Cover Detail
Show a cover letter.
GET /api/covers/47660/?format=api
{ "id": 47660, "url": "http://patches.dpdk.org/api/covers/47660/?format=api", "web_url": "http://patches.dpdk.org/project/dpdk/cover/1541074741-41368-1-git-send-email-viacheslavo@mellanox.com/", "project": { "id": 1, "url": "http://patches.dpdk.org/api/projects/1/?format=api", "name": "DPDK", "link_name": "dpdk", "list_id": "dev.dpdk.org", "list_email": "dev@dpdk.org", "web_url": "http://core.dpdk.org", "scm_url": "git://dpdk.org/dpdk", "webscm_url": "http://git.dpdk.org/dpdk", "list_archive_url": "https://inbox.dpdk.org/dev", "list_archive_url_format": "https://inbox.dpdk.org/dev/{}", "commit_url_format": "" }, "msgid": "<1541074741-41368-1-git-send-email-viacheslavo@mellanox.com>", "list_archive_url": "https://inbox.dpdk.org/dev/1541074741-41368-1-git-send-email-viacheslavo@mellanox.com", "date": "2018-11-01T12:19:21", "name": "[v3,00/13] net/mlx5: e-switch VXLAN encap/decap hardware offload", "submitter": { "id": 1102, "url": "http://patches.dpdk.org/api/people/1102/?format=api", "name": "Slava Ovsiienko", "email": "viacheslavo@mellanox.com" }, "mbox": "http://patches.dpdk.org/project/dpdk/cover/1541074741-41368-1-git-send-email-viacheslavo@mellanox.com/mbox/", "series": [ { "id": 2204, "url": "http://patches.dpdk.org/api/series/2204/?format=api", "web_url": "http://patches.dpdk.org/project/dpdk/list/?series=2204", "date": "2018-11-01T12:19:21", "name": "net/mlx5: e-switch VXLAN encap/decap hardware offload", "version": 3, "mbox": "http://patches.dpdk.org/series/2204/mbox/" } ], "comments": "http://patches.dpdk.org/api/covers/47660/comments/", "headers": { "Return-Path": "<dev-bounces@dpdk.org>", "X-Original-To": "patchwork@dpdk.org", "Delivered-To": "patchwork@dpdk.org", "Received": [ "from [92.243.14.124] (localhost [127.0.0.1])\n\tby dpdk.org (Postfix) with ESMTP id 2DF541B127;\n\tThu, 1 Nov 2018 13:19:25 +0100 (CET)", "from EUR04-DB3-obe.outbound.protection.outlook.com\n\t(mail-eopbgr60085.outbound.protection.outlook.com [40.107.6.85])\n\tby dpdk.org (Postfix) with ESMTP id A3F3C1B10E\n\tfor <dev@dpdk.org>; Thu, 1 Nov 2018 13:19:23 +0100 (CET)", "from AM4PR05MB3265.eurprd05.prod.outlook.com (10.171.186.150) by\n\tAM4PR05MB1681.eurprd05.prod.outlook.com (10.165.245.152) with\n\tMicrosoft SMTP Server (version=TLS1_2,\n\tcipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id\n\t15.20.1273.25; Thu, 1 Nov 2018 12:19:21 +0000", "from AM4PR05MB3265.eurprd05.prod.outlook.com\n\t([fe80::544b:a68d:e6a5:ba6e]) by\n\tAM4PR05MB3265.eurprd05.prod.outlook.com\n\t([fe80::544b:a68d:e6a5:ba6e%2]) with mapi id 15.20.1273.030;\n\tThu, 1 Nov 2018 12:19:21 +0000" ], "DKIM-Signature": "v=1; a=rsa-sha256; c=relaxed/relaxed; d=Mellanox.com;\n\ts=selector1;\n\th=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck;\n\tbh=gs6aoQKA8qsCbbtze1z/LmGuFq96zadfbk7IVaAT/8M=;\n\tb=EcBj7O9Zpb3YKe8fDadt2exkAUofisweaWzwG48b5CvxKYiq49+/t7E0ZpDRRkD2VhPTOznkfy63H8H/t5Q2//xawfhDuxNrvPD8SifgWGTbPkFKj+1H0Erkne+cT7QKKZJwk0hXMGMdVi6p2kiGWgys0a9X8w020ujRtYB43Zo=", "From": "Slava