From patchwork Mon Apr 1 22:46:49 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: William Tu X-Patchwork-Id: 1073463 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Authentication-Results: ozlabs.org; spf=pass (mailfrom) smtp.mailfrom=openvswitch.org (client-ip=140.211.169.12; helo=mail.linuxfoundation.org; envelope-from=ovs-dev-bounces@openvswitch.org; receiver=) Authentication-Results: ozlabs.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: ozlabs.org; dkim=fail reason="signature verification failed" (2048-bit key; unprotected) header.d=gmail.com header.i=@gmail.com header.b="Lj6bc7YS"; dkim-atps=neutral Received: from mail.linuxfoundation.org (mail.linuxfoundation.org [140.211.169.12]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id 44Y6v1478Cz9sSV for ; Tue, 2 Apr 2019 09:48:13 +1100 (AEDT) Received: from mail.linux-foundation.org (localhost [127.0.0.1]) by mail.linuxfoundation.org (Postfix) with ESMTP id EF8952154; Mon, 1 Apr 2019 22:47:30 +0000 (UTC) X-Original-To: dev@openvswitch.org Delivered-To: ovs-dev@mail.linuxfoundation.org Received: from smtp1.linuxfoundation.org (smtp1.linux-foundation.org [172.17.192.35]) by mail.linuxfoundation.org (Postfix) with ESMTPS id 54C261AC3 for ; Mon, 1 Apr 2019 22:47:29 +0000 (UTC) X-Greylist: whitelisted by SQLgrey-1.7.6 Received: from mail-wm1-f65.google.com (mail-wm1-f65.google.com [209.85.128.65]) by smtp1.linuxfoundation.org (Postfix) with ESMTPS id BEBD0712 for ; Mon, 1 Apr 2019 22:47:28 +0000 (UTC) Received: by mail-wm1-f65.google.com with SMTP id z24so1205816wmi.5 for ; Mon, 01 Apr 2019 15:47:28 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=e1mvbbvjoUkvZH14Xd3cGozfREmfo5j/0uIatOTXpFw=; b=Lj6bc7YSLeCgnJ5XEH80w0QKE1bsUhcQqD4xbd5J6Q5IiMuSUo8BRhC8noQFLGTeZc EweebVvQCyLOJ6IpPMGFmNR/ruqOsUAj4pdf6Z/MN0LxCvM9UBmjD196KOtGWCkfe2af kUUHB1mJxb12Kkh5lwz1ephiqi5d7AGe2KJIs4HmqQMOHi3asAcmMNXjwyEC6wJ2XVtn 1/WpBE1W7maseDtdkBX/1prUrK07LRw6TLqDOw8Y4QIRYCSPeARXJ7Bh+v3SrwXwXWKb sElWOoI+Bnoqc+uSWDHhVdF7ZtVUz6yUZLKX6Irft3B1pkRCcYIV2XllBl2Vc4WismgO uX/Q== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=e1mvbbvjoUkvZH14Xd3cGozfREmfo5j/0uIatOTXpFw=; b=XgGkciNb5KCXT21YVvN9zZX3i3wLCD+o0LSFQUe3wLzuUhWKE+Mj6IPpcMVRL36dL3 unv9044fIKcHo8YMQtP8JvMA8NBBbdpdPCt00w8z4gWdqPsYpRkYk0/A6EsVjdHj7i7T NAl7sFa8ltWxtLoMuidYjvY35fO9Loj5WIV4XPhtbz780y1dW+C7TzLeM9t4jl38/TX6 Mw/NG39xevuRidyJkVCXBdD9VMT8vA9d+lFZcGOmii77eLyejeeup6osE29L14vt9wt1 Kd+ssj5pauSfghI43InU7BIzWmI3ng8A7gzs+zlJQPspVexTfQH62eKj+scEtnHjOYIv vQ1Q== X-Gm-Message-State: APjAAAUT1GfSY6lwgZyzFhywF7NGx0GLS7sc+IlSL2SLBu/pGM/QydHt K62Zd7atJLFUZtSmnk64A+Hk4I+W X-Google-Smtp-Source: APXvYqzqEuxD1xNZmbMDSKL37ne7ptz8NXs5MZOco3UN/enqiFIBUpH4O+E88OaZwl44F5BNBy1zpg== X-Received: by 2002:a7b:cbd6:: with SMTP id n22mr1332353wmi.57.1554158847169; Mon, 01 Apr 2019 15:47:27 -0700 (PDT) Received: from sc9-mailhost2.vmware.com ([66.170.99.2]) by smtp.gmail.com with ESMTPSA id w9sm27720055wmi.0.2019.04.01.15.47.24 (version=TLS1_2 cipher=ECDHE-RSA-AES128-SHA bits=128/128); Mon, 01 Apr 2019 15:47:26 -0700 (PDT) From: William Tu To: dev@openvswitch.org, trozet@redhat.com, bmcfall@redhat.com, echaudro@redhat.com, magnus.karlsson@gmail.com, bjorn.topel@gmail.com, tuc@vmware.com Date: Mon, 1 Apr 2019 15:46:49 -0700 Message-Id: <1554158812-44622-2-git-send-email-u9012063@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1554158812-44622-1-git-send-email-u9012063@gmail.com> References: <1554158812-44622-1-git-send-email-u9012063@gmail.com> X-Spam-Status: No, score=-1.7 required=5.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_ENVFROM_END_DIGIT,FREEMAIL_FROM, RCVD_IN_DNSWL_NONE autolearn=no version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on smtp1.linux-foundation.org Subject: [ovs-dev] [PATCH RFCv4 1/4] Add libbpf build support. X-BeenThere: ovs-dev@openvswitch.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Sender: ovs-dev-bounces@openvswitch.org Errors-To: ovs-dev-bounces@openvswitch.org Using AF_XDP depends on libbpf for managing eBPF/XDP program and maps. This patch checks if libbpf and if_xdp (XDP interface) are availble to determine if we should enable AF_XDP support. Signed-off-by: Yi-Hung Wei Co-authored-by: William Tu --- acinclude.m4 | 20 ++++++++++++++++++++ configure.ac | 1 + lib/automake.mk | 1 + 3 files changed, 22 insertions(+) diff --git a/acinclude.m4 b/acinclude.m4 index 9d6b682266e1..d0cd8aac2c1d 100644 --- a/acinclude.m4 +++ b/acinclude.m4 @@ -205,6 +205,26 @@ AC_DEFUN([OVS_CHECK_LINUX_TC], [ [Define to 1 if TCA_PEDIT_KEY_EX_HDR_TYPE_UDP is available.])]) ]) +dnl OVS_CHECK_LINUX_AF_XDP +dnl +dnl Check both Linux kernel AF_XDP and libbpf support +AC_DEFUN([OVS_CHECK_LINUX_AF_XDP], [ + AC_MSG_CHECKING([whether AF_XDP is supported]) + AC_CHECK_HEADER([bpf/libbpf.h], + [HAVE_LIBBPF=yes], + [HAVE_LIBBPF==no]) + AC_CHECK_HEADER([linux/if_xdp.h], + [HAVE_IF_XDP=yes], + [HAVE_IF_XDP=no]) + AM_CONDITIONAL([SUPPORT_AF_XDP], + [test "$HAVE_LIBBPF" = yes && test "$HAVE_IF_XDP" = yes]) + AM_COND_IF([SUPPORT_AF_XDP], [ + AC_DEFINE([HAVE_AF_XDP], [1], [Define to 1 if linux/if_xdp.h is available.]) + LIBBPF_LDADD=" -lbpf -lelf" + AC_SUBST([LIBBPF_LDADD]) + ]) +]) + dnl OVS_CHECK_DPDK dnl dnl Configure DPDK source tree diff --git a/configure.ac b/configure.ac index 8a05870a43dc..5ce7dc9b33d2 100644 --- a/configure.ac +++ b/configure.ac @@ -99,6 +99,7 @@ OVS_CHECK_SPHINX OVS_CHECK_DOT OVS_CHECK_IF_DL OVS_CHECK_STRTOK_R +OVS_CHECK_LINUX_AF_XDP AC_CHECK_DECLS([sys_siglist], [], [], [[#include ]]) AC_CHECK_MEMBERS([struct stat.st_mtim.tv_nsec, struct stat.st_mtimensec], [], [], [[#include ]]) diff --git a/lib/automake.mk b/lib/automake.mk index b1ff495ff1f3..db868933c37a 100644 --- a/lib/automake.mk +++ b/lib/automake.mk @@ -9,6 +9,7 @@ lib_LTLIBRARIES += lib/libopenvswitch.la lib_libopenvswitch_la_LIBADD = $(SSL_LIBS) lib_libopenvswitch_la_LIBADD += $(CAPNG_LDADD) +lib_libopenvswitch_la_LIBADD += $(LIBBPF_LDADD) if WIN32 lib_libopenvswitch_la_LIBADD += ${PTHREAD_LIBS} From patchwork Mon Apr 1 22:46:50 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: William Tu X-Patchwork-Id: 1073464 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Authentication-Results: ozlabs.org; spf=pass (mailfrom) smtp.mailfrom=openvswitch.org (client-ip=140.211.169.12; helo=mail.linuxfoundation.org; envelope-from=ovs-dev-bounces@openvswitch.org; receiver=) Authentication-Results: ozlabs.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: ozlabs.org; dkim=fail reason="signature verification failed" (2048-bit key; unprotected) header.d=gmail.com header.i=@gmail.com header.b="qEAewAT2"; dkim-atps=neutral Received: from mail.linuxfoundation.org (mail.linuxfoundation.org [140.211.169.12]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id 44Y6vh1xy9z9sSV for ; Tue, 2 Apr 2019 09:48:48 +1100 (AEDT) Received: from mail.linux-foundation.org (localhost [127.0.0.1]) by mail.linuxfoundation.org (Postfix) with ESMTP id BDBDD2262; Mon, 1 Apr 2019 22:47:34 +0000 (UTC) X-Original-To: dev@openvswitch.org Delivered-To: ovs-dev@mail.linuxfoundation.org Received: from smtp1.linuxfoundation.org (smtp1.linux-foundation.org [172.17.192.35]) by mail.linuxfoundation.org (Postfix) with ESMTPS id 27C941DE9 for ; Mon, 1 Apr 2019 22:47:34 +0000 (UTC) X-Greylist: whitelisted by SQLgrey-1.7.6 Received: from mail-wr1-f68.google.com (mail-wr1-f68.google.com [209.85.221.68]) by smtp1.linuxfoundation.org (Postfix) with ESMTPS id B5F1D7A6 for ; Mon, 1 Apr 2019 22:47:31 +0000 (UTC) Received: by mail-wr1-f68.google.com with SMTP id w10so14083064wrm.4 for ; Mon, 01 Apr 2019 15:47:31 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:subject:date:message-id:in-reply-to:references; bh=o1e6aLMR2xxuUYKpWLY1jHctPpQfg0upSwVaoO6v66o=; b=qEAewAT2B87rGn9WoWh6vu7YtaiEnyi1owgwFCjtoNQvDe2iEgS2R3xzmWq1J/GxI+ BuhvU9yqlCRBU+Sk0L0LExwMtmaYbV/drfrgMQijVZ4gW8ZOdUFi9A5ymrnUkQEn6nYP j93YI5R2H/9rSWO5GfxFfG6vq+XhbpaSHpInSYrXXS0tZMSeouM66Ncj1Jyr3DrxUn/z ok5LXNuLL5rPRJ/HFSrJVG2v2k3h9RqgIq03E42ZR8JWd8en1lh+iRzqi5Nv+nEuVEXz AF/Xnjm7p1ZlTXdYbs4EPsR9S3A2Oz01WRChQ9s836MXn84jl/mv9keQ4l1+zXngkFvw GPig== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:subject:date:message-id:in-reply-to :references; bh=o1e6aLMR2xxuUYKpWLY1jHctPpQfg0upSwVaoO6v66o=; b=tn5zmmcOjlovT5CtVzS/sfjjR4gp5UvKbZsCPwGw5UX85SQfilrpSNZnCTttlpdEz7 bysHJMt2ENvgpxOGae1K3ONGytAbvQUxiULkzgD7IfJxWC1Cn/kU8BovP5gwnY4aMa1x ghgAjc3rCdxBMsb7TP7yvdhkU9uZvEeOC/1izzTDEJpGdV4E+f7PI0mEtaNEY9CXR66S OIC6/vsG/CeQyHZgGpC+FDJ6VUHLTdiaKjWe5kMxokKCgQ6pSslKOtQHPyM9wUwQFVpv P1FmGFZHitbcPzrBZ2/I4mLiznrKOArOGof+FdyOIPdeI4hgZhtjT4EW38sP+PBthCSV ydgA== X-Gm-Message-State: APjAAAXfJSpw576sl+nP0O8KYSH+HyBWaNzWN28HqQVXbfYzL4uNiv1R Dt1prNeDIVwXphfwKTMAvaFUtBKW X-Google-Smtp-Source: APXvYqyfXo6HT8UI6tAHHpBeNtUFb7CbS2WzX09ZvJ2jlPBFn0WJqrwkZ+UrCUwLAYzejREApfaAAg== X-Received: by 2002:a5d:6a89:: with SMTP id s9mr44907174wru.58.1554158849747; Mon, 01 Apr 2019 15:47:29 -0700 (PDT) Received: from sc9-mailhost2.vmware.com ([66.170.99.2]) by smtp.gmail.com with ESMTPSA id w9sm27720055wmi.0.2019.04.01.15.47.27 (version=TLS1_2 cipher=ECDHE-RSA-AES128-SHA bits=128/128); Mon, 01 Apr 2019 15:47:29 -0700 (PDT) From: William Tu To: dev@openvswitch.org, trozet@redhat.com, bmcfall@redhat.com, echaudro@redhat.com, magnus.karlsson@gmail.com, bjorn.topel@gmail.com, tuc@vmware.com Date: Mon, 1 Apr 2019 15:46:50 -0700 Message-Id: <1554158812-44622-3-git-send-email-u9012063@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1554158812-44622-1-git-send-email-u9012063@gmail.com> References: <1554158812-44622-1-git-send-email-u9012063@gmail.com> X-Spam-Status: No, score=-1.7 required=5.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_ENVFROM_END_DIGIT,FREEMAIL_FROM, RCVD_IN_DNSWL_NONE autolearn=no version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on smtp1.linux-foundation.org Subject: [ovs-dev] [PATCH RFCv4 2/4] netdev-afxdp: add new netdev type for AF_XDP X-BeenThere: ovs-dev@openvswitch.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Sender: ovs-dev-bounces@openvswitch.org Errors-To: ovs-dev-bounces@openvswitch.org The patch creates a new netdev type called "afxdp" and re-uses some of the AF_XDP API implementation from xdpsock_user.c at linux sample code. By default, it binds a device's queue 0 and uses the generic XDP support to send and receive packets. Signed-off-by: William Tu --- lib/automake.mk | 6 +- lib/dp-packet.c | 12 ++ lib/dp-packet.h | 32 +++- lib/netdev-afxdp.c | 491 ++++++++++++++++++++++++++++++++++++++++++++++++++ lib/netdev-afxdp.h | 39 ++++ lib/netdev-linux.c | 78 +++++++- lib/netdev-provider.h | 1 + lib/netdev.c | 1 + lib/xdpsock.c | 179 ++++++++++++++++++ lib/xdpsock.h | 129 +++++++++++++ 10 files changed, 963 insertions(+), 5 deletions(-) create mode 100644 lib/netdev-afxdp.c create mode 100644 lib/netdev-afxdp.h create mode 100644 lib/xdpsock.c create mode 100644 lib/xdpsock.h diff --git a/lib/automake.mk b/lib/automake.mk index db868933c37a..2c7c59c3ded8 100644 --- a/lib/automake.mk +++ b/lib/automake.mk @@ -324,7 +324,11 @@ lib_libopenvswitch_la_SOURCES = \ lib/lldp/lldpd.c \ lib/lldp/lldpd.h \ lib/lldp/lldpd-structs.c \ - lib/lldp/lldpd-structs.h + lib/lldp/lldpd-structs.h \ + lib/xdpsock.c \ + lib/xdpsock.h \ + lib/netdev-afxdp.c \ + lib/netdev-afxdp.h if WIN32 lib_libopenvswitch_la_SOURCES += \ diff --git a/lib/dp-packet.c b/lib/dp-packet.c index 93b0e9c84793..5c9dc7ab31d7 100644 --- a/lib/dp-packet.c +++ b/lib/dp-packet.c @@ -121,6 +121,15 @@ dp_packet_uninit(struct dp_packet *b) * created as a dp_packet */ free_dpdk_buf((struct dp_packet*) b); #endif + } else if (b->source == DPBUF_AFXDP) { +#ifdef HAVE_AF_XDP + struct dp_packet_afxdp *xpacket; + + xpacket = dp_packet_cast_afxdp(b); + if (xpacket->mpool) + umem_elem_push(xpacket->mpool, dp_packet_base(b)); +#endif + return; } } } @@ -249,6 +258,8 @@ dp_packet_resize__(struct dp_packet *b, size_t new_headroom, size_t new_tailroom case DPBUF_STACK: OVS_NOT_REACHED(); + case DPBUF_AFXDP: + OVS_NOT_REACHED(); case DPBUF_STUB: b->source = DPBUF_MALLOC; new_base = xmalloc(new_allocated); @@ -434,6 +445,7 @@ dp_packet_steal_data(struct dp_packet *b) { void *p; ovs_assert(b->source != DPBUF_DPDK); + ovs_assert(b->source != DPBUF_AFXDP); if (b->source == DPBUF_MALLOC && dp_packet_data(b) == dp_packet_base(b)) { p = dp_packet_data(b); diff --git a/lib/dp-packet.h b/lib/dp-packet.h index c297a8f509ab..90282414eee9 100644 --- a/lib/dp-packet.h +++ b/lib/dp-packet.h @@ -30,6 +30,7 @@ #include "packets.h" #include "util.h" #include "flow.h" +#include "lib/xdpsock.h" #ifdef __cplusplus extern "C" { @@ -42,10 +43,10 @@ enum OVS_PACKED_ENUM dp_packet_source { DPBUF_DPDK, /* buffer data is from DPDK allocated memory. * ref to dp_packet_init_dpdk() in dp-packet.c. */ + DPBUF_AFXDP, /* buffer data from XDP frame */ }; #define DP_PACKET_CONTEXT_SIZE 64 - /* Buffer for holding packet data. A dp_packet is automatically reallocated * as necessary if it grows too large for the available memory. * By default the packet type is set to Ethernet (PT_ETH). @@ -80,6 +81,20 @@ struct dp_packet { }; }; +struct dp_packet_afxdp { + struct umem_pool *mpool; + struct dp_packet packet; +}; + +#if HAVE_AF_XDP +static struct dp_packet_afxdp * +dp_packet_cast_afxdp(const struct dp_packet *d OVS_UNUSED) +{ + ovs_assert(d->source == DPBUF_AFXDP); + return CONTAINER_OF(d, struct dp_packet_afxdp, packet); +} +#endif + static inline void *dp_packet_data(const struct dp_packet *); static inline void dp_packet_set_data(struct dp_packet *, void *); static inline void *dp_packet_base(const struct dp_packet *); @@ -174,7 +189,20 @@ dp_packet_delete(struct dp_packet *b) free_dpdk_buf((struct dp_packet*) b); return; } - + if (b->source == DPBUF_AFXDP) { +#ifdef HAVE_AF_XDP + struct dp_packet_afxdp *xpacket; + + /* if a packet is received from afxdp port, + * and tx to a system port. Then we need to + * push the rx umem back here + */ + xpacket = dp_packet_cast_afxdp(b); + if (xpacket->mpool) + umem_elem_push(xpacket->mpool, dp_packet_base(b)); +#endif + return; + } dp_packet_uninit(b); free(b); } diff --git a/lib/netdev-afxdp.c b/lib/netdev-afxdp.c new file mode 100644 index 000000000000..b3bf2f044860 --- /dev/null +++ b/lib/netdev-afxdp.c @@ -0,0 +1,491 @@ +/* + * Copyright (c) 2018, 2019 Nicira, Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at: + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +#include +#include "netdev-linux.h" +#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 + +#include "coverage.h" +#include "dp-packet.h" +#include "dpif-netlink.h" +#include "dpif-netdev.h" +#include "openvswitch/dynamic-string.h" +#include "fatal-signal.h" +#include "hash.h" +#include "openvswitch/hmap.h" +#include "netdev-provider.h" +#include "netdev-tc-offloads.h" +#include "netdev-vport.h" +#include "netlink-notifier.h" +#include "netlink-socket.h" +#include "netlink.h" +#include "netnsid.h" +#include "openvswitch/ofpbuf.h" +#include "openflow/openflow.h" +#include "ovs-atomic.h" +#include "packets.h" +#include "openvswitch/poll-loop.h" +#include "rtnetlink.h" +#include "openvswitch/shash.h" +#include "socket-util.h" +#include "sset.h" +#include "tc.h" +#include "timer.h" +#include "unaligned.h" +#include "openvswitch/vlog.h" +#include "util.h" +#include "xdpsock.h" +#include "netdev-afxdp.h" + +#ifdef HAVE_AF_XDP +#ifndef SOL_XDP +#define SOL_XDP 283 +#endif +#ifndef AF_XDP +#define AF_XDP 44 +#endif +#ifndef PF_XDP +#define PF_XDP AF_XDP +#endif + +VLOG_DEFINE_THIS_MODULE(netdev_afxdp); +static struct vlog_rate_limit rl = VLOG_RATE_LIMIT_INIT(5, 20); + +#define UMEM2DESC(elem, base) ((uint64_t)((char *)elem - (char *)base)) +#define UMEM2XPKT(base, i) \ + (struct dp_packet_afxdp *)((char *)base + i * sizeof(struct dp_packet_afxdp)) + +static uint32_t opt_xdp_bind_flags = XDP_COPY; +static uint32_t opt_xdp_flags = XDP_FLAGS_UPDATE_IF_NOEXIST | XDP_FLAGS_SKB_MODE; +static uint32_t prog_id; + +static struct xsk_umem_info *xsk_configure_umem(void *buffer, uint64_t size) +{ + struct xsk_umem_info *umem; + int ret; + int i; + + umem = calloc(1, sizeof(*umem)); + if (!umem) { + VLOG_FATAL("xsk config umem failed (%s)", ovs_strerror(errno)); + } + + ret = xsk_umem__create(&umem->umem, buffer, size, &umem->fq, &umem->cq, + NULL); + + if (ret) { + VLOG_FATAL("xsk umem create failed (%s)", ovs_strerror(errno)); + } + + umem->buffer = buffer; + + /* set-up umem pool */ + umem_pool_init(&umem->mpool, NUM_FRAMES); + + for (i = NUM_FRAMES - 1; i >= 0; i--) { + struct umem_elem *elem; + + elem = (struct umem_elem *)((char *)umem->buffer + + i * FRAME_SIZE); + umem_elem_push(&umem->mpool, elem); + } + + /* set-up metadata */ + xpacket_pool_init(&umem->xpool, NUM_FRAMES); + + VLOG_DBG("%s xpacket pool from %p to %p", __func__, + umem->xpool.array, + (char *)umem->xpool.array + + NUM_FRAMES * sizeof(struct dp_packet_afxdp)); + + for (i = NUM_FRAMES - 1; i >= 0; i--) { + struct dp_packet_afxdp *xpacket; + struct dp_packet *packet; + + xpacket = UMEM2XPKT(umem->xpool.array, i); + xpacket->mpool = &umem->mpool; + + packet = &xpacket->packet; + packet->source = DPBUF_AFXDP; + } + + return umem; +} + +static struct xsk_socket_info * +xsk_configure_socket(struct xsk_umem_info *umem, uint32_t ifindex, + uint32_t queue_id) +{ + struct xsk_socket_config cfg; + struct xsk_socket_info *xsk; + char devname[IF_NAMESIZE]; + uint32_t idx; + int ret; + int i; + + xsk = calloc(1, sizeof(*xsk)); + if (!xsk) { + VLOG_FATAL("xsk create failed (%s)", ovs_strerror(errno)); + } + + xsk->umem = umem; + cfg.rx_size = CONS_NUM_DESCS; + cfg.tx_size = PROD_NUM_DESCS; + cfg.libbpf_flags = 0; + cfg.xdp_flags = opt_xdp_flags; + cfg.bind_flags = opt_xdp_bind_flags; + + if (if_indextoname(ifindex, devname) == NULL) { + VLOG_FATAL("ifindex %d devname failed (%s)", + ifindex, ovs_strerror(errno)); + } + + ret = xsk_socket__create(&xsk->xsk, devname, queue_id, umem->umem, + &xsk->rx, &xsk->tx, &cfg); + if (ret) { + VLOG_FATAL("xsk create failed (%s)", ovs_strerror(errno)); + } + + /* make sure the XDP program is there */ + ret = bpf_get_link_xdp_id(ifindex, &prog_id, opt_xdp_flags); + if (ret) { + VLOG_FATAL("get XDP prog ID failed (%s)", ovs_strerror(errno)); + } + + ret = xsk_ring_prod__reserve(&xsk->umem->fq, + PROD_NUM_DESCS, + &idx); + if (ret != PROD_NUM_DESCS) { + VLOG_FATAL("fq set-up failed (%s)", ovs_strerror(errno)); + } + + for (i = 0; + i < PROD_NUM_DESCS * FRAME_SIZE; + i += FRAME_SIZE) { + struct umem_elem *elem; + uint64_t addr; + + elem = umem_elem_pop(&xsk->umem->mpool); + addr = UMEM2DESC(elem, xsk->umem->buffer); + + *xsk_ring_prod__fill_addr(&xsk->umem->fq, idx++) = addr; + } + + xsk_ring_prod__submit(&xsk->umem->fq, + PROD_NUM_DESCS); + return xsk; +} + +struct xsk_socket_info * +xsk_configure(int ifindex, int xdp_queue_id) +{ + struct xsk_socket_info *xsk; + struct xsk_umem_info *umem; + void *bufs; + int ret; + + ret = posix_memalign(&bufs, getpagesize(), + NUM_FRAMES * FRAME_SIZE); + ovs_assert(!ret); + + /* Create sockets... */ + umem = xsk_configure_umem(bufs, + NUM_FRAMES * FRAME_SIZE); + xsk = xsk_configure_socket(umem, ifindex, xdp_queue_id); + return xsk; +} + +static void OVS_UNUSED vlog_hex_dump(const void *buf, size_t count) +{ + struct ds ds = DS_EMPTY_INITIALIZER; + ds_put_hex_dump(&ds, buf, count, 0, false); + VLOG_DBG_RL(&rl, "%s", ds_cstr(&ds)); + ds_destroy(&ds); +} + +void +xsk_destroy(struct xsk_socket_info *xsk, uint32_t ifindex) +{ + struct xsk_umem *umem = xsk->umem->umem; + uint32_t curr_prog_id = 0; + + xsk_socket__delete(xsk->xsk); + (void)xsk_umem__delete(umem); + + /* cleanup umem pool */ + umem_pool_cleanup(&xsk->umem->mpool); + + /* cleanup metadata pool */ + xpacket_pool_cleanup(&xsk->umem->xpool); + + /* remove_xdp_program() */ + if (bpf_get_link_xdp_id(ifindex, &curr_prog_id, opt_xdp_flags)) { + bpf_set_link_xdp_fd(ifindex, -1, opt_xdp_flags); + } + if (prog_id == curr_prog_id) { + bpf_set_link_xdp_fd(ifindex, -1, opt_xdp_flags); + } else if (!curr_prog_id) { + VLOG_WARN("couldn't find a prog id on a given interface"); + } else { + VLOG_WARN("program on interface changed, not removing"); + } + + return; +} + +static inline void +print_xsk_stat(struct xsk_socket_info *xsk OVS_UNUSED) { + struct xdp_statistics stat; + socklen_t optlen; + + optlen = sizeof(stat); + ovs_assert(getsockopt(xsk_socket__fd(xsk->xsk), SOL_XDP, XDP_STATISTICS, + &stat, &optlen) == 0); + + VLOG_DBG_RL(&rl, "rx dropped %llu, rx_invalid %llu, tx_invalid %llu", + stat.rx_dropped, stat.rx_invalid_descs, stat.tx_invalid_descs); + return; +} + +/* Receive packet from AF_XDP socket */ +int +netdev_linux_rxq_xsk(struct xsk_socket_info *xsk, + struct dp_packet_batch *batch) +{ + unsigned int rcvd, i; + uint32_t idx_rx = 0, idx_fq = 0; + int ret = 0; + + unsigned int non_afxdp; + + /* See if there is any packet on RX queue, + * if yes, idx_rx is the index having the packet. + */ + rcvd = xsk_ring_cons__peek(&xsk->rx, BATCH_SIZE, &idx_rx); + if (!rcvd) { + return 0; + } + + /* Form a dp_packet batch from descriptor in RX queue */ + for (i = 0; i < rcvd; i++) { + uint64_t addr = xsk_ring_cons__rx_desc(&xsk->rx, idx_rx)->addr; + uint32_t len = xsk_ring_cons__rx_desc(&xsk->rx, idx_rx)->len; + char *pkt = xsk_umem__get_data(xsk->umem->buffer, addr); + uint64_t index; + + struct dp_packet_afxdp *xpacket; + struct dp_packet *packet; + + index = addr >> FRAME_SHIFT; + xpacket = UMEM2XPKT(xsk->umem->xpool.array, index); + + packet = &xpacket->packet; + xpacket->mpool = &xsk->umem->mpool; + + vlog_hex_dump(pkt, 20); + if (packet->source != DPBUF_AFXDP) { + /* FIXME: might be a bug */ + VLOG_WARN_RL(&rl, "invalid packet: %d", non_afxdp++); + continue; + } + + /* Initialize the struct dp_packet */ + dp_packet_set_base(packet, pkt); + dp_packet_set_data(packet, pkt); + dp_packet_set_size(packet, len); + + /* Add packet into batch, increase batch->count */ + dp_packet_batch_add(batch, packet); + + idx_rx++; + } + + /* We've consume rcvd packets in RX, now re-fill the + * same number back to FILL queue. + */ + for (i = 0; i < rcvd; i++) { + uint64_t index; + struct umem_elem *elem; + + ret = xsk_ring_prod__reserve(&xsk->umem->fq, 1, &idx_fq); + while (ret == 0) { + /* The FILL queue is full, so retry. (or skip)? */ + ret = xsk_ring_prod__reserve(&xsk->umem->fq, 1, &idx_fq); + } + + /* Get one free umem, program it into FILL queue */ + elem = umem_elem_pop(&xsk->umem->mpool); + index = (uint64_t)((char *)elem - (char *)xsk->umem->buffer); + ovs_assert((index & FRAME_SHIFT_MASK) == 0); + *xsk_ring_prod__fill_addr(&xsk->umem->fq, idx_fq) = index; + + idx_fq++; + } + xsk_ring_prod__submit(&xsk->umem->fq, rcvd); + + /* Release the RX queue */ + xsk_ring_cons__release(&xsk->rx, rcvd); + xsk->rx_npkts += rcvd; + + print_xsk_stat(xsk); + return 0; +} + +static void kick_tx(struct xsk_socket_info *xsk) +{ + int ret; + + ret = sendto(xsk_socket__fd(xsk->xsk), NULL, 0, MSG_DONTWAIT, NULL, 0); + if (ret >= 0 || errno == ENOBUFS || errno == EAGAIN || errno == EBUSY) + return; + VLOG_FATAL("kick tx failed (%s)", ovs_strerror(errno)); +} + +/* + * A dp_packet might come from + * 1) AFXDP buffer + * 2) non-AFXDP buffer, ex: send from tap device + */ +int +netdev_linux_afxdp_batch_send(struct xsk_socket_info *xsk, + struct dp_packet_batch *batch) +{ + uint32_t tx_done, idx_cq = 0; + struct dp_packet *packet; + uint32_t idx; + int j; + + /* Make sure we have enough TX descs */ + if (xsk_ring_prod__reserve(&xsk->tx, batch->count, &idx) == 0) { + return -EAGAIN; + } + + DP_PACKET_BATCH_FOR_EACH (i, packet, batch) { + struct dp_packet_afxdp *xpacket; + struct umem_elem *elem; + uint64_t index; + + elem = umem_elem_pop(&xsk->umem->mpool); + if (!elem) { + VLOG_ERR_RL(&rl, "no available elem!"); + return -EAGAIN; + } + + memcpy(elem, dp_packet_data(packet), dp_packet_size(packet)); + + index = (uint64_t)((char *)elem - (char *)xsk->umem->buffer); + xsk_ring_prod__tx_desc(&xsk->tx, idx + i)->addr = index; + xsk_ring_prod__tx_desc(&xsk->tx, idx + i)->len + = dp_packet_size(packet); + + vlog_hex_dump(dp_packet_data(packet), 20); + + if (packet->source == DPBUF_AFXDP) { + xpacket = dp_packet_cast_afxdp(packet); + umem_elem_push(xpacket->mpool, dp_packet_base(packet)); + /* Avoid freeing it twice at dp_packet_uninit */ + xpacket->mpool = NULL; + } + } + xsk_ring_prod__submit(&xsk->tx, batch->count); + xsk->outstanding_tx += batch->count; + +retry: + kick_tx(xsk); + + /* Process CQ */ + tx_done = xsk_ring_cons__peek(&xsk->umem->cq, batch->count, &idx_cq); + if (tx_done > 0) { + xsk->outstanding_tx -= tx_done; + xsk->tx_npkts += tx_done; + VLOG_DBG_RL(&rl, "%s complete %d tx", __func__, tx_done); + } + + /* Recycle