[mirror_qemu.git] / hw / net / vhost_net.c

/*
 * vhost-net support
 *
 * Copyright Red Hat, Inc. 2010
 *
 * Authors:
 *  Michael S. Tsirkin <mst@redhat.com>
 *
 * This work is licensed under the terms of the GNU GPL, version 2.  See
 * the COPYING file in the top-level directory.
 *
 * Contributions after 2012-01-13 are licensed under the terms of the
 * GNU GPL, version 2 or (at your option) any later version.
 */

#include "qemu/osdep.h"
#include "net/net.h"
#include "net/tap.h"
#include "net/vhost-user.h"
#include "net/vhost-vdpa.h"

#include "standard-headers/linux/vhost_types.h"
#include "hw/virtio/virtio-net.h"
#include "net/vhost_net.h"
#include "qemu/error-report.h"
#include "qemu/main-loop.h"

#include <sys/socket.h>
#include <net/if.h>
#include <netinet/in.h>


#include "standard-headers/linux/virtio_ring.h"
#include "hw/virtio/vhost.h"
#include "hw/virtio/virtio-bus.h"


/* Features supported by host kernel. */
static const int kernel_feature_bits[] = {
    VIRTIO_F_NOTIFY_ON_EMPTY,
    VIRTIO_RING_F_INDIRECT_DESC,
    VIRTIO_RING_F_EVENT_IDX,
    VIRTIO_NET_F_MRG_RXBUF,
    VIRTIO_F_VERSION_1,
    VIRTIO_NET_F_MTU,
    VIRTIO_F_IOMMU_PLATFORM,
    VIRTIO_F_RING_PACKED,
    VHOST_INVALID_FEATURE_BIT
};

/* Features supported by others. */
static const int user_feature_bits[] = {
    VIRTIO_F_NOTIFY_ON_EMPTY,
    VIRTIO_RING_F_INDIRECT_DESC,
    VIRTIO_RING_F_EVENT_IDX,

    VIRTIO_F_ANY_LAYOUT,
    VIRTIO_F_VERSION_1,
    VIRTIO_NET_F_CSUM,
    VIRTIO_NET_F_GUEST_CSUM,
    VIRTIO_NET_F_GSO,
    VIRTIO_NET_F_GUEST_TSO4,
    VIRTIO_NET_F_GUEST_TSO6,
    VIRTIO_NET_F_GUEST_ECN,
    VIRTIO_NET_F_GUEST_UFO,
    VIRTIO_NET_F_HOST_TSO4,
    VIRTIO_NET_F_HOST_TSO6,
    VIRTIO_NET_F_HOST_ECN,
    VIRTIO_NET_F_HOST_UFO,
    VIRTIO_NET_F_MRG_RXBUF,
    VIRTIO_NET_F_MTU,
    VIRTIO_F_IOMMU_PLATFORM,
    VIRTIO_F_RING_PACKED,

    /* This bit implies RARP isn't sent by QEMU out of band */
    VIRTIO_NET_F_GUEST_ANNOUNCE,

    VIRTIO_NET_F_MQ,

    VHOST_INVALID_FEATURE_BIT
};

static const int *vhost_net_get_feature_bits(struct vhost_net *net)
{
    const int *feature_bits = 0;

    switch (net->nc->info->type) {
    case NET_CLIENT_DRIVER_TAP:
        feature_bits = kernel_feature_bits;
        break;
    case NET_CLIENT_DRIVER_VHOST_USER:
        feature_bits = user_feature_bits;
        break;
#ifdef CONFIG_VHOST_NET_VDPA
    case NET_CLIENT_DRIVER_VHOST_VDPA:
        feature_bits = vdpa_feature_bits;
        break;
#endif
    default:
        error_report("Feature bits not defined for this type: %d",
                net->nc->info->type);
        break;
    }

    return feature_bits;
}

uint64_t vhost_net_get_features(struct vhost_net *net, uint64_t features)
{
    return vhost_get_features(&net->dev, vhost_net_get_feature_bits(net),
            features);
}
int vhost_net_get_config(struct vhost_net *net,  uint8_t *config,
                         uint32_t config_len)
{
    return vhost_dev_get_config(&net->dev, config, config_len);
}
int vhost_net_set_config(struct vhost_net *net, const uint8_t *data,
                         uint32_t offset, uint32_t size, uint32_t flags)
{
    return vhost_dev_set_config(&net->dev, data, offset, size, flags);
}

void vhost_net_ack_features(struct vhost_net *net, uint64_t features)
{
    net->dev.acked_features = net->dev.backend_features;
    vhost_ack_features(&net->dev, vhost_net_get_feature_bits(net), features);
}

uint64_t vhost_net_get_max_queues(VHostNetState *net)
{
    return net->dev.max_queues;
}

