From patchwork Wed Nov 25 15:21:19 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Maxime Coquelin X-Patchwork-Id: 84548 X-Patchwork-Delegate: maxime.coquelin@redhat.com Return-Path: X-Original-To: patchwork@inbox.dpdk.org Delivered-To: patchwork@inbox.dpdk.org Received: from dpdk.org (dpdk.org [92.243.14.124]) by inbox.dpdk.org (Postfix) with ESMTP id 1BF07A0545; Wed, 25 Nov 2020 16:21:48 +0100 (CET) Received: from [92.243.14.124] (localhost [127.0.0.1]) by dpdk.org (Postfix) with ESMTP id A288CC954; Wed, 25 Nov 2020 16:21:45 +0100 (CET) Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [63.128.21.124]) by dpdk.org (Postfix) with ESMTP id EE1EFC940 for ; Wed, 25 Nov 2020 16:21:43 +0100 (CET) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1606317702; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=Dq+n6gPcQRs4AZBZ2mmEInp7J0Jj5J4Hrxw48/t/EOo=; b=dxVTM16YRi/fxZfwa7v/2dk7xd7MpLBhoacjVnC2/NfxwLEXqv8E7J/dYYTj9/tZBL2PKJ 7aevbSURf/8bp9xD9EPsHHVUjWQIYGH76QnFiQyCOT/MgSJz02DhmSyDkXr7VdI2Kh12Gx SS3DH3DwgBT47zG7Ukyx3fAoLI4yIpE= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-328-2LWxW4HTPEK_Kqpibtvo2w-1; Wed, 25 Nov 2020 10:21:39 -0500 X-MC-Unique: 2LWxW4HTPEK_Kqpibtvo2w-1 Received: from smtp.corp.redhat.com (int-mx05.intmail.prod.int.phx2.redhat.com [10.5.11.15]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id B9A5C19251D4; Wed, 25 Nov 2020 15:21:26 +0000 (UTC) Received: from localhost.localdomain (unknown [10.36.110.64]) by smtp.corp.redhat.com (Postfix) with ESMTP id F141F5D6AC; Wed, 25 Nov 2020 15:21:24 +0000 (UTC) From: Maxime Coquelin To: dev@dpdk.org, chenbo.xia@intel.com, amorenoz@redhat.com, jasowang@redhat.com, david.marchand@redhat.com Cc: Maxime Coquelin , stable@dpdk.org Date: Wed, 25 Nov 2020 16:21:19 +0100 Message-Id: <20201125152120.183691-2-maxime.coquelin@redhat.com> In-Reply-To: <20201125152120.183691-1-maxime.coquelin@redhat.com> References: <20201125152120.183691-1-maxime.coquelin@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.15 Authentication-Results: relay.mimecast.com; auth=pass smtp.auth=CUSA124A263 smtp.mailfrom=maxime.coquelin@redhat.com X-Mimecast-Spam-Score: 0 X-Mimecast-Originator: redhat.com Subject: [dpdk-dev] [PATCH 21.02 v2 1/2] net/virtio: fix missing backend features negotiation 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" This patch adds missing backend features negotiation for in Vhost-vDPA. Without it, IOTLB messages v2 could be sent by Virtio-user PMD while not supported by the backend. Fixes: 6b901437056e ("net/virtio: introduce vhost-vDPA backend") Cc: stable@dpdk.org Signed-off-by: Maxime Coquelin --- drivers/net/virtio/virtio_user/vhost.h | 4 ++++ drivers/net/virtio/virtio_user/vhost_vdpa.c | 14 ++++++++++++++ drivers/net/virtio/virtio_user/virtio_user_dev.c | 14 ++++++++++---- 3 files changed, 28 insertions(+), 4 deletions(-) diff --git a/drivers/net/virtio/virtio_user/vhost.h b/drivers/net/virtio/virtio_user/vhost.h index 210a3704e7..c1dcc50b58 100644 --- a/drivers/net/virtio/virtio_user/vhost.h +++ b/drivers/net/virtio/virtio_user/vhost.h @@ -86,6 +86,10 @@ enum vhost_user_request { VHOST_USER_MAX }; +#ifndef VHOST_BACKEND_F_IOTLB_MSG_V2 +#define VHOST_BACKEND_F_IOTLB_MSG_V2 1 +#endif + extern const char * const vhost_msg_strings[VHOST_USER_MAX]; struct vhost_memory_region { diff --git a/drivers/net/virtio/virtio_user/vhost_vdpa.c b/drivers/net/virtio/virtio_user/vhost_vdpa.c index c7b9349fc8..b6c81d6f17 100644 --- a/drivers/net/virtio/virtio_user/vhost_vdpa.c +++ b/drivers/net/virtio/virtio_user/vhost_vdpa.c @@ -35,6 +35,8 @@ #define VHOST_VDPA_SET_STATUS _IOW(VHOST_VIRTIO, 0x72, __u8) #define VHOST_VDPA_SET_VRING_ENABLE _IOW(VHOST_VIRTIO, 0x75, \ struct vhost_vring_state) +#define VHOST_SET_BACKEND_FEATURES _IOW(VHOST_VIRTIO, 0x25, __u64) +#define VHOST_GET_BACKEND_FEATURES _IOR(VHOST_VIRTIO, 0x26, __u64) static uint64_t vhost_req_user_to_vdpa[] = { [VHOST_USER_SET_OWNER] = VHOST_SET_OWNER, @@ -51,6 +53,8 @@ static uint64_t vhost_req_user_to_vdpa[] = { [VHOST_USER_SET_STATUS] = VHOST_VDPA_SET_STATUS, [VHOST_USER_GET_STATUS] = VHOST_VDPA_GET_STATUS, [VHOST_USER_SET_VRING_ENABLE] = VHOST_VDPA_SET_VRING_ENABLE, + [VHOST_USER_GET_PROTOCOL_FEATURES] = VHOST_GET_BACKEND_FEATURES, + [VHOST_USER_SET_PROTOCOL_FEATURES] = VHOST_SET_BACKEND_FEATURES, }; /* no alignment requirement */ @@ -86,6 +90,11 @@ vhost_vdpa_dma_map(struct virtio_user_dev *dev, void *addr, { struct vhost_msg msg = {}; + if (!(dev->protocol_features & (1ULL << VHOST_BACKEND_F_IOTLB_MSG_V2))) { + PMD_DRV_LOG(ERR, "IOTLB_MSG_V2 not supported by the backend."); + return -1; + } + msg.type = VHOST_IOTLB_MSG_V2; msg.iotlb.type = VHOST_IOTLB_UPDATE; msg.iotlb.iova = iova; @@ -108,6 +117,11 @@ vhost_vdpa_dma_unmap(struct virtio_user_dev *dev, __rte_unused void *addr, { struct vhost_msg msg = {}; + if (!(dev->protocol_features & (1ULL << VHOST_BACKEND_F_IOTLB_MSG_V2))) { + PMD_DRV_LOG(ERR, "IOTLB_MSG_V2 not supported by the backend."); + return -1; + } + msg.type = VHOST_IOTLB_MSG_V2; msg.iotlb.type = VHOST_IOTLB_INVALIDATE; msg.iotlb.iova = iova; diff --git a/drivers/net/virtio/virtio_user/virtio_user_dev.c b/drivers/net/virtio/virtio_user/virtio_user_dev.c index 053f0267ca..96bc6b232d 100644 --- a/drivers/net/virtio/virtio_user/virtio_user_dev.c +++ b/drivers/net/virtio/virtio_user/virtio_user_dev.c @@ -439,11 +439,13 @@ virtio_user_dev_setup(struct virtio_user_dev *dev) 1ULL << VIRTIO_F_RING_PACKED | \ 1ULL << VHOST_USER_F_PROTOCOL_FEATURES) -#define VIRTIO_USER_SUPPORTED_PROTOCOL_FEATURES \ +#define VHOST_USER_SUPPORTED_PROTOCOL_FEATURES \ (1ULL << VHOST_USER_PROTOCOL_F_MQ | \ 1ULL << VHOST_USER_PROTOCOL_F_REPLY_ACK | \ 1ULL << VHOST_USER_PROTOCOL_F_STATUS) +#define VHOST_VDPA_SUPPORTED_PROTOCOL_FEATURES \ + (1ULL << VHOST_BACKEND_F_IOTLB_MSG_V2) int virtio_user_dev_init(struct virtio_user_dev *dev, char *path, int queues, int cq, int queue_size, const char *mac, char **ifname, @@ -462,9 +464,13 @@ virtio_user_dev_init(struct virtio_user_dev *dev, char *path, int queues, dev->mac_specified = 