Ovsiienko <viacheslavo@mellanox.com>", "To": "Shahaf Shuler <shahafs@mellanox.com>", "CC": "\"dev@dpdk.org\" <dev@dpdk.org>, Yongseok Koh <yskoh@mellanox.com>, Slava\n\tOvsiienko <viacheslavo@mellanox.com>", "Thread-Topic": "[PATCH v3 00/13] net/mlx5: e-switch VXLAN encap/decap hardware\n\toffload", "Thread-Index": "AQHUcd0eafwZ1MomRkGFeHttp19GCw==", "Date": "Thu, 1 Nov 2018 12:19:21 +0000", "Message-ID": "<1541074741-41368-1-git-send-email-viacheslavo@mellanox.com>", "References": "<1539612815-47199-1-git-send-email-viacheslavo@mellanox.com>", "In-Reply-To": "<1539612815-47199-1-git-send-email-viacheslavo@mellanox.com>", "Accept-Language": "en-US", "Content-Language": "en-US", "X-MS-Has-Attach": "", "X-MS-TNEF-Correlator": "", "x-clientproxiedby": "LO2P265CA0079.GBRP265.PROD.OUTLOOK.COM\n\t(2603:10a6:600:8::19) To AM4PR05MB3265.eurprd05.prod.outlook.com\n\t(2603:10a6:205:4::22)", "authentication-results": "spf=none (sender IP is )\n\tsmtp.mailfrom=viacheslavo@mellanox.com; ", "x-ms-exchange-messagesentrepresentingtype": "1", "x-originating-ip": "[37.142.13.130]", "x-ms-publictraffictype": "Email", "x-microsoft-exchange-diagnostics": "1; AM4PR05MB1681;\n\t6:c+nL5fmHpOyH3ZJMOS4226Pc83CLaabWneJYEwyjPOYsXg6G0wzyArM5GUZIthL7gKuoa/8PlYgB+4XdNt4mhyHN1UEZati9so8PFj1FJqK1Aa5PIdlcMq2uFPPLMraqs1JuJA9f8ABUL3cGZwFq/c5j6aRXKyC+GRY2FpQahepCT9xPmMWUHJhThS1lRrC3bcUU24NrWwjZHq24+J3CCNqPk5kk7WWQMFUbidi24S6uw6Hegq8cQxpICc82v3fVlCl/1I7R0ndxQAqsn6VysdjnyrszyoMqmpXPMgSvF9TIjfqOxlgCGTPeRwtB/XtgjxK13pFya2Giwfqs2+pWNE/6scOyaah48bCBCFZfUqxvG9uVbbP2c0yAp4m0k8u+uvj/Q0H1Y4Ah8WTMOdcctzLMPXm6iargL/7b4eBasVGbkHaq9s0+Yq9RayPETPgEhXe304tqf2utP04jCUQj+A==;\n\t5:n8OmWh43tziyCOvu4H0izxl9711h5aDe+KUzcTYzQ4ChzVDOfCungomWqSjqG9LwK5+JJeGw0ftBDbR7w8w4kfN8t94ig7bZ7DOrEuc1S6C2xtfpsZlDxut+1iK9EK1jR/v6oD4dzS1eNlVdq6B41yLtnWEPRGVHYdN/bAidp94=;\n\t7:aC4lguY36rzIPbMjVzboUfj95u6EgQEasijokZDqULDkBg3/LgbPGL+dsjv2L7ytUecAeMV7GukOw3uGQDS8N3jxLwN83WsqxfDnCiLtw776hUGe8jTwOvulFj1kkkzvR9z8nxGg9UwfZDsbfK+h3Q==", "x-ms-office365-filtering-correlation-id": "b975f7cf-f396-445b-77f4-08d63ff440c8", "x-ms-office365-filtering-ht": "Tenant", "x-microsoft-antispam": "BCL:0; PCL:0;\n\tRULEID:(7020095)(4652040)(8989299)(4534185)(4627221)(201703031133081)(201702281549075)(8990200)(5600074)(711020)(4618075)(2017052603328)(7153060)(7193020);\n\tSRVR:AM4PR05MB1681; ", "x-ms-traffictypediagnostic": "AM4PR05MB1681:", "x-microsoft-antispam-prvs": "<AM4PR05MB1681212E17C68BA992634674D2CE0@AM4PR05MB1681.eurprd05.prod.outlook.com>", "x-exchange-antispam-report-test": "UriScan:;", "x-ms-exchange-senderadcheck": "1", "x-exchange-antispam-report-cfa-test": "BCL:0; PCL:0;\n\tRULEID:(8211001083)(6040522)(2401047)(8121501046)(5005006)(3002001)(93006095)(93001095)(10201501046)(3231382)(944501410)(4982022)(52105095)(6055026