back to umem pool */ + for (j = 0; j < tx_done; j++) { + struct umem_elem *elem; + uint64_t addr; + + addr = *xsk_ring_cons__comp_addr(&xsk->umem->cq, idx_cq++); + + elem = (struct umem_elem *)((char *)xsk->umem->buffer + addr); + umem_elem_push(&xsk->umem->mpool, elem); + } + xsk_ring_cons__release(&xsk->umem->cq, tx_done); + + if (xsk->outstanding_tx > PROD_NUM_DESCS / 2) { + /* If there are still a lot not transmitted, + * try harder. + */ + goto retry; + } + + return 0; +} + +#else +struct xsk_socket_info * +xsk_configure(int ifindex OVS_UNUSED, int xdp_queue_id OVS_UNUSED) +{ + return NULL; +} + +void +xsk_destroy(struct xsk_socket_info *xsk OVS_UNUSED, uint32_t ifindex OVS_UNUSED) +{ + return; +} + +int +netdev_linux_rxq_xsk(struct xsk_socket_info *xsk OVS_UNUSED, + struct dp_packet_batch *batch OVS_UNUSED) +{ + return 0; +} + +int +netdev_linux_afxdp_batch_send(struct xsk_socket_info *xsk OVS_UNUSED, + struct dp_packet_batch *batch OVS_UNUSED) +{ + return 0; +} +#endif diff --git a/lib/netdev-afxdp.h b/lib/netdev-afxdp.h new file mode 100644 index 000000000000..2f455aee2bd4 --- /dev/null +++ b/lib/netdev-afxdp.h @@ -0,0 +1,39 @@ +/* + * Copyright (c) 2018 Nicira, Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at: + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +#ifndef NETDEV_AFXDP_H +#define NETDEV_AFXDP_H 1 + +#include +#include + +/* These functions are Linux AF_XDP specific, so they should be used directly + * only by Linux-specific code. */ +struct netdev; +struct xsk_socket_info; +struct xdp_umem; +struct dp_packet_batch; + +struct xsk_socket_info *xsk_configure(int ifindex, int xdp_queue_id); +void xsk_destroy(struct xsk_socket_info *xsk, uint32_t ifindex); + +int netdev_linux_rxq_xsk(struct xsk_socket_info *xsk, + struct dp_packet_batch *batch); + +int netdev_linux_afxdp_batch_send(struct xsk_socket_info *xsk, + struct dp_packet_batch *batch); + +#endif /* netdev-afxdp.h */ diff --git a/lib/netdev-linux.c b/lib/netdev-linux.c index 25d037cb643f..36e9882d22a0 100644 --- a/lib/netdev-linux.c +++ b/lib/netdev-linux.c @@ -74,6 +74,7 @@ #include "unaligned.h" #include "openvswitch/vlog.h" #include "util.h" +#include "netdev-afxdp.h" VLOG_DEFINE_THIS_MODULE(netdev_linux); @@ -523,6 +524,7 @@ struct netdev_linux { /* LAG information. */ bool is_lag_master; /* True if the netdev is a LAG master. */ + struct xsk_socket_info *xsk[1]; /* af_xdp socket: use only one queue */ }; struct netdev_rxq_linux { @@ -572,6 +574,12 @@ is_netdev_linux_class(const struct netdev_class *netdev_class) } static bool +is_afxdp_netdev(const struct netdev *netdev) +{ + return netdev_get_class(netdev) == &netdev_afxdp_class; +} + +static bool is_tap_netdev(const struct netdev *netdev) { return netdev_get_class(netdev) == &netdev_tap_class; @@ -1073,6 +1081,18 @@ netdev_linux_destruct(struct netdev *netdev_) atomic_count_dec(&miimon_cnt); } + if (is_afxdp_netdev(netdev_)) { + int ifindex; + int ret; + + ret = get_ifindex(netdev_, &ifindex); + if (ret) { + VLOG_ERR("get ifindex error"); + } else { + xsk_destroy(netdev->xsk[0], ifindex); + } + } + ovs_mutex_destroy(&netdev->mutex); } @@ -1102,6 +1122,30 @@ netdev_linux_rxq_construct(struct netdev_rxq *rxq_) rx->is_tap = is_tap_netdev(netdev_); if (rx->is_tap) { rx->fd = netdev->tap_fd; + } else if (is_afxdp_netdev(netdev_)) { + struct rlimit r = {RLIM_INFINITY, RLIM_INFINITY}; + int ifindex, num_socks = 0; + int xdp_queue_id = 0; + struct xsk_socket_info *xsk; + + if (setrlimit(RLIMIT_MEMLOCK, &r)) { + VLOG_ERR("ERROR: setrlimit(RLIMIT_MEMLOCK) \"%s\"\n", + ovs_strerror(errno)); + ovs_assert(0); + } + + VLOG_DBG("%s: %s: queue=%d configuring xdp sock", + __func__, netdev_->name, xdp_queue_id); + + /* Get ethernet device index. */ + error = get_ifindex(&netdev->up, &ifindex); + if (error) { + goto error; + } + + xsk = xsk_configure(ifindex, xdp_queue_id); + netdev->xsk[num_socks++] = xsk; + rx->fd = xsk_socket__fd(xsk->xsk); /* for netdev layer to poll */ } else { struct sockaddr_ll sll; int ifindex, val; @@ -1307,9 +1351,14 @@ netdev_linux_rxq_recv(struct netdev_rxq *rxq_, struct dp_packet_batch *batch, { struct netdev_rxq_linux *rx = netdev_rxq_linux_cast(rxq_); struct netdev *netdev = rx->up.netdev; - struct dp_packet *buffer; + struct dp_packet *buffer = NULL; ssize_t retval; int mtu; + struct netdev_linux *netdev_ = netdev_linux_cast(netdev); + + if (is_afxdp_netdev(netdev)) { + return netdev_linux_rxq_xsk(netdev_->xsk[0], batch); + } if (netdev_linux_get_mtu__(netdev_linux_cast(netdev), &mtu)) { mtu = ETH_PAYLOAD_MAX; @@ -1318,6 +1367,7 @@ netdev_linux_rxq_recv(struct netdev_rxq *rxq_, struct dp_packet_batch *batch, /* Assume Ethernet port. No need to set packet_type. */ buffer = dp_packet_new_with_headroom(VLAN_ETH_HEADER_LEN + mtu, DP_NETDEV_HEADROOM); + retval = (rx->is_tap ? netdev_linux_rxq_recv_tap(rx->fd, buffer) : netdev_linux_rxq_recv_sock(rx->fd, buffer)); @@ -1328,6 +1378,13 @@ netdev_linux_rxq_recv(struct netdev_rxq *rxq_, struct dp_packet_batch *batch, netdev_rxq_get_name(rxq_), ovs_strerror(errno)); } dp_packet_delete(buffer); + } else if (is_afxdp_netdev(netdev)) { + dp_packet_batch_init_packet_fields(batch); + + if (batch->count != 0) + VLOG_DBG("%s AFXDP recv %lu packets", __func__, batch->count); + + return retval; } else { dp_packet_batch_init_packet(batch, buffer); } @@ -1469,7 +1526,8 @@ netdev_linux_send(struct netdev *netdev_, int qid OVS_UNUSED, int error = 0; int sock = 0; - if (!is_tap_netdev(netdev_)) { + if (!is_tap_netdev(netdev_) && + !is_afxdp_netdev(netdev_)) { if (netdev_linux_netnsid_is_remote(netdev_linux_cast(netdev_))) { error = EOPNOTSUPP; goto free_batch; @@ -1488,6 +1546,10 @@ netdev_linux_send(struct netdev *netdev_, int qid OVS_UNUSED, } error = netdev_linux_sock_batch_send(sock, ifindex, batch); + } else if (is_afxdp_netdev(netdev_)) { + struct netdev_linux *netdev = netdev_linux_cast(netdev_); + + error = netdev_linux_afxdp_batch_send(netdev->xsk[0], batch); } else { error = netdev_linux_tap_batch_send(netdev_, batch); } @@ -3205,6 +3267,7 @@ const struct netdev_class netdev_linux_class = { NETDEV_LINUX_CLASS_COMMON, LINUX_FLOW_OFFLOAD_API, .type = "system", + .is_pmd = false, .construct = netdev_linux_construct, .get_stats = netdev_linux_get_stats, .get_features = netdev_linux_get_features, @@ -3215,6 +3278,7 @@ const struct netdev_class netdev_linux_class = { const struct netdev_class netdev_tap_class = { NETDEV_LINUX_CLASS_COMMON, .type = "tap", + .is_pmd = false, .construct = netdev_linux_construct_tap, .get_stats = netdev_tap_get_stats, .get_features = netdev_linux_get_features, @@ -3224,10 +3288,20 @@ const struct netdev_class netdev_tap_class = { const struct netdev_class netdev_internal_class = { NETDEV_LINUX_CLASS_COMMON, .type = "internal", + .is_pmd = false, .construct = netdev_linux_construct, .get_stats = netdev_internal_get_stats, .get_status = netdev_internal_get_status, }; + +const struct netdev_class netdev_afxdp_class = { + NETDEV_LINUX_CLASS_COMMON, + .type = "afxdp", + .is_pmd = true, + .construct = netdev_linux_construct, + .get_stats = netdev_linux_get_stats, + .get_status = netdev_linux_get_status, +}; #define CODEL_N_QUEUES 0x0000 diff --git a/lib/netdev-provider.h b/lib/netdev-provider.h index fb0c27e6e8e8..5bf041316503 100644 --- a/lib/netdev-provider.h +++ b/lib/netdev-provider.h @@ -902,6 +902,7 @@ extern const struct netdev_class netdev_linux_class; #endif extern const struct netdev_class netdev_internal_class; extern const struct netdev_class netdev_tap_class; +extern const struct netdev_class netdev_afxdp_class; #ifdef __cplusplus } diff --git a/lib/netdev.c b/lib/netdev.c index 45b50f26c3f9..5197864ca1c5 100644 --- a/lib/netdev.c +++ b/lib/netdev.c @@ -146,6 +146,7 @@ netdev_initialize(void) netdev_register_provider(&netdev_linux_class); netdev_register_provider(&netdev_internal_class); netdev_register_provider(&netdev_tap_class); + netdev_register_provider(&netdev_afxdp_class); netdev_vport_tunnel_register(); #endif #if defined(__FreeBSD__) || defined(__NetBSD__) diff --git a/lib/xdpsock.c b/lib/xdpsock.c new file mode 100644 index 000000000000..c50927b5f152 --- /dev/null +++ b/lib/xdpsock.c @@ -0,0 +1,179 @@ +/* + * Copyright (c) 2018, 2019 Nicira, Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at: + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include "openvswitch/vlog.h" +#include "async-append.h" +#include "coverage.h" +#include "dirs.h" +#include "openvswitch/dynamic-string.h" +#include "openvswitch/ofpbuf.h" +#include "ovs-thread.h" +#include "sat-math.h" +#include "socket-util.h" +#include "svec.h" +#include "syslog-direct.h" +#include "syslog-libc.h" +#include "syslog-provider.h" +#include "timeval.h" +#include "unixctl.h" +#include "util.h" +#include "ovs-atomic.h" +#include "xdpsock.h" +#include "openvswitch/compiler.h" +#include "dp-packet.h" + +#ifdef HAVE_AF_XDP +void +__umem_elem_push_n(struct umem_pool *umemp OVS_UNUSED, void **addrs, int n) +{ + void *ptr; + + if (OVS_UNLIKELY(umemp->index + n > umemp->size)) { + OVS_NOT_REACHED(); + } + + ptr = &umemp->array[umemp->index]; + memcpy(ptr, addrs, n * sizeof(void *)); + umemp->index += n; +} + +inline void +__umem_elem_push(struct umem_pool *umemp OVS_UNUSED, void *addr) +{ + umemp->array[umemp->index++] = addr; +} + +void +umem_elem_push(struct umem_pool *umemp OVS_UNUSED, void *addr) +{ + + if (OVS_UNLIKELY(umemp->index >= umemp->size)) { + /* stack is full */ + /* it's possible that one umem gets pushed twice, + * because actions=1,2,3... multiple ports? + */ + OVS_NOT_REACHED(); + } + + ovs_assert(((uint64_t)addr & FRAME_SHIFT_MASK) == 0); + + ovs_mutex_lock(&umemp->mutex); + __umem_elem_push(umemp, addr); + ovs_mutex_unlock(&umemp->mutex); +} + +void +__umem_elem_pop_n(struct umem_pool *umemp OVS_UNUSED, void **addrs, int n) +{ + void *ptr; + + umemp->index -= n; + + if (OVS_UNLIKELY(umemp->index < 0)) { + OVS_NOT_REACHED(); + } + + ptr = &umemp->array[umemp->index]; + memcpy(addrs, ptr, n * sizeof(void *)); +} + +inline void * +__umem_elem_pop(struct umem_pool *umemp OVS_UNUSED) +{ + return umemp->array[--umemp->index]; +} + +void * +umem_elem_pop(struct umem_pool *umemp OVS_UNUSED) +{ + void *ptr; + + ovs_mutex_lock(&umemp->mutex); + ptr = __umem_elem_pop(umemp); + ovs_mutex_unlock(&umemp->mutex); + + return ptr; +} + +void ** +__umem_pool_alloc(unsigned int size) +{ + void *bufs; + + ovs_assert(posix_memalign(&bufs, getpagesize(), + size * sizeof(void *)) == 0); + memset(bufs, 0, size * sizeof(void *)); + return (void **)bufs; +} + +unsigned int +umem_elem_count(struct umem_pool *mpool) +{ + return mpool->index; +} + +int +umem_pool_init(struct umem_pool *umemp OVS_UNUSED, unsigned int size) +{ + umemp->array = __umem_pool_alloc(size); + if (!umemp->array) + OVS_NOT_REACHED(); + + umemp->size = size; + umemp->index = 0; + ovs_mutex_init(&umemp->mutex); + return 0; +} + +void +umem_pool_cleanup(struct umem_pool *umemp OVS_UNUSED) +{ + free(umemp->array); +} + +/* AF_XDP metadata init/destroy */ +int +xpacket_pool_init(struct xpacket_pool *xp, unsigned int size) +{ + void *bufs; + + ovs_assert(posix_memalign(&bufs, getpagesize(), + size * sizeof(struct dp_packet_afxdp)) == 0); + memset(bufs, 0, size * sizeof(struct dp_packet_afxdp)); + + xp->array = bufs; + xp->size = size; + return 0; +} + +void +xpacket_pool_cleanup(struct xpacket_pool *xp) +{ + free(xp->array); +} +#else /* !HAVE_AF_XDP below */ +#endif diff --git a/lib/xdpsock.h b/lib/xdpsock.h new file mode 100644 index 000000000000..8df8fa45105f --- /dev/null +++ b/lib/xdpsock.h @@ -0,0 +1,129 @@ +/* + * Copyright (c) 2018, 2019 Nicira, Inc. + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at: + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +#ifndef XDPSOCK_H +#define XDPSOCK_H 1 +#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 "ovs-atomic.h" +#include "openvswitch/thread.h" + +/* bpf/xsk.h uses the following macros not defined in OVS, + * so re-define them before include. + */ +#define unlikely OVS_UNLIKELY +#define likely OVS_LIKELY +#define barrier() __asm__ __volatile__("": : :"memory") +#define smp_rmb() barrier() +#define smp_wmb() barrier() +#include + +#define FRAME_HEADROOM 0 +#define FRAME_SIZE XSK_UMEM__DEFAULT_FRAME_SIZE +#define BATCH_SIZE NETDEV_MAX_BURST +#define FRAME_SHIFT XSK_UMEM__DEFAULT_FRAME_SHIFT +#define FRAME_SHIFT_MASK ((1< X-Patchwork-Id: 1073466 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Authentication-Results: ozlabs.org; spf=pass (mailfrom) smtp.mailfrom=openvswitch.org (client-ip=140.211.169.12; helo=mail.linuxfoundation.org; envelope-from=ovs-dev-bounces@openvswitch.org; receiver=) Authentication-Results: ozlabs.