From patchwork Fri Jun 8 03:22:23 2018 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Tiwei Bie X-Patchwork-Id: 40803 X-Patchwork-Delegate: maxime.coquelin@redhat.com Return-Path: X-Original-To: patchwork@dpdk.org Delivered-To: patchwork@dpdk.org Received: from [92.243.14.124] (localhost [127.0.0.1]) by dpdk.org (Postfix) with ESMTP id 921EC1B28E; Fri, 8 Jun 2018 05:22:22 +0200 (CEST) Received: from mga17.intel.com (mga17.intel.com [192.55.52.151]) by dpdk.org (Postfix) with ESMTP id 988D1D49E for ; Fri, 8 Jun 2018 05:22:16 +0200 (CEST) X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from fmsmga008.fm.intel.com ([10.253.24.58]) by fmsmga107.fm.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 07 Jun 2018 20:22:14 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.49,489,1520924400"; d="scan'208";a="46168861" Received: from debian.sh.intel.com ([10.67.104.203]) by fmsmga008.fm.intel.com with ESMTP; 07 Jun 2018 20:22:14 -0700 From: Tiwei Bie To: maxime.coquelin@redhat.com, dev@dpdk.org Date: Fri, 8 Jun 2018 11:22:23 +0800 Message-Id: <20180608032224.22049-2-tiwei.bie@intel.com> X-Mailer: git-send-email 2.17.0 In-Reply-To: <20180608032224.22049-1-tiwei.bie@intel.com> References: <20180608032224.22049-1-tiwei.bie@intel.com> Subject: [dpdk-dev] [PATCH v3 1/2] vhost: support host notifier X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" When a vDPA device is attached, vhost user will try to register host notifiers to QEMU to allow notifications to be delivered between the driver in the guest and the vDPA device in the host directly. Signed-off-by: Tiwei Bie --- lib/librte_vhost/rte_vhost.h | 8 ++ lib/librte_vhost/vhost.c | 3 + lib/librte_vhost/vhost.h | 1 + lib/librte_vhost/vhost_user.c | 146 ++++++++++++++++++++++++++++++++++ lib/librte_vhost/vhost_user.h | 13 ++- 5 files changed, 170 insertions(+), 1 deletion(-) diff --git a/lib/librte_vhost/rte_vhost.h b/lib/librte_vhost/rte_vhost.h index 7f0cb9bc8..b02673d4a 100644 --- a/lib/librte_vhost/rte_vhost.h +++ b/lib/librte_vhost/rte_vhost.h @@ -58,6 +58,14 @@ extern "C" { #define VHOST_USER_PROTOCOL_F_CRYPTO_SESSION 7 #endif +#ifndef VHOST_USER_PROTOCOL_F_SLAVE_SEND_FD +#define VHOST_USER_PROTOCOL_F_SLAVE_SEND_FD 10 +#endif + +#ifndef VHOST_USER_PROTOCOL_F_HOST_NOTIFIER +#define VHOST_USER_PROTOCOL_F_HOST_NOTIFIER 11 +#endif + /** Indicate whether protocol features negotiation is supported. */ #ifndef VHOST_USER_F_PROTOCOL_FEATURES #define VHOST_USER_F_PROTOCOL_FEATURES 30 diff --git a/lib/librte_vhost/vhost.c b/lib/librte_vhost/vhost.c index afded4952..6ca5beb13 100644 --- a/lib/librte_vhost/vhost.c +++ b/lib/librte_vhost/vhost.c @@ -291,6 +291,7 @@ vhost_new_device(void) dev->flags = VIRTIO_DEV_BUILTIN_VIRTIO_NET; dev->slave_req_fd = -1; dev->vdpa_dev_id = -1; + rte_spinlock_init(&dev->slave_req_lock); return i; } @@ -346,6 +347,8 @@ vhost_detach_vdpa_device(int vid) if (dev == NULL) return; + vhost_user_host_notifier_ctrl(vid, false); + dev->vdpa_dev_id = -1; } diff --git a/lib/librte_vhost/vhost.h b/lib/librte_vhost/vhost.h index 58c425a5c..b10ac124b 100644 --- a/lib/librte_vhost/vhost.h +++ b/lib/librte_vhost/vhost.h @@ -301,6 +301,7 @@ struct virtio_net { struct guest_page *guest_pages; int slave_req_fd; + rte_spinlock_t slave_req_lock; /* * Device id to identify a specific backend device. diff --git a/lib/librte_vhost/vhost_user.c b/lib/librte_vhost/vhost_user.c index 947290fc3..a28fc3495 100644 --- a/lib/librte_vhost/vhost_user.c +++ b/lib/librte_vhost/vhost_user.c @@ -1384,6 +1384,22 @@ send_vhost_reply(int sockfd, struct VhostUserMsg *msg) return send_vhost_message(sockfd, msg, NULL, 0); } +static int +send_vhost_slave_message(struct virtio_net *dev, struct VhostUserMsg *msg, + int *fds, int fd_num) +{ + int ret; + + if (msg->flags & VHOST_USER_NEED_REPLY) + rte_spinlock_lock(&dev->slave_req_lock); + + ret = send_vhost_message(dev->slave_req_fd, msg, fds, fd_num); + if (ret < 0 && (msg->flags & VHOST_USER_NEED_REPLY)) + rte_spinlock_unlock(&dev->slave_req_lock); + + return ret; +} + /* * Allocate a queue pair if it hasn't been allocated yet */ @@ -1705,11 +1721,45 @@ vhost_user_msg_handler(int vid, int fd) if (vdpa_dev->ops->dev_conf) vdpa_dev->ops->dev_conf(dev->vid); dev->flags |= VIRTIO_DEV_VDPA_CONFIGURED; + if (vhost_user_host_notifier_ctrl(dev->vid, true) != 0) { + RTE_LOG(INFO, VHOST_CONFIG, + "(%d) software relay is used for vDPA, performance may be low.\n", + dev->vid); + } } return 0; } +static int process_slave_message_reply(struct virtio_net *dev, + const VhostUserMsg *msg) +{ + VhostUserMsg msg_reply; + int ret; + + if ((msg->flags & VHOST_USER_NEED_REPLY) == 0) + return 0; + + if (read_vhost_message(dev->slave_req_fd, &msg_reply) < 0) { + ret = -1; + goto out; + } + + if (msg_reply.request.slave != msg->request.slave) { + RTE_LOG(ERR, VHOST_CONFIG, + "Received unexpected msg type (%u), expected %u\n", + msg_reply.request.slave, msg->request.slave); + ret = -1; + goto out; + } + + ret = msg_reply.payload.u64 ? -1 : 0; + +out: + rte_spinlock_unlock(&dev->slave_req_lock); + return ret; +} + int vhost_user_iotlb_miss(struct virtio_net *dev, uint64_t iova, uint8_t perm) { @@ -1735,3 +1785,99 @@ vhost_user_iotlb_miss(struct virtio_net *dev, uint64_t iova, uint8_t perm) return 0; } + +static int vhost_user_slave_set_vring_host_notifier(struct virtio_net *dev, + int index, int fd, + uint64_t offset, + uint64_t size) +{ + int *fdp = NULL; + size_t fd_num = 0; + int ret; + struct VhostUserMsg msg = { + .request.slave = VHOST_USER_SLAVE_VRING_HOST_NOTIFIER_MSG, + .flags = VHOST_USER_VERSION | VHOST_USER_NEED_REPLY, + .size = sizeof(msg.payload.area), + .payload.area = { + .u64 = index & VHOST_USER_VRING_IDX_MASK, + .size = size, + .offset = offset, + }, + }; + + if (fd < 0) + msg.payload.area.u64 |= VHOST_USER_VRING_NOFD_MASK; + else { + fdp = &fd; + fd_num = 1; + } + + ret = send_vhost_slave_message(dev, &msg, fdp, fd_num); + if (ret < 0) { + RTE_LOG(ERR, VHOST_CONFIG, + "Failed to set host notifier (%d)\n", ret); + return ret; + } + + return process_slave_message_reply(dev, &msg); +} + +int vhost_user_host_notifier_ctrl(int vid, bool enable) +{ + struct virtio_net *dev; + struct rte_vdpa_device *vdpa_dev; + int vfio_device_fd, did, ret = 0; + uint64_t offset, size; + unsigned int i; + + dev = get_device(vid); + if (!dev) + return -ENODEV; + + did = dev->vdpa_dev_id; + if (did < 0) + return -EINVAL; + + if (!