From patchwork Tue Nov 7 15:59:28 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Mark Kavanagh X-Patchwork-Id: 835356 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=) 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 3yWZ0G5hVpz9s3w for ; Wed, 8 Nov 2017 03:00:46 +1100 (AEDT) Received: from mail.linux-foundation.org (localhost [127.0.0.1]) by mail.linuxfoundation.org (Postfix) with ESMTP id 799CBB0A; Tue, 7 Nov 2017 15:59:36 +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 DD9E9C89 for ; Tue, 7 Nov 2017 15:59:33 +0000 (UTC) X-Greylist: domain auto-whitelisted by SQLgrey-1.7.6 Received: from mga07.intel.com (mga07.intel.com [134.134.136.100]) by smtp1.linuxfoundation.org (Postfix) with ESMTPS id 759CD3DB for ; Tue, 7 Nov 2017 15:59:33 +0000 (UTC) Received: from fmsmga006.fm.intel.com ([10.253.24.20]) by orsmga105.jf.intel.com with ESMTP; 07 Nov 2017 07:59:33 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.44,359,1505804400"; d="scan'208";a="173123610" Received: from silpixa00380299.ir.intel.com ([10.237.222.17]) by fmsmga006.fm.intel.com with ESMTP; 07 Nov 2017 07:59:31 -0800 From: Mark Kavanagh To: dev@openvswitch.org Date: Tue, 7 Nov 2017 15:59:28 +0000 Message-Id: <1510070368-242276-3-git-send-email-mark.b.kavanagh@intel.com> X-Mailer: git-send-email 1.9.3 In-Reply-To: <1510070368-242276-1-git-send-email-mark.b.kavanagh@intel.com> References: <1510070368-242276-1-git-send-email-mark.b.kavanagh@intel.com> X-Spam-Status: No, score=-2.3 required=5.0 tests=RCVD_IN_DNSWL_MED, RP_MATCHES_RCVD autolearn=disabled version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on smtp1.linux-foundation.org Cc: maxime.coquelin@redhat.com Subject: [ovs-dev] [RFC PATCH 2/2] netdev-dpdk: add support for vhost IOMMU feature 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 DPDK v17.11 introduces support for the vHost IOMMU feature. This is a security feature, that restricts the vhost memory that a virtio device may access. This feature also enables the vhost REPLY_ACK protocol, the implementation of which is known to work in newer versions of QEMU (i.e. v2.10.0), but is buggy in older versions (v2.7.0 - v2.9.0, inclusive). As such, the feature is disabled by default in (and should remain so, for the aforementioned older QEMU verions). This patch adds a new vhost port option, vhost-iommu-support, to allow enablement of the vhost IOMMU feature: $ ovs-vsctl add-port br0 vhost-client-1 \ -- set Interface vhost-client-1 type=dpdkvhostuserclient \ options:vhost-server-path=$VHOST_USER_SOCKET_PATH \ options:vhost-iommu-support=true Note that support for this feature is only implemented for vhost user client ports (since vhost user ports are considered deprecated). Signed-off-by: Mark Kavanagh --- Documentation/topics/dpdk/vhost-user.rst | 19 +++++++++++++++++++ NEWS | 1 + lib/netdev-dpdk.c | 26 +++++++++++++++++++++----- vswitchd/vswitch.xml | 10 ++++++++++ 4 files changed, 51 insertions(+), 5 deletions(-) diff --git a/Documentation/topics/dpdk/vhost-user.rst b/Documentation/topics/dpdk/vhost-user.rst index fdf6ae1..ba70e45 100644 --- a/Documentation/topics/dpdk/vhost-user.rst +++ b/Documentation/topics/dpdk/vhost-user.rst @@ -250,6 +250,25 @@ Once the vhost-user-client ports have been added to the switch, they must be added to the guest. Like vhost-user ports, there are two ways to do this: using QEMU directly, or using libvirt. Only the QEMU case is covered here. +vhost-user client IOMMU +~~~~~~~~~~~~~~~~~~~~~~~ +It is possible to enable IOMMU support for vHost User client ports. This is +a feature which restricts the vhost memory that a virtio device can access, and +as such is useful in deployments in which security is a concern. IOMMU mode may +be enabled on the command line:: + + $ ovs-vsctl add-port br0 vhost-client-1 \ + -- set Interface vhost-client-1 type=dpdkvhostuserclient \ + options:vhost-server-path=$VHOST_USER_SOCKET_PATH \ + options:vhost-iommu-support=true + +.. important:: + + Enabling the IOMMU feature also enables the vhost user reply-ack protocol; + this is known to work on QEMU v2.10.0, but is buggy on older versions + (2.7.0 - 2.9.0, inclusive). Consequently, the IOMMU feaure is disabled by + default (and should remain so if using the aforementioned versions of QEMU). + Adding vhost-user-client ports to the guest (QEMU) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ diff --git a/NEWS b/NEWS index 6acd8bd..de53d2f 100644 --- a/NEWS +++ b/NEWS @@ -13,6 +13,7 @@ Post-v2.8.0 * Add support for compiling OVS with the latest Linux 4.13 kernel - DPDK: * Add support for DPDK v17.11 + * Add support for vHost IOMMU feature v2.8.0 - 31 Aug 2017 -------------------- diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c index ed5bf62..d214fac 100644 --- a/lib/netdev-dpdk.c +++ b/lib/netdev-dpdk.c @@ -1424,15 +1424,22 @@ netdev_dpdk_vhost_client_set_config(struct netdev *netdev, { struct netdev_dpdk *dev = netdev_dpdk_cast(netdev); const char *path; + bool iommu_enable; ovs_mutex_lock(&dev->mutex); if (!(dev->vhost_driver_flags & RTE_VHOST_USER_CLIENT)) { path = smap_get(args, "vhost-server-path"); - if (path && strcmp(path, dev->vhost_id)) { + if (path && strcmp(path, dev->vhost_id)) strcpy(dev->vhost_id, path); - netdev_request_reconfigure(netdev); - } } + + iommu_enable = smap_get_bool(args, "vhost-iommu-support", false); + if (iommu_enable) + dev->vhost_driver_flags |= RTE_VHOST_USER_IOMMU_SUPPORT; + else + dev->vhost_driver_flags &= ~RTE_VHOST_USER_IOMMU_SUPPORT; + + netdev_request_reconfigure(netdev); ovs_mutex_unlock(&dev->mutex); return 0; @@ -3326,9 +3333,18 @@ netdev_dpdk_vhost_client_reconfigure(struct netdev *netdev) */ if (!(dev->vhost_driver_flags & RTE_VHOST_USER_CLIENT) && strlen(dev->vhost_id)) { + /* Enable vhost IOMMU, if it was requested. + * XXX: the 'flags' variable is required, as not all vhost backend + * features are currently supported by OvS; in time, it should be + * possible to invoke rte_vhost_driver_register(), passing + * dev->vhost_driver_flags directly as a parameter to same. + */ + uint64_t flags = RTE_VHOST_USER_CLIENT; + if (dev->vhost_driver_flags & RTE_VHOST_USER_IOMMU_SUPPORT) + flags |= RTE_VHOST_USER_IOMMU_SUPPORT; + /* Register client-mode device */ - err = rte_vhost_driver_register(dev->vhost_id, - RTE_VHOST_USER_CLIENT); + err = rte_vhost_driver_register(dev->vhost_id, flags); if (err) { VLOG_ERR("vhost-user device setup failure for device %s\n", dev->vhost_id); diff --git a/vswitchd/vswitch.xml b/vswitchd/vswitch.xml index d7f6839..df5a42d 100644 --- a/vswitchd/vswitch.xml +++ b/vswitchd/vswitch.xml @@ -2649,6 +2649,16 @@ ovs-vsctl add-port br0 p0 -- set Interface p0 type=patch options:peer=p1 \

+ +

+ The value specifies whether IOMMU support is enabled for a vHost User + client mode device that has been or will be created by QEMU. + Only supported by dpdkvhostuserclient interfaces. If not specified or + an incorrect value is specified, defaults to 'false'. +

+ +