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: ozlabs.org; dkim=fail reason="signature verification failed" (2048-bit key; unprotected) header.d=gmail.com header.i=@gmail.com header.b="Y3rRTl5M"; dkim-atps=neutral Received: from mail.linuxfoundation.org (mail.linuxfoundation.org [140.211.169.12]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id 44Y6wz6C1Cz9sSj for ; Tue, 2 Apr 2019 09:49:55 +1100 (AEDT) Received: from mail.linux-foundation.org (localhost [127.0.0.1]) by mail.linuxfoundation.org (Postfix) with ESMTP id 9BC1A228F; Mon, 1 Apr 2019 22:47:39 +0000 (UTC) X-Original-To: dev@openvswitch.org Delivered-To: ovs-dev@mail.linuxfoundation.org Received: from smtp1.linuxfoundation.org (smtp1.linux-foundation.org [172.17.192.35]) by mail.linuxfoundation.org (Postfix) with ESMTPS id 44DB0227A for ; Mon, 1 Apr 2019 22:47:38 +0000 (UTC) X-Greylist: whitelisted by SQLgrey-1.7.6 Received: from mail-wm1-f46.google.com (mail-wm1-f46.google.com [209.85.128.46]) by smtp1.linuxfoundation.org (Postfix) with ESMTPS id 882F6712 for ; Mon, 1 Apr 2019 22:47:34 +0000 (UTC) Received: by mail-wm1-f46.google.com with SMTP id 4so1126051wmf.1 for ; Mon, 01 Apr 2019 15:47:34 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:subject:date:message-id:in-reply-to:references; bh=RKVK+Dqs+ximp5JProj2MAQzx4SP2KPNQwvR98HJIuY=; b=Y3rRTl5MJjMaLBKAFHo+QwD3+xUKrvcfItJ38Zrurxtajk5N1aHNZpEp9HtaBdotKc ED2eTthBgp6PtBEO5tfHveEWFgmaNvZxSyWyZ562KigNth761/xZArC4Mdg7HIBk9Ddh EuGlD3hL+Mn6WVaTsfWgp64DeGDXdbT/riv82N54dHtwEHMVmMcyiNJt2v9yYUz+1M23 +9KN7hAkGSDD7raB9uKrMN5i1BRXa8AaMz0dfZ5Lxlw6mlRn32MnR90qyk33ASa7Hwo6 DJCxznMG+Nla8UMJAt8oGABI4J1mMlQR36ZbfvTnl81D2SWjJpzBzzcXlfdYVf7z6BDF D3KA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:subject:date:message-id:in-reply-to :references; bh=RKVK+Dqs+ximp5JProj2MAQzx4SP2KPNQwvR98HJIuY=; b=ghs8QrTrHoz2qK5yrdMWSK+yj3fd9pKEOeaxTg2vKd/DKpqt7ajFvtdr0jjy5yZ/j1 lo//h8cncOWMuAph3DJm4IY8dmMkRgGjDN1Qj1FtJMtSKmAJP0jUmyOFh2A+l0EaEn1+ EpUPOvZcSP2R3zWzKU74AR7SCVkCFlA2EBPcwnt/ktNsaQ/qfyNA50wwVikVImcnUt0m 3Pv3y9Wwtq4gn0Zxe3gWs96pWIUs0W8e1mi6y44noT+ImVHBQjg9g0ePN4XLgI0GK90Z axBeMCExP+BivTCi3B/P9zLByq3NiKv7OTRfLYWuU/8t7XOF+r7b1Ztv4U1cxfhpv60v wF4g== X-Gm-Message-State: APjAAAXLexrygUlpJqsF1sKUxp4S2/A3jLi7XCcDo8aPPF1G1kZ8shog B4OZRc0YnMoHrHsIAONoNI9mpyLI X-Google-Smtp-Source: APXvYqzAgmGz4Foudi4/uhtaIGwp9UQ5Iq/q8pjdmYfEIaJ+oJp1BBWw+LkoGXxwJdTzBXIloq492Q== X-Received: by 2002:a1c:988d:: with SMTP id a135mr1341833wme.24.1554158852043; Mon, 01 Apr 2019 15:47:32 -0700 (PDT) Received: from sc9-mailhost2.vmware.com ([66.170.99.2]) by smtp.gmail.com with ESMTPSA id w9sm27720055wmi.0.2019.04.01.15.47.29 (version=TLS1_2 cipher=ECDHE-RSA-AES128-SHA bits=128/128); Mon, 01 Apr 2019 15:47:31 -0700 (PDT) From: William Tu To: dev@openvswitch.org, trozet@redhat.com, bmcfall@redhat.com, echaudro@redhat.com, magnus.karlsson@gmail.com, bjorn.topel@gmail.com, tuc@vmware.com Date: Mon, 1 Apr 2019 15:46:51 -0700 Message-Id: <1554158812-44622-4-git-send-email-u9012063@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1554158812-44622-1-git-send-email-u9012063@gmail.com> References: <1554158812-44622-1-git-send-email-u9012063@gmail.com> X-Spam-Status: No, score=-1.7 required=5.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_ENVFROM_END_DIGIT,FREEMAIL_FROM, RCVD_IN_DNSWL_NONE autolearn=no version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on smtp1.linux-foundation.org Subject: [ovs-dev] [PATCH RFCv4 3/4] tests: add AF_XDP netdev test cases. X-BeenThere: ovs-dev@openvswitch.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Sender: ovs-dev-bounces@openvswitch.org Errors-To: ovs-dev-bounces@openvswitch.org The patch adds the test framework for OVS using afxdp. Most of the test cases are slightly modified from the existing test cases at system-traffic.at. All the veth creations, ADD_VETH, are replaced by using AF_XDP veth, with new macro ADD_VETH_AFXDP. So packet I/O is based on AF_XDP socket interface. Run `make check-afxdp` to kick start the testing. Signed-off-by: William Tu --- lib/dpif-netdev.c | 2 +- tests/automake.mk | 17 + tests/system-afxdp-macros.at | 153 +++++++ tests/system-afxdp-testsuite.at | 26 ++ tests/system-afxdp-traffic.at | 978 ++++++++++++++++++++++++++++++++++++++++ 5 files changed, 1175 insertions(+), 1 deletion(-) create mode 100644 tests/system-afxdp-macros.at create mode 100644 tests/system-afxdp-testsuite.at create mode 100644 tests/system-afxdp-traffic.at diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c index 481ef50498fd..3f8881a22fe8 100644 --- a/lib/dpif-netdev.c +++ b/lib/dpif-netdev.c @@ -4607,7 +4607,7 @@ rxq_scheduling(struct dp_netdev *dp, bool pinned) OVS_REQUIRES(dp->port_mutex) continue; } rxqs[i]->pmd = rr_numa_get_pmd(non_local_numa, assign_cyc); - VLOG_WARN("There's no available (non-isolated) pmd thread " + VLOG_INFO("There's no available (non-isolated) pmd thread " "on numa node %d. Queue %d on port \'%s\' will " "be assigned to the pmd on core %d " "(numa node %d). Expect reduced performance.", diff --git a/tests/automake.mk b/tests/automake.mk index 92d56b29dedd..c4c0481db293 100644 --- a/tests/automake.mk +++ b/tests/automake.mk @@ -4,11 +4,13 @@ EXTRA_DIST += \ $(SYSTEM_TESTSUITE_AT) \ $(SYSTEM_KMOD_TESTSUITE_AT) \ $(SYSTEM_USERSPACE_TESTSUITE_AT) \ + $(SYSTEM_AFXDP_TESTSUITE_AT) \ $(SYSTEM_OFFLOADS_TESTSUITE_AT) \ $(SYSTEM_DPDK_TESTSUITE_AT) \ $(TESTSUITE) \ $(SYSTEM_KMOD_TESTSUITE) \ $(SYSTEM_USERSPACE_TESTSUITE) \ + $(SYSTEM_AFXDP_TESTSUITE) \ $(SYSTEM_OFFLOADS_TESTSUITE) \ $(SYSTEM_DPDK_TESTSUITE) \ tests/atlocal.in \ @@ -152,6 +154,11 @@ SYSTEM_USERSPACE_TESTSUITE_AT = \ tests/system-userspace-macros.at \ tests/system-userspace-packet-type-aware.at +SYSTEM_AFXDP_TESTSUITE_AT = \ + tests/system-afxdp-testsuite.at \ + tests/system-afxdp-traffic.at \ + tests/system-afxdp-macros.at + SYSTEM_TESTSUITE_AT = \ tests/system-common-macros.at \ tests/system-ovn.at \ @@ -176,6 +183,7 @@ TESTSUITE = $(srcdir)/tests/testsuite TESTSUITE_PATCH = $(srcdir)/tests/testsuite.patch SYSTEM_KMOD_TESTSUITE = $(srcdir)/tests/system-kmod-testsuite SYSTEM_USERSPACE_TESTSUITE = $(srcdir)/tests/system-userspace-testsuite +SYSTEM_AFXDP_TESTSUITE = $(srcdir)/tests/system-afxdp-testsuite SYSTEM_OFFLOADS_TESTSUITE = $(srcdir)/tests/system-offloads-testsuite SYSTEM_DPDK_TESTSUITE = $(srcdir)/tests/system-dpdk-testsuite DISTCLEANFILES += tests/atconfig tests/atlocal @@ -302,6 +310,11 @@ check-system-userspace: all set $(SHELL) '$(SYSTEM_USERSPACE_TESTSUITE)' -C tests AUTOTEST_PATH='$(AUTOTEST_PATH)'; \ "$$@" $(TESTSUITEFLAGS) -j1 || (test X'$(RECHECK)' = Xyes && "$$@" --recheck) +check-afxdp: all + $(MAKE) install + set $(SHELL) '$(SYSTEM_AFXDP_TESTSUITE)' -C tests AUTOTEST_PATH='$(AUTOTEST_PATH)' $(TESTSUITEFLAGS) -j1; \ + "$$@" || (test X'$(RECHECK)' = Xyes && "$$@" --recheck) + check-offloads: all set $(SHELL) '$(SYSTEM_OFFLOADS_TESTSUITE)' -C tests AUTOTEST_PATH='$(AUTOTEST_PATH)'; \ "$$@" $(TESTSUITEFLAGS) -j1 || (test X'$(RECHECK)' = Xyes && "$$@" --recheck) @@ -334,6 +347,10 @@ $(SYSTEM_USERSPACE_TESTSUITE): package.m4 $(SYSTEM_TESTSUITE_AT) $(SYSTEM_USERSP $(AM_V_GEN)$(AUTOTEST) -I '$(srcdir)' -o $@.tmp $@.at $(AM_V_at)mv $@.tmp $@ +$(SYSTEM_AFXDP_TESTSUITE): package.m4 $(SYSTEM_TESTSUITE_AT) $(SYSTEM_AFXDP_TESTSUITE_AT) $(COMMON_MACROS_AT) + $(AM_V_GEN)$(AUTOTEST) -I '$(srcdir)' -o $@.tmp $@.at + $(AM_V_at)mv $@.tmp $@ + $(SYSTEM_OFFLOADS_TESTSUITE): package.m4 $(SYSTEM_TESTSUITE_AT) $(SYSTEM_OFFLOADS_TESTSUITE_AT) $(COMMON_MACROS_AT) $(AM_V_GEN)$(AUTOTEST) -I '$(srcdir)' -o $@.tmp $@.at $(AM_V_at)mv $@.tmp $@ diff --git a/tests/system-afxdp-macros.at b/tests/system-afxdp-macros.at new file mode 100644 index 000000000000..2c58c2d6554b --- /dev/null +++ b/tests/system-afxdp-macros.at @@ -0,0 +1,153 @@ +# _ADD_BR([name]) +# +# Expands into the proper ovs-vsctl commands to create a bridge with the +# appropriate type and properties +m4_define([_ADD_BR], [[add-br $1 -- set Bridge $1 datapath_type=netdev protocols=OpenFlow10,OpenFlow11,OpenFlow12,OpenFlow13,OpenFlow14,OpenFlow15 fail-mode=secure ]]) + +# OVS_TRAFFIC_VSWITCHD_START([vsctl-args], [vsctl-output], [=override]) +# +# Creates a database and starts ovsdb-server, starts ovs-vswitchd +# connected to that database, calls ovs-vsctl to create a bridge named +# br0 with predictable settings, passing 'vsctl-args' as additional +# commands to ovs-vsctl. If 'vsctl-args' causes ovs-vsctl to provide +# output (e.g. because it includes "create" commands) then 'vsctl-output' +# specifies the expected output after filtering through uuidfilt. +m4_define([OVS_TRAFFIC_VSWITCHD_START], + [ + export OVS_PKGDATADIR=$(`pwd`) + _OVS_VSWITCHD_START([--disable-system]) + AT_CHECK([ovs-vsctl -- _ADD_BR([br0]) -- $1 m4_if([$2], [], [], [| uuidfilt])], [0], [$2]) +]) + +# OVS_TRAFFIC_VSWITCHD_STOP([WHITELIST], [extra_cmds]) +# +# Gracefully stops ovs-vswitchd and ovsdb-server, checking their log files +# for messages with severity WARN or higher and signaling an error if any +# is present. The optional WHITELIST may contain shell-quoted "sed" +# commands to delete any warnings that are actually expected, e.g.: +# +# OVS_TRAFFIC_VSWITCHD_STOP(["/expected error/d"]) +# +# 'extra_cmds' are shell commands to be executed afte OVS_VSWITCHD_STOP() is +# invoked. They can be used to perform additional cleanups such as name space +# removal. +m4_define([OVS_TRAFFIC_VSWITCHD_STOP], + [OVS_VSWITCHD_STOP([dnl +$1";/netdev_linux.*obtaining netdev stats via vport failed/d +/dpif_netlink.*Generic Netlink family 'ovs_datapath' does not exist. The Open vSwitch kernel module is probably not loaded./d +/dpif_netdev(revalidator.*)|ERR|internal error parsing flow key/d +/dpif(revalidator.*)|WARN|netdev@ovs-netdev: failed to put/d +"]) + AT_CHECK([:; $2]) + ]) + +m4_define([ADD_VETH_AFXDP], + [ AT_CHECK([ip link add $1 type veth peer name ovs-$1 || return 77]) + CONFIGURE_AFXDP_VETH_OFFLOADS([$1]) + AT_CHECK([ip link set $1 netns $2]) + AT_CHECK([ip link set dev ovs-$1 up]) + AT_CHECK([ovs-vsctl add-port $3 ovs-$1 -- \ + set interface ovs-$1 external-ids:iface-id="$1" type="afxdp"]) + NS_CHECK_EXEC([$2], [ip addr add $4 dev $1 $7]) + NS_CHECK_EXEC([$2], [ip link set dev $1 up]) + if test -n "$5"; then + NS_CHECK_EXEC([$2], [ip link set dev $1 address $5]) + fi + if test -n "$6"; then + NS_CHECK_EXEC([$2], [ip route add default via $6]) + fi + on_exit 'ip link del ovs-$1' + ] +) + +# CONFIGURE_AFXDP_VETH_OFFLOADS([VETH]) +# +# Disable TX offloads and VLAN offloads for veths used in AF_XDP. +m4_define([CONFIGURE_AFXDP_VETH_OFFLOADS], + [AT_CHECK([ethtool -K $1 tx off], [0], [ignore], [ignore]) + AT_CHECK([ethtool -K $1 rxvlan off], [0], [ignore], [ignore]) + AT_CHECK([ethtool -K $1 txvlan off], [0], [ignore], [ignore]) + ] +) + +# CONFIGURE_VETH_OFFLOADS([VETH]) +# +# Disable TX offloads for veths. The userspace datapath uses the AF_PACKET +# socket to receive packets for veths. Unfortunately, the AF_PACKET socket +# doesn't play well with offloads: +# 1. GSO packets are received without segmentation and therefore discarded. +# 2. Packets with offloaded partial checksum are received with the wrong +# checksum, therefore discarded by the receiver. +# +# By disabling tx offloads in the non-OVS side of the veth peer we make sure +# that the AF_PACKET socket will not receive bad packets. +# +# This is a workaround, and should be removed when offloads are properly +# supported in netdev-linux. +m4_define([CONFIGURE_VETH_OFFLOADS], + [AT_CHECK([ethtool -K $1 tx off], [0], [ignore], [ignore])] +) + +# CHECK_CONNTRACK() +# +# Perform requirements checks for running conntrack tests. +# +m4_define([CHECK_CONNTRACK], + [AT_SKIP_IF([test $HAVE_PYTHON = no])] +) + +# CHECK_CONNTRACK_ALG() +# +# Perform requirements checks for running conntrack ALG tests. The userspace +# supports FTP and TFTP. +# +m4_define([CHECK_CONNTRACK_ALG]) + +# CHECK_CONNTRACK_FRAG() +# +# Perform