uint64_t vhost_net_get_acked_features(VHostNetState *net)
{
    return net->dev.acked_features;
}

static int vhost_net_get_fd(NetClientState *backend)
{
    switch (backend->info->type) {
    case NET_CLIENT_DRIVER_TAP:
        return tap_get_fd(backend);
    default:
        fprintf(stderr, "vhost-net requires tap backend\n");
        return -ENOSYS;
    }
}

struct vhost_net *vhost_net_init(VhostNetOptions *options)
{
    int r;
    bool backend_kernel = options->backend_type == VHOST_BACKEND_TYPE_KERNEL;
    struct vhost_net *net = g_new0(struct vhost_net, 1);
    uint64_t features = 0;

    if (!options->net_backend) {
        fprintf(stderr, "vhost-net requires net backend to be setup\n");
        goto fail;
    }
    net->nc = options->net_backend;

    net->dev.max_queues = 1;
    net->dev.nvqs = 2;
    net->dev.vqs = net->vqs;

    if (backend_kernel) {
        r = vhost_net_get_fd(options->net_backend);
        if (r < 0) {
            goto fail;
        }
        net->dev.backend_features = qemu_has_vnet_hdr(options->net_backend)
            ? 0 : (1ULL << VHOST_NET_F_VIRTIO_NET_HDR);
        net->backend = r;
        net->dev.protocol_features = 0;
    } else {
        net->dev.backend_features = 0;
        net->dev.protocol_features = 0;
        net->backend = -1;

        /* vhost-user needs vq_index to initiate a specific queue pair */
        net->dev.vq_index = net->nc->queue_index * net->dev.nvqs;
    }

    r = vhost_dev_init(&net->dev, options->opaque,
                       options->backend_type, options->busyloop_timeout);
    if (r < 0) {
        goto fail;
    }
    if (backend_kernel) {
        if (!qemu_has_vnet_hdr_len(options->net_backend,
                               sizeof(struct virtio_net_hdr_mrg_rxbuf))) {
            net->dev.features &= ~(1ULL << VIRTIO_NET_F_MRG_RXBUF);
        }
        if (~net->dev.features & net->dev.backend_features) {
            fprintf(stderr, "vhost lacks feature mask %" PRIu64
                   " for backend\n",
                   (uint64_t)(~net->dev.features & net->dev.backend_features));
            goto fail;
        }
    }

    /* Set sane init value. Override when guest acks. */
#ifdef CONFIG_VHOST_NET_USER
    if (net->nc->info->type == NET_CLIENT_DRIVER_VHOST_USER) {
        features = vhost_user_get_acked_features(net->nc);
        if (~net->dev.features & features) {
            fprintf(stderr, "vhost lacks feature mask %" PRIu64
                    " for backend\n",
                    (uint64_t)(~net->dev.features & features));
            goto fail;
        }
    }
#endif

    vhost_net_ack_features(net, features);

    return net;

fail:
    vhost_dev_cleanup(&net->dev);
    g_free(net);
    return NULL;
}

static void vhost_net_set_vq_index(struct vhost_net *net, int vq_index)
{
    net->dev.vq_index = vq_index;
}

static int vhost_net_start_one(struct vhost_net *net,
                               VirtIODevice *dev)
{
    struct vhost_vring_file file = { };
    int r;

    net->dev.nvqs = 2;
    net->dev.vqs = net->vqs;

    r = vhost_dev_enable_notifiers(&net->dev, dev);
    if (r < 0) {
        goto fail_notifiers;
    }

    r = vhost_dev_start(&net->dev, dev);
    if (r < 0) {
        goto fail_start;
    }

    if (net->nc->info->poll) {
        net->nc->info->poll(net->nc, false);
    }

    if (net->nc->info->type == NET_CLIENT_DRIVER_TAP) {
        qemu_set_fd_handler(net->backend, NULL, NULL, NULL);
        file.fd = net->backend;
        for (file.index = 0; file.index < net->dev.nvqs; ++file.index) {
            if (!virtio_queue_enabled(dev, net->dev.vq_index +
                                      file.index)) {
                /* Queue might not be ready for start */
                continue;
            }
            r = vhost_net_set_backend(&net->dev, &file);
            if (r < 0) {
                r = -errno;
                goto fail;
            }
        }
    }
    return 0;
fail:
    file.fd = -1;
    if (net->nc->info->type == NET_CLIENT_DRIVER_TAP) {
        while (file.index-- > 0) {
            if (!virtio_queue_enabled(dev, net->dev.vq_index +
                                      file.index)) {
                /* Queue might not be ready for start */
                continue;
            }
            int r = vhost_net_set_backend(&net->dev, &file);
            assert(r >= 0);
        }
    }
    if (net->nc->info->poll) {
        net->nc->info->poll(net->nc, true);
    }
    vhost_dev_stop(&net->dev, dev);
fail_start:
    vhost_dev_disable_notifiers(&net->dev, dev);
fail_notifiers:
    return r;
}

