Series comparison

-[Qemu-devel] [PULL 0/6] Net patches
+[PULL 00/15] Net patches
-The following changes since commit e607bbee553cfe73072870cef458cfa4e78133e2:
+The following changes since commit f45cc81911adc7726e8a2801986b6998b91b816e:
-  Merge remote-tracking branch 'remotes/edgar/tags/edgar/xilinx-next-2018-01-26.for-upstream' into staging (2018-01-26 14:24:25 +0000)
+  Merge remote-tracking branch 'remotes/cschoenebeck/tags/pull-9p-20220307' into staging (2022-03-08 09:06:57 +0000)
 are available in the git repository at:
   https://github.com/jasowang/qemu.git tags/net-pull-request
-for you to fetch changes up to bf4835a4d5338bb7424827715df22570a8adc67c:
+for you to fetch changes up to a10dd1e279fc56cebc7e738925e0db85d0cea089:
-  MAINTAINERS: update Dmitry Fleytman email (2018-01-29 16:05:38 +0800)
+  vdpa: Expose VHOST_F_LOG_ALL on SVQ (2022-03-08 21:18:41 +0800)
 ----------------------------------------------------------------
 ----------------------------------------------------------------
-Mao Zhongyi (2):
+Eugenio Pérez (14):
-      colo: modified the payload compare function
+      vhost: Add VhostShadowVirtqueue
-      colo: compare the packet based on the tcp sequence number
+      vhost: Add Shadow VirtQueue kick forwarding capabilities
       vhost: Add Shadow VirtQueue call forwarding capabilities
       vhost: Add vhost_svq_valid_features to shadow vq
       virtio: Add vhost_svq_get_vring_addr
       vdpa: adapt vhost_ops callbacks to svq
       vhost: Shadow virtqueue buffers forwarding
       util: Add iova_tree_alloc_map
       util: add iova_tree_find_iova
       vhost: Add VhostIOVATree
       vdpa: Add custom IOTLB translations to SVQ
       vdpa: Adapt vhost_vdpa_get_vring_base to SVQ
       vdpa: Never set log_base addr if SVQ is enabled
       vdpa: Expose VHOST_F_LOG_ALL on SVQ
-Philippe Mathieu-Daudé (1):
+Jason Wang (1):
-      MAINTAINERS: update Dmitry Fleytman email
+      virtio-net: fix map leaking on error during receive
-Thomas Huth (3):
+ hw/net/virtio-net.c                |   1 +
-      net: Allow hubports to connect to other netdevs
+ hw/virtio/meson.build              |   2 +-
-      net: Allow netdevs to be used with 'hostfwd_add' and 'hostfwd_remove'
+ hw/virtio/vhost-iova-tree.c        | 110 +++++++
-      qemu-doc: Get rid of "vlan=X" example in the documentation
+ hw/virtio/vhost-iova-tree.h        |  27 ++
+ hw/virtio/vhost-shadow-virtqueue.c | 637 +++++++++++++++++++++++++++++++++++++
- MAINTAINERS        |   8 +-
+ hw/virtio/vhost-shadow-virtqueue.h |  87 +++++
- hmp-commands.hx    |   4 +-
+ hw/virtio/vhost-vdpa.c             | 525 +++++++++++++++++++++++++++++-
- net/colo-compare.c | 411 +++++++++++++++++++++++++++++++++--------------------
+ include/hw/virtio/vhost-vdpa.h     |   8 +
- net/colo.c         |   9 ++
+ include/qemu/iova-tree.h           |  38 ++-
- net/colo.h         |  15 ++
+ util/iova-tree.c                   | 169 ++++++++++
- net/hub.c          |  27 +++-
+files changed, 1587 insertions(+), 17 deletions(-)
- net/hub.h          |   3 +-
+ create mode 100644 hw/virtio/vhost-iova-tree.c
- net/net.c          |   2 +-
+ create mode 100644 hw/virtio/vhost-iova-tree.h
- net/slirp.c        |  33 +++--
+ create mode 100644 hw/virtio/vhost-shadow-virtqueue.c
- net/trace-events   |   2 +-
+ create mode 100644 hw/virtio/vhost-shadow-virtqueue.h
  qapi/net.json      |   4 +-
  qemu-options.hx    |  12 +-
 files changed, 347 insertions(+), 183 deletions(-)

-New patch
+[PULL 01/15] virtio-net: fix map leaking on error during receive
+Commit bedd7e93d0196 ("virtio-net: fix use after unmap/free for sg")
+tries to fix the use after free of the sg by caching the virtqueue
+elements in an array and unmap them at once after receiving the
+packets, But it forgot to unmap the cached elements on error which
+will lead to leaking of mapping and other unexpected results.
+Fixing this by detaching the cached elements on error. This addresses
+CVE-2022-26353.
+Reported-by: Victor Tom <vv474172261@gmail.com>
+Cc: qemu-stable@nongnu.org
+Fixes: CVE-2022-26353
+Fixes: bedd7e93d0196 ("virtio-net: fix use after unmap/free for sg")
+Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
+Signed-off-by: Jason Wang <jasowang@redhat.com>
+---
+ hw/net/virtio-net.c | 1 +
+file changed, 1 insertion(+)
+diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c
+index XXXXXXX..XXXXXXX 100644
+--- a/hw/net/virtio-net.c
++++ b/hw/net/virtio-net.c
+@@ -XXX,XX +XXX,XX @@ static ssize_t virtio_net_receive_rcu(NetClientState *nc, const uint8_t *buf,
+ err:
+     for (j = 0; j < i; j++) {
++        virtqueue_detach_element(q->rx_vq, elems[j], lens[j]);
+         g_free(elems[j]);
+     }
+--
+.7.4

-New patch
+[PULL 02/15] vhost: Add VhostShadowVirtqueue
+From: Eugenio Pérez <eperezma@redhat.com>
+Vhost shadow virtqueue (SVQ) is an intermediate jump for virtqueue
+notifications and buffers, allowing qemu to track them. While qemu is
+forwarding the buffers and virtqueue changes, it is able to commit the
+memory it's being dirtied, the same way regular qemu's VirtIO devices
+do.
+This commit only exposes basic SVQ allocation and free. Next patches of
+the series add functionality like notifications and buffers forwarding.
+Acked-by: Michael S. Tsirkin <mst@redhat.com>
+Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
+Signed-off-by: Jason Wang <jasowang@redhat.com>
+---
+ hw/virtio/meson.build              |  2 +-
+ hw/virtio/vhost-shadow-virtqueue.c | 62 ++++++++++++++++++++++++++++++++++++++
+ hw/virtio/vhost-shadow-virtqueue.h | 28 +++++++++++++++++
+files changed, 91 insertions(+), 1 deletion(-)
+ create mode 100644 hw/virtio/vhost-shadow-virtqueue.c
+ create mode 100644 hw/virtio/vhost-shadow-virtqueue.h
+diff --git a/hw/virtio/meson.build b/hw/virtio/meson.build
+index XXXXXXX..XXXXXXX 100644
+--- a/hw/virtio/meson.build
++++ b/hw/virtio/meson.build
+@@ -XXX,XX +XXX,XX @@ softmmu_ss.add(when: 'CONFIG_ALL', if_true: files('vhost-stub.c'))
+ virtio_ss = ss.source_set()
+ virtio_ss.add(files('virtio.c'))
+-virtio_ss.add(when: 'CONFIG_VHOST', if_true: files('vhost.c', 'vhost-backend.c'))
++virtio_ss.add(when: 'CONFIG_VHOST', if_true: files('vhost.c', 'vhost-backend.c', 'vhost-shadow-virtqueue.c'))
+ virtio_ss.add(when: 'CONFIG_VHOST_USER', if_true: files('vhost-user.c'))
+ virtio_ss.add(when: 'CONFIG_VHOST_VDPA', if_true: files('vhost-vdpa.c'))
+ virtio_ss.add(when: 'CONFIG_VIRTIO_BALLOON', if_true: files('virtio-balloon.c'))
+diff --git a/hw/virtio/vhost-shadow-virtqueue.c b/hw/virtio/vhost-shadow-virtqueue.c
+new file mode 100644
+index XXXXXXX..XXXXXXX
+--- /dev/null
++++ b/hw/virtio/vhost-shadow-virtqueue.c
+@@ -XXX,XX +XXX,XX @@
++/*
++ * vhost shadow virtqueue
++ *
++ * SPDX-FileCopyrightText: Red Hat, Inc. 2021
++ * SPDX-FileContributor: Author: Eugenio Pérez <eperezma@redhat.com>
++ *
++ * SPDX-License-Identifier: GPL-2.0-or-later
++ */
++
++#include "qemu/osdep.h"
++#include "hw/virtio/vhost-shadow-virtqueue.h"
++
++#include "qemu/error-report.h"
++
++/**
++ * Creates vhost shadow virtqueue, and instructs the vhost device to use the
++ * shadow methods and file descriptors.
++ *
++ * Returns the new virtqueue or NULL.