requirements checks for running conntrack fragmentations tests. +# The userspace doesn't support fragmentation yet, so skip the tests. +m4_define([CHECK_CONNTRACK_FRAG], +[ + AT_SKIP_IF([:]) +]) + +# CHECK_CONNTRACK_LOCAL_STACK() +# +# Perform requirements checks for running conntrack tests with local stack. +# While the kernel connection tracker automatically passes all the connection +# tracking state from an internal port to the OpenvSwitch kernel module, there +# is simply no way of doing that with the userspace, so skip the tests. +m4_define([CHECK_CONNTRACK_LOCAL_STACK], +[ + AT_SKIP_IF([:]) +]) + +# CHECK_CONNTRACK_NAT() +# +# Perform requirements checks for running conntrack NAT tests. The userspace +# datapath supports NAT. +# +m4_define([CHECK_CONNTRACK_NAT]) + +# CHECK_CT_DPIF_FLUSH_BY_CT_TUPLE() +# +# Perform requirements checks for running ovs-dpctl flush-conntrack by +# conntrack 5-tuple test. The userspace datapath does not support +# this feature yet. +m4_define([CHECK_CT_DPIF_FLUSH_BY_CT_TUPLE], +[ + AT_SKIP_IF([:]) +]) + +# CHECK_CT_DPIF_SET_GET_MAXCONNS() +# +# Perform requirements checks for running ovs-dpctl ct-set-maxconns or +# ovs-dpctl ct-get-maxconns. The userspace datapath does support this feature. +m4_define([CHECK_CT_DPIF_SET_GET_MAXCONNS]) + +# CHECK_CT_DPIF_GET_NCONNS() +# +# Perform requirements checks for running ovs-dpctl ct-get-nconns. The +# userspace datapath does support this feature. +m4_define([CHECK_CT_DPIF_GET_NCONNS]) diff --git a/tests/system-afxdp-testsuite.at b/tests/system-afxdp-testsuite.at new file mode 100644 index 000000000000..538c0d15d556 --- /dev/null +++ b/tests/system-afxdp-testsuite.at @@ -0,0 +1,26 @@ +AT_INIT + +AT_COPYRIGHT([Copyright (c) 2018 Nicira, Inc. + +Licensed under the Apache License, Version 2.0 (the "License"); +you may not use this file except in compliance with the License. +You may obtain a copy of the License at: + + http://www.apache.org/licenses/LICENSE-2.0 + +Unless required by applicable law or agreed to in writing, software +distributed under the License is distributed on an "AS IS" BASIS, +WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +See the License for the specific language governing permissions and +limitations under the License.]) + +m4_ifdef([AT_COLOR_TESTS], [AT_COLOR_TESTS]) + +m4_include([tests/ovs-macros.at]) +m4_include([tests/ovsdb-macros.at]) +m4_include([tests/ofproto-macros.at]) +m4_include([tests/system-afxdp-macros.at]) +m4_include([tests/system-common-macros.at]) + +m4_include([tests/system-afxdp-traffic.at]) +m4_include([tests/system-ovn.at]) diff --git a/tests/system-afxdp-traffic.at b/tests/system-afxdp-traffic.at new file mode 100644 index 000000000000..26f72acf48ef --- /dev/null +++ b/tests/system-afxdp-traffic.at @@ -0,0 +1,978 @@ +AT_BANNER([AF_XDP netdev datapath-sanity]) + +AT_SETUP([datapath - ping between two ports]) +OVS_TRAFFIC_VSWITCHD_START() + +ulimit -l unlimited + +ADD_NAMESPACES(at_ns0, at_ns1) +AT_CHECK([ovs-ofctl add-flow br0 "actions=normal"]) + +ADD_VETH_AFXDP(p0, at_ns0, br0, "10.1.1.1/24") +ADD_VETH_AFXDP(p1, at_ns1, br0, "10.1.1.2/24") + +NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 10.1.1.2 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([datapath - ping between two ports on vlan]) +OVS_TRAFFIC_VSWITCHD_START() + +AT_CHECK([ovs-ofctl add-flow br0 "actions=normal"]) + +ADD_NAMESPACES(at_ns0, at_ns1) + +ADD_VETH_AFXDP(p0, at_ns0, br0, "10.1.1.1/24") +ADD_VETH_AFXDP(p1, at_ns1, br0, "10.1.1.2/24") + +ADD_VLAN(p0, at_ns0, 100, "10.2.2.1/24") +ADD_VLAN(p1, at_ns1, 100, "10.2.2.2/24") + +NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 10.2.2.2 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([datapath - ping6 between two ports]) +OVS_TRAFFIC_VSWITCHD_START() + +AT_CHECK([ovs-ofctl add-flow br0 "actions=normal"]) + +ADD_NAMESPACES(at_ns0, at_ns1) + +ADD_VETH_AFXDP(p0, at_ns0, br0, "fc00::1/96") +ADD_VETH_AFXDP(p1, at_ns1, br0, "fc00::2/96") + +dnl Linux seems to take a little time to get its IPv6 stack in order. Without +dnl waiting, we get occasional failures due to the following error: +dnl "connect: Cannot assign requested address" +OVS_WAIT_UNTIL([ip netns exec at_ns0 ping6 -c 1 fc00::2]) + +NS_CHECK_EXEC([at_ns0], [ping6 -q -c 3 -i 0.3 -w 6 fc00::2 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([datapath - ping6 between two ports on vlan]) +OVS_TRAFFIC_VSWITCHD_START() + +AT_CHECK([ovs-ofctl add-flow br0 "actions=normal"]) + +ADD_NAMESPACES(at_ns0, at_ns1) + +ADD_VETH_AFXDP(p0, at_ns0, br0, "fc00::1/96") +ADD_VETH_AFXDP(p1, at_ns1, br0, "fc00::2/96") + +ADD_VLAN(p0, at_ns0, 100, "fc00:1::1/96") +ADD_VLAN(p1, at_ns1, 100, "fc00:1::2/96") + +dnl Linux seems to take a little time to get its IPv6 stack in order. Without +dnl waiting, we get occasional failures due to the following error: +dnl "connect: Cannot assign requested address" +OVS_WAIT_UNTIL([ip netns exec at_ns0 ping6 -c 1 fc00:1::2]) + +NS_CHECK_EXEC([at_ns0], [ping6 -q -c 3 -i 0.3 -w 2 fc00:1::2 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) +NS_CHECK_EXEC([at_ns0], [ping6 -s 1600 -q -c 3 -i 0.3 -w 2 fc00:1::2 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) +NS_CHECK_EXEC([at_ns0], [ping6 -s 3200 -q -c 3 -i 0.3 -w 2 fc00:1::2 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([datapath - ping over vxlan tunnel]) +OVS_CHECK_VXLAN() + +OVS_TRAFFIC_VSWITCHD_START() +ADD_BR([br-underlay]) + +AT_CHECK([ovs-ofctl add-flow br0 "actions=normal"]) +AT_CHECK([ovs-ofctl add-flow br-underlay "actions=normal"]) + +ADD_NAMESPACES(at_ns0) + +dnl Set up underlay link from host into the namespace using veth pair. +ADD_VETH_AFXDP(p0, at_ns0, br-underlay, "172.31.1.1/24") +AT_CHECK([ip addr add dev br-underlay "172.31.1.100/24"]) +AT_CHECK([ip link set dev br-underlay up]) + + +dnl Set up tunnel endpoints on OVS outside the namespace and with a native +dnl linux device inside the namespace. +ADD_OVS_TUNNEL([vxlan], [br0], [at_vxlan0], [172.31.1.1], [10.1.1.100/24]) +ADD_NATIVE_TUNNEL([vxlan], [at_vxlan1], [at_ns0], [172.31.1.100], [10.1.1.1/24], + [id 0 dstport 4789]) + +AT_CHECK([ovs-appctl ovs/route/add 10.1.1.100/24 br0], [0], [OK +]) +AT_CHECK([ovs-appctl ovs/route/add 172.31.1.92/24 br-underlay], [0], [OK +]) + +dnl First, check the underlay +NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 172.31.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +dnl Okay, now check the overlay with different packet sizes +NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) +NS_CHECK_EXEC([at_ns0], [ping -s 1600 -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) +NS_CHECK_EXEC([at_ns0], [ping -s 3200 -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([datapath - ping over vxlan6 tunnel]) +OVS_CHECK_VXLAN_UDP6ZEROCSUM() + +OVS_TRAFFIC_VSWITCHD_START() +ADD_BR([br-underlay]) + +AT_CHECK([ovs-ofctl add-flow br0 "actions=normal"]) +AT_CHECK([ovs-ofctl add-flow br-underlay "actions=normal"]) + +ADD_NAMESPACES(at_ns0) + +dnl Set up underlay link from host into the namespace using veth pair. +ADD_VETH_AFXDP(p0, at_ns0, br-underlay, "fc00::1/64", [], [], "nodad") +AT_CHECK([ip addr add dev br-underlay "fc00::100/64" nodad]) +AT_CHECK([ip link set dev br-underlay up]) + +dnl Set up tunnel endpoints on OVS outside the namespace and with a native +dnl linux device inside the namespace. +ADD_OVS_TUNNEL6([vxlan], [br0], [at_vxlan0], [fc00::1], [10.1.1.100/24]) +ADD_NATIVE_TUNNEL6([vxlan], [at_vxlan1], [at_ns0], [fc00::100], [10.1.1.1/24], + [id 0 dstport 4789 udp6zerocsumtx udp6zerocsumrx]) + +AT_CHECK([ovs-appctl ovs/route/add 10.1.1.100/24 br0], [0], [OK +]) +AT_CHECK([ovs-appctl ovs/route/add fc00::100/64 br-underlay], [0], [OK +]) + +OVS_WAIT_UNTIL([ip netns exec at_ns0 ping6 -c 1 fc00::100]) + +dnl First, check the underlay +NS_CHECK_EXEC([at_ns0], [ping6 -q -c 3 -i 0.3 -w 2 fc00::100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +dnl Okay, now check the overlay with different packet sizes +NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) +NS_CHECK_EXEC([at_ns0], [ping -s 1600 -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) +NS_CHECK_EXEC([at_ns0], [ping -s 3200 -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([datapath - ping over gre tunnel]) +OVS_CHECK_GRE() + +OVS_TRAFFIC_VSWITCHD_START() +ADD_BR([br-underlay]) + +AT_CHECK([ovs-ofctl add-flow br0 "actions=normal"]) +AT_CHECK([ovs-ofctl add-flow br-underlay "actions=normal"]) + +ADD_NAMESPACES(at_ns0) + +dnl Set up underlay link from host into the namespace using veth pair. +ADD_VETH_AFXDP(p0, at_ns0, br-underlay, "172.31.1.1/24") +AT_CHECK([ip addr add dev br-underlay "172.31.1.100/24"]) +AT_CHECK([ip link set dev br-underlay up]) + +dnl Set up tunnel endpoints on OVS outside the namespace and with a native +dnl linux device inside the namespace. +ADD_OVS_TUNNEL([gre], [br0], [at_gre0], [172.31.1.1], [10.1.1.100/24]) +ADD_NATIVE_TUNNEL([gretap], [ns_gre0], [at_ns0], [172.31.1.100], [10.1.1.1/24]) + +AT_CHECK([ovs-appctl ovs/route/add 10.1.1.100/24 br0], [0], [OK +]) +AT_CHECK([ovs-appctl ovs/route/add 172.31.1.92/24 br-underlay], [0], [OK +]) + +dnl First, check the underlay +NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 172.31.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +dnl Okay, now check the overlay with different packet sizes +NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) +NS_CHECK_EXEC([at_ns0], [ping -s 1600 -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) +NS_CHECK_EXEC([at_ns0], [ping -s 3200 -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([datapath - ping over erspan v1 tunnel]) +OVS_CHECK_GRE() +OVS_CHECK_ERSPAN() + +OVS_TRAFFIC_VSWITCHD_START() +ADD_BR([br-underlay]) + +AT_CHECK([ovs-ofctl add-flow br0 "actions=normal"]) +AT_CHECK([ovs-ofctl add-flow br-underlay "actions=normal"]) + +ADD_NAMESPACES(at_ns0) + +dnl Set up underlay link from host into the namespace using veth pair. +ADD_VETH_AFXDP(p0, at_ns0, br-underlay, "172.31.1.1/24") +AT_CHECK([ip addr add dev br-underlay "172.31.1.100/24"]) +AT_CHECK([ip link set dev br-underlay up]) + +dnl Set up tunnel endpoints on OVS outside the namespace and with a native +dnl linux device inside the namespace. +ADD_OVS_TUNNEL([erspan], [br0], [at_erspan0], [172.31.1.1], [10.1.1.100/24], [options:key=1 options:erspan_ver=1 options:erspan_idx=7]) +ADD_NATIVE_TUNNEL([erspan], [ns_erspan0], [at_ns0], [172.31.1.100], [10.1.1.1/24], [seq key 1 erspan_ver 1 erspan 7]) + +AT_CHECK([ovs-appctl ovs/route/add 10.1.1.100/24 br0], [0], [OK +]) +AT_CHECK([ovs-appctl ovs/route/add 172.31.1.92/24 br-underlay], [0], [OK +]) + +dnl First, check the underlay +NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 172.31.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +dnl Okay, now check the overlay with different packet sizes +dnl NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +NS_CHECK_EXEC([at_ns0], [ping -s 1200 -i 0.3 -c 3 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([datapath - ping over erspan v2 tunnel]) +OVS_CHECK_GRE() +OVS_CHECK_ERSPAN() + +OVS_TRAFFIC_VSWITCHD_START() +ADD_BR([br-underlay]) + +AT_CHECK([ovs-ofctl add-flow br0 "actions=normal"]) +AT_CHECK([ovs-ofctl add-flow br-underlay "actions=normal"]) + +ADD_NAMESPACES(at_ns0) + +dnl Set up underlay link from host into the namespace using veth pair. +ADD_VETH_AFXDP(p0, at_ns0, br-underlay, "172.31.1.1/24") +AT_CHECK([ip addr add dev br-underlay "172.31.1.100/24"]) +AT_CHECK([ip link set dev br-underlay up]) + +dnl Set up tunnel endpoints on OVS outside the namespace and with a native +dnl linux device inside the namespace. +ADD_OVS_TUNNEL([erspan], [br0], [at_erspan0], [172.31.1.1], [10.1.1.100/24], [options:key=1 options:erspan_ver=2 options:erspan_dir=1 options:erspan_hwid=0x7]) +ADD_NATIVE_TUNNEL([erspan], [ns_erspan0], [at_ns0], [172.31.1.100], [10.1.1.1/24], [seq key 1 erspan_ver 2 erspan_dir egress erspan_hwid 7]) + +AT_CHECK([ovs-appctl ovs/route/add 10.1.1.100/24 br0], [0], [OK +]) +AT_CHECK([ovs-appctl ovs/route/add 172.31.1.92/24 br-underlay], [0], [OK +]) + +dnl First, check the underlay +NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 172.31.