static void vhost_net_stop_one(struct vhost_net *net,
                               VirtIODevice *dev)
{
    struct vhost_vring_file file = { .fd = -1 };

    if (net->nc->info->type == NET_CLIENT_DRIVER_TAP) {
        for (file.index = 0; file.index < net->dev.nvqs; ++file.index) {
            int r = vhost_net_set_backend(&net->dev, &file);
            assert(r >= 0);
        }
    }
    if (net->nc->info->poll) {
        net->nc->info->poll(net->nc, true);
    }
    vhost_dev_stop(&net->dev, dev);
    vhost_dev_disable_notifiers(&net->dev, dev);
}

int vhost_net_start(VirtIODevice *dev, NetClientState *ncs,
                    int total_queues)
{
    BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(dev)));
    VirtioBusState *vbus = VIRTIO_BUS(qbus);
    VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(vbus);
    struct vhost_net *net;
    int r, e, i;
    NetClientState *peer;

    if (!k->set_guest_notifiers) {
        error_report("binding does not support guest notifiers");
        return -ENOSYS;
    }

    for (i = 0; i < total_queues; i++) {

        peer = qemu_get_peer(ncs, i);
        net = get_vhost_net(peer);
        vhost_net_set_vq_index(net, i * 2);

        /* Suppress the masking guest notifiers on vhost user
         * because vhost user doesn't interrupt masking/unmasking
         * properly.
         */
        if (net->nc->info->type == NET_CLIENT_DRIVER_VHOST_USER) {
            dev->use_guest_notifier_mask = false;
        }
     }

    r = k->set_guest_notifiers(qbus->parent, total_queues * 2, true);
    if (r < 0) {
        error_report("Error binding guest notifier: %d", -r);
        goto err;
    }

    for (i = 0; i < total_queues; i++) {
        peer = qemu_get_peer(ncs, i);
        r = vhost_net_start_one(get_vhost_net(peer), dev);

        if (r < 0) {
            goto err_start;
        }

        if (peer->vring_enable) {
            /* restore vring enable state */
            r = vhost_set_vring_enable(peer, peer->vring_enable);

            if (r < 0) {
                goto err_start;
            }
        }
    }

    return 0;

err_start:
    while (--i >= 0) {
        peer = qemu_get_peer(ncs , i);
        vhost_net_stop_one(get_vhost_net(peer), dev);
    }
    e = k->set_guest_notifiers(qbus->parent, total_queues * 2, false);
    if (e < 0) {
        fprintf(stderr, "vhost guest notifier cleanup failed: %d\n", e);
        fflush(stderr);
    }
err:
    return r;
}

void vhost_net_stop(VirtIODevice *dev, NetClientState *ncs,
                    int total_queues)
{
    BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(dev)));
    VirtioBusState *vbus = VIRTIO_BUS(qbus);
    VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(vbus);
    int i, r;

    for (i = 0; i < total_queues; i++) {
        vhost_net_stop_one(get_vhost_net(ncs[i].peer), dev);
    }

    r = k->set_guest_notifiers(qbus->parent, total_queues * 2, false);
    if (r < 0) {
        fprintf(stderr, "vhost guest notifier cleanup failed: %d\n", r);
        fflush(stderr);
    }
    assert(r >= 0);
}

void vhost_net_cleanup(struct vhost_net *net)
{
    vhost_dev_cleanup(&net->dev);
}

int vhost_net_notify_migration_done(struct vhost_net *net, char* mac_addr)
{
    const VhostOps *vhost_ops = net->dev.vhost_ops;

    assert(vhost_ops->backend_type == VHOST_BACKEND_TYPE_USER);
    assert(vhost_ops->vhost_migration_done);

    return vhost_ops->vhost_migration_done(&net->dev, mac_addr);
}

bool vhost_net_virtqueue_pending(VHostNetState *net, int idx)
{
    return vhost_virtqueue_pending(&net->dev, idx);
}

void vhost_net_virtqueue_mask(VHostNetState *net, VirtIODevice *dev,
                              int idx, bool mask)
{
    vhost_virtqueue_mask(&net->dev, dev, idx, mask);
}