)(148016)(149066)(150057)(6041310)(20161123560045)(20161123564045)(20161123562045)(20161123558120)(201703131423095)(201702281528075)(20161123555045)(201703061421075)(201703061406153)(201708071742011)(7699051)(76991095);\n\tSRVR:AM4PR05MB1681; BCL:0; PCL:0; RULEID:; SRVR:AM4PR05MB1681; ", "x-forefront-prvs": "0843C17679", "x-forefront-antispam-report": "SFV:NSPM;\n\tSFS:(10009020)(136003)(39860400002)(366004)(396003)(346002)(376002)(199004)(189003)(97736004)(3846002)(6116002)(561944003)(6862004)(478600001)(5250100002)(25786009)(4326008)(71190400001)(71200400001)(86362001)(316002)(966005)(8676002)(68736007)(81156014)(8936002)(81166006)(106356001)(105586002)(2906002)(14454004)(36756003)(7736002)(305945005)(107886003)(54906003)(37006003)(6436002)(6486002)(102836004)(99286004)(66066001)(6512007)(53936002)(2900100001)(6306002)(476003)(6636002)(52116002)(76176011)(386003)(186003)(256004)(5024004)(446003)(26005)(5660300001)(14444005)(6506007)(11346002)(486006)(2616005)(21314003);\n\tDIR:OUT; SFP:1101; SCL:1; SRVR:AM4PR05MB1681;\n\tH:AM4PR05MB3265.eurprd05.prod.outlook.com; FPR:; SPF:None; LANG:en;\n\tPTR:InfoNoRecords; A:1; MX:1; ", "received-spf": "None (protection.outlook.com: mellanox.com does not designate\n\tpermitted sender hosts)", "x-microsoft-antispam-message-info": "/9HFwCQ7Y4VSV6Ropjus5cA1b0nGqOeWU0U5k6KT+ydEix+JR/lH1V0k5UV10daeSUxNyHLSeNS4QwvDLTa4AGDNBqM/+Fw5oaJl5aV0Jz5SaaBLwiOye+PKSQUbGg3MRe97AlVpjYDiCkY+jETbED52haM8kry+VARbeyPBzWv+3+FOLDe/8NQX7Y33w9U0/0JZNvQUgKaxq18JavfjgriMFQHjyR/IbCrznXq5m7HIxwxdHMq7G1j10bVthI0uTM8RKuTEQhwAji5PrwBRgbvef+ZEDOrSeJPahebMNAf7HEZEUWSzkHnyA8m5lV0h773s1GaMBw9eufEKgoGP/aW7Iz3XJWIXSGQu3WJgZo0=", "spamdiagnosticoutput": "1:99", "spamdiagnosticmetadata": "NSPM", "Content-Type": "text/plain; charset=\"iso-8859-1\"", "Content-Transfer-Encoding": "quoted-printable", "MIME-Version": "1.0", "X-OriginatorOrg": "Mellanox.com", "X-MS-Exchange-CrossTenant-Network-Message-Id": "b975f7cf-f396-445b-77f4-08d63ff440c8", "X-MS-Exchange-CrossTenant-originalarrivaltime": "01 Nov 2018 12:19:21.4703\n\t(UTC)", "X-MS-Exchange-CrossTenant-fromentityheader": "Hosted", "X-MS-Exchange-CrossTenant-id": "a652971c-7d2e-4d9b-a6a4-d149256f461b", "X-MS-Exchange-Transport-CrossTenantHeadersStamped": "AM4PR05MB1681", "Subject": "[dpdk-dev] [PATCH v3 00/13] net/mlx5: e-switch VXLAN encap/decap\n\thardware offload", "X-BeenThere": "dev@dpdk.org", "X-Mailman-Version": "2.1.15", "Precedence": "list", "List-Id": "DPDK patches and discussions <dev.dpdk.org>", "List-Unsubscribe": "<https://mails.dpdk.org/options/dev>,\n\t<mailto:dev-request@dpdk.org?subject=unsubscribe>", "List-Archive": "<http://mails.dpdk.org/archives/dev/>", "List-Post": "<mailto:dev@dpdk.org>", "List-Help": "<mailto:dev-request@dpdk.org?subject=help>", "List-Subscribe": "<https://mails.dpdk.org/listinfo/dev>,\n\t<mailto:dev-request@dpdk.org?subject=subscribe>", "Errors-To": "dev-bounces@dpdk.org", "Sender": "\"dev\" <dev-bounces@dpdk.org>" }, "content": "This patchset adds the VXLAN encapsulation/decapsulation hardware\noffload feature for E-Switch.