From patchwork Fri Jul 16 02:59:19 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Jiang, Cheng1" X-Patchwork-Id: 95929 X-Patchwork-Delegate: maxime.coquelin@redhat.com Return-Path: X-Original-To: patchwork@inbox.dpdk.org Delivered-To: patchwork@inbox.dpdk.org Received: from mails.dpdk.org (mails.dpdk.org [217.70.189.124]) by inbox.dpdk.org (Postfix) with ESMTP id BA039A0C4A; Fri, 16 Jul 2021 05:16:04 +0200 (CEST) Received: from [217.70.189.124] (localhost [127.0.0.1]) by mails.dpdk.org (Postfix) with ESMTP id 4C655412AC; Fri, 16 Jul 2021 05:16:00 +0200 (CEST) Received: from mga09.intel.com (mga09.intel.com [134.134.136.24]) by mails.dpdk.org (Postfix) with ESMTP id E398B40143 for ; Fri, 16 Jul 2021 05:15:57 +0200 (CEST) X-IronPort-AV: E=McAfee;i="6200,9189,10046"; a="210653762" X-IronPort-AV: E=Sophos;i="5.84,244,1620716400"; d="scan'208";a="210653762" Received: from fmsmga002.fm.intel.com ([10.253.24.26]) by orsmga102.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 15 Jul 2021 20:15:57 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.84,244,1620716400"; d="scan'208";a="506308141" Received: from dpdk_jiangcheng.sh.intel.com ([10.67.119.149]) by fmsmga002.fm.intel.com with ESMTP; 15 Jul 2021 20:15:55 -0700 From: Cheng Jiang To: maxime.coquelin@redhat.com, Chenbo.Xia@intel.com Cc: dev@dpdk.org, jiayu.hu@intel.com, yvonnex.yang@intel.com, Cheng Jiang Date: Fri, 16 Jul 2021 02:59:19 +0000 Message-Id: <20210716025923.27164-2-cheng1.jiang@intel.com> X-Mailer: git-send-email 2.29.2 In-Reply-To: <20210716025923.27164-1-cheng1.jiang@intel.com> References: <20210602042802.31943-1-cheng1.jiang@intel.com> <20210716025923.27164-1-cheng1.jiang@intel.com> MIME-Version: 1.0 Subject: [dpdk-dev] [PATCH v4 1/5] vhost: fix async vhost ops return type X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.29 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" The async vhost ops callback should return -1 when there are something wrong in the callback, so the return type should be changed into int32_t. The issue in vhost example is also fixed in this patch. Signed-off-by: Cheng Jiang --- examples/vhost/ioat.c | 4 +-- examples/vhost/ioat.h | 4 +-- lib/vhost/rte_vhost_async.h | 4 +-- lib/vhost/virtio_net.c | 58 ++++++++++++++++++++++++++++++++----- 4 files changed, 56 insertions(+), 14 deletions(-) diff --git a/examples/vhost/ioat.c b/examples/vhost/ioat.c index 2a2c2d7202..457f8171f0 100644 --- a/examples/vhost/ioat.c +++ b/examples/vhost/ioat.c @@ -122,7 +122,7 @@ open_ioat(const char *value) return ret; } -uint32_t +int32_t ioat_transfer_data_cb(int vid, uint16_t queue_id, struct rte_vhost_async_desc *descs, struct rte_vhost_async_status *opaque_data, uint16_t count) @@ -168,7 +168,7 @@ ioat_transfer_data_cb(int vid, uint16_t queue_id, return i_desc; } -uint32_t +int32_t ioat_check_completed_copies_cb(int vid, uint16_t queue_id, struct rte_vhost_async_status *opaque_data, uint16_t max_packets) diff --git a/examples/vhost/ioat.h b/examples/vhost/ioat.h index 1aa28ed6a3..b57b5645b0 100644 --- a/examples/vhost/ioat.h +++ b/examples/vhost/ioat.h @@ -27,12 +27,12 @@ struct dma_for_vhost { #ifdef RTE_RAW_IOAT int open_ioat(const char *value); -uint32_t +int32_t ioat_transfer_data_cb(int vid, uint16_t queue_id, struct rte_vhost_async_desc *descs, struct rte_vhost_async_status *opaque_data, uint16_t count); -uint32_t +int32_t ioat_check_completed_copies_cb(int vid, uint16_t