VHostNetState *get_vhost_net(NetClientState *nc)
{
    VHostNetState *vhost_net = 0;

    if (!nc) {
        return 0;
    }

    switch (nc->info->type) {
    case NET_CLIENT_DRIVER_TAP:
        vhost_net = tap_get_vhost_net(nc);
        break;
#ifdef CONFIG_VHOST_NET_USER
    case NET_CLIENT_DRIVER_VHOST_USER:
        vhost_net = vhost_user_get_vhost_net(nc);
        assert(vhost_net);
        break;
#endif
#ifdef CONFIG_VHOST_NET_VDPA
    case NET_CLIENT_DRIVER_VHOST_VDPA:
        vhost_net = vhost_vdpa_get_vhost_net(nc);
        assert(vhost_net);
        break;
#endif
    default:
        break;
    }

    return vhost_net;
}

int vhost_set_vring_enable(NetClientState *nc, int enable)
{
    VHostNetState *net = get_vhost_net(nc);
    const VhostOps *vhost_ops = net->dev.vhost_ops;

    nc->vring_enable = enable;

    if (vhost_ops && vhost_ops->vhost_set_vring_enable) {
        return vhost_ops->vhost_set_vring_enable(&net->dev, enable);
    }

    return 0;
}

int vhost_net_set_mtu(struct vhost_net *net, uint16_t mtu)
{
    const VhostOps *vhost_ops = net->dev.vhost_ops;