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +dnl Okay, now check the overlay with different packet sizes +dnl NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +NS_CHECK_EXEC([at_ns0], [ping -s 1200 -i 0.3 -c 3 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([datapath - ping over ip6erspan v1 tunnel]) +OVS_CHECK_GRE() +OVS_CHECK_ERSPAN() + +OVS_TRAFFIC_VSWITCHD_START() +ADD_BR([br-underlay]) + +AT_CHECK([ovs-ofctl add-flow br0 "actions=normal"]) +AT_CHECK([ovs-ofctl add-flow br-underlay "actions=normal"]) + +ADD_NAMESPACES(at_ns0) + +dnl Set up underlay link from host into the namespace using veth pair. +ADD_VETH_AFXDP(p0, at_ns0, br-underlay, "fc00:100::1/96", [], [], nodad) +AT_CHECK([ip addr add dev br-underlay "fc00:100::100/96" nodad]) +AT_CHECK([ip link set dev br-underlay up]) + +dnl Set up tunnel endpoints on OVS outside the namespace and with a native +dnl linux device inside the namespace. +ADD_OVS_TUNNEL6([ip6erspan], [br0], [at_erspan0], [fc00:100::1], [10.1.1.100/24], + [options:key=123 options:erspan_ver=1 options:erspan_idx=0x7]) +ADD_NATIVE_TUNNEL6([ip6erspan], [ns_erspan0], [at_ns0], [fc00:100::100], + [10.1.1.1/24], [local fc00:100::1 seq key 123 erspan_ver 1 erspan 7]) + +AT_CHECK([ovs-appctl ovs/route/add 10.1.1.100/24 br0], [0], [OK +]) +AT_CHECK([ovs-appctl ovs/route/add fc00:100::1/96 br-underlay], [0], [OK +]) + +OVS_WAIT_UNTIL([ip netns exec at_ns0 ping6 -c 2 fc00:100::100]) + +dnl First, check the underlay +NS_CHECK_EXEC([at_ns0], [ping6 -q -c 3 -i 0.3 -w 2 fc00:100::100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +dnl Okay, now check the overlay with different packet sizes +NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([datapath - ping over ip6erspan v2 tunnel]) +OVS_CHECK_GRE() +OVS_CHECK_ERSPAN() + +OVS_TRAFFIC_VSWITCHD_START() +ADD_BR([br-underlay]) + +AT_CHECK([ovs-ofctl add-flow br0 "actions=normal"]) +AT_CHECK([ovs-ofctl add-flow br-underlay "actions=normal"]) + +ADD_NAMESPACES(at_ns0) + +dnl Set up underlay link from host into the namespace using veth pair. +ADD_VETH_AFXDP(p0, at_ns0, br-underlay, "fc00:100::1/96", [], [], nodad) +AT_CHECK([ip addr add dev br-underlay "fc00:100::100/96" nodad]) +AT_CHECK([ip link set dev br-underlay up]) + +dnl Set up tunnel endpoints on OVS outside the namespace and with a native +dnl linux device inside the namespace. +ADD_OVS_TUNNEL6([ip6erspan], [br0], [at_erspan0], [fc00:100::1], [10.1.1.100/24], + [options:key=121 options:erspan_ver=2 options:erspan_dir=0 options:erspan_hwid=0x7]) +ADD_NATIVE_TUNNEL6([ip6erspan], [ns_erspan0], [at_ns0], [fc00:100::100], + [10.1.1.1/24], + [local fc00:100::1 seq key 121 erspan_ver 2 erspan_dir ingress erspan_hwid 0x7]) + +AT_CHECK([ovs-appctl ovs/route/add 10.1.1.100/24 br0], [0], [OK +]) +AT_CHECK([ovs-appctl ovs/route/add fc00:100::1/96 br-underlay], [0], [OK +]) + +OVS_WAIT_UNTIL([ip netns exec at_ns0 ping6 -c 2 fc00:100::100]) + +dnl First, check the underlay +NS_CHECK_EXEC([at_ns0], [ping6 -q -c 3 -i 0.3 -w 2 fc00:100::100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +dnl Okay, now check the overlay with different packet sizes +NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([datapath - ping over geneve tunnel]) +OVS_CHECK_GENEVE() + +OVS_TRAFFIC_VSWITCHD_START() +ADD_BR([br-underlay]) + +AT_CHECK([ovs-ofctl add-flow br0 "actions=normal"]) +AT_CHECK([ovs-ofctl add-flow br-underlay "actions=normal"]) + +ADD_NAMESPACES(at_ns0) + +dnl Set up underlay link from host into the namespace using veth pair. +ADD_VETH_AFXDP(p0, at_ns0, br-underlay, "172.31.1.1/24") +AT_CHECK([ip addr add dev br-underlay "172.31.1.100/24"]) +AT_CHECK([ip link set dev br-underlay up]) + +dnl Set up tunnel endpoints on OVS outside the namespace and with a native +dnl linux device inside the namespace. +ADD_OVS_TUNNEL([geneve], [br0], [at_gnv0], [172.31.1.1], [10.1.1.100/24]) +ADD_NATIVE_TUNNEL([geneve], [ns_gnv0], [at_ns0], [172.31.1.100], [10.1.1.1/24], + [vni 0]) + +AT_CHECK([ovs-appctl ovs/route/add 10.1.1.100/24 br0], [0], [OK +]) +AT_CHECK([ovs-appctl ovs/route/add 172.31.1.100/24 br-underlay], [0], [OK +]) + +dnl First, check the underlay +NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 172.31.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +dnl Okay, now check the overlay with different packet sizes +NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) +NS_CHECK_EXEC([at_ns0], [ping -s 1600 -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) +NS_CHECK_EXEC([at_ns0], [ping -s 3200 -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([datapath - ping over geneve6 tunnel]) +OVS_CHECK_GENEVE_UDP6ZEROCSUM() + +OVS_TRAFFIC_VSWITCHD_START() +ADD_BR([br-underlay]) + +AT_CHECK([ovs-ofctl add-flow br0 "actions=normal"]) +AT_CHECK([ovs-ofctl add-flow br-underlay "actions=normal"]) + +ADD_NAMESPACES(at_ns0) + +dnl Set up underlay link from host into the namespace using veth pair. +ADD_VETH_AFXDP(p0, at_ns0, br-underlay, "fc00::1/64", [], [], "nodad") +AT_CHECK([ip addr add dev br-underlay "fc00::100/64" nodad]) +AT_CHECK([ip link set dev br-underlay up]) + +dnl Set up tunnel endpoints on OVS outside the namespace and with a native +dnl linux device inside the namespace. +ADD_OVS_TUNNEL6([geneve], [br0], [at_gnv0], [fc00::1], [10.1.1.100/24]) +ADD_NATIVE_TUNNEL6([geneve], [ns_gnv0], [at_ns0], [fc00::100], [10.1.1.1/24], + [vni 0 udp6zerocsumtx udp6zerocsumrx]) + +AT_CHECK([ovs-appctl ovs/route/add 10.1.1.100/24 br0], [0], [OK +]) +AT_CHECK([ovs-appctl ovs/route/add fc00::100/64 br-underlay], [0], [OK +]) + +OVS_WAIT_UNTIL([ip netns exec at_ns0 ping6 -c 1 fc00::100]) + +dnl First, check the underlay +NS_CHECK_EXEC([at_ns0], [ping6 -q -c 3 -i 0.3 -w 2 fc00::100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +dnl Okay, now check the overlay with different packet sizes +NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) +NS_CHECK_EXEC([at_ns0], [ping -s 1600 -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) +NS_CHECK_EXEC([at_ns0], [ping -s 3200 -q -c 3 -i 0.3 -w 2 10.1.1.100 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([datapath - clone action]) +OVS_TRAFFIC_VSWITCHD_START() + +ADD_NAMESPACES(at_ns0, at_ns1, at_ns2) + +ADD_VETH_AFXDP(p0, at_ns0, br0, "10.1.1.1/24") +ADD_VETH_AFXDP(p1, at_ns1, br0, "10.1.1.2/24") + +AT_CHECK([ovs-vsctl -- set interface ovs-p0 ofport_request=1 \ + -- set interface ovs-p1 ofport_request=2]) + +AT_DATA([flows.txt], [dnl +priority=1 actions=NORMAL +priority=10 in_port=1,ip,actions=clone(mod_dl_dst(50:54:00:00:00:0a),set_field:192.168.3.3->ip_dst), output:2 +priority=10 in_port=2,ip,actions=clone(mod_dl_src(ae:c6:7e:54:8d:4d),mod_dl_dst(50:54:00:00:00:0b),set_field:192.168.4.4->ip_dst, controller), output:1 +]) +AT_CHECK([ovs-ofctl add-flows br0 flows.txt]) + +AT_CHECK([ovs-ofctl monitor br0 65534 invalid_ttl --detach --no-chdir --pidfile 2> ofctl_monitor.log]) +NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 10.1.1.2 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +AT_CHECK([cat ofctl_monitor.log | STRIP_MONITOR_CSUM], [0], [dnl +icmp,vlan_tci=0x0000,dl_src=ae:c6:7e:54:8d:4d,dl_dst=50:54:00:00:00:0b,nw_src=10.1.1.2,nw_dst=192.168.4.4,nw_tos=0,nw_ecn=0,nw_ttl=64,icmp_type=0,icmp_code=0 icmp_csum: +icmp,vlan_tci=0x0000,dl_src=ae:c6:7e:54:8d:4d,dl_dst=50:54:00:00:00:0b,nw_src=10.1.1.2,nw_dst=192.168.4.4,nw_tos=0,nw_ecn=0,nw_ttl=64,icmp_type=0,icmp_code=0 icmp_csum: +icmp,vlan_tci=0x0000,dl_src=ae:c6:7e:54:8d:4d,dl_dst=50:54:00:00:00:0b,nw_src=10.1.1.2,nw_dst=192.168.4.4,nw_tos=0,nw_ecn=0,nw_ttl=64,icmp_type=0,icmp_code=0 icmp_csum: +]) + +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([datapath - basic truncate action]) +AT_SKIP_IF([test $HAVE_NC = no]) +OVS_TRAFFIC_VSWITCHD_START() +AT_CHECK([ovs-ofctl del-flows br0]) + +dnl Create p0 and ovs-p0(1) +ADD_NAMESPACES(at_ns0) +ADD_VETH_AFXDP(p0, at_ns0, br0, "10.1.1.1/24") +NS_CHECK_EXEC([at_ns0], [ip link set dev p0 address e6:66:c1:11:11:11]) +NS_CHECK_EXEC([at_ns0], [arp -s 10.1.1.2 e6:66:c1:22:22:22]) + +dnl Create p1(3) and ovs-p1(2), packets received from ovs-p1 will appear in p1 +AT_CHECK([ip link add p1 type veth peer name ovs-p1]) +on_exit 'ip link del ovs-p1' +AT_CHECK([ip link set dev ovs-p1 up]) +AT_CHECK([ip link set dev p1 up]) +AT_CHECK([ovs-vsctl add-port br0 ovs-p1 -- set interface ovs-p1 ofport_request=2]) +dnl Use p1 to check the truncated packet +AT_CHECK([ovs-vsctl add-port br0 p1 -- set interface p1 ofport_request=3]) + +dnl Create p2(5) and ovs-p2(4) +AT_CHECK([ip link add p2 type veth peer name ovs-p2]) +on_exit 'ip link del ovs-p2' +AT_CHECK([ip link set dev ovs-p2 up]) +AT_CHECK([ip link set dev p2 up]) +AT_CHECK([ovs-vsctl add-port br0 ovs-p2 -- set interface ovs-p2 ofport_request=4]) +dnl Use p2 to check the truncated packet +AT_CHECK([ovs-vsctl add-port br0 p2 -- set interface p2 ofport_request=5]) + +dnl basic test +AT_CHECK([ovs-ofctl del-flows br0]) +AT_DATA([flows.txt], [dnl +in_port=3 dl_dst=e6:66:c1:22:22:22 actions=drop +in_port=5 dl_dst=e6:66:c1:22:22:22 actions=drop +in_port=1 dl_dst=e6:66:c1:22:22:22 actions=output(port=2,max_len=100),output:4 +]) +AT_CHECK([ovs-ofctl add-flows br0 flows.txt]) + +dnl use this file as payload file for ncat +AT_CHECK([dd if=/dev/urandom of=payload200.bin bs=200 count=1 2> /dev/null]) +on_exit 'rm -f payload200.bin' +NS_CHECK_EXEC([at_ns0], [nc $NC_EOF_OPT -u 10.1.1.2 1234 < payload200.bin]) + +dnl packet with truncated size +AT_CHECK([ovs-appctl revalidator/purge], [0]) +AT_CHECK([ovs-ofctl dump-flows br0 table=0 | grep "in_port=3" | sed -n 's/.*\(n\_bytes=[[0-9]]*\).*/\1/p'], [0], [dnl +n_bytes=100 +]) +dnl packet with original size +AT_CHECK([ovs-appctl revalidator/purge], [0]) +AT_CHECK([ovs-ofctl dump-flows br0 table=0 | grep "in_port=5" | sed -n 's/.*\(n\_bytes=[[0-9]]*\).*/\1/p'], [0], [dnl +n_bytes=242 +]) + +dnl more complicated output actions +AT_CHECK([ovs-ofctl del-flows br0]) +AT_DATA([flows.txt], [dnl +in_port=3 dl_dst=e6:66:c1:22:22:22 actions=drop +in_port=5 dl_dst=e6:66:c1:22:22:22 actions=drop +in_port=1 dl_dst=e6:66:c1:22:22:22 actions=output(port=2,max_len=100),output:4,output(port=2,max_len=100),output(port=4,max_len=100),output:2,output(port=4,max_len=200),output(port=2,max_len=65535) +]) +AT_CHECK([ovs-ofctl add-flows br0 flows.txt]) + +NS_CHECK_EXEC([at_ns0], [nc $NC_EOF_OPT -u 10.1.1.2 1234 < payload200.bin]) + +dnl 100 + 100 + 242 + min(65535,242) = 684 +AT_CHECK([ovs-appctl revalidator/purge], [0]) +AT_CHECK([ovs-ofctl dump-flows br0 table=0 | grep "in_port=3" | sed -n 's/.*\(n\_bytes=[[0-9]]*\).*/\1/p'], [0], [dnl +n_bytes=684 +]) +dnl 242 + 100 + min(242,200) = 542 +AT_CHECK([ovs-ofctl dump-flows br0 table=0 | grep "in_port=5" | sed -n 's/.*\(n\_bytes=[[0-9]]*\).*/\1/p'], [0], [dnl +n_bytes=542 +]) + +dnl SLOW_ACTION: disable kernel datapath truncate support +dnl Repeat the test above, but exercise the SLOW_ACTION code path +AT_CHECK([ovs-appctl dpif/set-dp-features br0 trunc false], [0]) + +dnl SLOW_ACTION test1: check datapatch actions +AT_CHECK([ovs-ofctl del-flows br0]) +AT_CHECK([ovs-ofctl add-flows br0 flows.txt]) + +AT_CHECK([ovs-appctl ofproto/trace br0 "in_port=1,dl_type=0x800,dl_src=e6:66:c1:11:11:11,dl_dst=e6:66:c1:22:22:22,nw_src=192.168.0.1,nw_dst=192.168.0.2,nw_proto=6,tp_src=8,tp_dst=9"], [0], [stdout]) +AT_CHECK([tail -3 stdout], [0], +[Datapath actions: trunc(100),3,5,trunc(100),3,trunc(100),5,3,trunc(200),5,trunc(65535),3 +This flow is handled by the userspace slow path because it: + - Uses action(s) not supported by datapath. +]) + +dnl SLOW_ACTION test2: check actual packet truncate +AT_CHECK([ovs-ofctl del-flows br0]) +AT_CHECK([ovs-ofctl add-flows br0 flows.txt]) +NS_CHECK_EXEC([at_ns0], [nc $NC_EOF_OPT -u 10.1.1.2 1234 < payload200.bin]) + +dnl 100 + 100 + 242 + min(65535,242) = 684 +AT_CHECK([ovs-appctl revalidator/purge], [0]) +AT_CHECK([ovs-ofctl dump-flows br0 table=0 | grep "in_port=3" | sed -n 's/.*\(n\_bytes=[[0-9]]*\).*/\1/p'], [0], [dnl +n_bytes=684 +]) + +dnl 242 + 100 + min(242,200) = 542 +AT_CHECK([ovs-ofctl dump-flows br0 table=0 | grep "in_port=5" | sed -n 's/.*\(n\_bytes=[[0-9]]*\).