From patchwork Thu Apr 7 04:56:24 2016 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Jason Wang X-Patchwork-Id: 8768621 Return-Path: X-Original-To: patchwork-qemu-devel@patchwork.kernel.org Delivered-To: patchwork-parsemail@patchwork1.web.kernel.org Received: from mail.kernel.org (mail.kernel.org [198.145.29.136]) by patchwork1.web.kernel.org (Postfix) with ESMTP id EC6F99F36E for ; Thu, 7 Apr 2016 04:56:46 +0000 (UTC) Received: from mail.kernel.org (localhost [127.0.0.1]) by mail.kernel.org (Postfix) with ESMTP id F0C7F201BB for ; Thu, 7 Apr 2016 04:56:45 +0000 (UTC) Received: from lists.gnu.org (lists.gnu.org [208.118.235.17]) (using TLSv1 with cipher AES256-SHA (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 91887201B4 for ; Thu, 7 Apr 2016 04:56:44 +0000 (UTC) Received: from localhost ([::1]:47638 helo=lists.gnu.org) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1ao1zo-0002q8-2U for patchwork-qemu-devel@patchwork.kernel.org; Thu, 07 Apr 2016 00:56:44 -0400 Received: from eggs.gnu.org ([2001:4830:134:3::10]:39116) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1ao1zd-0002po-WD for qemu-devel@nongnu.org; Thu, 07 Apr 2016 00:56:35 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1ao1zY-0005Wm-Pu for qemu-devel@nongnu.org; Thu, 07 Apr 2016 00:56:33 -0400 Received: from mx1.redhat.com ([209.132.183.28]:49754) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1ao1zY-0005Wi-Hg for qemu-devel@nongnu.org; Thu, 07 Apr 2016 00:56:28 -0400 Received: from int-mx10.intmail.prod.int.phx2.redhat.com (int-mx10.intmail.prod.int.phx2.redhat.com [10.5.11.23]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 1FFA481F07; Thu, 7 Apr 2016 04:56:28 +0000 (UTC) Received: from jason-ThinkPad-T430s.redhat.com (vpn1-4-5.pek2.redhat.com [10.72.4.5]) by int-mx10.intmail.prod.int.phx2.redhat.com (8.14.4/8.14.4) with ESMTP id u374uPfB027676; Thu, 7 Apr 2016 00:56:25 -0400 From: Jason Wang To: mst@redhat.com, qemu-devel@nongnu.org Date: Thu, 7 Apr 2016 12:56:24 +0800 Message-Id: <1460004984-3784-1-git-send-email-jasowang@redhat.com> X-Scanned-By: MIMEDefang 2.68 on 10.5.11.23 X-detected-operating-system: by eggs.gnu.org: GNU/Linux 2.2.x-3.x [generic] X-Received-From: 209.132.183.28 Cc: famz@redhat.com, Jason Wang , gkurz@linux.vnet.ibm.com Subject: [Qemu-devel] [PATCH V3] tap: vhost busy polling support X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.14 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+patchwork-qemu-devel=patchwork.kernel.org@nongnu.org Sender: qemu-devel-bounces+patchwork-qemu-devel=patchwork.kernel.org@nongnu.org X-Spam-Status: No, score=-6.9 required=5.0 tests=BAYES_00, RCVD_IN_DNSWL_HI, UNPARSEABLE_RELAY autolearn=unavailable version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on mail.