++ *
++ * In case of error, reason is reported through error_report.
++ */
++VhostShadowVirtqueue *vhost_svq_new(void)
++{
++    g_autofree VhostShadowVirtqueue *svq = g_new0(VhostShadowVirtqueue, 1);
++    int r;
++
++    r = event_notifier_init(&svq->hdev_kick, 0);
++    if (r != 0) {
++        error_report("Couldn't create kick event notifier: %s (%d)",
++                     g_strerror(errno), errno);
++        goto err_init_hdev_kick;
++    }
++
++    r = event_notifier_init(&svq->hdev_call, 0);
++    if (r != 0) {
++        error_report("Couldn't create call event notifier: %s (%d)",
++                     g_strerror(errno), errno);
++        goto err_init_hdev_call;
++    }
++
++    return g_steal_pointer(&svq);
++
++err_init_hdev_call:
++    event_notifier_cleanup(&svq->hdev_kick);
++
++err_init_hdev_kick:
++    return NULL;
++}
++
++/**
++ * Free the resources of the shadow virtqueue.
++ *
++ * @pvq: gpointer to SVQ so it can be used by autofree functions.
++ */
++void vhost_svq_free(gpointer pvq)
++{
++    VhostShadowVirtqueue *vq = pvq;
++    event_notifier_cleanup(&vq->hdev_kick);
++    event_notifier_cleanup(&vq->hdev_call);
++    g_free(vq);
++}
+diff --git a/hw/virtio/vhost-shadow-virtqueue.h b/hw/virtio/vhost-shadow-virtqueue.h
+new file mode 100644
+index XXXXXXX..XXXXXXX
+--- /dev/null
++++ b/hw/virtio/vhost-shadow-virtqueue.h
+@@ -XXX,XX +XXX,XX @@
++/*
++ * vhost shadow virtqueue
++ *
++ * SPDX-FileCopyrightText: Red Hat, Inc. 2021
++ * SPDX-FileContributor: Author: Eugenio Pérez <eperezma@redhat.com>
++ *
++ * SPDX-License-Identifier: GPL-2.0-or-later
++ */
++
++#ifndef VHOST_SHADOW_VIRTQUEUE_H
++#define VHOST_SHADOW_VIRTQUEUE_H
++
++#include "qemu/event_notifier.h"
++
++/* Shadow virtqueue to relay notifications */
++typedef struct VhostShadowVirtqueue {
++    /* Shadow kick notifier, sent to vhost */
++    EventNotifier hdev_kick;
++    /* Shadow call notifier, sent to vhost */
++    EventNotifier hdev_call;
++} VhostShadowVirtqueue;
++
++VhostShadowVirtqueue *vhost_svq_new(void);
++
++void vhost_svq_free(gpointer vq);
++G_DEFINE_AUTOPTR_CLEANUP_FUNC(VhostShadowVirtqueue, vhost_svq_free);
++
++#endif
+--
+.7.4

-New patch
+[PULL 03/15] vhost: Add Shadow VirtQueue kick forwarding capabilities
+From: Eugenio Pérez <eperezma@redhat.com>
 At this mode no buffer forwarding will be performed in SVQ mode: Qemu
 will just forward the guest's kicks to the device.
 Host memory notifiers regions are left out for simplicity, and they will
 not be addressed in this series.
 Acked-by: Michael S. Tsirkin <mst@redhat.com>
 Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
 Signed-off-by: Jason Wang <jasowang@redhat.com>
 ---
  hw/virtio/vhost-shadow-virtqueue.c |  56 ++++++++++++++
  hw/virtio/vhost-shadow-virtqueue.h |  14 ++++
  hw/virtio/vhost-vdpa.c             | 145 ++++++++++++++++++++++++++++++++++++-
  include/hw/virtio/vhost-vdpa.h     |   4 +
 files changed, 217 insertions(+), 2 deletions(-)
 diff --git a/hw/virtio/vhost-shadow-virtqueue.c b/hw/virtio/vhost-shadow-virtqueue.c
 index XXXXXXX..XXXXXXX 100644
 --- a/hw/virtio/vhost-shadow-virtqueue.c
 +++ b/hw/virtio/vhost-shadow-virtqueue.c
@@ -XXX,XX +XXX,XX @@
  #include "hw/virtio/vhost-shadow-virtqueue.h"
  #include "qemu/error-report.h"
 +#include "qemu/main-loop.h"
 +#include "linux-headers/linux/vhost.h"
 +
 +/**
 + * Forward guest notifications.
 + *
 + * @n: guest kick event notifier, the one that guest set to notify svq.
 + */
 +static void vhost_handle_guest_kick(EventNotifier *n)
 +{
 +    VhostShadowVirtqueue *svq = container_of(n, VhostShadowVirtqueue,
 +                                             svq_kick);
 +    event_notifier_test_and_clear(n);
 +    event_notifier_set(&svq->hdev_kick);
 +}
 +
 +/**
 + * Set a new file descriptor for the guest to kick the SVQ and notify for avail
 + *
 + * @svq: The svq
 + * @svq_kick_fd: The svq kick fd
 + *
 + * Note that the SVQ will never close the old file descriptor.
 + */
 +void vhost_svq_set_svq_kick_fd(VhostShadowVirtqueue *svq, int svq_kick_fd)
 +{
 +    EventNotifier *svq_kick = &svq->svq_kick;
 +    bool poll_stop = VHOST_FILE_UNBIND != event_notifier_get_fd(svq_kick);
 +    bool poll_start = svq_kick_fd != VHOST_FILE_UNBIND;
 +
 +    if (poll_stop) {
 +        event_notifier_set_handler(svq_kick, NULL);
 +    }
 +
 +    /*
 +     * event_notifier_set_handler already checks for guest's notifications if
 +     * they arrive at the new file descriptor in the switch, so there is no
 +     * need to explicitly check for them.
 +     */
 +    if (poll_start) {
 +        event_notifier_init_fd(svq_kick, svq_kick_fd);
 +        event_notifier_set(svq_kick);
 +        event_notifier_set_handler(svq_kick, vhost_handle_guest_kick);
 +    }
 +}
 +
 +/**
 + * Stop the shadow virtqueue operation.
 + * @svq: Shadow Virtqueue
 + */
 +void vhost_svq_stop(VhostShadowVirtqueue *svq)
 +{
 +    event_notifier_set_handler(&svq->svq_kick, NULL);
 +}
  /**
   * Creates vhost shadow virtqueue, and instructs the vhost device to use the
@@ -XXX,XX +XXX,XX @@ VhostShadowVirtqueue *vhost_svq_new(void)
          goto err_init_hdev_call;
      }
 +    event_notifier_init_fd(&svq->svq_kick, VHOST_FILE_UNBIND);
      return g_steal_pointer(&svq);
  err_init_hdev_call:
@@ -XXX,XX +XXX,XX @@ err_init_hdev_kick:
  void vhost_svq_free(gpointer pvq)
  {
      VhostShadowVirtqueue *vq = pvq;
 +    vhost_svq_stop(vq);
      event_notifier_cleanup(&vq->hdev_kick);
      event_notifier_cleanup(&vq->hdev_call);
      g_free(vq);
 diff --git a/hw/virtio/vhost-shadow-virtqueue.h b/hw/virtio/vhost-shadow-virtqueue.h
 index XXXXXXX..XXXXXXX 100644
 --- a/hw/virtio/vhost-shadow-virtqueue.h
 +++ b/hw/virtio/vhost-shadow-virtqueue.h
@@ -XXX,XX +XXX,XX @@ typedef struct VhostShadowVirtqueue {
      EventNotifier hdev_kick;
      /* Shadow call notifier, sent to vhost */
      EventNotifier hdev_call;
 +
 +    /*
 +     * Borrowed virtqueue's guest to host notifier. To borrow it in this event
 +     * notifier allows to recover the VhostShadowVirtqueue from the event loop
 +     * easily. If we use the VirtQueue's one, we don't have an easy way to
 +     * retrieve VhostShadowVirtqueue.
 +     *
 +     * So shadow virtqueue must not clean it, or we would lose VirtQueue one.
