[qemu.git] / hw / block / vhost-user-blk.c

/*
 * vhost-user-blk host device
 *
 * Copyright(C) 2017 Intel Corporation.
 *
 * Authors:
 *  Changpeng Liu <[email protected]>
 *
 * Largely based on the "vhost-user-scsi.c" and "vhost-scsi.c" implemented by:
 * Felipe Franciosi <[email protected]>
 * Stefan Hajnoczi <[email protected]>
 * Nicholas Bellinger <[email protected]>
 *
 * This work is licensed under the terms of the GNU LGPL, version 2 or later.
 * See the COPYING.LIB file in the top-level directory.
 *
 */

#include "qemu/osdep.h"
#include "qapi/error.h"
#include "qemu/error-report.h"
#include "qemu/cutils.h"
#include "hw/qdev-core.h"
#include "hw/qdev-properties.h"
#include "hw/qdev-properties-system.h"
#include "hw/virtio/vhost.h"
#include "hw/virtio/vhost-user-blk.h"
#include "hw/virtio/virtio.h"
#include "hw/virtio/virtio-bus.h"
#include "hw/virtio/virtio-access.h"
#include "sysemu/sysemu.h"
#include "sysemu/runstate.h"

static const int user_feature_bits[] = {
    VIRTIO_BLK_F_SIZE_MAX,
    VIRTIO_BLK_F_SEG_MAX,
    VIRTIO_BLK_F_GEOMETRY,
    VIRTIO_BLK_F_BLK_SIZE,
    VIRTIO_BLK_F_TOPOLOGY,
    VIRTIO_BLK_F_MQ,
    VIRTIO_BLK_F_RO,
    VIRTIO_BLK_F_FLUSH,
    VIRTIO_BLK_F_CONFIG_WCE,
    VIRTIO_BLK_F_DISCARD,
    VIRTIO_BLK_F_WRITE_ZEROES,
    VIRTIO_F_VERSION_1,
    VIRTIO_RING_F_INDIRECT_DESC,
    VIRTIO_RING_F_EVENT_IDX,
    VIRTIO_F_NOTIFY_ON_EMPTY,
    VIRTIO_F_RING_PACKED,
    VIRTIO_F_IOMMU_PLATFORM,
    VHOST_INVALID_FEATURE_BIT
};

static void vhost_user_blk_event(void *opaque, QEMUChrEvent event);

static void vhost_user_blk_update_config(VirtIODevice *vdev, uint8_t *config)
{
    VHostUserBlk *s = VHOST_USER_BLK(vdev);

    /* Our num_queues overrides the device backend */
    virtio_stw_p(vdev, &s->blkcfg.num_queues, s->num_queues);

    memcpy(config, &s->blkcfg, sizeof(struct virtio_blk_config));
}

static void vhost_user_blk_set_config(VirtIODevice *vdev, const uint8_t *config)
{
    VHostUserBlk *s = VHOST_USER_BLK(vdev);
    struct virtio_blk_config *blkcfg = (struct virtio_blk_config *)config;
    int ret;

    if (blkcfg->wce == s->blkcfg.wce) {
        return;
    }

    ret = vhost_dev_set_config(&s->dev, &blkcfg->wce,
                               offsetof(struct virtio_blk_config, wce),
                               sizeof(blkcfg->wce),
                               VHOST_SET_CONFIG_TYPE_MASTER);
    if (ret) {
        error_report("set device config space failed");
        return;
    }

    s->blkcfg.wce = blkcfg->wce;
}

static int vhost_user_blk_handle_config_change(struct vhost_dev *dev)
{
    int ret;
    struct virtio_blk_config blkcfg;
    VHostUserBlk *s = VHOST_USER_BLK(dev->vdev);

    ret = vhost_dev_get_config(dev, (uint8_t *)&blkcfg,
                               sizeof(struct virtio_blk_config));
    if (ret < 0) {
        error_report("get config space failed");
        return -1;
    }

    /* valid for resize only */
    if (blkcfg.capacity != s->blkcfg.capacity) {
        s->blkcfg.capacity = blkcfg.capacity;
        memcpy(dev->vdev->config, &s->blkcfg, sizeof(struct virtio_blk_config));
        virtio_notify_config(dev->vdev);
    }

    return 0;
}

const VhostDevConfigOps blk_ops = {
    .vhost_dev_config_notifier = vhost_user_blk_handle_config_change,
};

static int vhost_user_blk_start(VirtIODevice *vdev)
{
    VHostUserBlk *s = VHOST_USER_BLK(vdev);
    BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
    VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
    int i, ret;

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

    ret = vhost_dev_enable_notifiers(&s->dev, vdev);
    if (ret < 0) {
        error_report("Error enabling host notifiers: %d", -ret);
        return ret;
    }

    ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, true);
    if (ret < 0) {
        error_report("Error binding guest notifier: %d", -ret);
        goto err_host_notifiers;
    }

    s->dev.acked_features = vdev->guest_features;

    ret = vhost_dev_prepare_inflight(&s->dev, vdev);
    if (ret < 0) {
        error_report("Error set inflight format: %d", -ret);
        goto err_guest_notifiers;
    }

    if (!s->inflight->addr) {
        ret = vhost_dev_get_inflight(&s->dev, s->queue_size, s->inflight);
        if (ret < 0) {
            error_report("Error get inflight: %d", -ret);
            goto err_guest_notifiers;
        }
    }

    ret = vhost_dev_set_inflight(&s->dev, s->inflight);
    if (ret < 0) {
        error_report("Error set inflight: %d", -ret);
        goto err_guest_notifiers;
    }

    ret = vhost_dev_start(&s->dev, vdev);
    if (ret < 0) {
        error_report("Error starting vhost: %d", -ret);
        goto err_guest_notifiers;
    }
    s->started_vu = true;