(dev->features & (1ULL << VIRTIO_F_VERSION_1)) || + !(dev->features & (1ULL << VHOST_USER_F_PROTOCOL_FEATURES)) || + !(dev->protocol_features & + (1ULL << VHOST_USER_PROTOCOL_F_SLAVE_REQ)) || + !(dev->protocol_features & + (1ULL << VHOST_USER_PROTOCOL_F_SLAVE_SEND_FD)) || + !(dev->protocol_features & + (1ULL << VHOST_USER_PROTOCOL_F_HOST_NOTIFIER))) + return -ENOTSUP; + + vdpa_dev = rte_vdpa_get_device(did); + + RTE_FUNC_PTR_OR_ERR_RET(vdpa_dev->ops->get_vfio_device_fd, -ENOTSUP); + RTE_FUNC_PTR_OR_ERR_RET(vdpa_dev->ops->get_notify_area, -ENOTSUP); + + vfio_device_fd = vdpa_dev->ops->get_vfio_device_fd(vid); + if (vfio_device_fd < 0) + return -ENOTSUP; + + if (enable) { + for (i = 0; i < dev->nr_vring; i++) { + if (vdpa_dev->ops->get_notify_area(vid, i, &offset, + &size) < 0) { + ret = -ENOTSUP; + goto disable; + } + + if (vhost_user_slave_set_vring_host_notifier(dev, i, + vfio_device_fd, offset, size) < 0) { + ret = -EFAULT; + goto disable; + } + } + } else { +disable: + for (i = 0; i < dev->nr_vring; i++) { + vhost_user_slave_set_vring_host_notifier(dev, i, -1, + 0, 0); + } + } + + return ret; +} diff --git a/lib/librte_vhost/vhost_user.h b/lib/librte_vhost/vhost_user.h index 1ad5cf467..42166adf2 100644 --- a/lib/librte_vhost/vhost_user.h +++ b/lib/librte_vhost/vhost_user.h @@ -20,7 +20,9 @@ (1ULL << VHOST_USER_PROTOCOL_F_REPLY_ACK) | \ (1ULL << VHOST_USER_PROTOCOL_F_NET_MTU) | \ (1ULL << VHOST_USER_PROTOCOL_F_SLAVE_REQ) | \ - (1ULL << VHOST_USER_PROTOCOL_F_CRYPTO_SESSION)) + (1ULL << VHOST_USER_PROTOCOL_F_CRYPTO_SESSION) | \ + (1ULL << VHOST_USER_PROTOCOL_F_SLAVE_SEND_FD) | \ + (1ULL << VHOST_USER_PROTOCOL_F_HOST_NOTIFIER)) typedef enum VhostUserRequest { VHOST_USER_NONE = 0, @@ -54,6 +56,7 @@ typedef enum VhostUserRequest { typedef enum VhostUserSlaveRequest { VHOST_USER_SLAVE_NONE = 0, VHOST_USER_SLAVE_IOTLB_MSG = 1, + VHOST_USER_SLAVE_VRING_HOST_NOTIFIER_MSG = 3, VHOST_USER_SLAVE_MAX } VhostUserSlaveRequest; @@ -99,6 +102,12 @@ typedef struct VhostUserCryptoSessionParam { uint8_t auth_key_buf[VHOST_USER_CRYPTO_MAX_HMAC_KEY_LENGTH]; } VhostUserCryptoSessionParam; +typedef struct VhostUserVringArea { + uint64_t u64; + uint64_t size; + uint64_t offset; +} VhostUserVringArea; + typedef struct VhostUserMsg { union { uint32_t master; /* a VhostUserRequest value */ @@ -120,6 +129,7 @@ typedef struct VhostUserMsg { VhostUserLog log; struct vhost_iotlb_msg iotlb; VhostUserCryptoSessionParam crypto_session; + VhostUserVringArea area; } payload; int fds[VHOST_MEMORY_MAX_NREGIONS]; } __attribute((packed)) VhostUserMsg; @@ -133,6 +143,7 @@ typedef struct VhostUserMsg { /* vhost_user.c */ int vhost_user_msg_handler(int vid, int fd); int vhost_user_iotlb_miss(struct virtio_net *dev, uint64_t iova, uint8_t perm); +int vhost_user_host_notifier_ctrl(int vid, bool enable); /* socket.c */ int read_fd_message(int sockfd, char *buf, int buflen, int *fds, int fd_num);