 +     */
 +    EventNotifier svq_kick;
  } VhostShadowVirtqueue;
 +void vhost_svq_set_svq_kick_fd(VhostShadowVirtqueue *svq, int svq_kick_fd);
 +
 +void vhost_svq_stop(VhostShadowVirtqueue *svq);
 +
  VhostShadowVirtqueue *vhost_svq_new(void);
  void vhost_svq_free(gpointer vq);
 diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c
 index XXXXXXX..XXXXXXX 100644
 --- a/hw/virtio/vhost-vdpa.c
 +++ b/hw/virtio/vhost-vdpa.c
@@ -XXX,XX +XXX,XX @@
  #include "hw/virtio/vhost.h"
  #include "hw/virtio/vhost-backend.h"
  #include "hw/virtio/virtio-net.h"
 +#include "hw/virtio/vhost-shadow-virtqueue.h"
  #include "hw/virtio/vhost-vdpa.h"
  #include "exec/address-spaces.h"
  #include "qemu/main-loop.h"
  #include "cpu.h"
  #include "trace.h"
  #include "qemu-common.h"
 +#include "qapi/error.h"
  /*
   * Return one past the end of the end of section. Be careful with uint64_t
@@ -XXX,XX +XXX,XX @@ static bool vhost_vdpa_one_time_request(struct vhost_dev *dev)
      return v->index != 0;
  }
 +static int vhost_vdpa_init_svq(struct vhost_dev *hdev, struct vhost_vdpa *v,
 +                               Error **errp)
 +{
 +    g_autoptr(GPtrArray) shadow_vqs = NULL;
 +
 +    if (!v->shadow_vqs_enabled) {
 +        return 0;
 +    }
 +
 +    shadow_vqs = g_ptr_array_new_full(hdev->nvqs, vhost_svq_free);
 +    for (unsigned n = 0; n < hdev->nvqs; ++n) {
 +        g_autoptr(VhostShadowVirtqueue) svq = vhost_svq_new();
 +
 +        if (unlikely(!svq)) {
 +            error_setg(errp, "Cannot create svq %u", n);
 +            return -1;
 +        }
 +        g_ptr_array_add(shadow_vqs, g_steal_pointer(&svq));
 +    }
 +
 +    v->shadow_vqs = g_steal_pointer(&shadow_vqs);
 +    return 0;
 +}
 +
  static int vhost_vdpa_init(struct vhost_dev *dev, void *opaque, Error **errp)
  {
      struct vhost_vdpa *v;
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_init(struct vhost_dev *dev, void *opaque, Error **errp)
      dev->opaque =  opaque ;
      v->listener = vhost_vdpa_memory_listener;
      v->msg_type = VHOST_IOTLB_MSG_V2;
 +    ret = vhost_vdpa_init_svq(dev, v, errp);
 +    if (ret) {
 +        goto err;
 +    }
      vhost_vdpa_get_iova_range(v);
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_init(struct vhost_dev *dev, void *opaque, Error **errp)
                                 VIRTIO_CONFIG_S_DRIVER);
      return 0;
 +
 +err:
 +    ram_block_discard_disable(false);
 +    return ret;
  }
  static void vhost_vdpa_host_notifier_uninit(struct vhost_dev *dev,
@@ -XXX,XX +XXX,XX @@ err:
  static void vhost_vdpa_host_notifiers_init(struct vhost_dev *dev)
  {
 +    struct vhost_vdpa *v = dev->opaque;
      int i;
 +    if (v->shadow_vqs_enabled) {
 +        /* FIXME SVQ is not compatible with host notifiers mr */
 +        return;
 +    }
 +
      for (i = dev->vq_index; i < dev->vq_index + dev->nvqs; i++) {
          if (vhost_vdpa_host_notifier_init(dev, i)) {
              goto err;
@@ -XXX,XX +XXX,XX @@ err:
      return;
  }
 +static void vhost_vdpa_svq_cleanup(struct vhost_dev *dev)
 +{
 +    struct vhost_vdpa *v = dev->opaque;
 +    size_t idx;
 +
 +    if (!v->shadow_vqs) {
 +        return;
 +    }
 +
 +    for (idx = 0; idx < v->shadow_vqs->len; ++idx) {
 +        vhost_svq_stop(g_ptr_array_index(v->shadow_vqs, idx));
 +    }
 +    g_ptr_array_free(v->shadow_vqs, true);
 +}
 +
  static int vhost_vdpa_cleanup(struct vhost_dev *dev)
  {
      struct vhost_vdpa *v;
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_cleanup(struct vhost_dev *dev)
      trace_vhost_vdpa_cleanup(dev, v);
      vhost_vdpa_host_notifiers_uninit(dev, dev->nvqs);
      memory_listener_unregister(&v->listener);
 +    vhost_vdpa_svq_cleanup(dev);
      dev->opaque = NULL;
      ram_block_discard_disable(false);
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_get_device_id(struct vhost_dev *dev,
      return ret;
  }
 +static void vhost_vdpa_reset_svq(struct vhost_vdpa *v)
 +{
 +    if (!v->shadow_vqs_enabled) {
 +        return;
 +    }
 +
 +    for (unsigned i = 0; i < v->shadow_vqs->len; ++i) {
 +        VhostShadowVirtqueue *svq = g_ptr_array_index(v->shadow_vqs, i);
 +        vhost_svq_stop(svq);
 +    }
 +}
 +
  static int vhost_vdpa_reset_device(struct vhost_dev *dev)
  {
 +    struct vhost_vdpa *v = dev->opaque;
      int ret;
      uint8_t status = 0;
 +    vhost_vdpa_reset_svq(v);
 +
      ret = vhost_vdpa_call(dev, VHOST_VDPA_SET_STATUS, &status);
      trace_vhost_vdpa_reset_device(dev, status);
      return ret;
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_get_config(struct vhost_dev *dev, uint8_t *config,
      return ret;
   }
 +static int vhost_vdpa_set_vring_dev_kick(struct vhost_dev *dev,
 +                                         struct vhost_vring_file *file)
 +{
 +    trace_vhost_vdpa_set_vring_kick(dev, file->index, file->fd);
 +    return vhost_vdpa_call(dev, VHOST_SET_VRING_KICK, file);
 +}
 +
 +/**
 + * Set the shadow virtqueue descriptors to the device
 + *
 + * @dev: The vhost device model
 + * @svq: The shadow virtqueue
 + * @idx: The index of the virtqueue in the vhost device
 + * @errp: Error
 + */
 +static bool vhost_vdpa_svq_setup(struct vhost_dev *dev,
 +                                 VhostShadowVirtqueue *svq,
 +                                 unsigned idx,
 +                                 Error **errp)
 +{
 +    struct vhost_vring_file file = {
 +        .index = dev->vq_index + idx,
 +    };
 +    const EventNotifier *event_notifier = &svq->hdev_kick;
 +    int r;
 +
 +    file.fd = event_notifier_get_fd(event_notifier);
 +    r = vhost_vdpa_set_vring_dev_kick(dev, &file);
 +    if (unlikely(r != 0)) {
 +        error_setg_errno(errp, -r, "Can't set device kick fd");
 +    }
 +
 +    return r == 0;
 +}
 +
 +static bool vhost_vdpa_svqs_start(struct vhost_dev *dev)
 +{
 +    struct vhost_vdpa *v = dev->opaque;
 +    Error *err = NULL;
 +    unsigned i;
 +
 +    if (!v->shadow_vqs) {
 +        return true;
 +    }
 +
 +    for (i = 0; i < v->shadow_vqs->len; ++i) {
 +        VhostShadowVirtqueue *svq = g_ptr_array_index(v->shadow_vqs, i);
 +        bool ok = vhost_vdpa_svq_setup(dev, svq, i, &err);
 +        if (unlikely(!ok)) {
 +            error_reportf_err(err, "Cannot setup SVQ %u: ", i);
 +            return false;
 +        }
 +    }
 +
 +    return true;
 +}
 +
  static int vhost_vdpa_dev_start(struct vhost_dev *dev, bool started)
  {
      struct vhost_vdpa *v = dev->opaque;
 +    bool ok;
      trace_vhost_vdpa_dev_start(dev, started);
      if (started) {
          vhost_vdpa_host_notifiers_init(dev);
 +        ok = vhost_vdpa_svqs_start(dev);
 +        if (unlikely(!ok)) {
 +            return -1;
 +        }
          vhost_vdpa_set_vring_ready(dev);
      } else {
          vhost_vdpa_host_notifiers_uninit(dev, dev->nvqs);
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_get_vring_base(struct vhost_dev *dev,
  static int vhost_vdpa_set_vring_kick(struct vhost_dev *dev,
                                         struct vhost_vring_file *file)
  {
 -    trace_vhost_vdpa_set_vring_kick(dev, file->index, file->fd);
 -    return vhost_vdpa_call(dev, VHOST_SET_VRING_KICK, file);
 +    struct vhost_vdpa *v = dev->opaque;
 +    int vdpa_idx = file->index - dev->vq_index;
 +
 +    if (v->shadow_vqs_enabled) {
 +        VhostShadowVirtqueue *svq = g_ptr_array_index(v->shadow_vqs, vdpa_idx);
 +        vhost_svq_set_svq_kick_fd(svq, file->fd);
 +        return 0;
 +    } else {
 +        return vhost_vdpa_set_vring_dev_kick(dev, file);
 +    }
  }
  static int vhost_vdpa_set_vring_call(struct vhost_dev *dev,
 diff --git a/include/hw/virtio/vhost-vdpa.h b/include/hw/virtio/vhost-vdpa.h
 index XXXXXXX..XXXXXXX 100644
 --- a/include/hw/virtio/vhost-vdpa.h
 +++ b/include/hw/virtio/vhost-vdpa.h
@@ -XXX,XX +XXX,XX @@
  #ifndef HW_VIRTIO_VHOST_VDPA_H
  #define HW_VIRTIO_VHOST_VDPA_H
 +#include <gmodule.h>
 +
  #include "hw/virtio/virtio.h"
  #include "standard-headers/linux/vhost_types.h"
@@ -XXX,XX +XXX,XX @@ typedef struct vhost_vdpa {
      bool iotlb_batch_begin_sent;
      MemoryListener listener;
      struct vhost_vdpa_iova_range iova_range;
 +    bool shadow_vqs_enabled;
 +    GPtrArray *shadow_vqs;
      struct vhost_dev *dev;
      VhostVDPAHostNotifier notifier[VIRTIO_QUEUE_MAX];
  } VhostVDPA;
 --
 .7.4

-New patch
+[PULL 04/15] vhost: Add Shadow VirtQueue call forwarding capabilities
+From: Eugenio Pérez <eperezma@redhat.com>
+This will make qemu aware of the device used buffers, allowing it to
+write the guest memory with its contents if needed.