*/\1/p'], [0], [dnl +n_bytes=542 +]) + +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + + +AT_BANNER([conntrack]) + +AT_SETUP([conntrack - controller]) +CHECK_CONNTRACK() +OVS_TRAFFIC_VSWITCHD_START() +AT_CHECK([ovs-appctl vlog/set dpif:dbg dpif_netdev:dbg ofproto_dpif_upcall:dbg]) + +ADD_NAMESPACES(at_ns0, at_ns1) + +ADD_VETH_AFXDP(p0, at_ns0, br0, "10.1.1.1/24") +ADD_VETH_AFXDP(p1, at_ns1, br0, "10.1.1.2/24") + +dnl Allow any traffic from ns0->ns1. Only allow nd, return traffic from ns1->ns0. +AT_DATA([flows.txt], [dnl +priority=1,action=drop +priority=10,arp,action=normal +priority=100,in_port=1,udp,action=ct(commit),controller +priority=100,in_port=2,ct_state=-trk,udp,action=ct(table=0) +priority=100,in_port=2,ct_state=+trk+est,udp,action=controller +]) + +AT_CHECK([ovs-ofctl --bundle add-flows br0 flows.txt]) + +AT_CAPTURE_FILE([ofctl_monitor.log]) +AT_CHECK([ovs-ofctl monitor br0 65534 invalid_ttl --detach --no-chdir --pidfile 2> ofctl_monitor.log]) + +dnl Send an unsolicited reply from port 2. This should be dropped. +AT_CHECK([ovs-ofctl -O OpenFlow13 packet-out br0 2 ct\(table=0\) '50540000000a50540000000908004500001c000000000011a4cd0a0101020a0101010002000100080000']) + +dnl OK, now start a new connection from port 1. +AT_CHECK([ovs-ofctl -O OpenFlow13 packet-out br0 1 ct\(commit\),controller '50540000000a50540000000908004500001c000000000011a4cd0a0101010a0101020001000200080000']) + +dnl Now try a reply from port 2. +AT_CHECK([ovs-ofctl -O OpenFlow13 packet-out br0 2 ct\(table=0\) '50540000000a50540000000908004500001c000000000011a4cd0a0101020a0101010002000100080000']) + +dnl Check this output. We only see the latter two packets, not the first. +AT_CHECK([cat ofctl_monitor.log], [0], [dnl +NXT_PACKET_IN2 (xid=0x0): total_len=42 in_port=1 (via action) data_len=42 (unbuffered) +udp,vlan_tci=0x0000,dl_src=50:54:00:00:00:09,dl_dst=50:54:00:00:00:0a,nw_src=10.1.1.1,nw_dst=10.1.1.2,nw_tos=0,nw_ecn=0,nw_ttl=0,tp_src=1,tp_dst=2 udp_csum:0 +NXT_PACKET_IN2 (xid=0x0): cookie=0x0 total_len=42 ct_state=est|rpl|trk,ct_nw_src=10.1.1.1,ct_nw_dst=10.1.1.2,ct_nw_proto=17,ct_tp_src=1,ct_tp_dst=2,ip,in_port=2 (via action) data_len=42 (unbuffered) +udp,vlan_tci=0x0000,dl_src=50:54:00:00:00:09,dl_dst=50:54:00:00:00:0a,nw_src=10.1.1.2,nw_dst=10.1.1.1,nw_tos=0,nw_ecn=0,nw_ttl=0,tp_src=2,tp_dst=1 udp_csum:0 +]) + +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([conntrack - force commit]) +CHECK_CONNTRACK() +OVS_TRAFFIC_VSWITCHD_START() +AT_CHECK([ovs-appctl vlog/set dpif:dbg dpif_netdev:dbg ofproto_dpif_upcall:dbg]) + +ADD_NAMESPACES(at_ns0, at_ns1) + +ADD_VETH_AFXDP(p0, at_ns0, br0, "10.1.1.1/24") +ADD_VETH_AFXDP(p1, at_ns1, br0, "10.1.1.2/24") + +AT_DATA([flows.txt], [dnl +priority=1,action=drop +priority=10,arp,action=normal +priority=100,in_port=1,udp,action=ct(force,commit),controller +priority=100,in_port=2,ct_state=-trk,udp,action=ct(table=0) +priority=100,in_port=2,ct_state=+trk+est,udp,action=ct(force,commit,table=1) +table=1,in_port=2,ct_state=+trk,udp,action=controller +]) + +AT_CHECK([ovs-ofctl --bundle add-flows br0 flows.txt]) + +AT_CAPTURE_FILE([ofctl_monitor.log]) +AT_CHECK([ovs-ofctl monitor br0 65534 invalid_ttl --detach --no-chdir --pidfile 2> ofctl_monitor.log]) + +dnl Send an unsolicited reply from port 2. This should be dropped. +AT_CHECK([ovs-ofctl -O OpenFlow13 packet-out br0 "in_port=2 packet=50540000000a50540000000908004500001c000000000011a4cd0a0101020a0101010002000100080000 actions=resubmit(,0)"]) + +dnl OK, now start a new connection from port 1. +AT_CHECK([ovs-ofctl -O OpenFlow13 packet-out br0 "in_port=1 packet=50540000000a50540000000908004500001c000000000011a4cd0a0101010a0101020001000200080000 actions=resubmit(,0)"]) + +dnl Now try a reply from port 2. +AT_CHECK([ovs-ofctl -O OpenFlow13 packet-out br0 "in_port=2 packet=50540000000a50540000000908004500001c000000000011a4cd0a0101020a0101010002000100080000 actions=resubmit(,0)"]) + +AT_CHECK([ovs-appctl revalidator/purge], [0]) + +dnl Check this output. We only see the latter two packets, not the first. +AT_CHECK([cat ofctl_monitor.log], [0], [dnl +NXT_PACKET_IN2 (xid=0x0): cookie=0x0 total_len=42 in_port=1 (via action) data_len=42 (unbuffered) +udp,vlan_tci=0x0000,dl_src=50:54:00:00:00:09,dl_dst=50:54:00:00:00:0a,nw_src=10.1.1.1,nw_dst=10.1.1.2,nw_tos=0,nw_ecn=0,nw_ttl=0,tp_src=1,tp_dst=2 udp_csum:0 +NXT_PACKET_IN2 (xid=0x0): table_id=1 cookie=0x0 total_len=42 ct_state=new|trk,ct_nw_src=10.1.1.2,ct_nw_dst=10.1.1.1,ct_nw_proto=17,ct_tp_src=2,ct_tp_dst=1,ip,in_port=2 (via action) data_len=42 (unbuffered) +udp,vlan_tci=0x0000,dl_src=50:54:00:00:00:09,dl_dst=50:54:00:00:00:0a,nw_src=10.1.1.2,nw_dst=10.1.1.1,nw_tos=0,nw_ecn=0,nw_ttl=0,tp_src=2,tp_dst=1 udp_csum:0 +]) + +dnl +dnl Check that the directionality has been changed by force commit. +dnl +AT_CHECK([ovs-appctl dpctl/dump-conntrack | grep "orig=.src=10\.1\.1\.2,"], [], [dnl +udp,orig=(src=10.1.1.2,dst=10.1.1.1,sport=2,dport=1),reply=(src=10.1.1.1,dst=10.1.1.2,sport=1,dport=2) +]) + +dnl OK, now send another packet from port 1 and see that it switches again +AT_CHECK([ovs-ofctl -O OpenFlow13 packet-out br0 "in_port=1 packet=50540000000a50540000000908004500001c000000000011a4cd0a0101010a0101020001000200080000 actions=resubmit(,0)"]) +AT_CHECK([ovs-appctl revalidator/purge], [0]) + +AT_CHECK([ovs-appctl dpctl/dump-conntrack | grep "orig=.src=10\.1\.1\.1,"], [], [dnl +udp,orig=(src=10.1.1.1,dst=10.1.1.2,sport=1,dport=2),reply=(src=10.1.1.2,dst=10.1.1.1,sport=2,dport=1) +]) + +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([conntrack - ct flush by 5-tuple]) +CHECK_CONNTRACK() +OVS_TRAFFIC_VSWITCHD_START() + +ADD_NAMESPACES(at_ns0, at_ns1) + +ADD_VETH_AFXDP(p0, at_ns0, br0, "10.1.1.1/24") +ADD_VETH_AFXDP(p1, at_ns1, br0, "10.1.1.2/24") + +AT_DATA([flows.txt], [dnl +priority=1,action=drop +priority=10,arp,action=normal +priority=100,in_port=1,udp,action=ct(commit),2 +priority=100,in_port=2,udp,action=ct(zone=5,commit),1 +priority=100,in_port=1,icmp,action=ct(commit),2 +priority=100,in_port=2,icmp,action=ct(zone=5,commit),1 +]) + +AT_CHECK([ovs-ofctl --bundle add-flows br0 flows.txt]) + +dnl Test UDP from port 1 +AT_CHECK([ovs-ofctl -O OpenFlow13 packet-out br0 "in_port=1 packet=50540000000a50540000000908004500001c000000000011a4cd0a0101010a0101020001000200080000 actions=resubmit(,0)"]) + +AT_CHECK([ovs-appctl dpctl/dump-conntrack | grep "orig=.src=10\.1\.1\.1,"], [], [dnl +udp,orig=(src=10.1.1.1,dst=10.1.1.2,sport=1,dport=2),reply=(src=10.1.1.2,dst=10.1.1.1,sport=2,dport=1) +]) + +AT_CHECK([ovs-appctl dpctl/flush-conntrack 'ct_nw_src=10.1.1.2,ct_nw_dst=10.1.1.1,ct_nw_proto=17,ct_tp_src=2,ct_tp_dst=1']) + +AT_CHECK([ovs-appctl dpctl/dump-conntrack | grep "orig=.src=10\.1\.1\.1,"], [1], [dnl +]) + +dnl Test UDP from port 2 +AT_CHECK([ovs-ofctl -O OpenFlow13 packet-out br0 "in_port=2 packet=50540000000a50540000000908004500001c000000000011a4cd0a0101020a0101010002000100080000 actions=resubmit(,0)"]) + +AT_CHECK([ovs-appctl dpctl/dump-conntrack | grep "orig=.src=10\.1\.1\.2,"], [0], [dnl +udp,orig=(src=10.1.1.2,dst=10.1.1.1,sport=2,dport=1),reply=(src=10.1.1.1,dst=10.1.1.2,sport=1,dport=2),zone=5 +]) + +AT_CHECK([ovs-appctl dpctl/flush-conntrack zone=5 'ct_nw_src=10.1.1.1,ct_nw_dst=10.1.1.2,ct_nw_proto=17,ct_tp_src=1,ct_tp_dst=2']) + +AT_CHECK([ovs-appctl dpctl/dump-conntrack | FORMAT_CT(10.1.1.2)], [0], [dnl +]) + +dnl Test ICMP traffic +NS_CHECK_EXEC([at_ns1], [ping -q -c 3 -i 0.3 -w 2 10.1.1.1 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +AT_CHECK([ovs-appctl dpctl/dump-conntrack | grep "orig=.src=10\.1\.1\.2,"], [0], [stdout]) +AT_CHECK([cat stdout | FORMAT_CT(10.1.1.1)], [0],[dnl +icmp,orig=(src=10.1.1.2,dst=10.1.1.1,id=,type=8,code=0),reply=(src=10.1.1.1,dst=10.1.1.2,id=,type=0,code=0),zone=5 +]) + +ICMP_ID=`cat stdout | cut -d ',' -f4 | cut -d '=' -f2` +ICMP_TUPLE=ct_nw_src=10.1.1.2,ct_nw_dst=10.1.1.1,ct_nw_proto=1,icmp_id=$ICMP_ID,icmp_type=8,icmp_code=0 +AT_CHECK([ovs-appctl dpctl/flush-conntrack zone=5 $ICMP_TUPLE]) + +AT_CHECK([ovs-appctl dpctl/dump-conntrack | grep "orig=.src=10\.1\.1\.2,"], [1], [dnl +]) + +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([conntrack - IPv4 ping]) +CHECK_CONNTRACK() +OVS_TRAFFIC_VSWITCHD_START() + +ADD_NAMESPACES(at_ns0, at_ns1) + +ADD_VETH_AFXDP(p0, at_ns0, br0, "10.1.1.1/24") +ADD_VETH_AFXDP(p1, at_ns1, br0, "10.1.1.2/24") + +dnl Allow any traffic from ns0->ns1. Only allow nd, return traffic from ns1->ns0. +AT_DATA([flows.txt], [dnl +priority=1,action=drop +priority=10,arp,action=normal +priority=100,in_port=1,icmp,action=ct(commit),2 +priority=100,in_port=2,icmp,ct_state=-trk,action=ct(table=0) +priority=100,in_port=2,icmp,ct_state=+trk+est,action=1 +]) + +AT_CHECK([ovs-ofctl --bundle add-flows br0 flows.txt]) + +dnl Pings from ns0->ns1 should work fine. +NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 10.1.1.2 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +AT_CHECK([ovs-appctl dpctl/dump-conntrack | FORMAT_CT(10.1.1.2)], [0], [dnl +icmp,orig=(src=10.1.1.1,dst=10.1.1.2,id=,type=8,code=0),reply=(src=10.1.1.2,dst=10.1.1.1,id=,type=0,code=0) +]) + +AT_CHECK([ovs-appctl dpctl/flush-conntrack]) + +dnl Pings from ns1->ns0 should fail. +NS_CHECK_EXEC([at_ns1], [ping -q -c 3 -i 0.3 -w 2 10.1.1.1 | FORMAT_PING], [0], [dnl +7 packets transmitted, 0 received, 100% packet loss, time 0ms +]) + +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([conntrack - get_nconns and get/set_maxconns]) +CHECK_CONNTRACK() +CHECK_CT_DPIF_SET_GET_MAXCONNS() +CHECK_CT_DPIF_GET_NCONNS() +OVS_TRAFFIC_VSWITCHD_START() + +ADD_NAMESPACES(at_ns0, at_ns1) + +ADD_VETH_AFXDP(p0, at_ns0, br0, "10.1.1.1/24") +ADD_VETH_AFXDP(p1, at_ns1, br0, "10.1.1.2/24") + +dnl Allow any traffic from ns0->ns1. Only allow nd, return traffic from ns1->ns0. +AT_DATA([flows.txt], [dnl +priority=1,action=drop +priority=10,arp,action=normal +priority=100,in_port=1,icmp,action=ct(commit),2 +priority=100,in_port=2,icmp,ct_state=-trk,action=ct(table=0) +priority=100,in_port=2,icmp,ct_state=+trk+est,action=1 +]) + +AT_CHECK([ovs-ofctl --bundle add-flows br0 flows.txt]) + +dnl Pings from ns0->ns1 should work fine. +NS_CHECK_EXEC([at_ns0], [ping -q -c 3 -i 0.3 -w 2 10.1.1.2 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +AT_CHECK([ovs-appctl dpctl/dump-conntrack | FORMAT_CT(10.1.1.2)], [0], [dnl +icmp,orig=(src=10.1.1.1,dst=10.1.1.2,id=,type=8,code=0),reply=(src=10.1.1.2,dst=10.1.1.1,id=,type=0,code=0) +]) + +AT_CHECK([ovs-appctl dpctl/ct-set-maxconns one-bad-dp], [2], [], [dnl +ovs-vswitchd: maxconns missing or malformed (Invalid argument) +ovs-appctl: ovs-vswitchd: server returned an error +]) + +AT_CHECK([ovs-appctl dpctl/ct-set-maxconns a], [2], [], [dnl +ovs-vswitchd: maxconns missing or malformed (Invalid argument) +ovs-appctl: ovs-vswitchd: server returned an error +]) + +AT_CHECK([ovs-appctl dpctl/ct-set-maxconns one-bad-dp 10], [2], [], [dnl +ovs-vswitchd: datapath not found (Invalid argument) +ovs-appctl: ovs-vswitchd: server returned an error +]) + +AT_CHECK([ovs-appctl dpctl/ct-get-maxconns one-bad-dp], [2], [], [dnl +ovs-vswitchd: datapath not found (Invalid argument) +ovs-appctl: ovs-vswitchd: server returned an error +]) + +AT_CHECK([ovs-appctl dpctl/ct-get-nconns one-bad-dp], [2], [], [dnl +ovs-vswitchd: datapath not found (Invalid argument) +ovs-appctl: ovs-vswitchd: server returned an error +]) + +AT_CHECK([ovs-appctl dpctl/ct-get-nconns], [], [dnl +1 +]) + +AT_CHECK([ovs-appctl dpctl/ct-get-maxconns], [], [dnl +3000000 +]) + +AT_CHECK([ovs-appctl dpctl/ct-set-maxconns 10], [], [dnl +setting maxconns successful +]) + +AT_CHECK([ovs-appctl dpctl/ct-get-maxconns], [], [dnl +10 +]) + +AT_CHECK([ovs-appctl dpctl/flush-conntrack]) + +AT_CHECK([ovs-appctl dpctl/ct-get-nconns], [], [dnl +0 +]) + +AT_CHECK([ovs-appctl dpctl/ct-get-maxconns], [], [dnl +10 +]) + +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP + +AT_SETUP([conntrack - IPv6 ping]) +CHECK_CONNTRACK() +OVS_TRAFFIC_VSWITCHD_START() + +ADD_NAMESPACES(at_ns0, at_ns1) + +ADD_VETH_AFXDP(p0, at_ns0, br0, "fc00::1/96") +ADD_VETH_AFXDP(p1, at_ns1, br0, "fc00::2/96") + +AT_DATA([flows.txt], [dnl + +dnl ICMPv6 echo request and reply go to table 1. The rest of the traffic goes +dnl through normal action. +table=0,priority=10,icmp6,icmp_type=128,action=goto_table:1 +table=0,priority=10,icmp6,icmp_type=129,action=goto_table:1 +table=0,priority=1,action=normal + +dnl Allow everything from ns0->ns1. Only allow return traffic from ns1->ns0. +table=1,priority=100,in_port=1,icmp6,action=ct(commit),2 +table=1,priority=100,in_port=2,icmp6,ct_state=-trk,action=ct(table=0) +table=1,priority=100,in_port=2,icmp6,ct_state=+trk+est,action=1 +table=1,priority=1,action=drop +]) + +AT_CHECK([ovs-ofctl --bundle add-flows br0 flows.txt]) + +OVS_WAIT_UNTIL([ip netns exec at_ns0 ping6 -c 1 fc00::2]) + +dnl The above ping creates state in the connection tracker. We're not +dnl interested in that state. +AT_CHECK([ovs-appctl dpctl/flush-conntrack]) + +dnl Pings from ns1->ns0 should fail. +NS_CHECK_EXEC([at_ns1], [ping6 -q -c 3 -i 0.3 -w 2 fc00::1 | FORMAT_PING], [0], [dnl +7 packets transmitted, 0 received, 100% packet loss, time 0ms +]) + +dnl Pings from ns0->ns1 should work fine. +NS_CHECK_EXEC([at_ns0], [ping6 -q -c 3 -i 0.3 -w 2 fc00::2 | FORMAT_PING], [0], [dnl +3 packets transmitted, 3 received, 0% packet loss, time 0ms +]) + +AT_CHECK([ovs-appctl dpctl/dump-conntrack | FORMAT_CT(fc00::2)], [0], [dnl +icmpv6,orig=(src=fc00::1,dst=fc00::2,id=,type=128,code=0),reply=(src=fc00::2,dst=fc00::1,id=,type=129,code=0) +]) + +OVS_TRAFFIC_VSWITCHD_STOP +AT_CLEANUP From patchwork Mon Apr 1 22:46:52 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: William Tu X-Patchwork-Id: 1073465 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Authentication-Results: ozlabs.org; spf=pass (mailfrom) smtp.mailfrom=openvswitch.org (client-ip=140.211.169.12; helo=mail.linuxfoundation.org; envelope-from=ovs-dev-bounces@openvswitch.org; receiver=) Authentication-Results: ozlabs.