queue_id, struct rte_vhost_async_status *opaque_data, uint16_t max_packets); diff --git a/lib/vhost/rte_vhost_async.h b/lib/vhost/rte_vhost_async.h index 6faa31f5ad..bc81cd0caa 100644 --- a/lib/vhost/rte_vhost_async.h +++ b/lib/vhost/rte_vhost_async.h @@ -61,7 +61,7 @@ struct rte_vhost_async_channel_ops { * @return * number of descs processed */ - uint32_t (*transfer_data)(int vid, uint16_t queue_id, + int32_t (*transfer_data)(int vid, uint16_t queue_id, struct rte_vhost_async_desc *descs, struct rte_vhost_async_status *opaque_data, uint16_t count); @@ -78,7 +78,7 @@ struct rte_vhost_async_channel_ops { * @return * number of async descs completed */ - uint32_t (*check_completed_copies)(int vid, uint16_t queue_id, + int32_t (*check_completed_copies)(int vid, uint16_t queue_id, struct rte_vhost_async_status *opaque_data, uint16_t max_packets); }; diff --git a/lib/vhost/virtio_net.c b/lib/vhost/virtio_net.c index b93482587c..8156796a46 100644 --- a/lib/vhost/virtio_net.c +++ b/lib/vhost/virtio_net.c @@ -1528,6 +1528,7 @@ virtio_dev_rx_async_submit_split(struct virtio_net *dev, struct async_inflight_info *pkts_info = vq->async_pkts_info; uint32_t n_pkts = 0, pkt_err = 0; uint32_t num_async_pkts = 0, num_done_pkts = 0; + int32_t n_enq; struct { uint16_t pkt_idx; uint16_t last_avail_idx; @@ -1608,8 +1609,16 @@ virtio_dev_rx_async_submit_split(struct virtio_net *dev, if (unlikely(pkt_burst_idx >= VHOST_ASYNC_BATCH_THRESHOLD || ((VHOST_MAX_ASYNC_VEC >> 1) - segs_await < BUF_VECTOR_MAX))) { - n_pkts = vq->async_ops.transfer_data(dev->vid, + n_enq = vq->async_ops.transfer_data(dev->vid, queue_id, tdes, 0, pkt_burst_idx); + if (n_enq >= 0) { + n_pkts = n_enq; + } else { + VHOST_LOG_DATA(ERR, "(%d) %s: wrong opaque data for queue id %d.\n", + dev->vid, __func__, queue_id); + n_pkts = 0; + } + iovec_idx = 0; it_idx = 0; @@ -1632,8 +1641,15 @@ virtio_dev_rx_async_submit_split(struct virtio_net *dev, } if (pkt_burst_idx) { - n_pkts = vq->async_ops.transfer_data(dev->vid, - queue_id, tdes, 0, pkt_burst_idx); + n_enq = vq->async_ops.transfer_data(dev->vid, queue_id, tdes, 0, pkt_burst_idx); + if (n_enq >= 0) { + n_pkts = n_enq; + } else { + VHOST_LOG_DATA(ERR, "(%d) %s: wrong opaque data for queue id %d.\n", + dev->vid, __func__, queue_id); + n_pkts = 0; + } + vq->async_pkts_inflight_n += n_pkts; if (unlikely(n_pkts < pkt_burst_idx)) @@ -1903,6 +1919,7 @@ virtio_dev_rx_async_submit_packed(struct virtio_net *dev, uint16_t async_descs_idx = 0; uint16_t num_buffers; uint16_t num_descs; + int32_t n_enq; struct rte_vhost_iov_iter *it_pool = vq->it_pool; struct iovec *vec_pool = vq->vec_pool; @@ -1983,8 +2000,16 @@ virtio_dev_rx_async_submit_packed(struct virtio_net *dev, */ if (unlikely(pkt_burst_idx >= VHOST_ASYNC_BATCH_THRESHOLD || ((VHOST_MAX_ASYNC_VEC >> 1) - segs_await < BUF_VECTOR_MAX))) { - n_pkts = vq->async_ops.transfer_data(dev->vid, queue_id, - tdes, 0, pkt_burst_idx); + n_enq = vq->async_ops.transfer_data(dev->vid, + queue_id, tdes, 0, pkt_burst_idx); + if (n_enq >= 0) { + n_pkts = n_enq; + } else { + VHOST_LOG_DATA(ERR, "(%d) %s: wrong opaque data for queue id %d.\n", + dev->vid, __func__, queue_id); + n_pkts = 0; + } + iovec_idx = 0; it_idx = 0; segs_await = 0; @@ -2006,7 +2031,15 @@ virtio_dev_rx_async_submit_packed(struct virtio_net *dev, } while (pkt_idx < count); if (pkt_burst_idx) { - n_pkts = vq->async_ops.transfer_data(dev->vid, queue_id, tdes, 0, pkt_burst_idx); + n_enq = vq->async_ops.transfer_data(dev->vid, queue_id, tdes, 0, pkt_burst_idx); + if (n_enq >= 0) { + n_pkts = n_enq; + } else { + VHOST_LOG_DATA(ERR, "(%d) %s: wrong opaque data for queue id %d.