+Acked-by: Michael S. Tsirkin <mst@redhat.com>
+Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
+Signed-off-by: Jason Wang <jasowang@redhat.com>
+---
+ hw/virtio/vhost-shadow-virtqueue.c | 38 ++++++++++++++++++++++++++++++++++++++
+ hw/virtio/vhost-shadow-virtqueue.h |  4 ++++
+ hw/virtio/vhost-vdpa.c             | 31 +++++++++++++++++++++++++++++--
+files changed, 71 insertions(+), 2 deletions(-)
+diff --git a/hw/virtio/vhost-shadow-virtqueue.c b/hw/virtio/vhost-shadow-virtqueue.c
+index XXXXXXX..XXXXXXX 100644
+--- a/hw/virtio/vhost-shadow-virtqueue.c
++++ b/hw/virtio/vhost-shadow-virtqueue.c
+@@ -XXX,XX +XXX,XX @@ static void vhost_handle_guest_kick(EventNotifier *n)
+ }
+ /**
++ * Forward vhost notifications
++ *
++ * @n: hdev call event notifier, the one that device set to notify svq.
++ */
++static void vhost_svq_handle_call(EventNotifier *n)
++{
++    VhostShadowVirtqueue *svq = container_of(n, VhostShadowVirtqueue,
++                                             hdev_call);
++    event_notifier_test_and_clear(n);
++    event_notifier_set(&svq->svq_call);
++}
++
++/**
++ * Set the call notifier for the SVQ to call the guest
++ *
++ * @svq: Shadow virtqueue
++ * @call_fd: call notifier
++ *
++ * Called on BQL context.
++ */
++void vhost_svq_set_svq_call_fd(VhostShadowVirtqueue *svq, int call_fd)
++{
++    if (call_fd == VHOST_FILE_UNBIND) {
++        /*
++         * Fail event_notifier_set if called handling device call.
++         *
++         * SVQ still needs device notifications, since it needs to keep
++         * forwarding used buffers even with the unbind.
++         */
++        memset(&svq->svq_call, 0, sizeof(svq->svq_call));
++    } else {
++        event_notifier_init_fd(&svq->svq_call, call_fd);
++    }
++}
++
++/**
+  * Set a new file descriptor for the guest to kick the SVQ and notify for avail
+  *
+  * @svq: The svq
+@@ -XXX,XX +XXX,XX @@ VhostShadowVirtqueue *vhost_svq_new(void)
+     }
+     event_notifier_init_fd(&svq->svq_kick, VHOST_FILE_UNBIND);
++    event_notifier_set_handler(&svq->hdev_call, vhost_svq_handle_call);
+     return g_steal_pointer(&svq);
+ err_init_hdev_call:
+@@ -XXX,XX +XXX,XX @@ void vhost_svq_free(gpointer pvq)
+     VhostShadowVirtqueue *vq = pvq;
+     vhost_svq_stop(vq);
+     event_notifier_cleanup(&vq->hdev_kick);
++    event_notifier_set_handler(&vq->hdev_call, NULL);
+     event_notifier_cleanup(&vq->hdev_call);
+     g_free(vq);
+ }
+diff --git a/hw/virtio/vhost-shadow-virtqueue.h b/hw/virtio/vhost-shadow-virtqueue.h
+index XXXXXXX..XXXXXXX 100644
+--- a/hw/virtio/vhost-shadow-virtqueue.h
++++ b/hw/virtio/vhost-shadow-virtqueue.h
+@@ -XXX,XX +XXX,XX @@ typedef struct VhostShadowVirtqueue {
+      * So shadow virtqueue must not clean it, or we would lose VirtQueue one.
+      */
+     EventNotifier svq_kick;
++
++    /* Guest's call notifier, where the SVQ calls guest. */
++    EventNotifier svq_call;
+ } VhostShadowVirtqueue;
+ void vhost_svq_set_svq_kick_fd(VhostShadowVirtqueue *svq, int svq_kick_fd);
++void vhost_svq_set_svq_call_fd(VhostShadowVirtqueue *svq, int call_fd);
+ void vhost_svq_stop(VhostShadowVirtqueue *svq);
+diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c
+index XXXXXXX..XXXXXXX 100644
+--- a/hw/virtio/vhost-vdpa.c
++++ b/hw/virtio/vhost-vdpa.c
+@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_vring_dev_kick(struct vhost_dev *dev,
+     return vhost_vdpa_call(dev, VHOST_SET_VRING_KICK, file);
+ }
++static int vhost_vdpa_set_vring_dev_call(struct vhost_dev *dev,
++                                         struct vhost_vring_file *file)
++{
++    trace_vhost_vdpa_set_vring_call(dev, file->index, file->fd);
++    return vhost_vdpa_call(dev, VHOST_SET_VRING_CALL, file);
++}
++
+ /**
+  * Set the shadow virtqueue descriptors to the device
+  *
+@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_vring_dev_kick(struct vhost_dev *dev,
+  * @svq: The shadow virtqueue
+  * @idx: The index of the virtqueue in the vhost device
+  * @errp: Error
++ *
++ * Note that this function does not rewind kick file descriptor if cannot set
++ * call one.
+  */
+ static bool vhost_vdpa_svq_setup(struct vhost_dev *dev,
+                                  VhostShadowVirtqueue *svq,
+@@ -XXX,XX +XXX,XX @@ static bool vhost_vdpa_svq_setup(struct vhost_dev *dev,
+     r = vhost_vdpa_set_vring_dev_kick(dev, &file);
+     if (unlikely(r != 0)) {
+         error_setg_errno(errp, -r, "Can't set device kick fd");
++        return false;
++    }
++
++    event_notifier = &svq->hdev_call;
++    file.fd = event_notifier_get_fd(event_notifier);
++    r = vhost_vdpa_set_vring_dev_call(dev, &file);
++    if (unlikely(r != 0)) {
++        error_setg_errno(errp, -r, "Can't set device call fd");
+     }
+     return r == 0;
+@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_vring_kick(struct vhost_dev *dev,
+ static int vhost_vdpa_set_vring_call(struct vhost_dev *dev,
+                                        struct vhost_vring_file *file)
+ {
+-    trace_vhost_vdpa_set_vring_call(dev, file->index, file->fd);
+-    return vhost_vdpa_call(dev, VHOST_SET_VRING_CALL, file);
++    struct vhost_vdpa *v = dev->opaque;
++
++    if (v->shadow_vqs_enabled) {
++        int vdpa_idx = file->index - dev->vq_index;
++        VhostShadowVirtqueue *svq = g_ptr_array_index(v->shadow_vqs, vdpa_idx);
++
++        vhost_svq_set_svq_call_fd(svq, file->fd);
++        return 0;
++    } else {
++        return vhost_vdpa_set_vring_dev_call(dev, file);
++    }
+ }
+ static int vhost_vdpa_get_features(struct vhost_dev *dev,
+--
+.7.4

-New patch
+[PULL 05/15] vhost: Add vhost_svq_valid_features to shadow vq
+From: Eugenio Pérez <eperezma@redhat.com>
+This allows SVQ to negotiate features with the guest and the device. For
+the device, SVQ is a driver. While this function bypasses all
+non-transport features, it needs to disable the features that SVQ does
+not support when forwarding buffers. This includes packed vq layout,
+indirect descriptors or event idx.
+Future changes can add support to offer more features to the guest,
+since the use of VirtQueue gives this for free. This is left out at the
+moment for simplicity.