0; dev->frontend_features = 0; dev->unsupported_features = ~VIRTIO_USER_SUPPORTED_FEATURES; - dev->protocol_features = VIRTIO_USER_SUPPORTED_PROTOCOL_FEATURES; dev->backend_type = backend_type; + if (dev->backend_type == VIRTIO_USER_BACKEND_VHOST_USER) + dev->protocol_features = VHOST_USER_SUPPORTED_PROTOCOL_FEATURES; + else if (dev->backend_type == VIRTIO_USER_BACKEND_VHOST_VDPA) + dev->protocol_features = VHOST_VDPA_SUPPORTED_PROTOCOL_FEATURES; + parse_mac(dev, mac); if (*ifname) { @@ -497,8 +503,8 @@ virtio_user_dev_init(struct virtio_user_dev *dev, char *path, int queues, } - if (dev->device_features & - (1ULL << VHOST_USER_F_PROTOCOL_FEATURES)) { + if ((dev->device_features & (1ULL << VHOST_USER_F_PROTOCOL_FEATURES) || + dev->backend_type == VIRTIO_USER_BACKEND_VHOST_VDPA)) { if (dev->ops->send_request(dev, VHOST_USER_GET_PROTOCOL_FEATURES, &protocol_features)) From patchwork Wed Nov 25 15:21:20 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Maxime Coquelin X-Patchwork-Id: 84550 X-Patchwork-Delegate: maxime.coquelin@redhat.com Return-Path: X-Original-To: patchwork@inbox.dpdk.org Delivered-To: patchwork@inbox.dpdk.org Received: from dpdk.org (dpdk.org [92.243.14.124]) by inbox.dpdk.org (Postfix) with ESMTP id 6AD12A0545; Wed, 25 Nov 2020 16:22:25 +0100 (CET) Received: from [92.243.14.124] (localhost [127.0.0.1]) by dpdk.org (Postfix) with ESMTP id 1AD06C96C; Wed, 25 Nov 2020 16:21:50 +0100 (CET) Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [63.128.21.124]) by dpdk.org (Postfix) with ESMTP id B6827C95E for ; Wed, 25 Nov 2020 16:21:45 +0100 (CET) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1606317704; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=v28CF6ey3j8Yr3WUhRjbp1n4bMV9YEbhPWEQ0hUvD3s=; b=JLTKUoTi14hyPlh0ompDXAwa0dZjUke10sKtgRwwWZo61AK7ThXTzJvWNp/KVmFCFPczPV BGkqv7CwZAY2kYeepM7Wb89IsRcuH79HB/5rCi259QsBjKDyABokyXR7mtKR/V5STrB1/J q6HmAXJwanLnhb3VVercObnhOUCVLL4= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-318-JQLOXnS6O8-H6RsCPlLSKw-1; Wed, 25 Nov 2020 10:21:41 -0500 X-MC-Unique: JQLOXnS6O8-H6RsCPlLSKw-1 Received: from smtp.corp.redhat.com (int-mx05.intmail.prod.int.phx2.redhat.com [10.5.11.15]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id DD66DB810B; Wed, 25 Nov 2020 15:21:28 +0000 (UTC) Received: from localhost.localdomain (unknown [10.36.110.64]) by smtp.corp.redhat.com (Postfix) with ESMTP id 3B32D5D6AC; Wed, 25 Nov 2020 15:21:26 +0000 (UTC) From: Maxime Coquelin To: dev@dpdk.org, chenbo.xia@intel.com, amorenoz@redhat.com, jasowang@redhat.com, david.marchand@redhat.com Cc: Maxime Coquelin , stable@dpdk.org Date: Wed, 25 Nov 2020 16:21:20 +0100 Message-Id: <20201125152120.183691-3-maxime.coquelin@redhat.com> In-Reply-To: <20201125152120.183691-1-maxime.coquelin@redhat.com> References: <20201125152120.183691-1-maxime.coquelin@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.15 Authentication-Results: relay.mimecast.com; auth=pass smtp.auth=CUSA124A263 smtp.mailfrom=maxime.coquelin@redhat.com X-Mimecast-Spam-Score: 0 X-Mimecast-Originator: redhat.com Subject: [dpdk-dev] [PATCH 21.02 v2 2/2] net/virtio: fix memory init with vDPA backend 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" This patch fixes an overhead met with mlx5-vdpa Kernel driver, where for every page in the mapped area, all the memory tables gets updated. For example, with 2MB hugepages, a single IOTLB_UPDATE for a 1GB region causes 512 memory updates on mlx5-vdpa side. Using batching mode, the mlx5 driver will only trigger a single memory update for all the IOTLB updates that happen between the batch begin and batch end commands. Fixes: 6b901437056e ("net/virtio: introduce vhost-vDPA backend") Cc: stable@dpdk.org Signed-off-by: Maxime Coquelin Reviewed-by: Chenbo Xia --- drivers/net/virtio/virtio_user/vhost.h | 4 + drivers/net/virtio/virtio_user/vhost_vdpa.c | 106 +++++++++++++++++- .../net/virtio/virtio_user/virtio_user_dev.c | 3 +- 3 files changed, 107 insertions(+), 6 deletions(-) diff --git a/drivers/net/virtio/virtio_user/vhost.h b/drivers/net/virtio/virtio_user/vhost.h index c1dcc50b58..be286173b0 100644 --- a/drivers/net/virtio/virtio_user/vhost.h +++ b/drivers/net/virtio/virtio_user/vhost.h @@ -90,6 +90,10 @@ enum vhost_user_request { #define VHOST_BACKEND_F_IOTLB_MSG_V2 1 #endif +#ifndef VHOST_BACKEND_F_IOTLB_BATCH +#define VHOST_BACKEND_F_IOTLB_BATCH 2 +#endif + extern const char * const vhost_msg_strings[VHOST_USER_MAX]; struct vhost_memory_region { diff --git a/drivers/net/virtio/virtio_user/vhost_vdpa.c b/drivers/net/virtio/virtio_user/vhost_vdpa.c index b6c81d6f17..269bab2f8e 100644 --- a/drivers/net/virtio/virtio_user/vhost_vdpa.c +++ b/drivers/net/virtio/virtio_user/vhost_vdpa.c @@ -70,6 +70,8 @@ struct vhost_iotlb_msg { #define VHOST_IOTLB_UPDATE 2 #define VHOST_IOTLB_INVALIDATE 3 #define VHOST_IOTLB_ACCESS_FAIL 4 +#define VHOST_IOTLB_BATCH_BEGIN 5 +#define VHOST_IOTLB_BATCH_END 6 uint8_t type; }; @@ -84,6 +86,56 @@ struct vhost_msg { }; }; +static int +vhost_vdpa_iotlb_batch_begin(struct virtio_user_dev *dev) +{ + struct vhost_msg msg = {}; + + if (!(dev->protocol_features & (1ULL << VHOST_BACKEND_F_IOTLB_BATCH))) + return 0; + + if (!(dev->protocol_features & (1ULL << VHOST_BACKEND_F_IOTLB_MSG_V2))) { + PMD_DRV_LOG(ERR, "IOTLB_MSG_V2 not supported by the backend."); + return -1; + } + + msg.type = VHOST_IOTLB_MSG_V2; + msg.iotlb.type = VHOST_IOTLB_BATCH_BEGIN; + + if (write(dev->vhostfd, &msg, sizeof(msg)) != sizeof(msg)) { + PMD_DRV_LOG(ERR, "Failed to send IOTLB batch begin (%s)", + strerror(errno)); + return -1; + } + + return 0; +} + +static int +vhost_vdpa_iotlb_batch_end(struct virtio_user_dev *dev) +{ + struct vhost_msg msg = {}; + + if (!(dev->protocol_features & (1ULL << VHOST_BACKEND_F_IOTLB_BATCH))) + return 0; + + if (!