\n \nA typical use case of tunneling infrastructure is port representors \nin switchdev mode, with VXLAN traffic encapsulation performed on\ntraffic coming *from* a representor and decapsulation on traffic\ngoing *to* that representor, in order to transparently assign\na given VXLAN to VF traffic.\n\nSince these actions are supported at the E-Switch level, the \"transfer\" \nattribute must be set on such flow rules. They must also be combined\nwith a port redirection action to make sense.\n\nSince only ingress is supported, encapsulation flow rules are normally\napplied on a physical port and emit traffic to a port representor. \nThe opposite order is used for decapsulation.\n\nLike other mlx5 E-Switch flow rule actions, these ones are implemented\nthrough Linux's TC flower API. Since the Linux interface for VXLAN\nencap/decap involves virtual network devices (i.e. ip link add type\n\t\tvxlan [...]), the PMD dynamically spawns them on a needed basis\nthrough Netlink calls. These VXLAN implicitly created devices are\ncalled VTEPs (Virtual Tunnel End Points).\n\nVXLAN interfaces are dynamically created for each local port of\nouter networks and then used as targets for TC \"flower\" filters\nin order to perform encapsulation. For decapsulation the VXLAN\ndevices are created for each unique UDP-port. These VXLAN interfaces\nare system-wide, the only one device with given UDP port can exist \nin the system (the attempt of creating another device with the \nsame UDP local port returns EEXIST), so PMD should support the\nshared (between PMD instances) device database. \n\nRules samples consideraions:\n\n$PF \t\t- physical device, outer network\n$VF \t\t- representor for VF, outer/inner network\n$VXLAN\t\t- VTEP netdev name\n$PF_OUTER_IP \t- $PF IP (v4 or v6) within outer network\n$REMOTE_IP \t- remote peer IP (v4 or v6) within outer network\n$LOCAL_PORT\t- local UDP port\n$REMOTE_PORT\t- remote UDP port\n\nVXLAN VTEP creation with iproute2 (PMD does the same via Netlink):\n\n- for encapsulation:\n\n ip link add $VXLAN type vxlan dstport $LOCAL_PORT external dev $PF\n ip link set dev $VXLAN up\n tc qdisc del dev $VXLAN ingress\n tc qdisc add dev $VXLAN ingress\n\n$LOCAL_PORT for egress encapsulated traffic (note, this is not\nsource UDP port in the VXLAN header, it is just UDP port assigned\nto VTEP, no practical usage) is selected from available\tUDP ports\nautomatically in range 30000-60000.\n\n- for decapsulation:\n\n ip link add $VXLAN type vxlan dstport $LOCAL_PORT external\n ip link set dev $VXLAN up\n tc qdisc del dev $VXLAN ingress\n tc qdisc add dev $VXLAN ingress\n\n$LOCAL_PORT is UDP port receiving the VXLAN traffic from outer networks.