+Acked-by: Michael S. Tsirkin <mst@redhat.com>
+Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
+Signed-off-by: Jason Wang <jasowang@redhat.com>
+---
+ hw/virtio/vhost-shadow-virtqueue.c | 44 ++++++++++++++++++++++++++++++++++++++
+ hw/virtio/vhost-shadow-virtqueue.h |  2 ++
+ hw/virtio/vhost-vdpa.c             | 15 +++++++++++++
+files changed, 61 insertions(+)
+diff --git a/hw/virtio/vhost-shadow-virtqueue.c b/hw/virtio/vhost-shadow-virtqueue.c
+index XXXXXXX..XXXXXXX 100644
+--- a/hw/virtio/vhost-shadow-virtqueue.c
++++ b/hw/virtio/vhost-shadow-virtqueue.c
+@@ -XXX,XX +XXX,XX @@
+ #include "hw/virtio/vhost-shadow-virtqueue.h"
+ #include "qemu/error-report.h"
++#include "qapi/error.h"
+ #include "qemu/main-loop.h"
+ #include "linux-headers/linux/vhost.h"
+ /**
++ * Validate the transport device features that both guests can use with the SVQ
++ * and SVQs can use with the device.
++ *
++ * @dev_features: The features
++ * @errp: Error pointer
++ */
++bool vhost_svq_valid_features(uint64_t features, Error **errp)
++{
++    bool ok = true;
++    uint64_t svq_features = features;
++
++    for (uint64_t b = VIRTIO_TRANSPORT_F_START; b <= VIRTIO_TRANSPORT_F_END;
++         ++b) {
++        switch (b) {
++        case VIRTIO_F_ANY_LAYOUT:
++            continue;
++
++        case VIRTIO_F_ACCESS_PLATFORM:
++            /* SVQ trust in the host's IOMMU to translate addresses */
++        case VIRTIO_F_VERSION_1:
++            /* SVQ trust that the guest vring is little endian */
++            if (!(svq_features & BIT_ULL(b))) {
++                set_bit(b, &svq_features);
++                ok = false;
++            }
++            continue;
++
++        default:
++            if (svq_features & BIT_ULL(b)) {
++                clear_bit(b, &svq_features);
++                ok = false;
++            }
++        }
++    }
++
++    if (!ok) {
++        error_setg(errp, "SVQ Invalid device feature flags, offer: 0x%"PRIx64
++                         ", ok: 0x%"PRIx64, features, svq_features);
++    }
++    return ok;
++}
++
++/**
+  * Forward guest notifications.
+  *
+  * @n: guest kick event notifier, the one that guest set to notify svq.
+diff --git a/hw/virtio/vhost-shadow-virtqueue.h b/hw/virtio/vhost-shadow-virtqueue.h
+index XXXXXXX..XXXXXXX 100644
+--- a/hw/virtio/vhost-shadow-virtqueue.h
++++ b/hw/virtio/vhost-shadow-virtqueue.h
+@@ -XXX,XX +XXX,XX @@ typedef struct VhostShadowVirtqueue {
+     EventNotifier svq_call;
+ } VhostShadowVirtqueue;
++bool vhost_svq_valid_features(uint64_t features, Error **errp);
++
+ void vhost_svq_set_svq_kick_fd(VhostShadowVirtqueue *svq, int svq_kick_fd);
+ void vhost_svq_set_svq_call_fd(VhostShadowVirtqueue *svq, int call_fd);
+diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c
+index XXXXXXX..XXXXXXX 100644
+--- a/hw/virtio/vhost-vdpa.c
++++ b/hw/virtio/vhost-vdpa.c
+@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_init_svq(struct vhost_dev *hdev, struct vhost_vdpa *v,
+                                Error **errp)
+ {
+     g_autoptr(GPtrArray) shadow_vqs = NULL;
++    uint64_t dev_features, svq_features;
++    int r;
++    bool ok;
+     if (!v->shadow_vqs_enabled) {
+         return 0;
+     }
++    r = hdev->vhost_ops->vhost_get_features(hdev, &dev_features);
++    if (r != 0) {
++        error_setg_errno(errp, -r, "Can't get vdpa device features");
++        return r;
++    }
++
++    svq_features = dev_features;
++    ok = vhost_svq_valid_features(svq_features, errp);
++    if (unlikely(!ok)) {
++        return -1;
++    }
++
+     shadow_vqs = g_ptr_array_new_full(hdev->nvqs, vhost_svq_free);
+     for (unsigned n = 0; n < hdev->nvqs; ++n) {
+         g_autoptr(VhostShadowVirtqueue) svq = vhost_svq_new();
+--
+.7.4

-New patch
+[PULL 06/15] virtio: Add vhost_svq_get_vring_addr
+From: Eugenio Pérez <eperezma@redhat.com>
+It reports the shadow virtqueue address from qemu virtual address space.
+Since this will be different from the guest's vaddr, but the device can
+access it, SVQ takes special care about its alignment & lack of garbage
+data. It assumes that IOMMU will work in host_page_size ranges for that.
+Acked-by: Michael S. Tsirkin <mst@redhat.com>
+Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
+Signed-off-by: Jason Wang <jasowang@redhat.com>
+---
+ hw/virtio/vhost-shadow-virtqueue.c | 29 +++++++++++++++++++++++++++++
+ hw/virtio/vhost-shadow-virtqueue.h |  9 +++++++++
+files changed, 38 insertions(+)
+diff --git a/hw/virtio/vhost-shadow-virtqueue.c b/hw/virtio/vhost-shadow-virtqueue.c
+index XXXXXXX..XXXXXXX 100644
+--- a/hw/virtio/vhost-shadow-virtqueue.c
++++ b/hw/virtio/vhost-shadow-virtqueue.c
+@@ -XXX,XX +XXX,XX @@ void vhost_svq_set_svq_call_fd(VhostShadowVirtqueue *svq, int call_fd)
+ }
+ /**
++ * Get the shadow vq vring address.
++ * @svq: Shadow virtqueue
++ * @addr: Destination to store address
++ */
++void vhost_svq_get_vring_addr(const VhostShadowVirtqueue *svq,
++                              struct vhost_vring_addr *addr)
++{
++    addr->desc_user_addr = (uint64_t)svq->vring.desc;
++    addr->avail_user_addr = (uint64_t)svq->vring.avail;
++    addr->used_user_addr = (uint64_t)svq->vring.used;
++}
++
++size_t vhost_svq_driver_area_size(const VhostShadowVirtqueue *svq)
++{
++    size_t desc_size = sizeof(vring_desc_t) * svq->vring.num;
++    size_t avail_size = offsetof(vring_avail_t, ring) +
++                                             sizeof(uint16_t) * svq->vring.num;
++
++    return ROUND_UP(desc_size + avail_size, qemu_real_host_page_size);
++}
++
++size_t vhost_svq_device_area_size(const VhostShadowVirtqueue *svq)
++{
++    size_t used_size = offsetof(vring_used_t, ring) +
++                                    sizeof(vring_used_elem_t) * svq->vring.num;
++    return ROUND_UP(used_size, qemu_real_host_page_size);
++}
++
++/**
+  * Set a new file descriptor for the guest to kick the SVQ and notify for avail
+  *
+  * @svq: The svq
+diff --git a/hw/virtio/vhost-shadow-virtqueue.h b/hw/virtio/vhost-shadow-virtqueue.h
+index XXXXXXX..XXXXXXX 100644
+--- a/hw/virtio/vhost-shadow-virtqueue.h
++++ b/hw/virtio/vhost-shadow-virtqueue.h
+@@ -XXX,XX +XXX,XX @@
+ #define VHOST_SHADOW_VIRTQUEUE_H
+ #include "qemu/event_notifier.h"
++#include "hw/virtio/virtio.h"
++#include "standard-headers/linux/vhost_types.h"
+ /* Shadow virtqueue to relay notifications */
+ typedef struct VhostShadowVirtqueue {
++    /* Shadow vring */
++    struct vring vring;
++
+     /* Shadow kick notifier, sent to vhost */
+     EventNotifier hdev_kick;
+     /* Shadow call notifier, sent to vhost */
+@@ -XXX,XX +XXX,XX @@ bool vhost_svq_valid_features(uint64_t features, Error **errp);
+ void vhost_svq_set_svq_kick_fd(VhostShadowVirtqueue *svq, int svq_kick_fd);
+ void vhost_svq_set_svq_call_fd(VhostShadowVirtqueue *svq, int call_fd);
++void vhost_svq_get_vring_addr(const VhostShadowVirtqueue *svq,
++                              struct vhost_vring_addr *addr);
++size_t vhost_svq_driver_area_size(const VhostShadowVirtqueue *svq);
++size_t vhost_svq_device_area_size(const VhostShadowVirtqueue *svq);
+ void vhost_svq_stop(VhostShadowVirtqueue *svq);
+--
+.7.4

-New patch
+[PULL 07/15] vdpa: adapt vhost_ops callbacks to svq
+From: Eugenio Pérez <eperezma@redhat.com>
+First half of the buffers forwarding part, preparing vhost-vdpa
+callbacks to SVQ to offer it. QEMU cannot enable it at this moment, so
+this is effectively dead code at the moment, but it helps to reduce
+patch size.