\n", + dev->vid, __func__, queue_id); + n_pkts = 0; + } + vq->async_pkts_inflight_n += n_pkts; if (unlikely(n_pkts < pkt_burst_idx)) @@ -2091,6 +2124,7 @@ uint16_t rte_vhost_poll_enqueue_completed(int vid, uint16_t queue_id, uint16_t start_idx, pkts_idx, vq_size; struct async_inflight_info *pkts_info; uint16_t from, i; + int32_t n_poll; if (!dev) return 0; @@ -2118,9 +2152,17 @@ uint16_t rte_vhost_poll_enqueue_completed(int vid, uint16_t queue_id, start_idx = virtio_dev_rx_async_get_info_idx(pkts_idx, vq_size, vq->async_pkts_inflight_n); - if (count > vq->async_last_pkts_n) - n_pkts_cpl = vq->async_ops.check_completed_copies(vid, + if (count > vq->async_last_pkts_n) { + n_poll = vq->async_ops.check_completed_copies(vid, queue_id, 0, count - vq->async_last_pkts_n); + if (n_poll >= 0) { + n_pkts_cpl = n_poll; + } else { + VHOST_LOG_DATA(ERR, "(%d) %s: wrong opaque data for queue id %d.\n", + dev->vid, __func__, queue_id); + n_pkts_cpl = 0; + } + } n_pkts_cpl += vq->async_last_pkts_n; n_pkts_put = RTE_MIN(count, n_pkts_cpl); From patchwork Fri Jul 16 02:59:20 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Jiang, Cheng1" X-Patchwork-Id: 95930 X-Patchwork-Delegate: maxime.coquelin@redhat.com Return-Path: X-Original-To: patchwork@inbox.dpdk.org Delivered-To: patchwork@inbox.dpdk.org Received: from mails.dpdk.org (mails.dpdk.org [217.70.189.124]) by inbox.dpdk.org (Postfix) with ESMTP id DD0FEA0C4A; Fri, 16 Jul 2021 05:16:10 +0200 (CEST) Received: from [217.70.189.124] (localhost [127.0.0.1]) by mails.dpdk.org (Postfix) with ESMTP id 6F0C0412DF; Fri, 16 Jul 2021 05:16:03 +0200 (CEST) Received: from mga09.intel.com (mga09.intel.com [134.134.136.24]) by mails.dpdk.org (Postfix) with ESMTP id 557B9412C9 for ; Fri, 16 Jul 2021 05:16:01 +0200 (CEST) X-IronPort-AV: E=McAfee;i="6200,9189,10046"; a="210653768" X-IronPort-AV: E=Sophos;i="5.84,244,1620716400"; d="scan'208";a="210653768" Received: from fmsmga002.fm.intel.com ([10.253.24.26]) by orsmga102.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 15 Jul 2021 20:16:00 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.84,244,1620716400"; d="scan'208";a="506308149" Received: from dpdk_jiangcheng.sh.intel.com ([10.67.119.149]) by fmsmga002.fm.intel.com with ESMTP; 15 Jul 2021 20:15:58 -0700 From: Cheng Jiang To: maxime.coquelin@redhat.com, Chenbo.Xia@intel.com Cc: dev@dpdk.org, jiayu.hu@intel.com, yvonnex.yang@intel.com, Cheng Jiang Date: Fri, 16 Jul 2021 02:59:20 +0000 Message-Id: <20210716025923.27164-3-cheng1.jiang@intel.com> X-Mailer: git-send-email 2.29.2 In-Reply-To: <20210716025923.27164-1-cheng1.jiang@intel.com> References: <20210602042802.31943-1-cheng1.jiang@intel.com> <20210716025923.27164-1-cheng1.jiang@intel.com> MIME-Version: 1.0 Subject: [dpdk-dev] [PATCH v4 2/5] vhost: add unsafe API to drain pkts in async vhost X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.29 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" Applications need to stop DMA transfers and finish all the in-flight pkts when in VM memory hot-plug case and async vhost is used. This patch is to provide an unsafe API to drain in-flight pkts which are submitted to DMA engine in vhost async data path. Signed-off-by: Cheng Jiang --- lib/vhost/rte_vhost_async.h | 24 ++++++++++ lib/vhost/version.map | 3 ++ lib/vhost/virtio_net.c | 90 +++++++++++++++++++++++++++---------- 3 files changed, 94 insertions(+), 23 deletions(-) diff --git a/lib/vhost/rte_vhost_async.h b/lib/vhost/rte_vhost_async.h index bc81cd0caa..fd622631b2 100644 --- a/lib/vhost/rte_vhost_async.h +++ b/lib/vhost/rte_vhost_async.h @@ -193,4 +193,28 @@ __rte_experimental uint16_t rte_vhost_poll_enqueue_completed(int vid, uint16_t queue_id, struct rte_mbuf **pkts, uint16_t count); +/** + * This function checks async completion status and empty all pakcets + * for a specific vhost device queue. Packets which are inflight will + * be returned in an array. + * + * @note This function does not perform any locking + * + * @param vid + * id of vhost device to enqueue data + * @param queue_id + * queue id to enqueue data + * @param pkts + * blank array to get return packet pointer + * @param count + * size of the packet array + * @param times + * max number of poll attempts + * @return + * num of packets returned + */ +__rte_experimental +uint16_t rte_vhost_try_drain_queue_thread_unsafe(int vid, uint16_t queue_id, + struct rte_mbuf **pkts, uint16_t count, uint16_t times); + #endif /* _RTE_VHOST_ASYNC_H_ */ diff --git a/lib/vhost/version.map b/lib/vhost/version.map index 9103a23cd4..b8fc8770dd 100644 --- a/lib/vhost/version.map +++ b/lib/vhost/version.map @@ -79,4 +79,7 @@ EXPERIMENTAL { # added in 21.05 rte_vhost_get_negotiated_protocol_features; + + # added in 21.08 + rte_vhost_try_drain_queue_thread_unsafe; }; diff --git a/lib/vhost/virtio_net.c b/lib/vhost/virtio_net.c index 8156796a46..9f541679b9 100644 --- a/lib/vhost/virtio_net.c +++ b/lib/vhost/virtio_net.c @@ -2115,10 +2115,10 @@ write_back_completed_descs_packed(struct vhost_virtqueue *vq, } while (nr_left > 0); } -uint16_t rte_vhost_poll_enqueue_completed(int vid, uint16_t queue_id, +static __rte_always_inline uint16_t +vhost_poll_enqueue_completed(struct virtio_net *dev, uint16_t queue_id, struct rte_mbuf **pkts, uint16_t count) { - struct virtio_net *dev = get_device(vid); struct vhost_virtqueue *vq; uint16_t n_pkts_cpl = 0, n_pkts_put = 0, n_descs = 0, n_buffers = 0; uint16_t start_idx, pkts_idx, vq_size; @@ -2126,26 +2126,8 @@ uint16_t rte_vhost_poll_enqueue_completed(int vid, uint16_t queue_id, uint16_t from, i; int32_t n_poll; - if (!dev) - return 0; - - VHOST_LOG_DATA(DEBUG, "(%d) %s\n", dev->vid, __func__); - if (unlikely(!is_valid_virt_queue_idx(queue_id, 0, dev->nr_vring))) { - VHOST_LOG_DATA(ERR, "(%d) %s: invalid virtqueue idx %d.\n", - dev->vid, __func__, queue_id); - return 0; - } - vq = dev->virtqueue[queue_id]; - if (unlikely(!vq->async_registered)) { - VHOST_LOG_DATA(ERR, "(%d) %s: async not registered for queue id %d.\n", - dev->vid, __func__, queue_id); - return 0; - } - - rte_spinlock_lock(&vq->access_lock); - pkts_idx = vq->async_pkts_idx % vq->size; pkts_info = vq->async_pkts_info; vq_size = vq->size; @@ -2153,7 +2135,7 @@ uint16_t rte_vhost_poll_enqueue_completed(int vid, uint16_t queue_id, vq_size, vq->async_pkts_inflight_n); if (count > vq->async_last_pkts_n) { - n_poll = vq->async_ops.check_completed_copies(vid, + n_poll = vq->async_ops.check_completed_copies(dev->vid, queue_id, 0, count - vq->async_last_pkts_n); if (n_poll >= 0) { n_pkts_cpl = n_poll; @@ -2168,7 +2150,7 @@ uint16_t rte_vhost_poll_enqueue_completed(int vid, uint16_t queue_id, n_pkts_put = RTE_MIN(count, n_pkts_cpl); if (unlikely(n_pkts_put == 0)) { vq->async_last_pkts_n = n_pkts_cpl; - goto done; + return 0; } if (vq_is_packed(dev)) { @@ -2207,12 +2189,74 @@ uint16_t rte_vhost_poll_enqueue_completed(int vid, uint16_t queue_id, vq->last_async_desc_idx_split += n_descs; } -done: + return n_pkts_put; +} + +uint16_t rte_vhost_poll_enqueue_completed(int vid, uint16_t queue_id, + struct rte_mbuf **pkts, uint16_t count) +{ + struct virtio_net *dev = get_device(vid); + struct vhost_virtqueue *vq; + uint16_t n_pkts_put = 0; + + if (!dev) + return 0; + + VHOST_LOG_DATA(DEBUG, "(%d) %s\n", dev->vid, __func__); + if (unlikely(!is_valid_virt_queue_idx(queue_id, 0, dev->nr_vring))) { + VHOST_LOG_DATA(ERR, "(%d) %s: invalid virtqueue idx %d.\n", + dev->vid, __func__, queue_id); + return 0; + } + + vq = dev->virtqueue[queue_id]; + + if (unlikely(!vq->async_registered)) { + VHOST_LOG_DATA(ERR, "(%d) %s: async not registered for queue id %d.\n", + dev->vid, __func__, queue_id); + return 0; + } + + rte_spinlock_lock(&vq->access_lock); + + n_pkts_put = vhost_poll_enqueue_completed(dev, queue_id, pkts, count); + rte_spinlock_unlock(&vq->access_lock); return n_pkts_put; } +uint16_t rte_vhost_try_drain_queue_thread_unsafe(int vid, uint16_t queue_id, + struct rte_mbuf **pkts, uint16_t count, uint16_t times) +{ + struct virtio_net *dev = get_device(vid); + struct vhost_virtqueue *vq; + uint16_t n_cpl = 0; + + if (!dev) + return 0; + + VHOST_LOG_DATA(DEBUG, "(%d) %s\n", dev->vid, __func__); + if (unlikely(!is_valid_virt_queue_idx(queue_id, 0, dev->nr_vring))) { + VHOST_LOG_DATA(ERR, "(%d) %s: invalid virtqueue idx %d.\n", + dev->vid, __func__, queue_id); + return 0; + } + + vq = dev->virtqueue[queue_id]; + + if (unlikely(!vq->async_registered)) { + VHOST_LOG_DATA(ERR, "(%d) %s: async not registered for queue id %d.\n", + dev->vid, __func__, queue_id); + return 0; + } + + while ((n_cpl < count) && times--) + n_cpl += vhost_poll_enqueue_completed(dev, queue_id, pkts + n_cpl, count); + + return n_cpl; +} + static __rte_always_inline uint32_t virtio_dev_rx_async_submit(struct virtio_net *dev, uint16_t queue_id, struct rte_mbuf **pkts, uint32_t count, From patchwork Fri Jul 16 02:59:21 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Jiang, Cheng1" X-Patchwork-Id: 95931 X-Patchwork-Delegate: maxime.coquelin@redhat.com Return-Path: X-Original-To: patchwork@inbox.dpdk.org Delivered-To: patchwork@inbox.dpdk.org Received: from mails.dpdk.org (mails.dpdk.org [217.70.189.124]) by inbox.dpdk.org (Postfix) with ESMTP id D2D95A0C4A; Fri, 16 Jul 2021 05:16:16 +0200 (CEST) Received: from [217.70.189.124] (localhost [127.0.0.1]) by mails.dpdk.org (Postfix) with ESMTP id A08FA412F8; Fri, 16 Jul 2021 05:16:05 +0200 (CEST) Received: from mga09.intel.com (mga09.intel.com [134.134.136.24]) by mails.dpdk.org (Postfix) with ESMTP id 0C5BF412C9 for ; Fri, 16 Jul 2021 05:16:02 +0200 (CEST) X-IronPort-AV: E=McAfee;i="6200,9189,10046"; a="210653770" X-IronPort-AV: E=Sophos;i="5.84,244,1620716400"; d="scan'208";a="210653770" Received: from fmsmga002.fm.intel.com ([10.253.24.26]) by orsmga102.