(dev->protocol_features & (1ULL << VHOST_BACKEND_F_IOTLB_MSG_V2))) { + PMD_DRV_LOG(ERR, "IOTLB_MSG_V2 not supported by the backend."); + return -1; + } + + msg.type = VHOST_IOTLB_MSG_V2; + msg.iotlb.type = VHOST_IOTLB_BATCH_END; + + if (write(dev->vhostfd, &msg, sizeof(msg)) != sizeof(msg)) { + PMD_DRV_LOG(ERR, "Failed to send IOTLB batch end (%s)", + strerror(errno)); + return -1; + } + + return 0; +} + static int vhost_vdpa_dma_map(struct virtio_user_dev *dev, void *addr, uint64_t iova, size_t len) @@ -136,6 +188,39 @@ vhost_vdpa_dma_unmap(struct virtio_user_dev *dev, __rte_unused void *addr, return 0; } +static int +vhost_vdpa_dma_map_batch(struct virtio_user_dev *dev, void *addr, + uint64_t iova, size_t len) +{ + int ret; + + if (vhost_vdpa_iotlb_batch_begin(dev) < 0) + return -1; + + ret = vhost_vdpa_dma_map(dev, addr, iova, len); + + if (vhost_vdpa_iotlb_batch_end(dev) < 0) + return -1; + + return ret; +} + +static int +vhost_vdpa_dma_unmap_batch(struct virtio_user_dev *dev, void *addr, + uint64_t iova, size_t len) +{ + int ret; + + if (vhost_vdpa_iotlb_batch_begin(dev) < 0) + return -1; + + ret = vhost_vdpa_dma_unmap(dev, addr, iova, len); + + if (vhost_vdpa_iotlb_batch_end(dev) < 0) + return -1; + + return ret; +} static int vhost_vdpa_map_contig(const struct rte_memseg_list *msl, @@ -173,21 +258,32 @@ vhost_vdpa_map(const struct rte_memseg_list *msl, const struct rte_memseg *ms, static int vhost_vdpa_dma_map_all(struct virtio_user_dev *dev) { + int ret; + + if (vhost_vdpa_iotlb_batch_begin(dev) < 0) + return -1; + vhost_vdpa_dma_unmap(dev, NULL, 0, SIZE_MAX); if (rte_eal_iova_mode() == RTE_IOVA_VA) { /* with IOVA as VA mode, we can get away with mapping contiguous * chunks rather than going page-by-page. */ - int ret = rte_memseg_contig_walk_thread_unsafe( + ret = rte_memseg_contig_walk_thread_unsafe( vhost_vdpa_map_contig, dev); if (ret) - return ret; + goto batch_end; /* we have to continue the walk because we've skipped the * external segments during the config walk. */ } - return rte_memseg_walk_thread_unsafe(vhost_vdpa_map, dev); + ret = rte_memseg_walk_thread_unsafe(vhost_vdpa_map, dev); + +batch_end: + if (vhost_vdpa_iotlb_batch_end(dev) < 0) + return -1; + + return ret; } /* with below features, vhost vdpa does not need to do the checksum and TSO, @@ -307,6 +403,6 @@ struct virtio_user_backend_ops virtio_ops_vdpa = { .setup = vhost_vdpa_setup, .send_request = vhost_vdpa_ioctl, .enable_qp = vhost_vdpa_enable_queue_pair, - .dma_map = vhost_vdpa_dma_map, - .dma_unmap = vhost_vdpa_dma_unmap, + .dma_map = vhost_vdpa_dma_map_batch, + .dma_unmap = vhost_vdpa_dma_unmap_batch, }; diff --git a/drivers/net/virtio/virtio_user/virtio_user_dev.c b/drivers/net/virtio/virtio_user/virtio_user_dev.c index 96bc6b232d..095deeea82 100644 --- a/drivers/net/virtio/virtio_user/virtio_user_dev.c +++ b/drivers/net/virtio/virtio_user/virtio_user_dev.c @@ -445,7 +445,8 @@ virtio_user_dev_setup(struct virtio_user_dev *dev) 1ULL << VHOST_USER_PROTOCOL_F_STATUS) #define VHOST_VDPA_SUPPORTED_PROTOCOL_FEATURES \ - (1ULL << VHOST_BACKEND_F_IOTLB_MSG_V2) + (1ULL << VHOST_BACKEND_F_IOTLB_MSG_V2 | \ + 1ULL << VHOST_BACKEND_F_IOTLB_BATCH) int virtio_user_dev_init(struct virtio_user_dev *dev, char *path, int queues, int cq, int queue_size, const char *mac, char **ifname,