    /* guest_notifier_mask/pending not used yet, so just unmask
     * everything here. virtio-pci will do the right thing by
     * enabling/disabling irqfd.
     */
    for (i = 0; i < s->dev.nvqs; i++) {
        vhost_virtqueue_mask(&s->dev, vdev, i, false);
    }

    return ret;

err_guest_notifiers:
    k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
err_host_notifiers:
    vhost_dev_disable_notifiers(&s->dev, vdev);
    return ret;
}

static void vhost_user_blk_stop(VirtIODevice *vdev)
{
    VHostUserBlk *s = VHOST_USER_BLK(vdev);
    BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
    VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
    int ret;

    if (!s->started_vu) {
        return;
    }
    s->started_vu = false;

    if (!k->set_guest_notifiers) {
        return;
    }

    vhost_dev_stop(&s->dev, vdev);

    ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
    if (ret < 0) {
        error_report("vhost guest notifier cleanup failed: %d", ret);
        return;
    }

    vhost_dev_disable_notifiers(&s->dev, vdev);
}

static void vhost_user_blk_set_status(VirtIODevice *vdev, uint8_t status)
{
    VHostUserBlk *s = VHOST_USER_BLK(vdev);
    bool should_start = virtio_device_started(vdev, status);
    int ret;

    if (!vdev->vm_running) {
        should_start = false;
    }

    if (!s->connected) {
        return;
    }

    if (s->dev.started == should_start) {
        return;
    }

    if (should_start) {
        ret = vhost_user_blk_start(vdev);
        if (ret < 0) {
            error_report("vhost-user-blk: vhost start failed: %s",
                         strerror(-ret));
            qemu_chr_fe_disconnect(&s->chardev);
        }
    } else {
        vhost_user_blk_stop(vdev);
    }

}

static uint64_t vhost_user_blk_get_features(VirtIODevice *vdev,
                                            uint64_t features,
                                            Error **errp)
{
    VHostUserBlk *s = VHOST_USER_BLK(vdev);

    /* Turn on pre-defined features */
    virtio_add_feature(&features, VIRTIO_BLK_F_SEG_MAX);
    virtio_add_feature(&features, VIRTIO_BLK_F_GEOMETRY);
    virtio_add_feature(&features, VIRTIO_BLK_F_TOPOLOGY);
    virtio_add_feature(&features, VIRTIO_BLK_F_BLK_SIZE);
    virtio_add_feature(&features, VIRTIO_BLK_F_FLUSH);
    virtio_add_feature(&features, VIRTIO_BLK_F_RO);
    virtio_add_feature(&features, VIRTIO_BLK_F_DISCARD);
    virtio_add_feature(&features, VIRTIO_BLK_F_WRITE_ZEROES);

    if (s->config_wce) {
        virtio_add_feature(&features, VIRTIO_BLK_F_CONFIG_WCE);
    }
    if (s->num_queues > 1) {
        virtio_add_feature(&features, VIRTIO_BLK_F_MQ);
    }

    return vhost_get_features(&s->dev, user_feature_bits, features);
}

static void vhost_user_blk_handle_output(VirtIODevice *vdev, VirtQueue *vq)
{
    VHostUserBlk *s = VHOST_USER_BLK(vdev);
    int i, ret;

    if (!vdev->start_on_kick) {
        return;
    }

    if (!s->connected) {
        return;
    }

    if (s->dev.started) {
        return;
    }

    /* Some guests kick before setting VIRTIO_CONFIG_S_DRIVER_OK so start
     * vhost here instead of waiting for .set_status().
     */
    ret = vhost_user_blk_start(vdev);
    if (ret < 0) {
        error_report("vhost-user-blk: vhost start failed: %s",
                     strerror(-ret));
        qemu_chr_fe_disconnect(&s->chardev);
        return;
    }

    /* Kick right away to begin processing requests already in vring */
    for (i = 0; i < s->dev.nvqs; i++) {
        VirtQueue *kick_vq = virtio_get_queue(vdev, i);

        if (!virtio_queue_get_desc_addr(vdev, i)) {
            continue;
        }
        event_notifier_set(virtio_queue_get_host_notifier(kick_vq));
    }
}

static void vhost_user_blk_reset(VirtIODevice *vdev)
{
    VHostUserBlk *s = VHOST_USER_BLK(vdev);

    vhost_dev_free_inflight(s->inflight);
}

static int vhost_user_blk_connect(DeviceState *dev, Error **errp)
{
    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
    VHostUserBlk *s = VHOST_USER_BLK(vdev);
    int ret = 0;

    if (s->connected) {
        return 0;
    }
    s->connected = true;

    s->dev.num_queues = s->num_queues;
    s->dev.nvqs = s->num_queues;
    s->dev.vqs = s->vhost_vqs;
    s->dev.vq_index = 0;
    s->dev.backend_features = 0;

    vhost_dev_set_config_notifier(&s->dev, &blk_ops);

    ret = vhost_dev_init(&s->dev, &s->vhost_user, VHOST_BACKEND_TYPE_USER, 0);
    if (ret < 0) {
        error_setg_errno(errp, -ret, "vhost initialization failed");
        return ret;
    }