\n\nAll ingress UDP traffic with given UDP destination port from ALL existing\nnetdevs is routed by kernel to the $VXLAN net device. While applying the\nrule the kernel checks the IP parameter withing rule, determines the\nappropriate underlaying PF and tryes to setup the rule hardware offload.\n\nVXLAN encapsulation \n\nVXLAN encap rules are applied to the VF ingress traffic and have the \nVTEP as actual redirection destinations instead of outer PF.\nThe encapsulation rule should provide:\n- redirection action VF->PF\n- VF port ID\n- some inner network parameters (MACs) \n- the tunnel outer source IP (v4/v6), (IS A MUST)\n- the tunnel outer destination IP (v4/v6), (IS A MUST).\n- VNI - Virtual Network Identifier (IS A MUST)\n\nVXLAN encapsulation rule sample for tc utility:\n\n tc filter add dev $VF protocol all parent ffff: flower skip_sw \\\n\taction tunnel_key set dst_port $REMOTE_PORT \\\n\tsrc_ip $PF_OUTER_IP dst_ip $REMOTE_IP id $VNI \\\n\taction mirred egress redirect dev $VXLAN\n\nVXLAN encapsulation rule sample for testpmd:\n\n- Setting up outer properties of VXLAN tunnel:\n\n set vxlan ip-version ipv4 vni $VNI \\\n\tudp-src $IGNORED udp-dst $REMOTE_PORT \\\n\tip-src $PF_OUTER_IP ip-dst $REMOTE_IP \\\n \teth-src $IGNORED eth-dst $REMOTE_MAC\n\n- Creating a flow rule on port ID 4 performing VXLAN encapsulation\n with the abovementioned properties and directing the resulting\n traffic to port ID 0:\n\n flow create 4 ingress transfer pattern eth src is $INNER_MAC / end\n\tactions vxlan_encap / port_id id 0 / end\n\nThere is no direct way found to provide kernel with all required\nencapsulatioh header parameters. The encapsulation VTEP is created\nattached to the outer interface and assumed as default path for\negress encapsulated traffic. The outer tunnel IP address are\nassigned to interface using Netlink, the implicit route is\ncreated like this:\n\n ip addr add <src_ip> peer <dst_ip> dev <outer> scope link\n\nThe peer address option provides implicit route, and scope link\nattribute reduces the risk of conflicts. At initialization time all\nlocal scope link addresses are flushed from the outer network device.\n\nThe destination MAC address is provided via permenent neigh rule:\n\n ip neigh add dev <outer> lladdr <dst_mac> to <dst_ip> nud permanent\n\nAt initialization time all neigh rules of permanent type are flushed\nfrom the outer network device. \n\nVXLAN decapsulation \n\nVXLAN decap rules are applied to the ingress traffic of VTEP ($VXLAN)\ndevice instead of PF. The decapsulation rule should provide:\n- redirection action PF->VF\n- VF port ID as redirection destination\n- $VXLAN device as ingress traffic source\n- the tunnel outer source IP (v4/v6), (optional)\n- the tunnel outer destination IP (v4/v6), (IS A MUST)\n- the tunnel local UDP port (IS A MUST, PMD looks for appropriate VTEP\n with given local UDP port)\n- VNI - Virtual Network Identifier (IS A MUST)\n\nVXLAN decap rule sample for tc utility: \n\n tc filter add dev $VXLAN protocol all parent