From patchwork Thu Jun 1 20:08:11 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Maxime Coquelin X-Patchwork-Id: 127933 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 970E542C08; Thu, 1 Jun 2023 22:10:58 +0200 (CEST) Received: from mails.dpdk.org (localhost [127.0.0.1]) by mails.dpdk.org (Postfix) with ESMTP id A74F043018; Thu, 1 Jun 2023 22:09:52 +0200 (CEST) Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.133.124]) by mails.dpdk.org (Postfix) with ESMTP id 4535942FFE for ; Thu, 1 Jun 2023 22:09:50 +0200 (CEST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1685650189; 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=GB/AU+Vbvek3rhMkceixizYLiCXsskLFM3Hzutrpcc4=; b=JRXL6Fnxe1G2u/3cqkFoqyNC6aMmpg16olKpVQDMFwTy2QbUb3entvOwmzgq501pSNbHPs eiCv1vmxCUm4IXAwANzs9R14dYjB6OAZPaZKUUrT+LxbKgFC8sOM96BUjHvhOCGx2VL8tY eiG8RLHkngZZ3dyNcku7y7NU+mEJbPU= Received: from mimecast-mx02.redhat.com (mx3-rdu2.redhat.com [66.187.233.73]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id us-mta-658--Uu0AXkeNqqkiHJWB77tUw-1; Thu, 01 Jun 2023 16:09:45 -0400 X-MC-Unique: -Uu0AXkeNqqkiHJWB77tUw-1 Received: from smtp.corp.redhat.com (int-mx04.intmail.prod.int.rdu2.redhat.com [10.11.54.4]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx02.redhat.com (Postfix) with ESMTPS id 19B2F380673A; Thu, 1 Jun 2023 20:09:43 +0000 (UTC) Received: from max-t490s.redhat.com (unknown [10.39.208.25]) by smtp.corp.redhat.com (Postfix) with ESMTP id 677502029F6D; Thu, 1 Jun 2023 20:09:40 +0000 (UTC) From: Maxime Coquelin To: dev@dpdk.org, chenbo.xia@intel.com, david.marchand@redhat.com, mkp@redhat.com, fbl@redhat.com, jasowang@redhat.com, cunming.liang@intel.com, xieyongji@bytedance.com, echaudro@redhat.com, eperezma@redhat.com, amorenoz@redhat.com, lulu@redhat.com Cc: Maxime Coquelin Subject: [PATCH v4 25/26] vhost: add multiqueue support to VDUSE Date: Thu, 1 Jun 2023 22:08:11 +0200 Message-Id: <20230601200812.672233-26-maxime.coquelin@redhat.com> In-Reply-To: <20230601200812.672233-1-maxime.coquelin@redhat.com> References: <20230601200812.672233-1-maxime.coquelin@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 3.1 on 10.11.54.4 X-Mimecast-Spam-Score: 0 X-Mimecast-Originator: redhat.com 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 This patch enables control queue support in order to support multiqueue. Signed-off-by: Maxime Coquelin Reviewed-by: Chenbo Xia --- lib/vhost/vduse.c | 83 +++++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 76 insertions(+), 7 deletions(-) diff --git a/lib/vhost/vduse.c b/lib/vhost/vduse.c index a10dc24d38..699cfed9e3 100644 --- a/lib/vhost/vduse.c +++ b/lib/vhost/vduse.c @@ -21,6 +21,7 @@ #include "iotlb.h" #include "vduse.h" #include "vhost.h" +#include "virtio_net_ctrl.h" #define VHOST_VDUSE_API_VERSION 0 #define VDUSE_CTRL_PATH "/dev/vduse/control" @@ -41,7 +42,9 @@ (1ULL << VIRTIO_NET_F_GUEST_ECN) | \ (1ULL << VIRTIO_RING_F_INDIRECT_DESC) | \ (1ULL << VIRTIO_F_IN_ORDER) | \ - (1ULL << VIRTIO_F_IOMMU_PLATFORM)) + (1ULL << VIRTIO_F_IOMMU_PLATFORM) | \ + (1ULL << VIRTIO_NET_F_CTRL_VQ) | \ + (1ULL << VIRTIO_NET_F_MQ)) struct vduse { struct fdset fdset; @@ -141,6 +144,25 @@ static struct vhost_backend_ops vduse_backend_ops = { .inject_irq = vduse_inject_irq, }; +static void +vduse_control_queue_event(int fd, void *arg, int *remove __rte_unused) +{ + struct virtio_net *dev = arg; + uint64_t buf; + int ret; + + ret = read(fd, &buf, sizeof(buf)); + if (ret < 0) { + VHOST_LOG_CONFIG(dev->ifname, ERR, "Failed to read control queue event: %s\n", + strerror(errno)); + return; + } + + VHOST_LOG_CONFIG(dev->ifname, DEBUG, "Control queue kicked\n"); + if (virtio_net_ctrl_handle(dev)) + VHOST_LOG_CONFIG(dev->ifname, ERR, "Failed to handle ctrl request\n"); +} + static void vduse_vring_setup(struct virtio_net *dev, unsigned int index) { @@ -212,6 +234,22 @@ vduse_vring_setup(struct virtio_net *dev, unsigned int index) vq->kickfd = VIRTIO_UNINITIALIZED_EVENTFD; return; } + + if (vq == dev->cvq) { + ret = fdset_add(&vduse.fdset, vq->kickfd, vduse_control_queue_event, NULL, dev); + if (ret) { + VHOST_LOG_CONFIG(dev->ifname, ERR, + "Failed to setup kickfd handler for VQ %u: %s\n", + index, strerror(errno)); + vq_efd.fd = VDUSE_EVENTFD_DEASSIGN; + ioctl(dev->vduse_dev_fd, VDUSE_VQ_SETUP_KICKFD, &vq_efd); + close(vq->kickfd); + vq->kickfd = VIRTIO_UNINITIALIZED_EVENTFD; + } + fdset_pipe_notify(&vduse.fdset); + vhost_enable_guest_notification(dev, vq, 1); + VHOST_LOG_CONFIG(dev->ifname, INFO, "Ctrl queue event handler installed\n"); + } } static void @@ -258,6 +296,9 @@ vduse_device_start(struct virtio_net *dev) for (i = 0; i < dev->nr_vring; i++) { struct vhost_virtqueue *vq = dev->virtqueue[i]; + if (vq == dev->cvq) + continue; + if (dev->notify_ops->vring_state_changed) dev->notify_ops->vring_state_changed(dev->vid, i, vq->enabled); } @@ -334,9 +375,11 @@ vduse_device_create(const char *path) { int control_fd, dev_fd, vid, ret; pthread_t fdset_tid; - uint32_t i; + uint32_t i, max_queue_pairs, total_queues; struct virtio_net *dev; + struct virtio_net_config vnet_config = { 0 }; uint64_t ver = VHOST_VDUSE_API_VERSION; + uint64_t features = VDUSE_NET_SUPPORTED_FEATURES; struct vduse_dev_config *dev_config = NULL; const char *name = path + strlen("/dev/vduse/"); @@ -376,22 +419,39 @@ vduse_device_create(const char *path) goto out_ctrl_close; } - dev_config = malloc(offsetof(struct vduse_dev_config, config)); + dev_config = malloc(offsetof(struct vduse_dev_config, config) + + sizeof(vnet_config)); if (!dev_config) { VHOST_LOG_CONFIG(name, ERR, "Failed to allocate VDUSE config\n"); ret = -1; goto out_ctrl_close; } + ret = rte_vhost_driver_get_queue_num(path, &max_queue_pairs); + if (ret < 0) { + VHOST_LOG_CONFIG(name, ERR, "Failed to get max queue pairs\n"); + goto out_free; + } + + VHOST_LOG_CONFIG(path, INFO, "VDUSE max queue pairs: %u\n", max_queue_pairs); + total_queues = max_queue_pairs * 2; + + if (max_queue_pairs == 1) + features &= ~(RTE_BIT64(VIRTIO_NET_F_CTRL_VQ) | RTE_BIT64(VIRTIO_NET_F_MQ)); + else + total_queues += 1; /* Includes ctrl queue */ + + vnet_config.max_virtqueue_pairs = max_queue_pairs; memset(dev_config, 0, sizeof(struct vduse_dev_config)); strncpy(dev_config->name, name, VDUSE_NAME_MAX - 1); dev_config->device_id = VIRTIO_ID_NET; dev_config->vendor_id = 0; - dev_config->features = VDUSE_NET_SUPPORTED_FEATURES; - dev_config->vq_num = 2; + dev_config->features = features; + dev_config->vq_num = total_queues; dev_config->vq_align = sysconf(_SC_PAGE_SIZE); - dev_config->config_size = 0; + dev_config->config_size = sizeof(struct virtio_net_config); + memcpy(dev_config->config, &vnet_config, sizeof(vnet_config)); ret = ioctl(control_fd, VDUSE_CREATE_DEV, dev_config); if (ret < 0) { @@ -433,7 +493,7 @@ vduse_device_create(const char *path) dev->vduse_dev_fd = dev_fd; vhost_setup_virtio_net(dev->vid, true, true, true, true); - for (i = 0; i < 2; i++) { + for (i = 0; i < total_queues; i++) { struct vduse_vq_config vq_cfg = { 0 }; ret = alloc_vring_queue(dev, i); @@ -452,6 +512,8 @@ vduse_device_create(const char *path) } } + dev->cvq = dev->virtqueue[max_queue_pairs * 2]; + ret = fdset_add(&vduse.fdset, dev->vduse_dev_fd, vduse_events_handler, NULL, dev); if (ret) { VHOST_LOG_CONFIG(name, ERR, "Failed to add fd %d to vduse fdset\n", @@ -498,6 +560,13 @@ vduse_device_destroy(const char *path) if (vid == RTE_MAX_VHOST_DEVICE) return -1; + if (dev->cvq && dev->cvq->kickfd >= 0) { + fdset_del(&vduse.fdset, dev->cvq->kickfd); + fdset_pipe_notify(&vduse.fdset); + close(dev->cvq->kickfd); + dev->cvq->kickfd = VIRTIO_UNINITIALIZED_EVENTFD; + } + fdset_del(&vduse.fdset, dev->vduse_dev_fd); fdset_pipe_notify(&vduse.fdset);