    /* restore vhost state */
    if (virtio_device_started(vdev, vdev->status)) {
        ret = vhost_user_blk_start(vdev);
        if (ret < 0) {
            error_setg_errno(errp, -ret, "vhost start failed");
            return ret;
        }
    }

    return 0;
}

static void vhost_user_blk_disconnect(DeviceState *dev)
{
    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
    VHostUserBlk *s = VHOST_USER_BLK(vdev);

    if (!s->connected) {
        return;
    }
    s->connected = false;

    vhost_user_blk_stop(vdev);

    vhost_dev_cleanup(&s->dev);
}

static void vhost_user_blk_chr_closed_bh(void *opaque)
{
    DeviceState *dev = opaque;
    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
    VHostUserBlk *s = VHOST_USER_BLK(vdev);

    vhost_user_blk_disconnect(dev);
    qemu_chr_fe_set_handlers(&s->chardev, NULL, NULL, vhost_user_blk_event,
                             NULL, opaque, NULL, true);
}

static void vhost_user_blk_event(void *opaque, QEMUChrEvent event)
{
    DeviceState *dev = opaque;
    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
    VHostUserBlk *s = VHOST_USER_BLK(vdev);
    Error *local_err = NULL;

    switch (event) {
    case CHR_EVENT_OPENED:
        if (vhost_user_blk_connect(dev, &local_err) < 0) {
            error_report_err(local_err);
            qemu_chr_fe_disconnect(&s->chardev);
            return;
        }
        break;
    case CHR_EVENT_CLOSED:
        if (!runstate_check(RUN_STATE_SHUTDOWN)) {
            /*
             * A close event may happen during a read/write, but vhost
             * code assumes the vhost_dev remains setup, so delay the
             * stop & clear.
             */
            AioContext *ctx = qemu_get_current_aio_context();

            qemu_chr_fe_set_handlers(&s->chardev, NULL, NULL, NULL, NULL,
                    NULL, NULL, false);
            aio_bh_schedule_oneshot(ctx, vhost_user_blk_chr_closed_bh, opaque);

            /*
             * Move vhost device to the stopped state. The vhost-user device
             * will be clean up and disconnected in BH. This can be useful in
             * the vhost migration code. If disconnect was caught there is an
             * option for the general vhost code to get the dev state without
             * knowing its type (in this case vhost-user).
             */
            s->dev.started = false;
        }
        break;
    case CHR_EVENT_BREAK:
    case CHR_EVENT_MUX_IN:
    case CHR_EVENT_MUX_OUT:
        /* Ignore */
        break;
    }
}

static void vhost_user_blk_device_realize(DeviceState *dev, Error **errp)
{
    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
    VHostUserBlk *s = VHOST_USER_BLK(vdev);
    int i, ret;

    if (!s->chardev.chr) {
        error_setg(errp, "chardev is mandatory");
        return;
    }

    if (s->num_queues == VHOST_USER_BLK_AUTO_NUM_QUEUES) {
        s->num_queues = 1;
    }
    if (!s->num_queues || s->num_queues > VIRTIO_QUEUE_MAX) {
        error_setg(errp, "invalid number of IO queues");
        return;
    }

    if (!s->queue_size) {
        error_setg(errp, "queue size must be non-zero");
        return;
    }
    if (s->queue_size > VIRTQUEUE_MAX_SIZE) {
        error_setg(errp, "queue size must not exceed %d",
                   VIRTQUEUE_MAX_SIZE);
        return;
    }

    if (!vhost_user_init(&s->vhost_user, &s->chardev, errp)) {
        return;
    }

    virtio_init(vdev, "virtio-blk", VIRTIO_ID_BLOCK,
                sizeof(struct virtio_blk_config));

    s->virtqs = g_new(VirtQueue *, s->num_queues);
    for (i = 0; i < s->num_queues; i++) {
        s->virtqs[i] = virtio_add_queue(vdev, s->queue_size,
                                        vhost_user_blk_handle_output);
    }

    s->inflight = g_new0(struct vhost_inflight, 1);
    s->vhost_vqs = g_new0(struct vhost_virtqueue, s->num_queues);
    s->connected = false;

    if (qemu_chr_fe_wait_connected(&s->chardev, errp) < 0) {
        goto virtio_err;
    }

    if (vhost_user_blk_connect(dev, errp) < 0) {
        qemu_chr_fe_disconnect(&s->chardev);
        goto virtio_err;
    }
    assert(s->connected);

    ret = vhost_dev_get_config(&s->dev, (uint8_t *)&s->blkcfg,
                               sizeof(struct virtio_blk_config));
    if (ret < 0) {
        error_setg(errp, "vhost-user-blk: get block config failed");
        goto vhost_err;
    }

    /* we're fully initialized, now we can operate, so add the handler */
    qemu_chr_fe_set_handlers(&s->chardev,  NULL, NULL,
                             vhost_user_blk_event, NULL, (void *)dev,
                             NULL, true);
    return;

vhost_err:
    vhost_dev_cleanup(&s->dev);
virtio_err:
    g_free(s->vhost_vqs);
    s->vhost_vqs = NULL;
    g_free(s->inflight);
    s->inflight = NULL;
    for (i = 0; i < s->num_queues; i++) {
        virtio_delete_queue(s->virtqs[i]);
    }
    g_free(s->virtqs);
    virtio_cleanup(vdev);
    vhost_user_cleanup(&s->vhost_user);
}

static void vhost_user_blk_device_unrealize(DeviceState *dev)
{
    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
    VHostUserBlk *s = VHOST_USER_BLK(dev);
    int i;

    virtio_set_status(vdev, 0);
    qemu_chr_fe_set_handlers(&s->chardev,  NULL, NULL, NULL,
                             NULL, NULL, NULL, false);
    vhost_dev_cleanup(&s->dev);
    vhost_dev_free_inflight(s->inflight);
    g_free(s->vhost_vqs);
    s->vhost_vqs = NULL;
    g_free(s->inflight);
    s->inflight = NULL;

    for (i = 0; i < s->num_queues; i++) {
        virtio_delete_queue(s->virtqs[i]);
    }
    g_free(s->virtqs);
    virtio_cleanup(vdev);
    vhost_user_cleanup(&s->vhost_user);
}

static void vhost_user_blk_instance_init(Object *obj)
{
    VHostUserBlk *s = VHOST_USER_BLK(obj);