ffff: flower skip_sw \\\n\tenc_src_ip $REMOTE_IP enc_dst_ip $PF_OUTER_IP enc_key_id $VNI \\\n\tnc_dst_port $LOCAL_PORT \\\n\taction tunnel_key unset action mirred egress redirect dev $VF\n\t\t\t\t\t\t\nVXLAN decap rule sample for testpmd: \n\n- Creating a flow on port ID 0 performing VXLAN decapsulation and directing\n the result to port ID 4 with checking inner properties:\n\n flow create 0 ingress transfer pattern / \n \tipv4 src is $REMOTE_IP dst $PF_LOCAL_IP /\n\tudp src is 9999 dst is $LOCAL_PORT / vxlan vni is $VNI / \n\teth src is 00:11:22:33:44:55 dst is $INNER_MAC / end\n actions vxlan_decap / port_id id 4 / end\n\nThe VXLAN encap/decap rules constrains (implied by current kernel support)\n\n- VXLAN decapsulation provided for PF->VF direction only\n- VXLAN encapsulation provided for VF->PF direction only\n- current implementation will support non-shared database of VTEPs\n (impossible simultaneous usage of the same UDP port by several\n instances of DPDK apps)\n\nSuggested-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>\nSigned-off-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>\n\n---\nv3:\n * patchset is resplitted into more dedicated parts\n * decapsulation rule takes MAC from inner eth item\n * appropriate RTE_BEx are replaced with runtime rte_cpu_xxx\n * E-Switch Flow counter deletion is fixed\n * VTEP management routines are refactored\n * found typos are corrected\n\nv2:\n * removed non-VXLAN related parts\n * multipart Netlink messages support\n * local IP and peer IP rules management\n * neigh IP address to MAC address rules\n * management rules cleanup at outer device initialization\n * attached devices cleanup at outer device initialization\n\nv1:\n * http://patches.dpdk.org/patch/45800/\n * Refactored code of initial experimental proposal\n\nv0:\n * http://patches.dpdk.org/cover/44080/\n * Initial proposal by Adrien Mazarguil <adrien.mazarguil@6wind.com>\n\nViacheslav Ovsiienko (13):\n net/mlx5: prepare makefile for adding e-switch VXLAN\n net/mlx5: prepare meson.build for adding e-switch VXLAN\n net/mlx5: add necessary definitions for e-switch VXLAN\n net/mlx5: add necessary structures for e-switch VXLAN\n net/mlx5: swap items/actions validations for e-switch rules\n net/mlx5: add e-switch VXLAN support to validation routine\n net/mlx5: add VXLAN support to flow prepare routine\n net/mlx5: add VXLAN support to flow translate routine\n net/mlx5: e-switch VXLAN netlink routines update\n net/mlx5: fix e-switch Flow counter deletion\n net/mlx5: add e-switch VXLAN tunnel devices management\n net/mlx5: add e-switch VXLAN encapsulation rules\n net/mlx5: add e-switch VXLAN rule cleanup routines\n\n drivers/net/mlx5/Makefile | 85 +\n drivers/net/mlx5/meson.build | 34 +\n drivers/net/mlx5/mlx5_flow.h | 11 +\n drivers/net/mlx5/mlx5_flow_tcf.c | 5118 +++++++++++++++++++++++++++++---------\n 4 files changed, 4107 insertions(+), 1141 deletions(-)" }