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 15 Jul 2021 20:16:02 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.84,244,1620716400"; d="scan'208";a="506308171" Received: from dpdk_jiangcheng.sh.intel.com ([10.67.119.149]) by fmsmga002.fm.intel.com with ESMTP; 15 Jul 2021 20:16:01 -0700 From: Cheng Jiang To: maxime.coquelin@redhat.com, Chenbo.Xia@intel.com Cc: dev@dpdk.org, jiayu.hu@intel.com, yvonnex.yang@intel.com Date: Fri, 16 Jul 2021 02:59:21 +0000 Message-Id: <20210716025923.27164-4-cheng1.jiang@intel.com> X-Mailer: git-send-email 2.29.2 In-Reply-To: <20210716025923.27164-1-cheng1.jiang@intel.com> References: <20210602042802.31943-1-cheng1.jiang@intel.com> <20210716025923.27164-1-cheng1.jiang@intel.com> MIME-Version: 1.0 Subject: [dpdk-dev] [PATCH v4 3/5] vhost: handle memory hotplug for async vhost X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.29 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" From: Jiayu Hu When the guest memory is hotplugged, the vhost application which enables DMA acceleration must stop DMA transfers before the vhost re-maps the guest memory. This patch is to notify the vhost application of stopping DMA transfers. Signed-off-by: Jiayu Hu --- lib/vhost/vhost_user.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/lib/vhost/vhost_user.c b/lib/vhost/vhost_user.c index 031c578e54..39e8432d1c 100644 --- a/lib/vhost/vhost_user.c +++ b/lib/vhost/vhost_user.c @@ -1275,6 +1275,15 @@ vhost_user_set_mem_table(struct virtio_net **pdev, struct VhostUserMsg *msg, vdpa_dev->ops->dev_close(dev->vid); dev->flags &= ~VIRTIO_DEV_VDPA_CONFIGURED; } + + /* notify the backend application to stop DMA transfers */ + if (dev->async_copy && dev->notify_ops->vring_state_changed) { + for (i = 0; i < dev->nr_vring; i++) { + dev->notify_ops->vring_state_changed(dev->vid, + i, 0); + } + } + free_mem_region(dev); rte_free(dev->mem); dev->mem = NULL; From patchwork Fri Jul 16 02:59:22 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Jiang, Cheng1" X-Patchwork-Id: 95932 X-Patchwork-Delegate: maxime.coquelin@redhat.com Return-Path: X-Original-To: patchwork@inbox.dpdk.org Delivered-To: patchwork@inbox.dpdk.org Received: from mails.dpdk.org (mails.dpdk.org [217.70.189.124]) by inbox.dpdk.org (Postfix) with ESMTP id 76F0CA0C4A; Fri, 16 Jul 2021 05:16:22 +0200 (CEST) Received: from [217.70.189.124] (localhost [127.0.0.1]) by mails.dpdk.org (Postfix) with ESMTP id D41214130B; Fri, 16 Jul 2021 05:16:06 +0200 (CEST) Received: from mga09.intel.com (mga09.intel.com [134.134.136.24]) by mails.dpdk.org (Postfix) with ESMTP id C1577412FF for ; Fri, 16 Jul 2021 05:16:05 +0200 (CEST) X-IronPort-AV: E=McAfee;i="6200,9189,10046"; a="210653772" X-IronPort-AV: E=Sophos;i="5.84,244,1620716400"; d="scan'208";a="210653772" Received: from fmsmga002.fm.intel.com ([10.253.24.26]) by orsmga102.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 15 Jul 2021 20:16:05 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.84,244,1620716400"; d="scan'208";a="506308187" Received: from dpdk_jiangcheng.sh.intel.com ([10.67.119.149]) by fmsmga002.fm.intel.com with ESMTP; 15 Jul 2021 20:16:03 -0700 From: Cheng Jiang To: maxime.coquelin@redhat.com, Chenbo.Xia@intel.com Cc: dev@dpdk.org, jiayu.hu@intel.com, yvonnex.yang@intel.com, Cheng Jiang Date: Fri, 16 Jul 2021 02:59:22 +0000 Message-Id: <20210716025923.27164-5-cheng1.jiang@intel.com> X-Mailer: git-send-email 2.29.2 In-Reply-To: <20210716025923.27164-1-cheng1.jiang@intel.com> References: <20210602042802.31943-1-cheng1.jiang@intel.com> <20210716025923.27164-1-cheng1.jiang@intel.com> MIME-Version: 1.0 Subject: [dpdk-dev] [PATCH v4 4/5] examples/vhost: handle memory hotplug for async vhost X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.29 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 the guest memory is hotplugged, the vhost application which enables DMA acceleration must stop DMA transfers before the vhost re-maps the guest memory. To accomplish that, we need to do these changes in the vhost sample: 1. add inflight pkt count. 