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: ozlabs.org; dkim=fail reason="signature verification failed" (2048-bit key; unprotected) header.d=gmail.com header.i=@gmail.com header.b="BsjVaybx"; dkim-atps=neutral Received: from mail.linuxfoundation.org (mail.linuxfoundation.org [140.211.169.12]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id 44Y6wX6CcFz9sSV for ; Tue, 2 Apr 2019 09:49:32 +1100 (AEDT) Received: from mail.linux-foundation.org (localhost [127.0.0.1]) by mail.linuxfoundation.org (Postfix) with ESMTP id BB4562286; Mon, 1 Apr 2019 22:47:38 +0000 (UTC) X-Original-To: dev@openvswitch.org Delivered-To: ovs-dev@mail.linuxfoundation.org Received: from smtp1.linuxfoundation.org (smtp1.linux-foundation.org [172.17.192.35]) by mail.linuxfoundation.org (Postfix) with ESMTPS id BE659226A for ; Mon, 1 Apr 2019 22:47:36 +0000 (UTC) X-Greylist: whitelisted by SQLgrey-1.7.6 Received: from mail-wr1-f41.google.com (mail-wr1-f41.google.com [209.85.221.41]) by smtp1.linuxfoundation.org (Postfix) with ESMTPS id E0FD27A6 for ; Mon, 1 Apr 2019 22:47:35 +0000 (UTC) Received: by mail-wr1-f41.google.com with SMTP id r4so14084999wrq.8 for ; Mon, 01 Apr 2019 15:47:35 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:subject:date:message-id:in-reply-to:references; bh=wdGFyeSqeD9atTsW0mC/VyqK8L+Z4F/d2EK9yXiOqf0=; b=BsjVaybxGq5A67deGFH4tPj2xLmsLIkMfW2KxYqPwvqJHUAYIQKK4fjgDT7aOR+2Jg EymyjotMfQAmk7FWQfju6qWY/N7OzaV48c+6w8LF8D+GVrD3SsJo+JBI8u63+FsRFsrx 1k7gQ/sOP1OdvZigXWlNDEdPyq5Kzc0LBNEFrAzZDMCBT7Y72NaY3obZB4wRYyYOy91B CcaL6e9f1nqrf+IWjEMd30kGcGjZB/RlDPRARm7im2pgw1ZIWgh+5u4YzOEsbtVcEbvO PIl9R4heHPGJeg0lYr5zDIBStyk3Y3J0ALFBF5JbMwtWdkKqK33Vv9WdXICJQ+IUp3Bj O8bg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:subject:date:message-id:in-reply-to :references; bh=wdGFyeSqeD9atTsW0mC/VyqK8L+Z4F/d2EK9yXiOqf0=; b=O1rZ9WgQ+l6whbtlyOEwaWZnJBWFlZY8nSayHWTbi7oT+Brzhx5LmKbxIzs1TEnw8U oTgYx6bhz1SScrGL8QrDrldx0sq+81H1iJ0UlIImkQL9nYRA9+ITY3wwAAxhO8tUlyiZ T/bjX5RE9Ce/qnctw4eAbbajQR+e3xSVUlabc0MuTcy0sJfPuEA57q+8nh0gpuwxkXd9 fAczXjLyCEzuoZZkPLqNhmCOvmCFf5hjX4BCC9PXPlmfroZaDRUQ//ZhtwAOEDsOTwtu VMUn4Q69wuGLJkyPbErsNuMBUuP2l3RRbMqg2JAfOsFGDwymVl8JwR3wVDKcGnoDFVB4 WGDg== X-Gm-Message-State: APjAAAXP0qXcvXqLQiUVFhLOmcHGxs9ItCyWSkVoqCchlr0duBVY368K Ty4R7CAYJUY+WZ9mg8bMjsD+Zd7A X-Google-Smtp-Source: APXvYqx7fIH2+pgwzkNmeiDo6Mp02vLRSlweA5An15bJBJLHnZmDcNxcxFuiAn2DEb3oVGl8INZO1w== X-Received: by 2002:a5d:63c5:: with SMTP id c5mr42613869wrw.82.1554158854326; Mon, 01 Apr 2019 15:47:34 -0700 (PDT) Received: from sc9-mailhost2.vmware.com ([66.170.99.2]) by smtp.gmail.com with ESMTPSA id w9sm27720055wmi.0.2019.04.01.15.47.32 (version=TLS1_2 cipher=ECDHE-RSA-AES128-SHA bits=128/128); Mon, 01 Apr 2019 15:47:33 -0700 (PDT) From: William Tu To: dev@openvswitch.org, trozet@redhat.com, bmcfall@redhat.com, echaudro@redhat.com, magnus.karlsson@gmail.com, bjorn.topel@gmail.com, tuc@vmware.com Date: Mon, 1 Apr 2019 15:46:52 -0700 Message-Id: <1554158812-44622-5-git-send-email-u9012063@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1554158812-44622-1-git-send-email-u9012063@gmail.com> References: <1554158812-44622-1-git-send-email-u9012063@gmail.com> X-Spam-Status: No, score=-1.7 required=5.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_ENVFROM_END_DIGIT,FREEMAIL_FROM, RCVD_IN_DNSWL_NONE autolearn=no version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on smtp1.linux-foundation.org Subject: [ovs-dev] [PATCH RFCv4 4/4] afxdp netdev: add documentation and configuration. X-BeenThere: ovs-dev@openvswitch.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Sender: ovs-dev-bounces@openvswitch.org Errors-To: ovs-dev-bounces@openvswitch.org The patch add bpf installation guide and configuration for linking to libbpf library. Signed-off-by: William Tu --- Documentation/automake.mk | 1 + Documentation/index.rst | 1 + Documentation/intro/install/bpf.rst | 182 ++++++++++++++++++++++++++++++++++ Documentation/intro/install/index.rst | 1 + 4 files changed, 185 insertions(+) create mode 100644 Documentation/intro/install/bpf.rst diff --git a/Documentation/automake.mk b/Documentation/automake.mk index 082438e09a33..72c58002a234 100644 --- a/Documentation/automake.mk +++ b/Documentation/automake.mk @@ -10,6 +10,7 @@ DOC_SOURCE = \ Documentation/intro/why-ovs.rst \ Documentation/intro/install/index.rst \ Documentation/intro/install/bash-completion.rst \ + Documentation/intro/install/bpf.rst \ Documentation/intro/install/debian.rst \ Documentation/intro/install/documentation.rst \ Documentation/intro/install/distributions.rst \ diff --git a/Documentation/index.rst b/Documentation/index.rst index 46261235c732..72b1eda80aae 100644 --- a/Documentation/index.rst +++ b/Documentation/index.rst @@ -59,6 +59,7 @@ vSwitch? Start here. :doc:`intro/install/windows` | :doc:`intro/install/xenserver` | :doc:`intro/install/dpdk` | + :doc:`intro/install/bpf` | :doc:`Installation FAQs ` - **Tutorials:** :doc:`tutorials/faucet` | diff --git a/Documentation/intro/install/bpf.rst b/Documentation/intro/install/bpf.rst new file mode 100644 index 000000000000..5f38db368c27 --- /dev/null +++ b/Documentation/intro/install/bpf.rst @@ -0,0 +1,182 @@ +.. + Licensed under the Apache License, Version 2.0 (the "License"); you may + not use this file except in compliance with the License. You may obtain + a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, WITHOUT + WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the + License for the specific language governing permissions and limitations + under the License. + + Convention for heading levels in Open vSwitch documentation: + + ======= Heading 0 (reserved for the title in a document) + ------- Heading 1 + ~~~~~~~ Heading 2 + +++++++ Heading 3 + ''''''' Heading 4 + + Avoid deeper levels because they do not render well. + +================================ +Open vSwitch with BPF/XDP/AF_XDP +================================ + +This document describes how to build and install Open vSwitch using an BPF +datapath. + +.. warning:: + The AF_XDP support of Open vSwitch is considered 'experimental'. + +Build requirements +------------------ + +In addition to the requirements described in :doc:`general`, building Open +vSwitch with AF_XDP will require the following: + +- Latest libbpf from kernel source tree (kernel 5.0.0 or later) + +- Linux kernel 4.19 or later + + The following Kconfig options must be enabled to run the AF_XDP netdev: + +``_CONFIG_BPF=y`` +``_CONFIG_BPF_SYSCALL=y`` +``_CONFIG_XDP_SOCKETS=y`` + + The following optional Kconfig options are also recommended: + +``_CONFIG_BPF_JIT=y`` +``_CONFIG_HAVE_BPF_JIT=y`` +``_CONFIG_XDP_SOCKETS_DIAG=y`` + +- Linux-tools from a recent Linux kernel + +Installing +---------- + +OVS can be installed using different methods. For OVS to use AF_XDP netdev, it +has to be configured with LIBBPF support. + +#. Clone a recent version of Linux bpf-next tree:: + + $ git clone git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next.git + +#. Go into the Linux source directory and build libbpf in the tools directory:: + + $ cd bpf-next/ + $ make -C tools/lib/bpf/ + $ cd tools/lib/bpf/ + $ make install + +#. Ensure the standard OVS requirements, described in + :ref:`general-build-reqs`, are installed + +#. Bootstrap, if required, as described in :ref:`general-bootstrapping` + +#. Configure the package:: + + $ ./configure + + where ``LINUX_TOOLS`` is the path to the Linux tools/ directory that was + compiled in step 2. + + .. note:: + Make sure the ldconfig is properly set-up to point to the libbpf.so + +#. Build and install OVS, as described in :ref:`general-building` + +Additional information can be found in :doc:`general`. + +Setup +----- + +Before running OVS with AF_XDP, make sure the libbpf and libelf are set-up right:: + + # ldd vswitchd/ovs-vswitchd + + .. note:: + For example, you should see: + libbpf.so => /lib/x86_64-linux-gnu/libbpf.so (0x00007eff2fc0a000) + libelf.so.1 => /usr/lib/x86_64-linux-gnu/libelf.so.1 (0x00007eff2f9f2000) + +Open vSwitch should be started using userspace datapath as described in :doc:`general`:: + + # ovs-vswitchd --disable-system + # ovs-vsctl -- add-br br0 -- set Bridge br0 datapath_type=netdev + + .. note:: + OVS AF_XDP netdev is using the userspace datapath, the same datapath + as used by OVS-DPDK. So it requires --disable-system for ovs-vswitchd + and datapath_type=netdev when adding a new bridge. + +When adding a AF_XDP netdev port to the bridge, specify the netdev type as afxdp:: + + # using namespace and veth peer as an example + # ip link add p0 type veth peer name afxdp-p0 + # ip link set p0 netns at_ns0 + # ip link set dev afxdp-p0 up + # ovs-vsctl add-port br0 afxdp-p0 -- \ + set interface afxdp-p0 external-ids:iface-id="p0" type="afxdp" + +To validate that the bridge has successfully instantiated, you can use the +ovs-vsctl utility:: + + # ovs-vsctl show + +Or enable afxdp debugging by:: + + # ovs-appctl vlog/set netdev_afxdp::dbg + +An end-to-end AF_XDP netdev test can be started by running, with failed test +results saved at tests/system-afxdp-testsuite.dir//:: + + # make check-afxdp + # cat tests/system-afxdp-testsuite.dir/06/system-afxdp-testsuite.log + +Example Script +-------------- + +Below is a script using namespaces and veth peer:: + + #!/bin/bash + # ovs-vswitchd --no-chdir --pidfile -vvconn -vofproto_dpif -vunixctl --disable-system --detach + # ovs-vsctl -- add-br br0 -- set Bridge br0 \ + # protocols=OpenFlow10,OpenFlow11,OpenFlow12,OpenFlow13,OpenFlow14,OpenFlow15 \ + # fail-mode=secure datapath_type=netdev + # ovs-vsctl -- add-br br0 -- set Bridge br0 datapath_type=netdev + + # ip netns add at_ns0 + # ovs-appctl vlog/set netdev_afxdp::dbg + + # ip link add p0 type veth peer name afxdp-p0 + # ip link set p0 netns at_ns0 + # ip link set dev afxdp-p0 up + # ovs-vsctl add-port br0 afxdp-p0 -- \ + # set interface afxdp-p0 external-ids:iface-id="p0" type="afxdp" + + # ip netns exec at_ns0 sh << NS_EXEC_HEREDOC + # ip addr add "10.1.1.1/24" dev p0 + # ip link set dev p0 up + # NS_EXEC_HEREDOC + + # ip netns add at_ns1 + # ip link add p1 type veth peer name afxdp-p1 + # ip link set p1 netns at_ns1 + # ip link set dev afxdp-p1 up + # ovs-vsctl add-port br0 afxdp-p1 -- \ + # set interface afxdp-p1 external-ids:iface-id="p1" type="afxdp" + # ip netns exec at_ns1 sh << NS_EXEC_HEREDOC + # ip addr add "10.1.1.2/24" dev p1 + # ip link set dev p1 up + # NS_EXEC_HEREDOC + + # ip netns exec at_ns0 ping -i .2 10.1.1.2 + +Bug Reporting +------------- + +Please report problems to bugs@openvswitch.org. diff --git a/Documentation/intro/install/index.rst b/Documentation/intro/install/index.rst index 3193c736cf17..e063548f2bb3 100644 --- a/Documentation/intro/install/index.rst +++ b/Documentation/intro/install/index.rst @@ -45,6 +45,7 @@ Installation from Source xenserver userspace dpdk + bpf Installation from Packages --------------------------