id.vendor_id, + (unsigned int) pci_dev->id.device_id); + + if (rte_eal_process_type() == RTE_PROC_PRIMARY) + rte_free(eventdev->data->dev_private); + + eventdev->attached = RTE_EVENTDEV_DETACHED; + eventdev_globals.nb_devs--; + + return -ENXIO; +} + +int +rte_eventdev_pmd_pci_remove(struct rte_pci_device *pci_dev) +{ + const struct rte_eventdev_driver *eventdrv; + struct rte_eventdev *eventdev; + char eventdev_name[RTE_EVENTDEV_NAME_MAX_LEN]; + int ret; + + if (pci_dev == NULL) + return -EINVAL; + + rte_eal_pci_device_name(&pci_dev->addr, eventdev_name, + sizeof(eventdev_name)); + + eventdev = rte_eventdev_pmd_get_named_dev(eventdev_name); + if (eventdev == NULL) + return -ENODEV; + + eventdrv = (const struct rte_eventdev_driver *)pci_dev->driver; + if (eventdrv == NULL) + return -ENODEV; + + /* Invoke PMD device uninit function */ + if (*eventdrv->eventdev_uninit) { + ret = (*eventdrv->eventdev_uninit)(eventdev); + if (ret) + return ret; + } + + /* Free event device */ + rte_eventdev_pmd_release(eventdev); + + if (rte_eal_process_type() == RTE_PROC_PRIMARY) + rte_free(eventdev->data->dev_private); + + eventdev->pci_dev = NULL; + eventdev->driver = NULL; + + return 0; +} diff --git a/lib/librte_eventdev/rte_eventdev_pmd.h b/lib/librte_eventdev/rte_eventdev_pmd.h new file mode 100644 index 0000000..e9d9b83 --- /dev/null +++ b/lib/librte_eventdev/rte_eventdev_pmd.h @@ -0,0 +1,504 @@ +/* + * + * Copyright(c) 2016 Cavium networks. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Cavium networks nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#ifndef _RTE_EVENTDEV_PMD_H_ +#define _RTE_EVENTDEV_PMD_H_ + +/** @file + * RTE Event PMD APIs + * + * @note + * These API are from event PMD only and user applications should not call + * them directly. + */ + +#ifdef __cplusplus +extern "C" { +#endif + +#include + +#include +#include +#include +#include +#include + +#include "rte_eventdev.h" + +#ifdef RTE_LIBRTE_EVENTDEV_DEBUG +#define RTE_PMD_DEBUG_TRACE(...) \ + rte_pmd_debug_trace(__func__, __VA_ARGS__) +#else +#define RTE_PMD_DEBUG_TRACE(...) +#endif + +/* Logging Macros */ +#define EDEV_LOG_ERR(fmt, args...) \ + RTE_LOG(ERR, EVENTDEV, "%s() line %u: " fmt "\n", \ + __func__, __LINE__, ## args) + +#ifdef RTE_LIBRTE_EVENTDEV_DEBUG +#define EDEV_LOG_DEBUG(fmt, args...) \ + RTE_LOG(DEBUG, EVENTDEV, "%s() line %u: " fmt "\n", \ + __func__, __LINE__, ## args) +#else +#define EDEV_LOG_DEBUG(fmt, args...) (void)0 +#endif + +/* Macros to check for valid device */ +#define RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, retval) do { \ + if (!rte_eventdev_pmd_is_valid_dev((dev_id))) { \ + EDEV_LOG_ERR("Invalid dev_id=%d\n", dev_id); \ + return retval; \ + } \ +} while (0) + +#define RTE_EVENTDEV_VALID_DEVID_OR_RET(dev_id) do { \ + if (!rte_eventdev_pmd_is_valid_dev((dev_id))) { \ + EDEV_LOG_ERR("Invalid dev_id=%d\n", dev_id); \ + return; \ + } \ +} while (0) + +#define RTE_EVENTDEV_DETACHED (0) +#define RTE_EVENTDEV_ATTACHED (1) + +/** + * Initialisation function of a event driver invoked for each matching + * event PCI device detected during the PCI probing phase. + * + * @param dev + * The dev pointer is the address of the *rte_eventdev* structure associated + * with the matching device and which has been [automatically] allocated in + * the *rte_event_devices* array. + * + * @return + * - 0: Success, the device is properly initialised by the driver. + * In particular, the driver MUST have set up the *dev_ops* pointer + * of the *dev* structure. + * - <0: Error code of the device initialisation failure. + */ +typedef int (*eventdev_init_t)(struct rte_eventdev *dev); + +/** + * Finalisation function of a driver invoked for each matching + * PCI device detected during the PCI closing phase. + * + * @param dev + * The dev pointer is the address of the *rte_eventdev* structure associated + * with the matching device and which has been [automatically] allocated in + * the *rte_event_devices* array. + * + * @return + * - 0: Success, the device is properly finalised by the driver. + * In particular, the driver MUST free the *dev_ops* pointer + * of the *dev* structure. + * - <0: Error code of the device initialisation failure. + */ +typedef int (*eventdev_uninit_t)(struct rte_eventdev *dev); + +/** + * The structure associated with a PMD driver. + * + * Each driver acts as a PCI driver and is represented by a generic + * *event_driver* structure that holds: + * + * - An *rte_pci_driver* structure (which must be the first field). + * + * - The *eventdev_init* function invoked for each matching PCI device. + * + * - The size of the private data to allocate for each matching device. + */ +struct rte_eventdev_driver { + struct rte_pci_driver pci_drv; /**< The PMD is also a PCI driver. */ + unsigned int dev_private_size; /**< Size of device private data. */ + + eventdev_init_t eventdev_init; /**< Device init function. */ + eventdev_uninit_t eventdev_uninit; /**< Device uninit function. */ +}; + +/** Global structure used for maintaining state of allocated event devices */ +struct rte_eventdev_global { + uint8_t nb_devs; /**< Number of devices found */ + uint8_t max_devs; /**< Max number of devices */ +}; + +extern struct rte_eventdev_global *rte_eventdev_globals; +/** Pointer to global event devices data structure. */ +extern struct rte_eventdev *rte_eventdevs; +/** The pool of rte_eventdev structures. */ + +/** + * Get the rte_eventdev structure device pointer for the named device. + * + * @param name + * device name to select the device structure. + * + * @return + * - The rte_eventdev structure pointer for the given device ID. + */ +static inline struct rte_eventdev * +rte_eventdev_pmd_get_named_dev(const char *name) +{ + struct rte_eventdev *dev; + unsigned int i; + + if (name == NULL) + return NULL; + + for (i = 0, dev = &rte_eventdevs[i]; + i < rte_eventdev_globals->max_devs; i++) { + if ((dev->attached == RTE_EVENTDEV_ATTACHED) && + (strcmp(dev->data->name, name) == 0)) + return dev; + } + + return NULL; +} + +/** + * Validate if the event device index is valid attached event device. + * + * @param dev_id + * Event device index. + * + * @return + * - If the device index is valid (1) or not (0). + */ +static inline unsigned +rte_eventdev_pmd_is_valid_dev(uint8_t dev_id) +{ + struct rte_eventdev *dev; + + if (dev_id >= rte_eventdev_globals->nb_devs) + return 0; + + dev = &rte_eventdevs[dev_id]; + if (dev->attached != RTE_EVENTDEV_ATTACHED) + return 0; + else + return 1; +} + +/** + * Definitions of all functions exported by a driver through the + * the generic structure of type *event_dev_ops* supplied in the + * *rte_eventdev* structure associated with a device. + */ + +/** + * Get device information of a device. + * + * @param dev + * Event device pointer + * @param dev_info + * Event device information structure + * + * @return + * Returns 0 on success + */ +typedef void (*eventdev_info_get_t)(struct rte_eventdev *dev, + struct rte_event_dev_info *dev_info); + +/** + * Configure a device. + * + * @param dev + * Event device pointer + * + * @return + * Returns 0 on success + */ +typedef int (*eventdev_configure_t)(struct rte_eventdev *dev); + +/** + * Start a configured device. + * + * @param dev + * Event device pointer + * + * @return + * Returns 0 on success + */ +typedef int (*eventdev_start_t)(struct rte_eventdev *dev); + +/** + * Stop a configured device. + * + * @param dev + * Event device pointer + */ +typedef void (*eventdev_stop_t)(struct rte_eventdev *dev); + +/** + * Close a configured device. + * + * @param dev + * Event device pointer + * + * @return + * - 0 on success + * - (-EAGAIN) if