    if (!vhost_ops->vhost_net_set_mtu) {
        return 0;
    }

    return vhost_ops->vhost_net_set_mtu(&net->dev, mtu);
}
Commit	Line	Data
d5970055 MT	1	/*
	2	* vhost-net support
	3	*
	4	* Copyright Red Hat, Inc. 2010
	5	*
	6	* Authors:
	7	* Michael S. Tsirkin <mst@redhat.com>
	8	*
	9	* This work is licensed under the terms of the GNU GPL, version 2. See
	10	* the COPYING file in the top-level directory.
6b620ca3 PB	11	*
	12	* Contributions after 2012-01-13 are licensed under the terms of the
	13	* GNU GPL, version 2 or (at your option) any later version.
d5970055 MT	14	*/
d5970055 MT	15
e8d40465	16	#include "qemu/osdep.h"
1422e32d	17	#include "net/net.h"
d5970055	18	#include "net/tap.h"
03ce5744	19	#include "net/vhost-user.h"
108a6481	20	#include "net/vhost-vdpa.h"
d5970055	21
18658a3c	22	#include "standard-headers/linux/vhost_types.h"
0d09e41a PB	23	#include "hw/virtio/virtio-net.h"
0d09e41a PB	24	#include "net/vhost_net.h"
1de7afc9	25	#include "qemu/error-report.h"
db725815	26	#include "qemu/main-loop.h"
d5970055	27
d5970055	28	#include <sys/socket.h>
d5970055 MT	29	#include <net/if.h>
	30	#include <netinet/in.h>
	31
d5970055	32
4fbe0f32	33	#include "standard-headers/linux/virtio_ring.h"
0d09e41a	34	#include "hw/virtio/vhost.h"
1c819449	35	#include "hw/virtio/virtio-bus.h"
d5970055	36
d5970055	37
2e6d46d7 NN	38	/* Features supported by host kernel. */
	39	static const int kernel_feature_bits[] = {
	40	VIRTIO_F_NOTIFY_ON_EMPTY,
	41	VIRTIO_RING_F_INDIRECT_DESC,
	42	VIRTIO_RING_F_EVENT_IDX,
	43	VIRTIO_NET_F_MRG_RXBUF,
b1506132	44	VIRTIO_F_VERSION_1,
45a368ad	45	VIRTIO_NET_F_MTU,
c471ad0e	46	VIRTIO_F_IOMMU_PLATFORM,
dfea7930	47	VIRTIO_F_RING_PACKED,
2e6d46d7 NN	48	VHOST_INVALID_FEATURE_BIT
	49	};
	50
5f4c01ca	51	/* Features supported by others. */
d122f1a2	52	static const int user_feature_bits[] = {
5f4c01ca NN	53	VIRTIO_F_NOTIFY_ON_EMPTY,
	54	VIRTIO_RING_F_INDIRECT_DESC,
	55	VIRTIO_RING_F_EVENT_IDX,
	56
	57	VIRTIO_F_ANY_LAYOUT,
b1506132	58	VIRTIO_F_VERSION_1,
5f4c01ca NN	59	VIRTIO_NET_F_CSUM,
	60	VIRTIO_NET_F_GUEST_CSUM,
	61	VIRTIO_NET_F_GSO,
	62	VIRTIO_NET_F_GUEST_TSO4,
	63	VIRTIO_NET_F_GUEST_TSO6,
	64	VIRTIO_NET_F_GUEST_ECN,
	65	VIRTIO_NET_F_GUEST_UFO,
	66	VIRTIO_NET_F_HOST_TSO4,
	67	VIRTIO_NET_F_HOST_TSO6,
	68	VIRTIO_NET_F_HOST_ECN,
	69	VIRTIO_NET_F_HOST_UFO,
	70	VIRTIO_NET_F_MRG_RXBUF,
45a368ad	71	VIRTIO_NET_F_MTU,
6dcdd06e	72	VIRTIO_F_IOMMU_PLATFORM,
dfea7930	73	VIRTIO_F_RING_PACKED,
5f4c01ca	74
72018d1e	75	/* This bit implies RARP isn't sent by QEMU out of band */
f6f56291 TC	76	VIRTIO_NET_F_GUEST_ANNOUNCE,
f6f56291 TC	77
5f4c01ca NN	78	VIRTIO_NET_F_MQ,
	79
	80	VHOST_INVALID_FEATURE_BIT
	81	};
	82
2e6d46d7	83	static const int vhost_net_get_feature_bits(struct vhost_net net)
d5970055	84	{
2e6d46d7 NN	85	const int *feature_bits = 0;
	86
	87	switch (net->nc->info->type) {
f394b2e2	88	case NET_CLIENT_DRIVER_TAP:
2e6d46d7 NN	89	feature_bits = kernel_feature_bits;
2e6d46d7 NN	90	break;
f394b2e2	91	case NET_CLIENT_DRIVER_VHOST_USER:
5f4c01ca NN	92	feature_bits = user_feature_bits;
5f4c01ca NN	93	break;
108a6481 CL	94	#ifdef CONFIG_VHOST_NET_VDPA
	95	case NET_CLIENT_DRIVER_VHOST_VDPA:
	96	feature_bits = vdpa_feature_bits;
	97	break;
	98	#endif
2e6d46d7 NN	99	default:
	100	error_report("Feature bits not defined for this type: %d",
	101	net->nc->info->type);
	102	break;
ca736c8e	103	}
2e6d46d7 NN	104
	105	return feature_bits;
	106	}
	107
9a2ba823	108	uint64_t vhost_net_get_features(struct vhost_net *net, uint64_t features)
2e6d46d7 NN	109	{
	110	return vhost_get_features(&net->dev, vhost_net_get_feature_bits(net),
	111	features);
d5970055	112	}
38140cc4 CL	113	int vhost_net_get_config(struct vhost_net net, uint8_t config,
	114	uint32_t config_len)
	115	{
	116	return vhost_dev_get_config(&net->dev, config, config_len);
	117	}
	118	int vhost_net_set_config(struct vhost_net net, const uint8_t data,
	119	uint32_t offset, uint32_t size, uint32_t flags)
	120	{