    device_add_bootindex_property(obj, &s->bootindex, "bootindex",
                                  "/disk@0,0", DEVICE(obj));
}

static const VMStateDescription vmstate_vhost_user_blk = {
    .name = "vhost-user-blk",
    .minimum_version_id = 1,
    .version_id = 1,
    .fields = (VMStateField[]) {
        VMSTATE_VIRTIO_DEVICE,
        VMSTATE_END_OF_LIST()
    },
};

static Property vhost_user_blk_properties[] = {
    DEFINE_PROP_CHR("chardev", VHostUserBlk, chardev),
    DEFINE_PROP_UINT16("num-queues", VHostUserBlk, num_queues,
                       VHOST_USER_BLK_AUTO_NUM_QUEUES),
    DEFINE_PROP_UINT32("queue-size", VHostUserBlk, queue_size, 128),
    DEFINE_PROP_BIT("config-wce", VHostUserBlk, config_wce, 0, true),
    DEFINE_PROP_END_OF_LIST(),
};

static void vhost_user_blk_class_init(ObjectClass *klass, void *data)
{
    DeviceClass *dc = DEVICE_CLASS(klass);
    VirtioDeviceClass *vdc = VIRTIO_DEVICE_CLASS(klass);

    device_class_set_props(dc, vhost_user_blk_properties);
    dc->vmsd = &vmstate_vhost_user_blk;
    set_bit(DEVICE_CATEGORY_STORAGE, dc->categories);
    vdc->realize = vhost_user_blk_device_realize;
    vdc->unrealize = vhost_user_blk_device_unrealize;
    vdc->get_config = vhost_user_blk_update_config;
    vdc->set_config = vhost_user_blk_set_config;
    vdc->get_features = vhost_user_blk_get_features;
    vdc->set_status = vhost_user_blk_set_status;
    vdc->reset = vhost_user_blk_reset;
}

static const TypeInfo vhost_user_blk_info = {
    .name = TYPE_VHOST_USER_BLK,
    .parent = TYPE_VIRTIO_DEVICE,
    .instance_size = sizeof(VHostUserBlk),
    .instance_init = vhost_user_blk_instance_init,
    .class_init = vhost_user_blk_class_init,
};

static void virtio_register_types(void)
{
    type_register_static(&vhost_user_blk_info);
}