+Acked-by: Michael S. Tsirkin <mst@redhat.com>
+Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
+Signed-off-by: Jason Wang <jasowang@redhat.com>
+---
+ hw/virtio/vhost-vdpa.c | 48 +++++++++++++++++++++++++++++++++++++++++-------
+file changed, 41 insertions(+), 7 deletions(-)
+diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c
+index XXXXXXX..XXXXXXX 100644
+--- a/hw/virtio/vhost-vdpa.c
++++ b/hw/virtio/vhost-vdpa.c
+@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_get_config(struct vhost_dev *dev, uint8_t *config,
+     return ret;
+  }
++static int vhost_vdpa_set_dev_vring_base(struct vhost_dev *dev,
++                                         struct vhost_vring_state *ring)
++{
++    trace_vhost_vdpa_set_vring_base(dev, ring->index, ring->num);
++    return vhost_vdpa_call(dev, VHOST_SET_VRING_BASE, ring);
++}
++
+ static int vhost_vdpa_set_vring_dev_kick(struct vhost_dev *dev,
+                                          struct vhost_vring_file *file)
+ {
+@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_vring_dev_call(struct vhost_dev *dev,
+     return vhost_vdpa_call(dev, VHOST_SET_VRING_CALL, file);
+ }
++static int vhost_vdpa_set_vring_dev_addr(struct vhost_dev *dev,
++                                         struct vhost_vring_addr *addr)
++{
++    trace_vhost_vdpa_set_vring_addr(dev, addr->index, addr->flags,
++                                addr->desc_user_addr, addr->used_user_addr,
++                                addr->avail_user_addr,
++                                addr->log_guest_addr);
++
++    return vhost_vdpa_call(dev, VHOST_SET_VRING_ADDR, addr);
++
++}
++
+ /**
+  * Set the shadow virtqueue descriptors to the device
+  *
+@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_log_base(struct vhost_dev *dev, uint64_t base,
+ static int vhost_vdpa_set_vring_addr(struct vhost_dev *dev,
+                                        struct vhost_vring_addr *addr)
+ {
+-    trace_vhost_vdpa_set_vring_addr(dev, addr->index, addr->flags,
+-                                    addr->desc_user_addr, addr->used_user_addr,
+-                                    addr->avail_user_addr,
+-                                    addr->log_guest_addr);
+-    return vhost_vdpa_call(dev, VHOST_SET_VRING_ADDR, addr);
++    struct vhost_vdpa *v = dev->opaque;
++
++    if (v->shadow_vqs_enabled) {
++        /*
++         * Device vring addr was set at device start. SVQ base is handled by
++         * VirtQueue code.
++         */
++        return 0;
++    }
++
++    return vhost_vdpa_set_vring_dev_addr(dev, addr);
+ }
+ static int vhost_vdpa_set_vring_num(struct vhost_dev *dev,
+@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_vring_num(struct vhost_dev *dev,
+ static int vhost_vdpa_set_vring_base(struct vhost_dev *dev,
+                                        struct vhost_vring_state *ring)
+ {
+-    trace_vhost_vdpa_set_vring_base(dev, ring->index, ring->num);
+-    return vhost_vdpa_call(dev, VHOST_SET_VRING_BASE, ring);
++    struct vhost_vdpa *v = dev->opaque;
++
++    if (v->shadow_vqs_enabled) {
++        /*
++         * Device vring base was set at device start. SVQ base is handled by
++         * VirtQueue code.
++         */
++        return 0;
++    }
++
++    return vhost_vdpa_set_dev_vring_base(dev, ring);
+ }
+ static int vhost_vdpa_get_vring_base(struct vhost_dev *dev,
+--
+.7.4

-[Qemu-devel] [PULL 3/6] net: Allow hubports to connect to other netdevs
+[PULL 08/15] vhost: Shadow virtqueue buffers forwarding
-From: Thomas Huth <thuth@redhat.com>
+From: Eugenio Pérez <eperezma@redhat.com>
-QEMU can emulate hubs to connect NICs and netdevs. This is currently
+Initial version of shadow virtqueue that actually forward buffers. There
-primarily used for the mis-named 'vlan' feature of the networking
+is no iommu support at the moment, and that will be addressed in future
-subsystem. Now the 'vlan' feature has been marked as deprecated, since
+patches of this series. Since all vhost-vdpa devices use forced IOMMU,
-its name is rather confusing and the users often rather mis-configure
+this means that SVQ is not usable at this point of the series on any
-their network when trying to use it. But while the 'vlan' parameter
+device.
-should be removed at one point in time, the basic idea of emulating
-a hub in QEMU is still good: It's useful for bundling up the output of
+For simplicity it only supports modern devices, that expects vring
-multiple NICs into one single l2tp netdev for example.
+in little endian, with split ring and no event idx or indirect
+descriptors. Support for them will not be added in this series.
-Now to be able to use the hubport feature without 'vlan's, there is one
-missing piece: The possibility to connect a hubport to a netdev, too.
+It reuses the VirtQueue code for the device part. The driver part is
-This patch adds this possibility by introducing a new "netdev=..."
+based on Linux's virtio_ring driver, but with stripped functionality
-parameter to the hubports.
+and optimizations so it's easier to review.
-To bundle up the output of multiple NICs into one socket netdev, you can
+However, forwarding buffers have some particular pieces: One of the most
-now run QEMU with these parameters for example:
+unexpected ones is that a guest's buffer can expand through more than
+one descriptor in SVQ. While this is handled gracefully by qemu's
-qemu-system-ppc64 ... -netdev socket,id=s1,connect=:11122 \
+emulated virtio devices, it may cause unexpected SVQ queue full. This
-    -netdev hubport,hubid=1,id=h1,netdev=s1 \
+patch also solves it by checking for this condition at both guest's
-    -netdev hubport,hubid=1,id=h2 -device e1000,netdev=h2 \
+kicks and device's calls. The code may be more elegant in the future if
-    -netdev hubport,hubid=1,id=h3 -device virtio-net-pci,netdev=h3
+SVQ code runs in its own iocontext.
-For using the socket netdev, you have got to start another QEMU as the
+Acked-by: Michael S. Tsirkin <mst@redhat.com>
-receiving side first, for example with network dumping enabled:
+Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
 qemu-system-x86_64 -M isapc -netdev socket,id=s0,listen=:11122 \
     -device ne2k_isa,netdev=s0 \
     -object filter-dump,id=f1,netdev=s0,file=/tmp/dump.dat
 After the ppc64 guest tried to boot from both NICs, you can see in the
 dump file (using Wireshark, for example), that the output of both NICs
 (the e1000 and the virtio-net-pci) has been successfully transfered
 via the socket netdev in this case.
 Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
 Signed-off-by: Thomas Huth <thuth@redhat.com>
 Signed-off-by: Jason Wang <jasowang@redhat.com>
 ---
- net/hub.c       | 27 +++++++++++++++++++++------
+ hw/virtio/vhost-shadow-virtqueue.c | 353 ++++++++++++++++++++++++++++++++++++-
- net/hub.h       |  3 ++-
+ hw/virtio/vhost-shadow-virtqueue.h |  26 +++
- net/net.c       |  2 +-
+ hw/virtio/vhost-vdpa.c             | 159 ++++++++++++++++-
- qapi/net.json   |  4 +++-
+files changed, 526 insertions(+), 12 deletions(-)
- qemu-options.hx |  8 +++++---
-files changed, 32 insertions(+), 12 deletions(-)
+diff --git a/hw/virtio/vhost-shadow-virtqueue.c b/hw/virtio/vhost-shadow-virtqueue.c
 diff --git a/net/hub.c b/net/hub.c
 index XXXXXXX..XXXXXXX 100644
---- a/net/hub.c
+--- a/hw/virtio/vhost-shadow-virtqueue.c
-+++ b/net/hub.c
++++ b/hw/virtio/vhost-shadow-virtqueue.c
 @@ -XXX,XX +XXX,XX @@
+ #include "qemu/error-report.h"
+ #include "qapi/error.h"
+ #include "qemu/main-loop.h"
++#include "qemu/log.h"
+ #include "linux-headers/linux/vhost.h"
+ /**
+@@ -XXX,XX +XXX,XX @@ bool vhost_svq_valid_features(uint64_t features, Error **errp)
+ }
+ /**
+- * Forward guest notifications.
++ * Number of descriptors that the SVQ can make available from the guest.