	121	return vhost_dev_set_config(&net->dev, data, offset, size, flags);
	122	}
d5970055	123
9a2ba823	124	void vhost_net_ack_features(struct vhost_net *net, uint64_t features)
d5970055	125	{
b49ae913	126	net->dev.acked_features = net->dev.backend_features;
2e6d46d7	127	vhost_ack_features(&net->dev, vhost_net_get_feature_bits(net), features);
d5970055 MT	128	}
d5970055 MT	129
e2051e9e YL	130	uint64_t vhost_net_get_max_queues(VHostNetState *net)
	131	{
	132	return net->dev.max_queues;
	133	}
	134
a463215b MAL	135	uint64_t vhost_net_get_acked_features(VHostNetState *net)
	136	{
	137	return net->dev.acked_features;
	138	}
	139
4e68f7a0	140	static int vhost_net_get_fd(NetClientState *backend)
d5970055 MT	141	{
d5970055 MT	142	switch (backend->info->type) {
f394b2e2	143	case NET_CLIENT_DRIVER_TAP:
d5970055 MT	144	return tap_get_fd(backend);
	145	default:
	146	fprintf(stderr, "vhost-net requires tap backend\n");
af3bba76	147	return -ENOSYS;
d5970055 MT	148	}
	149	}
	150
81647a65	151	struct vhost_net vhost_net_init(VhostNetOptions options)
d5970055 MT	152	{
d5970055 MT	153	int r;
1a1bfac9	154	bool backend_kernel = options->backend_type == VHOST_BACKEND_TYPE_KERNEL;
f1a0365b	155	struct vhost_net *net = g_new0(struct vhost_net, 1);
a463215b	156	uint64_t features = 0;
81647a65 NN	157
	158	if (!options->net_backend) {
	159	fprintf(stderr, "vhost-net requires net backend to be setup\n");
d5970055 MT	160	goto fail;
d5970055 MT	161	}
b931bfbf	162	net->nc = options->net_backend;
81647a65	163
e2051e9e	164	net->dev.max_queues = 1;
b931bfbf CO	165	net->dev.nvqs = 2;
b931bfbf CO	166	net->dev.vqs = net->vqs;
e2051e9e	167
1a1bfac9 NN	168	if (backend_kernel) {
	169	r = vhost_net_get_fd(options->net_backend);
	170	if (r < 0) {
	171	goto fail;
	172	}
	173	net->dev.backend_features = qemu_has_vnet_hdr(options->net_backend)
9a2ba823	174	? 0 : (1ULL << VHOST_NET_F_VIRTIO_NET_HDR);
1a1bfac9	175	net->backend = r;
dcb10c00	176	net->dev.protocol_features = 0;
1a1bfac9 NN	177	} else {
1a1bfac9 NN	178	net->dev.backend_features = 0;
dcb10c00	179	net->dev.protocol_features = 0;
1a1bfac9	180	net->backend = -1;
d5970055	181
b931bfbf CO	182	/* vhost-user needs vq_index to initiate a specific queue pair */
	183	net->dev.vq_index = net->nc->queue_index * net->dev.nvqs;
	184	}
f56a1247	185
81647a65	186	r = vhost_dev_init(&net->dev, options->opaque,
69e87b32	187	options->backend_type, options->busyloop_timeout);
d5970055 MT	188	if (r < 0) {
	189	goto fail;
	190	}
1a1bfac9	191	if (backend_kernel) {
d8e80ae3 DM	192	if (!qemu_has_vnet_hdr_len(options->net_backend,
d8e80ae3 DM	193	sizeof(struct virtio_net_hdr_mrg_rxbuf))) {
9a2ba823	194	net->dev.features &= ~(1ULL << VIRTIO_NET_F_MRG_RXBUF);
d8e80ae3	195	}
1a1bfac9 NN	196	if (~net->dev.features & net->dev.backend_features) {
	197	fprintf(stderr, "vhost lacks feature mask %" PRIu64
	198	" for backend\n",
	199	(uint64_t)(~net->dev.features & net->dev.backend_features));
1a1bfac9 NN	200	goto fail;
1a1bfac9 NN	201	}
d5970055	202	}
a463215b	203
d5970055	204	/* Set sane init value. Override when guest acks. */
56f41de7	205	#ifdef CONFIG_VHOST_NET_USER
f394b2e2	206	if (net->nc->info->type == NET_CLIENT_DRIVER_VHOST_USER) {
a463215b MAL	207	features = vhost_user_get_acked_features(net->nc);
	208	if (~net->dev.features & features) {
	209	fprintf(stderr, "vhost lacks feature mask %" PRIu64
	210	" for backend\n",
	211	(uint64_t)(~net->dev.features & features));
a463215b MAL	212	goto fail;
	213	}
	214	}
56f41de7	215	#endif
a463215b MAL	216
	217	vhost_net_ack_features(net, features);
	218
d5970055	219	return net;
f1a0365b	220
d5970055	221	fail:
f1a0365b	222	vhost_dev_cleanup(&net->dev);
7267c094	223	g_free(net);
d5970055 MT	224	return NULL;
	225	}
	226
cd7d1d26 JW	227	static void vhost_net_set_vq_index(struct vhost_net *net, int vq_index)
	228	{
	229	net->dev.vq_index = vq_index;
	230	}
	231
a9f98bb5	232	static int vhost_net_start_one(struct vhost_net *net,
cd7d1d26	233	VirtIODevice *dev)