type_init(virtio_register_types)
Commit	Line	Data
00343e4b CL	1	/*
	2	* vhost-user-blk host device
	3	*
	4	* Copyright(C) 2017 Intel Corporation.
	5	*
	6	* Authors:
	7	* Changpeng Liu <[email protected]>
	8	*
	9	* Largely based on the "vhost-user-scsi.c" and "vhost-scsi.c" implemented by:
	10	* Felipe Franciosi <[email protected]>
	11	* Stefan Hajnoczi <[email protected]>
	12	* Nicholas Bellinger <[email protected]>
	13	*
	14	* This work is licensed under the terms of the GNU LGPL, version 2 or later.
	15	* See the COPYING.LIB file in the top-level directory.
	16	*
	17	*/
	18
	19	#include "qemu/osdep.h"
	20	#include "qapi/error.h"
	21	#include "qemu/error-report.h"
00343e4b	22	#include "qemu/cutils.h"
00343e4b	23	#include "hw/qdev-core.h"
a27bd6c7	24	#include "hw/qdev-properties.h"
ce35e229	25	#include "hw/qdev-properties-system.h"
00343e4b CL	26	#include "hw/virtio/vhost.h"
	27	#include "hw/virtio/vhost-user-blk.h"
	28	#include "hw/virtio/virtio.h"
	29	#include "hw/virtio/virtio-bus.h"
	30	#include "hw/virtio/virtio-access.h"
2f780b6a	31	#include "sysemu/sysemu.h"
54d31236	32	#include "sysemu/runstate.h"
00343e4b CL	33
	34	static const int user_feature_bits[] = {
	35	VIRTIO_BLK_F_SIZE_MAX,
	36	VIRTIO_BLK_F_SEG_MAX,
	37	VIRTIO_BLK_F_GEOMETRY,
	38	VIRTIO_BLK_F_BLK_SIZE,
	39	VIRTIO_BLK_F_TOPOLOGY,
	40	VIRTIO_BLK_F_MQ,
	41	VIRTIO_BLK_F_RO,
	42	VIRTIO_BLK_F_FLUSH,
	43	VIRTIO_BLK_F_CONFIG_WCE,
caa1ee43 CL	44	VIRTIO_BLK_F_DISCARD,
caa1ee43 CL	45	VIRTIO_BLK_F_WRITE_ZEROES,
00343e4b CL	46	VIRTIO_F_VERSION_1,
	47	VIRTIO_RING_F_INDIRECT_DESC,
	48	VIRTIO_RING_F_EVENT_IDX,
	49	VIRTIO_F_NOTIFY_ON_EMPTY,
7556a320 KW	50	VIRTIO_F_RING_PACKED,
7556a320 KW	51	VIRTIO_F_IOMMU_PLATFORM,
00343e4b CL	52	VHOST_INVALID_FEATURE_BIT
	53	};
	54
dabefdd6 KW	55	static void vhost_user_blk_event(void *opaque, QEMUChrEvent event);
dabefdd6 KW	56
00343e4b CL	57	static void vhost_user_blk_update_config(VirtIODevice vdev, uint8_t config)
	58	{
	59	VHostUserBlk *s = VHOST_USER_BLK(vdev);
	60
535255b4 SH	61	/* Our num_queues overrides the device backend */
	62	virtio_stw_p(vdev, &s->blkcfg.num_queues, s->num_queues);
	63
00343e4b CL	64	memcpy(config, &s->blkcfg, sizeof(struct virtio_blk_config));
	65	}
	66
	67	static void vhost_user_blk_set_config(VirtIODevice vdev, const uint8_t config)
	68	{
	69	VHostUserBlk *s = VHOST_USER_BLK(vdev);
	70	struct virtio_blk_config blkcfg = (struct virtio_blk_config )config;
	71	int ret;
	72
	73	if (blkcfg->wce == s->blkcfg.wce) {
	74	return;
	75	}
	76
	77	ret = vhost_dev_set_config(&s->dev, &blkcfg->wce,
	78	offsetof(struct virtio_blk_config, wce),
	79	sizeof(blkcfg->wce),
	80	VHOST_SET_CONFIG_TYPE_MASTER);
	81	if (ret) {
	82	error_report("set device config space failed");
	83	return;
	84	}
	85
	86	s->blkcfg.wce = blkcfg->wce;
	87	}
	88
	89	static int vhost_user_blk_handle_config_change(struct vhost_dev *dev)
	90	{
	91	int ret;
	92	struct virtio_blk_config blkcfg;
	93	VHostUserBlk *s = VHOST_USER_BLK(dev->vdev);
	94
	95	ret = vhost_dev_get_config(dev, (uint8_t *)&blkcfg,
	96	sizeof(struct virtio_blk_config));
	97	if (ret < 0) {
	98	error_report("get config space failed");
	99	return -1;
	100	}
	101
	102	/* valid for resize only */
	103	if (blkcfg.capacity != s->blkcfg.capacity) {
	104	s->blkcfg.capacity = blkcfg.capacity;
	105	memcpy(dev->vdev->config, &s->blkcfg, sizeof(struct virtio_blk_config));
	106	virtio_notify_config(dev->vdev);
	107	}
	108
	109	return 0;
	110	}
	111
	112	const VhostDevConfigOps blk_ops = {
	113	.vhost_dev_config_notifier = vhost_user_blk_handle_config_change,
	114	};
	115
a57f0091	116	static int vhost_user_blk_start(VirtIODevice *vdev)
00343e4b CL	117	{
	118	VHostUserBlk *s = VHOST_USER_BLK(vdev);
	119	BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
	120	VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
	121	int i, ret;
	122
	123	if (!k->set_guest_notifiers) {
	124	error_report("binding does not support guest notifiers");
a57f0091	125	return -ENOSYS;
00343e4b CL	126	}
	127
	128	ret = vhost_dev_enable_notifiers(&s->dev, vdev);
	129	if (ret < 0) {
	130	error_report("Error enabling host notifiers: %d", -ret);
a57f0091	131	return ret;
00343e4b CL	132	}
	133
	134	ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, true);
	135	if (ret < 0) {
	136	error_report("Error binding guest notifier: %d", -ret);
	137	goto err_host_notifiers;
	138	}
	139
	140	s->dev.acked_features = vdev->guest_features;
a1fe0b8f	141
1b0063b3 JY	142	ret = vhost_dev_prepare_inflight(&s->dev, vdev);
	143	if (ret < 0) {
	144	error_report("Error set inflight format: %d", -ret);
	145	goto err_guest_notifiers;
	146	}
	147
a1fe0b8f XY	148	if (!s->inflight->addr) {
	149	ret = vhost_dev_get_inflight(&s->dev, s->queue_size, s->inflight);
	150	if (ret < 0) {