++ *
++ * @svq: The svq
++ */
++static uint16_t vhost_svq_available_slots(const VhostShadowVirtqueue *svq)
++{
++    return svq->vring.num - (svq->shadow_avail_idx - svq->shadow_used_idx);
++}
++
++static void vhost_vring_write_descs(VhostShadowVirtqueue *svq,
++                                    const struct iovec *iovec,
++                                    size_t num, bool more_descs, bool write)
++{
++    uint16_t i = svq->free_head, last = svq->free_head;
++    unsigned n;
++    uint16_t flags = write ? cpu_to_le16(VRING_DESC_F_WRITE) : 0;
++    vring_desc_t *descs = svq->vring.desc;
++
++    if (num == 0) {
++        return;
++    }
++
++    for (n = 0; n < num; n++) {
++        if (more_descs || (n + 1 < num)) {
++            descs[i].flags = flags | cpu_to_le16(VRING_DESC_F_NEXT);
++        } else {
++            descs[i].flags = flags;
++        }
++        descs[i].addr = cpu_to_le64((hwaddr)iovec[n].iov_base);
++        descs[i].len = cpu_to_le32(iovec[n].iov_len);
++
++        last = i;
++        i = cpu_to_le16(descs[i].next);
++    }
++
++    svq->free_head = le16_to_cpu(descs[last].next);
++}
++
++static bool vhost_svq_add_split(VhostShadowVirtqueue *svq,
++                                VirtQueueElement *elem,
++                                unsigned *head)
++{
++    unsigned avail_idx;
++    vring_avail_t *avail = svq->vring.avail;
++
++    *head = svq->free_head;
++
++    /* We need some descriptors here */
++    if (unlikely(!elem->out_num && !elem->in_num)) {
++        qemu_log_mask(LOG_GUEST_ERROR,
++            "Guest provided element with no descriptors");
++        return false;
++    }
++
++    vhost_vring_write_descs(svq, elem->out_sg, elem->out_num,
++                            elem->in_num > 0, false);
++    vhost_vring_write_descs(svq, elem->in_sg, elem->in_num, false, true);
++
++    /*
++     * Put the entry in the available array (but don't update avail->idx until
++     * they do sync).
++     */
++    avail_idx = svq->shadow_avail_idx & (svq->vring.num - 1);
++    avail->ring[avail_idx] = cpu_to_le16(*head);
++    svq->shadow_avail_idx++;
++
++    /* Update the avail index after write the descriptor */
++    smp_wmb();
++    avail->idx = cpu_to_le16(svq->shadow_avail_idx);
++
++    return true;
++}
++
++static bool vhost_svq_add(VhostShadowVirtqueue *svq, VirtQueueElement *elem)
++{
++    unsigned qemu_head;
++    bool ok = vhost_svq_add_split(svq, elem, &qemu_head);
++    if (unlikely(!ok)) {
++        return false;
++    }
++
++    svq->ring_id_maps[qemu_head] = elem;
++    return true;
++}
++
++static void vhost_svq_kick(VhostShadowVirtqueue *svq)
++{
++    /*
++     * We need to expose the available array entries before checking the used
++     * flags
++     */
++    smp_mb();
++    if (svq->vring.used->flags & VRING_USED_F_NO_NOTIFY) {
++        return;
++    }
++
++    event_notifier_set(&svq->hdev_kick);
++}
++
++/**
++ * Forward available buffers.
++ *
++ * @svq: Shadow VirtQueue
++ *
++ * Note that this function does not guarantee that all guest's available
++ * buffers are available to the device in SVQ avail ring. The guest may have
++ * exposed a GPA / GIOVA contiguous buffer, but it may not be contiguous in
++ * qemu vaddr.
++ *
++ * If that happens, guest's kick notifications will be disabled until the
++ * device uses some buffers.
++ */
++static void vhost_handle_guest_kick(VhostShadowVirtqueue *svq)
++{
++    /* Clear event notifier */
++    event_notifier_test_and_clear(&svq->svq_kick);
++
++    /* Forward to the device as many available buffers as possible */
++    do {
++        virtio_queue_set_notification(svq->vq, false);
++
++        while (true) {
++            VirtQueueElement *elem;
++            bool ok;
++
++            if (svq->next_guest_avail_elem) {
++                elem = g_steal_pointer(&svq->next_guest_avail_elem);
++            } else {
++                elem = virtqueue_pop(svq->vq, sizeof(*elem));
++            }
++
++            if (!elem) {
++                break;
++            }
++
++            if (elem->out_num + elem->in_num >
++                vhost_svq_available_slots(svq)) {
++                /*
++                 * This condition is possible since a contiguous buffer in GPA
++                 * does not imply a contiguous buffer in qemu's VA
++                 * scatter-gather segments. If that happens, the buffer exposed
++                 * to the device needs to be a chain of descriptors at this
++                 * moment.
++                 *
++                 * SVQ cannot hold more available buffers if we are here:
++                 * queue the current guest descriptor and ignore further kicks
++                 * until some elements are used.
++                 */
++                svq->next_guest_avail_elem = elem;
++                return;
++            }
++
++            ok = vhost_svq_add(svq, elem);
++            if (unlikely(!ok)) {
++                /* VQ is broken, just return and ignore any other kicks */
++                return;
++            }
++            vhost_svq_kick(svq);
++        }
++
++        virtio_queue_set_notification(svq->vq, true);
++    } while (!virtio_queue_empty(svq->vq));
++}
++
++/**
++ * Handle guest's kick.
+  *
+  * @n: guest kick event notifier, the one that guest set to notify svq.
   */
+-static void vhost_handle_guest_kick(EventNotifier *n)
- #include "qemu/osdep.h"
++static void vhost_handle_guest_kick_notifier(EventNotifier *n)
 +#include "qapi/error.h"
  #include "monitor/monitor.h"
  #include "net/net.h"
  #include "clients.h"
@@ -XXX,XX +XXX,XX @@ static NetClientInfo net_hub_port_info = {
      .cleanup = net_hub_port_cleanup,
  };
 -static NetHubPort *net_hub_port_new(NetHub *hub, const char *name)
 +static NetHubPort *net_hub_port_new(NetHub *hub, const char *name,
 +                                    NetClientState *hubpeer)
  {
-     NetClientState *nc;
+     VhostShadowVirtqueue *svq = container_of(n, VhostShadowVirtqueue,
-     NetHubPort *port;
+                                              svq_kick);
-@@ -XXX,XX +XXX,XX @@ static NetHubPort *net_hub_port_new(NetHub *hub, const char *name)
+     event_notifier_test_and_clear(n);
-         name = default_name;
+-    event_notifier_set(&svq->hdev_kick);
 +    vhost_handle_guest_kick(svq);
 +}
 +
 +static bool vhost_svq_more_used(VhostShadowVirtqueue *svq)
 +{
 +    if (svq->last_used_idx != svq->shadow_used_idx) {
 +        return true;
 +    }
 +
 +    svq->shadow_used_idx = cpu_to_le16(svq->vring.used->idx);
 +
 +    return svq->last_used_idx != svq->shadow_used_idx;
  }
  /**
 - * Forward vhost notifications
 + * Enable vhost device calls after disable them.
 + *
 + * @svq: The svq
 + *
 + * It returns false if there are pending used buffers from the vhost device,
 + * avoiding the possible races between SVQ checking for more work and enabling
 + * callbacks. True if SVQ used vring has no more pending buffers.
 + */
 +static bool vhost_svq_enable_notification(VhostShadowVirtqueue *svq)
 +{
 +    svq->vring.avail->flags &= ~cpu_to_le16(VRING_AVAIL_F_NO_INTERRUPT);
 +    /* Make sure the flag is written before the read of used_idx */
 +    smp_mb();
 +    return !vhost_svq_more_used(svq);
 +}
 +
 +static void vhost_svq_disable_notification(VhostShadowVirtqueue *svq)
 +{
 +    svq->vring.avail->flags |= cpu_to_le16(VRING_AVAIL_F_NO_INTERRUPT);
 +}
 +
 +static VirtQueueElement *vhost_svq_get_buf(VhostShadowVirtqueue *svq,
 +                                           uint32_t *len)
 +{
 +    vring_desc_t *descs = svq->vring.desc;
 +    const vring_used_t *used = svq->vring.used;
 +    vring_used_elem_t used_elem;
 +    uint16_t last_used;
 +
 +    if (!vhost_svq_more_used(svq)) {
 +        return NULL;
 +    }
 +
 +    /* Only get used array entries after they have been exposed by dev */
 +    smp_rmb();
 +    last_used = svq->last_used_idx & (svq->vring.num - 1);
 +    used_elem.id = le32_to_cpu(used->ring[last_used].id);
 +    used_elem.len = le32_to_cpu(used->ring[last_used].len);
 +
 +    svq->last_used_idx++;
 +    if (unlikely(used_elem.id >= svq->vring.num)) {
 +        qemu_log_mask(LOG_GUEST_ERROR, "Device %s says index %u is used",
 +                      svq->vdev->name, used_elem.id);
 +        return NULL;
 +    }
 +
 +    if (unlikely(!svq->ring_id_maps[used_elem.id])) {
 +        qemu_log_mask(LOG_GUEST_ERROR,
 +            "Device %s says index %u is used, but it was not available",
 +            svq->vdev->name, used_elem.id);
 +        return NULL;
 +    }
 +
 +    descs[used_elem.id].next = svq->free_head;
 +    svq->free_head = used_elem.id;
 +
 +    *len = used_elem.len;
 +    return g_steal_pointer(&svq->ring_id_maps[used_elem.id]);
 +}
 +
 +static void vhost_svq_flush(VhostShadowVirtqueue *svq,
 +                            bool check_for_avail_queue)
 +{
 +    VirtQueue *vq = svq->vq;
 +
 +    /* Forward as many used buffers as possible. */
 +    do {
 +        unsigned i = 0;
 +
 +        vhost_svq_disable_notification(svq);
 +        while (true) {
 +            uint32_t len;
 +            g_autofree VirtQueueElement *elem = vhost_svq_get_buf(svq, &len);
 +            if (!elem) {
 +                break;
 +            }
 +
 +            if (unlikely(i >= svq->vring.num)) {
 +                qemu_log_mask(LOG_GUEST_ERROR,
 +                         "More than %u used buffers obtained in a %u size SVQ",
 +                         i, svq->vring.num);
 +                virtqueue_fill(vq, elem, len, i);
 +                virtqueue_flush(vq, i);
 +                return;
 +            }
 +            virtqueue_fill(vq, elem, len, i++);
 +        }
 +
 +        virtqueue_flush(vq, i);
 +        event_notifier_set(&svq->svq_call);
 +
 +        if (check_for_avail_queue && svq->next_guest_avail_elem) {
 +            /*
 +             * Avail ring was full when vhost_svq_flush was called, so it's a
 +             * good moment to make more descriptors available if possible.