can't close as device is busy + */ +typedef int (*eventdev_close_t)(struct rte_eventdev *dev); + +/** + * Retrieve the default event queue configuration. + * + * @param dev + * Event device pointer + * @param queue_id + * Event queue index + * @param[out] queue_conf + * Event queue configuration structure + * + */ +typedef void (*eventdev_queue_default_conf_get_t)(struct rte_eventdev *dev, + uint8_t queue_id, struct rte_event_queue_conf *queue_conf); + +/** + * Setup an event queue. + * + * @param dev + * Event device pointer + * @param queue_id + * Event queue index + * @param queue_conf + * Event queue configuration structure + * + * @return + * Returns 0 on success. + */ +typedef int (*eventdev_queue_setup_t)(struct rte_eventdev *dev, + uint8_t queue_id, struct rte_event_queue_conf *queue_conf); + +/** + * Release memory resources allocated by given event queue. + * + * @param queue + * Event queue pointer + * + */ +typedef void (*eventdev_queue_release_t)(void *queue); + +/** + * Retrieve the default event port configuration. + * + * @param dev + * Event device pointer + * @param port_id + * Event port index + * @param[out] port_conf + * Event port configuration structure + * + */ +typedef void (*eventdev_port_default_conf_get_t)(struct rte_eventdev *dev, + uint8_t port_id, struct rte_event_port_conf *port_conf); + +/** + * Setup an event port. + * + * @param dev + * Event device pointer + * @param port_id + * Event port index + * @param port_conf + * Event port configuration structure + * + * @return + * Returns 0 on success. + */ +typedef int (*eventdev_port_setup_t)(struct rte_eventdev *dev, + uint8_t port_id, struct rte_event_port_conf *port_conf); + +/** + * Release memory resources allocated by given event port. + * + * @param port + * Event port pointer + * + */ +typedef void (*eventdev_port_release_t)(void *port); + +/** + * Link multiple source event queues to destination event port. + * + * @param port + * Event port pointer + * @param link + * An array of *nb_links* pointers to *rte_event_queue_link* structure + * @param nb_links + * The number of links to establish + * + * @return + * Returns 0 on success. + * + */ +typedef int (*eventdev_port_link_t)(void *port, + struct rte_event_queue_link link[], uint16_t nb_links); + +/** + * Unlink multiple source event queues from destination event port. + * + * @param port + * Event port pointer + * @param queues + * An array of *nb_unlinks* event queues to be unlinked from the event port. + * @param nb_unlinks + * The number of unlinks to establish + * + * @return + * Returns 0 on success. + * + */ +typedef int (*eventdev_port_unlink_t)(void *port, + uint8_t queues[], uint16_t nb_unlinks); + +/** + * Converts nanoseconds to *wait* value for rte_event_dequeue() + * + * @param dev + * Event device pointer + * @param ns + * Wait time in nanosecond + * @param[out] wait_ticks + * Value for the *wait* parameter in rte_event_dequeue() function + * + */ +typedef void (*eventdev_dequeue_wait_time_t)(struct rte_eventdev *dev, + uint64_t ns, uint64_t *wait_ticks); + +/** + * Dump internal information + * + * @param dev + * Event device pointer + * @param f + * A pointer to a file for output + * + */ +typedef void (*eventdev_dump_t)(struct rte_eventdev *dev, FILE *f); + +/** Event device operations function pointer table */ +struct rte_eventdev_ops { + eventdev_info_get_t dev_infos_get; /**< Get device info. */ + eventdev_configure_t dev_configure; /**< Configure device. */ + eventdev_start_t dev_start; /**< Start device. */ + eventdev_stop_t dev_stop; /**< Stop device. */ + eventdev_close_t dev_close; /**< Close device. */ + + eventdev_queue_default_conf_get_t queue_def_conf; + /**< Get default queue configuration. */ + eventdev_queue_setup_t queue_setup; + /**< Set up an event queue. */ + eventdev_queue_release_t queue_release; + /**< Release an event queue. */ + + eventdev_port_default_conf_get_t port_def_conf; + /**< Get default port configuration. */ + eventdev_port_setup_t port_setup; + /**< Set up an event port. */ + eventdev_port_release_t port_release; + /**< Release an event port. */ + + eventdev_port_link_t port_link; + /**< Link event queues to an event port. */ + eventdev_port_unlink_t port_unlink; + /**< Unlink event queues from an event port. */ + eventdev_dequeue_wait_time_t wait_time; + /**< Converts nanoseconds to *wait* value for rte_event_dequeue() */ + eventdev_dump_t dump; + /* Dump internal information */ +}; + +/** + * Allocates a new eventdev slot for an event device and returns the pointer + * to that slot for the driver to use. + * + * @param name + * Unique identifier name for each device + * @param socket_id + * Socket to allocate resources on. + * @return + * - Slot in the rte_dev_devices array for a new device; + */ +struct rte_eventdev * +rte_eventdev_pmd_allocate(const char *name, int socket_id); + +/** + * Release the specified eventdev device. + * + * @param eventdev + * The *eventdev* pointer is the address of the *rte_eventdev* structure. + * @return + * - 0 on success, negative on error + */ +int +rte_eventdev_pmd_release(struct rte_eventdev *eventdev); + +/** + * Creates a new virtual event device and returns the pointer to that device. + * + * @param name + * PMD type name + * @param dev_private_size + * Size of event PMDs private data + * @param socket_id + * Socket to allocate resources on. + * + * @return + * - Eventdev pointer if device is successfully created. + * - NULL if device cannot be created. + */ +struct rte_eventdev * +rte_eventdev_pmd_vdev_init(const char *name, size_t dev_private_size, + int socket_id); + + +/** + * Wrapper for use by pci drivers as a .probe function to attach to a event + * interface. + */ +int rte_eventdev_pmd_pci_probe(struct rte_pci_driver *pci_drv, + struct rte_pci_device *pci_dev); + +/** + * Wrapper for use by pci drivers as a .remove function to detach a event + * interface. + */ +int rte_eventdev_pmd_pci_remove(struct rte_pci_device *pci_dev); + +#ifdef __cplusplus +} +#endif + +#endif /* _RTE_EVENTDEV_PMD_H_ */ diff --git a/lib/librte_eventdev/rte_eventdev_version.map b/lib/librte_eventdev/rte_eventdev_version.map new file mode 100644 index 0000000..ef40aae --- /dev/null +++ b/lib/librte_eventdev/rte_eventdev_version.map @@ -0,0 +1,39 @@ +DPDK_17.02 { + global: + + rte_eventdevs; + + rte_event_dev_count; + rte_event_dev_get_dev_id; + rte_event_dev_socket_id; + rte_event_dev_info_get; + rte_event_dev_configure; + rte_event_dev_start; + rte_event_dev_stop; + rte_event_dev_close; + rte_event_dev_dump; + + rte_event_port_default_conf_get; + rte_event_port_setup; + rte_event_port_dequeue_depth; + rte_event_port_enqueue_depth; + rte_event_port_count; + rte_event_port_link; + rte_event_port_unlink; + rte_event_port_links_get; + + rte_event_queue_default_conf_get + rte_event_queue_setup; + rte_event_queue_count; + rte_event_queue_priority; + + rte_event_dequeue_wait_time; + + rte_eventdev_pmd_allocate; + rte_eventdev_pmd_release; + rte_eventdev_pmd_vdev_init; + rte_eventdev_pmd_pci_probe; + rte_eventdev_pmd_pci_remove; + + local: *; +}; diff --git a/mk/rte.app.mk b/mk/rte.app.mk index f75f0e2..716725a 100644 --- a/mk/rte.app.mk +++ b/mk/rte.app.mk @@ -93,6 +93,7 @@ _LDLIBS-$(CONFIG_RTE_LIBRTE_MBUF) += -lrte_mbuf _LDLIBS-$(CONFIG_RTE_LIBRTE_NET) += -lrte_net _LDLIBS-$(CONFIG_RTE_LIBRTE_ETHER) += -lrte_ethdev _LDLIBS-$(CONFIG_RTE_LIBRTE_CRYPTODEV) += -lrte_cryptodev +_LDLIBS-$(CONFIG_RTE_LIBRTE_EVENTDEV) += -lrte_eventdev _LDLIBS-$(CONFIG_RTE_LIBRTE_MEMPOOL) += -lrte_mempool _LDLIBS-$(CONFIG_RTE_LIBRTE_RING) += -lrte_ring _LDLIBS-$(CONFIG_RTE_LIBRTE_EAL) += -lrte_eal