	151	error_report("Error get inflight: %d", -ret);
	152	goto err_guest_notifiers;
	153	}
	154	}
	155
	156	ret = vhost_dev_set_inflight(&s->dev, s->inflight);
	157	if (ret < 0) {
	158	error_report("Error set inflight: %d", -ret);
	159	goto err_guest_notifiers;
	160	}
	161
00343e4b CL	162	ret = vhost_dev_start(&s->dev, vdev);
	163	if (ret < 0) {
	164	error_report("Error starting vhost: %d", -ret);
	165	goto err_guest_notifiers;
	166	}
f5b22d06	167	s->started_vu = true;
00343e4b CL	168
	169	/* guest_notifier_mask/pending not used yet, so just unmask
	170	* everything here. virtio-pci will do the right thing by
	171	* enabling/disabling irqfd.
	172	*/
	173	for (i = 0; i < s->dev.nvqs; i++) {
	174	vhost_virtqueue_mask(&s->dev, vdev, i, false);
	175	}
	176
a57f0091	177	return ret;
00343e4b CL	178
	179	err_guest_notifiers:
	180	k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
	181	err_host_notifiers:
	182	vhost_dev_disable_notifiers(&s->dev, vdev);
a57f0091	183	return ret;
00343e4b CL	184	}
	185
	186	static void vhost_user_blk_stop(VirtIODevice *vdev)
	187	{
	188	VHostUserBlk *s = VHOST_USER_BLK(vdev);
	189	BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
	190	VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
	191	int ret;
	192
f5b22d06 DS	193	if (!s->started_vu) {
	194	return;
	195	}
	196	s->started_vu = false;
	197
00343e4b CL	198	if (!k->set_guest_notifiers) {
	199	return;
	200	}
	201
	202	vhost_dev_stop(&s->dev, vdev);
	203
	204	ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
	205	if (ret < 0) {
	206	error_report("vhost guest notifier cleanup failed: %d", ret);
	207	return;
	208	}
	209
	210	vhost_dev_disable_notifiers(&s->dev, vdev);
	211	}
	212
	213	static void vhost_user_blk_set_status(VirtIODevice *vdev, uint8_t status)
	214	{
	215	VHostUserBlk *s = VHOST_USER_BLK(vdev);
e57f2c31	216	bool should_start = virtio_device_started(vdev, status);
77542d43	217	int ret;
00343e4b CL	218
	219	if (!vdev->vm_running) {
	220	should_start = false;
	221	}
	222
77542d43 XY	223	if (!s->connected) {
	224	return;
	225	}
	226
00343e4b CL	227	if (s->dev.started == should_start) {
	228	return;
	229	}
	230
	231	if (should_start) {
77542d43 XY	232	ret = vhost_user_blk_start(vdev);
	233	if (ret < 0) {
	234	error_report("vhost-user-blk: vhost start failed: %s",
	235	strerror(-ret));
	236	qemu_chr_fe_disconnect(&s->chardev);
	237	}
00343e4b CL	238	} else {
	239	vhost_user_blk_stop(vdev);
	240	}
	241
	242	}
	243
	244	static uint64_t vhost_user_blk_get_features(VirtIODevice *vdev,
	245	uint64_t features,
	246	Error **errp)
	247	{
	248	VHostUserBlk *s = VHOST_USER_BLK(vdev);
00343e4b CL	249
	250	/* Turn on pre-defined features */
	251	virtio_add_feature(&features, VIRTIO_BLK_F_SEG_MAX);
	252	virtio_add_feature(&features, VIRTIO_BLK_F_GEOMETRY);
	253	virtio_add_feature(&features, VIRTIO_BLK_F_TOPOLOGY);
	254	virtio_add_feature(&features, VIRTIO_BLK_F_BLK_SIZE);
	255	virtio_add_feature(&features, VIRTIO_BLK_F_FLUSH);
25b1d45a	256	virtio_add_feature(&features, VIRTIO_BLK_F_RO);
caa1ee43 CL	257	virtio_add_feature(&features, VIRTIO_BLK_F_DISCARD);
caa1ee43 CL	258	virtio_add_feature(&features, VIRTIO_BLK_F_WRITE_ZEROES);
00343e4b CL	259
	260	if (s->config_wce) {
	261	virtio_add_feature(&features, VIRTIO_BLK_F_CONFIG_WCE);
	262	}
00343e4b CL	263	if (s->num_queues > 1) {
	264	virtio_add_feature(&features, VIRTIO_BLK_F_MQ);
	265	}
	266
4a4ff4c5	267	return vhost_get_features(&s->dev, user_feature_bits, features);
00343e4b CL	268	}
	269
	270	static void vhost_user_blk_handle_output(VirtIODevice vdev, VirtQueue vq)
	271	{
110b9463	272	VHostUserBlk *s = VHOST_USER_BLK(vdev);
77542d43	273	int i, ret;
110b9463	274
f3facbe6	275	if (!vdev->start_on_kick) {
110b9463 YX	276	return;
	277	}
	278
77542d43 XY	279	if (!s->connected) {
	280	return;
	281	}
	282
110b9463 YX	283	if (s->dev.started) {
	284	return;
	285	}
	286
	287	/* Some guests kick before setting VIRTIO_CONFIG_S_DRIVER_OK so start
	288	* vhost here instead of waiting for .set_status().
	289	*/
77542d43 XY	290	ret = vhost_user_blk_start(vdev);
	291	if (ret < 0) {
	292	error_report("vhost-user-blk: vhost start failed: %s",
	293	strerror(-ret));
	294	qemu_chr_fe_disconnect(&s->chardev);
	295	return;
	296	}
00343e4b	297
110b9463 YX	298	/* Kick right away to begin processing requests already in vring */
	299	for (i = 0; i < s->dev.nvqs; i++) {
	300	VirtQueue *kick_vq = virtio_get_queue(vdev, i);
	301
	302	if (!virtio_queue_get_desc_addr(vdev, i)) {
	303	continue;
	304	}
	305	event_notifier_set(virtio_queue_get_host_notifier(kick_vq));
	306	}
00343e4b CL	307	}
00343e4b CL	308
a1fe0b8f XY	309	static void vhost_user_blk_reset(VirtIODevice *vdev)
	310	{
	311	VHostUserBlk *s = VHOST_USER_BLK(vdev);
	312