d5970055 MT	234	{
	235	struct vhost_vring_file file = { };
	236	int r;
b0b3db79	237
a9f98bb5 JW	238	net->dev.nvqs = 2;
a9f98bb5 JW	239	net->dev.vqs = net->vqs;
a9f98bb5	240
b0b3db79 MT	241	r = vhost_dev_enable_notifiers(&net->dev, dev);
	242	if (r < 0) {
	243	goto fail_notifiers;
	244	}
d5970055	245
d5970055 MT	246	r = vhost_dev_start(&net->dev, dev);
d5970055 MT	247	if (r < 0) {
b0b3db79	248	goto fail_start;
d5970055 MT	249	}
d5970055 MT	250
212d69f2 NN	251	if (net->nc->info->poll) {
	252	net->nc->info->poll(net->nc, false);
	253	}
	254
f394b2e2	255	if (net->nc->info->type == NET_CLIENT_DRIVER_TAP) {
1a1bfac9 NN	256	qemu_set_fd_handler(net->backend, NULL, NULL, NULL);
	257	file.fd = net->backend;
	258	for (file.index = 0; file.index < net->dev.nvqs; ++file.index) {
23bfaf77 JW	259	if (!virtio_queue_enabled(dev, net->dev.vq_index +
	260	file.index)) {
	261	/* Queue might not be ready for start */
	262	continue;
	263	}
950d94ba	264	r = vhost_net_set_backend(&net->dev, &file);
1a1bfac9 NN	265	if (r < 0) {
	266	r = -errno;
	267	goto fail;
	268	}
d5970055 MT	269	}
	270	}
	271	return 0;
	272	fail:
	273	file.fd = -1;
f394b2e2	274	if (net->nc->info->type == NET_CLIENT_DRIVER_TAP) {
1a1bfac9	275	while (file.index-- > 0) {
23bfaf77 JW	276	if (!virtio_queue_enabled(dev, net->dev.vq_index +
	277	file.index)) {
	278	/* Queue might not be ready for start */
	279	continue;
	280	}
950d94ba	281	int r = vhost_net_set_backend(&net->dev, &file);
1a1bfac9 NN	282	assert(r >= 0);
1a1bfac9 NN	283	}
d5970055	284	}
212d69f2 NN	285	if (net->nc->info->poll) {
	286	net->nc->info->poll(net->nc, true);
	287	}
d5970055	288	vhost_dev_stop(&net->dev, dev);
b0b3db79 MT	289	fail_start:
	290	vhost_dev_disable_notifiers(&net->dev, dev);
	291	fail_notifiers:
d5970055 MT	292	return r;
	293	}
	294
a9f98bb5 JW	295	static void vhost_net_stop_one(struct vhost_net *net,
a9f98bb5 JW	296	VirtIODevice *dev)
d5970055 MT	297	{
	298	struct vhost_vring_file file = { .fd = -1 };
	299
f394b2e2	300	if (net->nc->info->type == NET_CLIENT_DRIVER_TAP) {
1a1bfac9	301	for (file.index = 0; file.index < net->dev.nvqs; ++file.index) {
950d94ba	302	int r = vhost_net_set_backend(&net->dev, &file);
1a1bfac9 NN	303	assert(r >= 0);
1a1bfac9 NN	304	}
d5970055	305	}
212d69f2 NN	306	if (net->nc->info->poll) {
	307	net->nc->info->poll(net->nc, true);
	308	}
d5970055	309	vhost_dev_stop(&net->dev, dev);
b0b3db79	310	vhost_dev_disable_notifiers(&net->dev, dev);
d5970055 MT	311	}
d5970055 MT	312
a9f98bb5 JW	313	int vhost_net_start(VirtIODevice dev, NetClientState ncs,
	314	int total_queues)
	315	{
1c819449 FK	316	BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(dev)));
	317	VirtioBusState *vbus = VIRTIO_BUS(qbus);
	318	VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(vbus);
92fbc3e0	319	struct vhost_net *net;
3154d1e4	320	int r, e, i;
92fbc3e0	321	NetClientState *peer;
a9f98bb5	322
1c819449	323	if (!k->set_guest_notifiers) {
312fd5f2	324	error_report("binding does not support guest notifiers");
a4076440	325	return -ENOSYS;
a9f98bb5 JW	326	}
a9f98bb5 JW	327
3154d1e4	328	for (i = 0; i < total_queues; i++) {
5669655a	329
92fbc3e0 CL	330	peer = qemu_get_peer(ncs, i);
92fbc3e0 CL	331	net = get_vhost_net(peer);
5669655a VK	332	vhost_net_set_vq_index(net, i * 2);
	333
	334	/* Suppress the masking guest notifiers on vhost user
	335	* because vhost user doesn't interrupt masking/unmasking
	336	* properly.
	337	*/
f394b2e2	338	if (net->nc->info->type == NET_CLIENT_DRIVER_VHOST_USER) {
01edc230	339	dev->use_guest_notifier_mask = false;
5669655a VK	340	}
5669655a VK	341	}
a9f98bb5	342
1c819449	343	r = k->set_guest_notifiers(qbus->parent, total_queues * 2, true);
a9f98bb5	344	if (r < 0) {
312fd5f2	345	error_report("Error binding guest notifier: %d", -r);
3154d1e4	346	goto err;
a9f98bb5 JW	347	}
a9f98bb5 JW	348
cd7d1d26	349	for (i = 0; i < total_queues; i++) {
92fbc3e0 CL	350	peer = qemu_get_peer(ncs, i);