2. add vring_state_changed() callback. 3. add inflight pkt drain process in destroy_device() and vring_state_changed(). Signed-off-by: Cheng Jiang --- examples/vhost/main.c | 48 +++++++++++++++++++++++++++++++++++++++++-- examples/vhost/main.h | 1 + 2 files changed, 47 insertions(+), 2 deletions(-) diff --git a/examples/vhost/main.c b/examples/vhost/main.c index d2179eadb9..9014c999be 100644 --- a/examples/vhost/main.c +++ b/examples/vhost/main.c @@ -851,8 +851,11 @@ complete_async_pkts(struct vhost_dev *vdev) complete_count = rte_vhost_poll_enqueue_completed(vdev->vid, VIRTIO_RXQ, p_cpl, MAX_PKT_BURST); - if (complete_count) + if (complete_count) { free_pkts(p_cpl, complete_count); + __atomic_sub_fetch(&vdev->pkts_inflight, complete_count, __ATOMIC_SEQ_CST); + } + } static __rte_always_inline void @@ -895,6 +898,7 @@ drain_vhost(struct vhost_dev *vdev) complete_async_pkts(vdev); ret = rte_vhost_submit_enqueue_burst(vdev->vid, VIRTIO_RXQ, m, nr_xmit, m_cpu_cpl, &cpu_cpl_nr); + __atomic_add_fetch(&vdev->pkts_inflight, ret - cpu_cpl_nr, __ATOMIC_SEQ_CST); if (cpu_cpl_nr) free_pkts(m_cpu_cpl, cpu_cpl_nr); @@ -1226,6 +1230,9 @@ drain_eth_rx(struct vhost_dev *vdev) enqueue_count = rte_vhost_submit_enqueue_burst(vdev->vid, VIRTIO_RXQ, pkts, rx_count, m_cpu_cpl, &cpu_cpl_nr); + __atomic_add_fetch(&vdev->pkts_inflight, enqueue_count - cpu_cpl_nr, + __ATOMIC_SEQ_CST); + if (cpu_cpl_nr) free_pkts(m_cpu_cpl, cpu_cpl_nr); @@ -1397,8 +1404,15 @@ destroy_device(int vid) "(%d) device has been removed from data core\n", vdev->vid); - if (async_vhost_driver) + if (async_vhost_driver) { + uint16_t n_pkt = 0; + struct rte_mbuf *m_cpl[vdev->pkts_inflight]; + n_pkt = rte_vhost_try_drain_queue_thread_unsafe(vid, VIRTIO_RXQ, m_cpl, + vdev->pkts_inflight, 2); + + free_pkts(m_cpl, n_pkt); rte_vhost_async_channel_unregister(vid, VIRTIO_RXQ); + } rte_free(vdev); } @@ -1487,6 +1501,35 @@ new_device(int vid) return 0; } +static int +vring_state_changed(int vid, uint16_t queue_id, int enable) +{ + struct vhost_dev *vdev = NULL; + + TAILQ_FOREACH(vdev, &vhost_dev_list, global_vdev_entry) { + if (vdev->vid == vid) + break; + } + if (!vdev) + return -1; + + if (queue_id != VIRTIO_RXQ) + return 0; + + if (async_vhost_driver) { + if (!enable) { + uint16_t n_pkt; + struct rte_mbuf *m_cpl[vdev->pkts_inflight]; + + n_pkt = rte_vhost_try_drain_queue_thread_unsafe(vid, queue_id, + m_cpl, vdev->pkts_inflight, 2); + free_pkts(m_cpl, n_pkt); + } + } + + return 0; +} + /* * These callback allow devices to be added to the data core when configuration * has been fully complete. @@ -1495,6 +1538,7 @@ static const struct vhost_device_ops virtio_net_device_ops = { .new_device = new_device, .destroy_device = destroy_device, + .vring_state_changed = vring_state_changed, }; /* diff --git a/examples/vhost/main.h b/examples/vhost/main.h index 0ccdce4b4a..e7b1ac60a6 100644 --- a/examples/vhost/main.h +++ b/examples/vhost/main.h @@ -51,6 +51,7 @@ struct vhost_dev { uint64_t features; size_t hdr_len; uint16_t nr_vrings; + uint16_t pkts_inflight; struct rte_vhost_memory *mem; struct device_statistics