	313	vhost_dev_free_inflight(s->inflight);
	314	}
	315
5b9243d2	316	static int vhost_user_blk_connect(DeviceState dev, Error *errp)
77542d43 XY	317	{
	318	VirtIODevice *vdev = VIRTIO_DEVICE(dev);
	319	VHostUserBlk *s = VHOST_USER_BLK(vdev);
	320	int ret = 0;
	321
	322	if (s->connected) {
	323	return 0;
	324	}
	325	s->connected = true;
	326
c90bd505	327	s->dev.num_queues = s->num_queues;
77542d43	328	s->dev.nvqs = s->num_queues;
38e245a4	329	s->dev.vqs = s->vhost_vqs;
77542d43 XY	330	s->dev.vq_index = 0;
	331	s->dev.backend_features = 0;
	332
	333	vhost_dev_set_config_notifier(&s->dev, &blk_ops);
	334
	335	ret = vhost_dev_init(&s->dev, &s->vhost_user, VHOST_BACKEND_TYPE_USER, 0);
	336	if (ret < 0) {
5b9243d2	337	error_setg_errno(errp, -ret, "vhost initialization failed");
77542d43 XY	338	return ret;
	339	}
	340
	341	/* restore vhost state */
e57f2c31	342	if (virtio_device_started(vdev, vdev->status)) {
77542d43 XY	343	ret = vhost_user_blk_start(vdev);
77542d43 XY	344	if (ret < 0) {
5b9243d2	345	error_setg_errno(errp, -ret, "vhost start failed");
77542d43 XY	346	return ret;
	347	}
	348	}
	349
	350	return 0;
	351	}
	352
	353	static void vhost_user_blk_disconnect(DeviceState *dev)
	354	{
	355	VirtIODevice *vdev = VIRTIO_DEVICE(dev);
	356	VHostUserBlk *s = VHOST_USER_BLK(vdev);
	357
	358	if (!s->connected) {
	359	return;
	360	}
	361	s->connected = false;
	362
f5b22d06	363	vhost_user_blk_stop(vdev);
77542d43 XY	364
	365	vhost_dev_cleanup(&s->dev);
	366	}
	367
4bcad76f DS	368	static void vhost_user_blk_chr_closed_bh(void *opaque)
	369	{
	370	DeviceState *dev = opaque;
	371	VirtIODevice *vdev = VIRTIO_DEVICE(dev);
	372	VHostUserBlk *s = VHOST_USER_BLK(vdev);
	373
	374	vhost_user_blk_disconnect(dev);
dabefdd6 KW	375	qemu_chr_fe_set_handlers(&s->chardev, NULL, NULL, vhost_user_blk_event,
dabefdd6 KW	376	NULL, opaque, NULL, true);
4bcad76f DS	377	}
4bcad76f DS	378
dabefdd6	379	static void vhost_user_blk_event(void *opaque, QEMUChrEvent event)
77542d43 XY	380	{
	381	DeviceState *dev = opaque;
	382	VirtIODevice *vdev = VIRTIO_DEVICE(dev);
	383	VHostUserBlk *s = VHOST_USER_BLK(vdev);
5b9243d2	384	Error *local_err = NULL;
77542d43 XY	385
	386	switch (event) {
	387	case CHR_EVENT_OPENED:
5b9243d2 KW	388	if (vhost_user_blk_connect(dev, &local_err) < 0) {
5b9243d2 KW	389	error_report_err(local_err);
77542d43 XY	390	qemu_chr_fe_disconnect(&s->chardev);
	391	return;
	392	}
77542d43 XY	393	break;
77542d43 XY	394	case CHR_EVENT_CLOSED:
dabefdd6	395	if (!runstate_check(RUN_STATE_SHUTDOWN)) {
bc79c87b DP	396	/*
	397	* A close event may happen during a read/write, but vhost
	398	* code assumes the vhost_dev remains setup, so delay the
	399	* stop & clear.
	400	*/
4bcad76f DS	401	AioContext *ctx = qemu_get_current_aio_context();
	402
	403	qemu_chr_fe_set_handlers(&s->chardev, NULL, NULL, NULL, NULL,
	404	NULL, NULL, false);
	405	aio_bh_schedule_oneshot(ctx, vhost_user_blk_chr_closed_bh, opaque);
f5b22d06	406
bc79c87b DP	407	/*
	408	* Move vhost device to the stopped state. The vhost-user device
	409	* will be clean up and disconnected in BH. This can be useful in
	410	* the vhost migration code. If disconnect was caught there is an
	411	* option for the general vhost code to get the dev state without
	412	* knowing its type (in this case vhost-user).
	413	*/
	414	s->dev.started = false;
bc79c87b	415	}
77542d43	416	break;
669457f3 PMD	417	case CHR_EVENT_BREAK:
	418	case CHR_EVENT_MUX_IN:
	419	case CHR_EVENT_MUX_OUT:
	420	/* Ignore */
	421	break;
77542d43 XY	422	}
	423	}
	424
00343e4b CL	425	static void vhost_user_blk_device_realize(DeviceState dev, Error *errp)
	426	{
	427	VirtIODevice *vdev = VIRTIO_DEVICE(dev);
	428	VHostUserBlk *s = VHOST_USER_BLK(vdev);
	429	int i, ret;
	430
	431	if (!s->chardev.chr) {
5b9243d2	432	error_setg(errp, "chardev is mandatory");
00343e4b CL	433	return;
	434	}
	435
a4eef071 SH	436	if (s->num_queues == VHOST_USER_BLK_AUTO_NUM_QUEUES) {
	437	s->num_queues = 1;
	438	}
00343e4b	439	if (!s->num_queues \|\| s->num_queues > VIRTIO_QUEUE_MAX) {
5b9243d2	440	error_setg(errp, "invalid number of IO queues");
00343e4b CL	441	return;
	442	}
	443
	444	if (!s->queue_size) {
5b9243d2	445	error_setg(errp, "queue size must be non-zero");
00343e4b CL	446	return;
00343e4b CL	447	}
68bf7336	448	if (s->queue_size > VIRTQUEUE_MAX_SIZE) {
5b9243d2	449	error_setg(errp, "queue size must not exceed %d",
68bf7336 KW	450	VIRTQUEUE_MAX_SIZE);
	451	return;
	452	}
00343e4b	453
0b99f224	454	if (!vhost_user_init(&s->vhost_user, &s->chardev, errp)) {
4d0cf552 TB	455	return;
	456	}
	457
00343e4b CL	458	virtio_init(vdev, "virtio-blk", VIRTIO_ID_BLOCK,
	459	sizeof(struct virtio_blk_config));
	460
38e245a4	461	s->virtqs = g_new(VirtQueue *, s->num_queues);
00343e4b	462	for (i = 0; i < s->num_queues; i++) {