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP This patch add the capability of basic vhost net busy polling which is supported by recent kernel. User could configure the maximum number of us that could be spent on busy polling through a new property of tap "vhost-poll-us". Signed-off-by: Jason Wang Tested-by: Greg Kurz --- hw/net/vhost_net.c | 2 +- hw/scsi/vhost-scsi.c | 2 +- hw/virtio/vhost-backend.c | 8 ++++++++ hw/virtio/vhost.c | 40 ++++++++++++++++++++++++++++++++++++++- include/hw/virtio/vhost-backend.h | 3 +++ include/hw/virtio/vhost.h | 3 ++- include/net/vhost_net.h | 1 + net/tap.c | 10 ++++++++-- net/vhost-user.c | 1 + qapi-schema.json | 6 +++++- qemu-options.hx | 3 +++ 11 files changed, 72 insertions(+), 7 deletions(-) diff --git a/hw/net/vhost_net.c b/hw/net/vhost_net.c index 6e1032f..1840c73 100644 --- a/hw/net/vhost_net.c +++ b/hw/net/vhost_net.c @@ -166,7 +166,7 @@ struct vhost_net *vhost_net_init(VhostNetOptions *options) } r = vhost_dev_init(&net->dev, options->opaque, - options->backend_type); + options->backend_type, options->busyloop_timeout); if (r < 0) { goto fail; } diff --git a/hw/scsi/vhost-scsi.c b/hw/scsi/vhost-scsi.c index 9261d51..2a00f2f 100644 --- a/hw/scsi/vhost-scsi.c +++ b/hw/scsi/vhost-scsi.c @@ -248,7 +248,7 @@ static void vhost_scsi_realize(DeviceState *dev, Error **errp) s->dev.backend_features = 0; ret = vhost_dev_init(&s->dev, (void *)(uintptr_t)vhostfd, - VHOST_BACKEND_TYPE_KERNEL); + VHOST_BACKEND_TYPE_KERNEL, 0); if (ret < 0) { error_setg(errp, "vhost-scsi: vhost initialization failed: %s", strerror(-ret)); diff --git a/hw/virtio/vhost-backend.c b/hw/virtio/vhost-backend.c index b358902..d62372e 100644 --- a/hw/virtio/vhost-backend.c +++ b/hw/virtio/vhost-backend.c @@ -138,6 +138,12 @@ static int vhost_kernel_set_vring_call(struct vhost_dev *dev, return vhost_kernel_call(dev, VHOST_SET_VRING_CALL, file); } +static int vhost_kernel_set_vring_busyloop_timeout(struct vhost_dev *dev, + struct vhost_vring_state *s) +{ + return vhost_kernel_call(dev, VHOST_SET_VRING_BUSYLOOP_TIMEOUT, s); +} + static int vhost_kernel_set_features(struct vhost_dev *dev, uint64_t features) { @@ -185,6 +191,8 @@ static const VhostOps kernel_ops = { .vhost_get_vring_base = vhost_kernel_get_vring_base, .vhost_set_vring_kick = vhost_kernel_set_vring_kick, .vhost_set_vring_call = vhost_kernel_set_vring_call, + .vhost_set_vring_busyloop_timeout = + vhost_kernel_set_vring_busyloop_timeout, .vhost_set_features = vhost_kernel_set_features, .vhost_get_features = vhost_kernel_get_features, .vhost_set_owner = vhost_kernel_set_owner, diff --git a/hw/virtio/vhost.c b/hw/virtio/vhost.c index 4400718..ebf8b08 100644 --- a/hw/virtio/vhost.c +++ b/hw/virtio/vhost.c @@ -964,6 +964,28 @@ static void vhost_eventfd_del(MemoryListener *listener, { } +static int vhost_virtqueue_set_busyloop_timeout(struct vhost_dev *dev, + int n, uint32_t timeout) +{ + int vhost_vq_index = dev->vhost_ops->vhost_get_vq_index(dev, n); + struct vhost_vring_state state = { + .index = vhost_vq_index, + .num = timeout, + }; + int r; + + if (!dev->vhost_ops->vhost_set_vring_busyloop_timeout) { + return -EINVAL; + } + + r = dev->vhost_ops->vhost_set_vring_busyloop_timeout(dev, &state); + if (r) { + return r; + } + + return 0; +} + static int vhost_virtqueue_init(struct vhost_dev *dev, struct vhost_virtqueue *vq, int n) { @@ -994,7 +1016,7 @@ static void vhost_virtqueue_cleanup(struct vhost_virtqueue *vq) } int vhost_dev_init(struct vhost_dev *hdev, void *opaque, - VhostBackendType backend_type) + VhostBackendType backend_type, uint32_t busyloop_timeout) { uint64_t features; int i, r; @@ -1035,6 +1057,17 @@ int vhost_dev_init(struct vhost_dev *hdev, void *opaque, goto fail_vq; } } + + if (busyloop_timeout) { + for (i = 0; i < hdev->nvqs; ++i) { + r = vhost_virtqueue_set_busyloop_timeout(hdev, hdev->vq_index + i, + busyloop_timeout); + if (r < 0) { + goto fail_busyloop; + } + } + } + hdev->features = features; hdev->memory_listener = (MemoryListener) { @@ -1077,6 +1110,11 @@ int vhost_dev_init(struct vhost_dev *hdev, void *opaque, hdev->memory_changed = false; memory_listener_register(&hdev->memory_listener, &address_space_memory); return 0; +fail_busyloop: + while (--i >= 0) { + vhost_virtqueue_set_busyloop_timeout(hdev, hdev->vq_index + i, 0); + } + i = hdev->nvqs; fail_vq: while (--i >= 0) { vhost_virtqueue_cleanup(hdev->vqs + i); diff --git a/include/hw/virtio/vhost-backend.h b/include/hw/virtio/vhost-backend.h index 95fcc96..84e1cb7 100644 --- a/include/hw/virtio/vhost-backend.h +++ b/include/hw/virtio/vhost-backend.h @@ -57,6 +57,8 @@ typedef int (*vhost_set_vring_kick_op)(struct vhost_dev *dev, struct vhost_vring_file *file); typedef int (*vhost_set_vring_call_op)(struct vhost_dev *dev, struct vhost_vring_file *file); +typedef int (*vhost_set_vring_busyloop_timeout_op)(struct vhost_dev *dev, + struct vhost_vring_state *r); typedef int (*vhost_set_features_op)(struct vhost_dev *dev, uint64_t features); typedef int (*vhost_get_features_op)(struct vhost_dev *dev, @@ -91,6 +93,7 @@ typedef struct VhostOps { vhost_get_vring_base_op vhost_get_vring_base; vhost_set_vring_kick_op vhost_set_vring_kick; vhost_set_vring_call_op vhost_set_vring_call; + vhost_set_vring_busyloop_timeout_op vhost_set_vring_busyloop_timeout; vhost_set_features_op vhost_set_features; vhost_get_features_op vhost_get_features; vhost_set_owner_op vhost_set_owner; diff --git a/include/hw/virtio/vhost.h b/include/hw/virtio/vhost.h index b60d758..2106ed8 100644 --- a/include/hw/virtio/vhost.h +++ b/include/hw/virtio/vhost.h @@ -64,7 +64,8 @@ struct vhost_dev { }; int vhost_dev_init(struct vhost_dev *hdev, void *opaque, - VhostBackendType backend_type); + VhostBackendType backend_type, + uint32_t busyloop_timeout); void vhost_dev_cleanup(struct vhost_dev *hdev); int vhost_dev_start(struct vhost_dev *hdev, VirtIODevice *vdev); void vhost_dev_stop(struct vhost_dev *hdev, VirtIODevice *vdev); diff --git a/include/net/vhost_net.h b/include/net/vhost_net.h index 3389b41..8354b98 100644 --- a/include/net/vhost_net.h +++ b/include/net/vhost_net.h @@ -10,6 +10,7 @@ typedef struct vhost_net VHostNetState; typedef struct VhostNetOptions { VhostBackendType backend_type; NetClientState *net_backend; + uint32_t busyloop_timeout; void *opaque; } VhostNetOptions; diff --git a/net/tap.c b/net/tap.c index 740e8a2..7e28102 100644 --- a/net/tap.c +++ b/net/tap.c @@ -663,6 +663,11 @@ static void net_init_tap_one(const NetdevTapOptions *tap, NetClientState *peer, options.backend_type = VHOST_BACKEND_TYPE_KERNEL; options.net_backend = &s->nc; + if (tap->has_vhost_poll_us) { + options.busyloop_timeout = tap->vhost_poll_us; + } else { + options.busyloop_timeout = 0; + } if (vhostfdname) { vhostfd = monitor_fd_param(cur_mon, vhostfdname, &err); @@ -686,8 +691,9 @@ static void net_init_tap_one(const NetdevTapOptions *tap, NetClientState *peer, "vhost-net requested but could not be initialized"); return; } - } else if (vhostfdname) { - error_setg(errp, "vhostfd= is not valid without vhost"); + } else if (vhostfdname || tap->has_vhost_poll_us) { + error_setg(errp, "vhostfd(s)= or vhost_poll_us= is not valid" + " without vhost"); } } diff --git a/net/vhost-user.c b/net/vhost-user.c index 1b9e73a..b200182 100644 --- a/net/vhost-user.c +++ b/net/vhost-user.c @@ -80,6 +80,7 @@ static int vhost_user_start(int queues, NetClientState *ncs[]) options.net_backend = ncs[i]; options.opaque = s->chr; + options.busyloop_timeout = 0; s->vhost_net = vhost_net_init(&options); if (!s->vhost_net) { error_report("failed to init vhost_net for queue %d", i); diff --git a/qapi-schema.json b/qapi-schema.json index 54634c4..8afdedf 100644 --- a/qapi-schema.json +++ b/qapi-schema.json @@ -2531,6 +2531,9 @@ # # @queues: #optional number of queues to be created for multiqueue capable tap # +# @vhost-poll-us: #optional maximum number of microseconds that could +# be spent on busy polling for vhost net (since 2.7) +# # Since 1.2 ## { 'struct': 'NetdevTapOptions', @@ -2547,7 +2550,8 @@ '*vhostfd': 'str', '*vhostfds': 'str', '*vhostforce': 'bool', - '*queues': 'uint32'} } + '*queues': 'uint32', + '*vhost-poll-us': 'uint32'} } ## # @NetdevSocketOptions diff --git a/qemu-options.hx b/qemu-options.hx index 587de8f..22ddb82 100644 --- a/qemu-options.hx +++ b/qemu-options.hx @@ -1569,6 +1569,7 @@ DEF("netdev", HAS_ARG, QEMU_OPTION_netdev, "-netdev tap,id=str[,fd=h][,fds=x:y:...:z][,ifname=name][,script=file][,downscript=dfile]\n" " [,helper=helper][,sndbuf=nbytes][,vnet_hdr=on|off][,vhost=on|off]\n" " [,vhostfd=h][,vhostfds=x:y:...:z][,vhostforce=on|off][,queues=n]\n" + " [,vhost-poll-us=n]\n" " configure a host TAP network backend with ID 'str'\n" " use network scripts 'file' (default=" DEFAULT_NETWORK_SCRIPT ")\n" " to configure it and 'dfile' (default=" DEFAULT_NETWORK_DOWN_SCRIPT ")\n" @@ -1588,6 +1589,8 @@ DEF("netdev", HAS_ARG, QEMU_OPTION_netdev, " use 'vhostfd=h' to connect to an already opened vhost net device\n" " use 'vhostfds=x:y:...:z to connect to multiple already opened vhost net devices\n" " use 'queues=n' to specify the number of queues to be created for multiqueue TAP\n" + " use 'vhost-poll-us=n' to speciy the maximum number of microseconds that could be\n" + " spent on busy polling for vhost net\n" "-netdev bridge,id=str[,br=bridge][,helper=helper]\n" " configure a host TAP network backend with ID 'str' that is\n" " connected to a bridge (default=" DEFAULT_BRIDGE_INTERFACE ")\n"