stats; TAILQ_ENTRY(vhost_dev) global_vdev_entry; From patchwork Fri Jul 16 02:59:23 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Jiang, Cheng1" X-Patchwork-Id: 95933 X-Patchwork-Delegate: maxime.coquelin@redhat.com Return-Path: X-Original-To: patchwork@inbox.dpdk.org Delivered-To: patchwork@inbox.dpdk.org Received: from mails.dpdk.org (mails.dpdk.org [217.70.189.124]) by inbox.dpdk.org (Postfix) with ESMTP id C4064A0C4A; Fri, 16 Jul 2021 05:16:30 +0200 (CEST) Received: from [217.70.189.124] (localhost [127.0.0.1]) by mails.dpdk.org (Postfix) with ESMTP id 63ACA412DD; Fri, 16 Jul 2021 05:16:10 +0200 (CEST) Received: from mga09.intel.com (mga09.intel.com [134.134.136.24]) by mails.dpdk.org (Postfix) with ESMTP id E5FB5412F6 for ; Fri, 16 Jul 2021 05:16:08 +0200 (CEST) X-IronPort-AV: E=McAfee;i="6200,9189,10046"; a="210653779" X-IronPort-AV: E=Sophos;i="5.84,244,1620716400"; d="scan'208";a="210653779" Received: from fmsmga002.fm.intel.com ([10.253.24.26]) by orsmga102.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 15 Jul 2021 20:16:08 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.84,244,1620716400"; d="scan'208";a="506308195" Received: from dpdk_jiangcheng.sh.intel.com ([10.67.119.149]) by fmsmga002.fm.intel.com with ESMTP; 15 Jul 2021 20:16:07 -0700 From: Cheng Jiang To: maxime.coquelin@redhat.com, Chenbo.Xia@intel.com Cc: dev@dpdk.org, jiayu.hu@intel.com, yvonnex.yang@intel.com, Cheng Jiang Date: Fri, 16 Jul 2021 02:59:23 +0000 Message-Id: <20210716025923.27164-6-cheng1.jiang@intel.com> X-Mailer: git-send-email 2.29.2 In-Reply-To: <20210716025923.27164-1-cheng1.jiang@intel.com> References: <20210602042802.31943-1-cheng1.jiang@intel.com> <20210716025923.27164-1-cheng1.jiang@intel.com> MIME-Version: 1.0 Subject: [dpdk-dev] [PATCH v4 5/5] doc: update doc for try drain API in vhost lib X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.29 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" update the program guide and release notes for try drain API in vhost lib. Signed-off-by: Cheng Jiang --- doc/guides/prog_guide/vhost_lib.rst | 5 +++++ doc/guides/rel_notes/release_21_08.rst | 5 +++++ 2 files changed, 10 insertions(+) diff --git a/doc/guides/prog_guide/vhost_lib.rst b/doc/guides/prog_guide/vhost_lib.rst index d18fb98910..85aabc4a75 100644 --- a/doc/guides/prog_guide/vhost_lib.rst +++ b/doc/guides/prog_guide/vhost_lib.rst @@ -281,6 +281,11 @@ The following is an overview of some key Vhost API functions: Poll enqueue completion status from async data path. Completed packets are returned to applications through ``pkts``. +* ``rte_vhost_try_drain_queue_thread_unsafe(vid, queue_id, **pkts, count, times)`` + + Try to drain in-flight packets which are submitted to DMA engine in vhost async data + path. Completed packets are returned to applications through ``pkts``. + Vhost-user Implementations -------------------------- diff --git a/doc/guides/rel_notes/release_21_08.rst b/doc/guides/rel_notes/release_21_08.rst index 6a902ef9ac..c38e358cf9 100644 --- a/doc/guides/rel_notes/release_21_08.rst +++ b/doc/guides/rel_notes/release_21_08.rst @@ -117,6 +117,11 @@ New Features The experimental PMD power management API now supports managing multiple Ethernet Rx queues per lcore. +* **Added try drain API in vhost library.** + + Added an API which can try to drain the inflight packets submitted to DMA + engine in vhost async data path. + Removed Items -------------