38e245a4 PN	463	s->virtqs[i] = virtio_add_queue(vdev, s->queue_size,
38e245a4 PN	464	vhost_user_blk_handle_output);
00343e4b CL	465	}
00343e4b CL	466
a1fe0b8f	467	s->inflight = g_new0(struct vhost_inflight, 1);
38e245a4	468	s->vhost_vqs = g_new0(struct vhost_virtqueue, s->num_queues);
77542d43	469	s->connected = false;
a1fe0b8f	470
f2672971	471	if (qemu_chr_fe_wait_connected(&s->chardev, errp) < 0) {
00343e4b CL	472	goto virtio_err;
	473	}
	474
5b9243d2	475	if (vhost_user_blk_connect(dev, errp) < 0) {
dabefdd6 KW	476	qemu_chr_fe_disconnect(&s->chardev);
dabefdd6 KW	477	goto virtio_err;
77542d43	478	}
dabefdd6	479	assert(s->connected);
77542d43	480
00343e4b	481	ret = vhost_dev_get_config(&s->dev, (uint8_t *)&s->blkcfg,
77542d43	482	sizeof(struct virtio_blk_config));
00343e4b	483	if (ret < 0) {
dabefdd6 KW	484	error_setg(errp, "vhost-user-blk: get block config failed");
dabefdd6 KW	485	goto vhost_err;
00343e4b CL	486	}
00343e4b CL	487
dabefdd6	488	/* we're fully initialized, now we can operate, so add the handler */
0c99d722	489	qemu_chr_fe_set_handlers(&s->chardev, NULL, NULL,
dabefdd6	490	vhost_user_blk_event, NULL, (void *)dev,
0c99d722	491	NULL, true);
00343e4b CL	492	return;
00343e4b CL	493
dabefdd6 KW	494	vhost_err:
dabefdd6 KW	495	vhost_dev_cleanup(&s->dev);
00343e4b	496	virtio_err:
38e245a4	497	g_free(s->vhost_vqs);
0ac2e635	498	s->vhost_vqs = NULL;
a1fe0b8f	499	g_free(s->inflight);
0ac2e635	500	s->inflight = NULL;
13e54681	501	for (i = 0; i < s->num_queues; i++) {
38e245a4	502	virtio_delete_queue(s->virtqs[i]);
13e54681	503	}
38e245a4	504	g_free(s->virtqs);
00343e4b	505	virtio_cleanup(vdev);
0b99f224	506	vhost_user_cleanup(&s->vhost_user);
00343e4b CL	507	}
00343e4b CL	508
b69c3c21	509	static void vhost_user_blk_device_unrealize(DeviceState *dev)
00343e4b CL	510	{
	511	VirtIODevice *vdev = VIRTIO_DEVICE(dev);
	512	VHostUserBlk *s = VHOST_USER_BLK(dev);
13e54681	513	int i;
00343e4b	514
96cb5498	515	virtio_set_status(vdev, 0);
77542d43 XY	516	qemu_chr_fe_set_handlers(&s->chardev, NULL, NULL, NULL,
77542d43 XY	517	NULL, NULL, NULL, false);
00343e4b	518	vhost_dev_cleanup(&s->dev);
a1fe0b8f	519	vhost_dev_free_inflight(s->inflight);
38e245a4	520	g_free(s->vhost_vqs);
0ac2e635	521	s->vhost_vqs = NULL;
a1fe0b8f	522	g_free(s->inflight);
0ac2e635	523	s->inflight = NULL;
13e54681 PN	524
13e54681 PN	525	for (i = 0; i < s->num_queues; i++) {
38e245a4	526	virtio_delete_queue(s->virtqs[i]);
13e54681	527	}
38e245a4	528	g_free(s->virtqs);
00343e4b	529	virtio_cleanup(vdev);
0b99f224	530	vhost_user_cleanup(&s->vhost_user);
00343e4b CL	531	}
	532
	533	static void vhost_user_blk_instance_init(Object *obj)
	534	{
	535	VHostUserBlk *s = VHOST_USER_BLK(obj);
	536
	537	device_add_bootindex_property(obj, &s->bootindex, "bootindex",
40c2281c	538	"/disk@0,0", DEVICE(obj));
00343e4b CL	539	}
	540
	541	static const VMStateDescription vmstate_vhost_user_blk = {
	542	.name = "vhost-user-blk",
	543	.minimum_version_id = 1,
	544	.version_id = 1,
	545	.fields = (VMStateField[]) {
	546	VMSTATE_VIRTIO_DEVICE,
	547	VMSTATE_END_OF_LIST()
	548	},
	549	};
	550
	551	static Property vhost_user_blk_properties[] = {
	552	DEFINE_PROP_CHR("chardev", VHostUserBlk, chardev),
a4eef071 SH	553	DEFINE_PROP_UINT16("num-queues", VHostUserBlk, num_queues,
a4eef071 SH	554	VHOST_USER_BLK_AUTO_NUM_QUEUES),
00343e4b CL	555	DEFINE_PROP_UINT32("queue-size", VHostUserBlk, queue_size, 128),
00343e4b CL	556	DEFINE_PROP_BIT("config-wce", VHostUserBlk, config_wce, 0, true),
00343e4b CL	557	DEFINE_PROP_END_OF_LIST(),
	558	};
	559
	560	static void vhost_user_blk_class_init(ObjectClass klass, void data)
	561	{
	562	DeviceClass *dc = DEVICE_CLASS(klass);
	563	VirtioDeviceClass *vdc = VIRTIO_DEVICE_CLASS(klass);
	564
4f67d30b	565	device_class_set_props(dc, vhost_user_blk_properties);
00343e4b CL	566	dc->vmsd = &vmstate_vhost_user_blk;
	567	set_bit(DEVICE_CATEGORY_STORAGE, dc->categories);
	568	vdc->realize = vhost_user_blk_device_realize;
	569	vdc->unrealize = vhost_user_blk_device_unrealize;
	570	vdc->get_config = vhost_user_blk_update_config;
	571	vdc->set_config = vhost_user_blk_set_config;
	572	vdc->get_features = vhost_user_blk_get_features;
	573	vdc->set_status = vhost_user_blk_set_status;
a1fe0b8f	574	vdc->reset = vhost_user_blk_reset;
00343e4b CL	575	}
	576
	577	static const TypeInfo vhost_user_blk_info = {
	578	.name = TYPE_VHOST_USER_BLK,
	579	.parent = TYPE_VIRTIO_DEVICE,
	580	.instance_size = sizeof(VHostUserBlk),
	581	.instance_init = vhost_user_blk_instance_init,
	582	.class_init = vhost_user_blk_class_init,
	583	};
	584
	585	static void virtio_register_types(void)
	586	{
	587	type_register_static(&vhost_user_blk_info);
	588	}
	589
	590	type_init(virtio_register_types)