92fbc3e0 CL	351	r = vhost_net_start_one(get_vhost_net(peer), dev);
cd7d1d26 JW	352
	353	if (r < 0) {
	354	goto err_start;
	355	}
bfc6cf31	356
92fbc3e0	357	if (peer->vring_enable) {
bfc6cf31	358	/* restore vring enable state */
92fbc3e0	359	r = vhost_set_vring_enable(peer, peer->vring_enable);
bfc6cf31 MAL	360
	361	if (r < 0) {
	362	goto err_start;
	363	}
	364	}
cd7d1d26 JW	365	}
cd7d1d26 JW	366
a9f98bb5 JW	367	return 0;
a9f98bb5 JW	368
cd7d1d26	369	err_start:
a9f98bb5	370	while (--i >= 0) {
92fbc3e0 CL	371	peer = qemu_get_peer(ncs , i);
92fbc3e0 CL	372	vhost_net_stop_one(get_vhost_net(peer), dev);
a9f98bb5	373	}
cd7d1d26 JW	374	e = k->set_guest_notifiers(qbus->parent, total_queues * 2, false);
	375	if (e < 0) {
	376	fprintf(stderr, "vhost guest notifier cleanup failed: %d\n", e);
	377	fflush(stderr);
	378	}
3154d1e4	379	err:
a9f98bb5 JW	380	return r;
	381	}
	382
	383	void vhost_net_stop(VirtIODevice dev, NetClientState ncs,
	384	int total_queues)
	385	{
1c819449 FK	386	BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(dev)));
	387	VirtioBusState *vbus = VIRTIO_BUS(qbus);
	388	VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(vbus);
a9f98bb5 JW	389	int i, r;
a9f98bb5 JW	390
cd7d1d26 JW	391	for (i = 0; i < total_queues; i++) {
	392	vhost_net_stop_one(get_vhost_net(ncs[i].peer), dev);
	393	}
	394
1c819449	395	r = k->set_guest_notifiers(qbus->parent, total_queues * 2, false);
a9f98bb5 JW	396	if (r < 0) {
	397	fprintf(stderr, "vhost guest notifier cleanup failed: %d\n", r);
	398	fflush(stderr);
	399	}
	400	assert(r >= 0);
a9f98bb5 JW	401	}
a9f98bb5 JW	402
d5970055 MT	403	void vhost_net_cleanup(struct vhost_net *net)
	404	{
	405	vhost_dev_cleanup(&net->dev);
d5970055	406	}
f56a1247	407
3e866365 TC	408	int vhost_net_notify_migration_done(struct vhost_net net, char mac_addr)
	409	{
	410	const VhostOps *vhost_ops = net->dev.vhost_ops;
3e866365	411
51f7aca9 MAL	412	assert(vhost_ops->backend_type == VHOST_BACKEND_TYPE_USER);
51f7aca9 MAL	413	assert(vhost_ops->vhost_migration_done);
3e866365	414
51f7aca9	415	return vhost_ops->vhost_migration_done(&net->dev, mac_addr);
3e866365 TC	416	}
3e866365 TC	417
f56a1247 MT	418	bool vhost_net_virtqueue_pending(VHostNetState *net, int idx)
	419	{
	420	return vhost_virtqueue_pending(&net->dev, idx);
	421	}
	422
	423	void vhost_net_virtqueue_mask(VHostNetState net, VirtIODevice dev,
	424	int idx, bool mask)
	425	{
	426	vhost_virtqueue_mask(&net->dev, dev, idx, mask);
	427	}
ed8b4afe NN	428
	429	VHostNetState get_vhost_net(NetClientState nc)
	430	{
	431	VHostNetState *vhost_net = 0;
	432
	433	if (!nc) {
	434	return 0;
	435	}
	436
	437	switch (nc->info->type) {
f394b2e2	438	case NET_CLIENT_DRIVER_TAP:
ed8b4afe NN	439	vhost_net = tap_get_vhost_net(nc);
ed8b4afe NN	440	break;
56f41de7	441	#ifdef CONFIG_VHOST_NET_USER
f394b2e2	442	case NET_CLIENT_DRIVER_VHOST_USER:
03ce5744	443	vhost_net = vhost_user_get_vhost_net(nc);
1a5b68ce	444	assert(vhost_net);
03ce5744	445	break;
108a6481 CL	446	#endif
	447	#ifdef CONFIG_VHOST_NET_VDPA
	448	case NET_CLIENT_DRIVER_VHOST_VDPA:
	449	vhost_net = vhost_vdpa_get_vhost_net(nc);
	450	assert(vhost_net);
	451	break;
56f41de7	452	#endif
ed8b4afe NN	453	default:
	454	break;
	455	}
	456
	457	return vhost_net;
	458	}
7263a0ad CO	459
	460	int vhost_set_vring_enable(NetClientState *nc, int enable)
	461	{
	462	VHostNetState *net = get_vhost_net(nc);
bb12e761	463	const VhostOps *vhost_ops = net->dev.vhost_ops;
72b65f92	464
bfc6cf31 MAL	465	nc->vring_enable = enable;
bfc6cf31 MAL	466
ca10203c	467	if (vhost_ops && vhost_ops->vhost_set_vring_enable) {
21e70425	468	return vhost_ops->vhost_set_vring_enable(&net->dev, enable);
7263a0ad CO	469	}
	470
	471	return 0;
	472	}
	473
45a368ad MC	474	int vhost_net_set_mtu(struct vhost_net *net, uint16_t mtu)
	475	{
	476	const VhostOps *vhost_ops = net->dev.vhost_ops;
	477
	478	if (!vhost_ops->vhost_net_set_mtu) {
	479	return 0;
	480	}
	481
	482	return vhost_ops->vhost_net_set_mtu(&net->dev, mtu);
	483	}