 +             */
 +            vhost_handle_guest_kick(svq);
 +        }
 +    } while (!vhost_svq_enable_notification(svq));
 +}
 +
 +/**
 + * Forward used buffers.
   *
   * @n: hdev call event notifier, the one that device set to notify svq.
 + *
 + * Note that we are not making any buffers available in the loop, there is no
 + * way that it runs more than virtqueue size times.
   */
  static void vhost_svq_handle_call(EventNotifier *n)
  {
      VhostShadowVirtqueue *svq = container_of(n, VhostShadowVirtqueue,
                                               hdev_call);
      event_notifier_test_and_clear(n);
 -    event_notifier_set(&svq->svq_call);
 +    vhost_svq_flush(svq, true);
  }
  /**
@@ -XXX,XX +XXX,XX @@ void vhost_svq_set_svq_kick_fd(VhostShadowVirtqueue *svq, int svq_kick_fd)
      if (poll_start) {
          event_notifier_init_fd(svq_kick, svq_kick_fd);
          event_notifier_set(svq_kick);
 -        event_notifier_set_handler(svq_kick, vhost_handle_guest_kick);
 +        event_notifier_set_handler(svq_kick, vhost_handle_guest_kick_notifier);
 +    }
 +}
 +
 +/**
 + * Start the shadow virtqueue operation.
 + *
 + * @svq: Shadow Virtqueue
 + * @vdev: VirtIO device
 + * @vq: Virtqueue to shadow
 + */
 +void vhost_svq_start(VhostShadowVirtqueue *svq, VirtIODevice *vdev,
 +                     VirtQueue *vq)
 +{
 +    size_t desc_size, driver_size, device_size;
 +
 +    svq->next_guest_avail_elem = NULL;
 +    svq->shadow_avail_idx = 0;
 +    svq->shadow_used_idx = 0;
 +    svq->last_used_idx = 0;
 +    svq->vdev = vdev;
 +    svq->vq = vq;
 +
 +    svq->vring.num = virtio_queue_get_num(vdev, virtio_get_queue_index(vq));
 +    driver_size = vhost_svq_driver_area_size(svq);
 +    device_size = vhost_svq_device_area_size(svq);
 +    svq->vring.desc = qemu_memalign(qemu_real_host_page_size, driver_size);
 +    desc_size = sizeof(vring_desc_t) * svq->vring.num;
 +    svq->vring.avail = (void *)((char *)svq->vring.desc + desc_size);
 +    memset(svq->vring.desc, 0, driver_size);
 +    svq->vring.used = qemu_memalign(qemu_real_host_page_size, device_size);
 +    memset(svq->vring.used, 0, device_size);
 +    svq->ring_id_maps = g_new0(VirtQueueElement *, svq->vring.num);
 +    for (unsigned i = 0; i < svq->vring.num - 1; i++) {
 +        svq->vring.desc[i].next = cpu_to_le16(i + 1);
      }
+ }
--    nc = qemu_new_net_client(&net_hub_port_info, NULL, "hub", name);
-+    nc = qemu_new_net_client(&net_hub_port_info, hubpeer, "hub", name);
+@@ -XXX,XX +XXX,XX @@ void vhost_svq_set_svq_kick_fd(VhostShadowVirtqueue *svq, int svq_kick_fd)
-     port = DO_UPCAST(NetHubPort, nc, nc);
+ void vhost_svq_stop(VhostShadowVirtqueue *svq)
-     port->id = id;
+ {
-     port->hub = hub;
+     event_notifier_set_handler(&svq->svq_kick, NULL);
-@@ -XXX,XX +XXX,XX @@ static NetHubPort *net_hub_port_new(NetHub *hub, const char *name)
++    g_autofree VirtQueueElement *next_avail_elem = NULL;
 +
 +    if (!svq->vq) {
 +        return;
 +    }
 +
 +    /* Send all pending used descriptors to guest */
 +    vhost_svq_flush(svq, false);
 +
 +    for (unsigned i = 0; i < svq->vring.num; ++i) {
 +        g_autofree VirtQueueElement *elem = NULL;
 +        elem = g_steal_pointer(&svq->ring_id_maps[i]);
 +        if (elem) {
 +            virtqueue_detach_element(svq->vq, elem, 0);
 +        }
 +    }
 +
 +    next_avail_elem = g_steal_pointer(&svq->next_guest_avail_elem);
 +    if (next_avail_elem) {
 +        virtqueue_detach_element(svq->vq, next_avail_elem, 0);
 +    }
 +    svq->vq = NULL;
 +    g_free(svq->ring_id_maps);
 +    qemu_vfree(svq->vring.desc);
 +    qemu_vfree(svq->vring.used);
  }
  /**
-  * Create a port on a given hub
+diff --git a/hw/virtio/vhost-shadow-virtqueue.h b/hw/virtio/vhost-shadow-virtqueue.h
-+ * @hub_id: Number of the hub
+index XXXXXXX..XXXXXXX 100644
-  * @name: Net client name or NULL for default name.
+--- a/hw/virtio/vhost-shadow-virtqueue.h
-+ * @hubpeer: Peer to use (if "netdev=id" has been specified)
++++ b/hw/virtio/vhost-shadow-virtqueue.h
-  *
+@@ -XXX,XX +XXX,XX @@ typedef struct VhostShadowVirtqueue {
-  * If there is no existing hub with the given id then a new hub is created.
      /* Guest's call notifier, where the SVQ calls guest. */
      EventNotifier svq_call;
 +
 +    /* Virtio queue shadowing */
 +    VirtQueue *vq;
 +
 +    /* Virtio device */
 +    VirtIODevice *vdev;
 +
 +    /* Map for use the guest's descriptors */
 +    VirtQueueElement **ring_id_maps;
 +
 +    /* Next VirtQueue element that guest made available */
 +    VirtQueueElement *next_guest_avail_elem;
 +
 +    /* Next head to expose to the device */
 +    uint16_t shadow_avail_idx;
 +
 +    /* Next free descriptor */
 +    uint16_t free_head;
 +
 +    /* Last seen used idx */
 +    uint16_t shadow_used_idx;
 +
 +    /* Next head to consume from the device */
 +    uint16_t last_used_idx;
  } VhostShadowVirtqueue;
  bool vhost_svq_valid_features(uint64_t features, Error **errp);
@@ -XXX,XX +XXX,XX @@ void vhost_svq_get_vring_addr(const VhostShadowVirtqueue *svq,
  size_t vhost_svq_driver_area_size(const VhostShadowVirtqueue *svq);
  size_t vhost_svq_device_area_size(const VhostShadowVirtqueue *svq);
 +void vhost_svq_start(VhostShadowVirtqueue *svq, VirtIODevice *vdev,
 +                     VirtQueue *vq);
  void vhost_svq_stop(VhostShadowVirtqueue *svq);
  VhostShadowVirtqueue *vhost_svq_new(void);
 diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c
 index XXXXXXX..XXXXXXX 100644
 --- a/hw/virtio/vhost-vdpa.c
 +++ b/hw/virtio/vhost-vdpa.c
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_vring_dev_addr(struct vhost_dev *dev,
   * Note that this function does not rewind kick file descriptor if cannot set
   * call one.
   */
--NetClientState *net_hub_add_port(int hub_id, const char *name)
+-static bool vhost_vdpa_svq_setup(struct vhost_dev *dev,
-+NetClientState *net_hub_add_port(int hub_id, const char *name,
+-                                 VhostShadowVirtqueue *svq,
-+                                 NetClientState *hubpeer)
+-                                 unsigned idx,
 -                                 Error **errp)
 +static int vhost_vdpa_svq_set_fds(struct vhost_dev *dev,
 +                                  VhostShadowVirtqueue *svq,
 +                                  unsigned idx,
 +                                  Error **errp)
  {
-     NetHub *hub;
+     struct vhost_vring_file file = {
-     NetHubPort *port;
+         .index = dev->vq_index + idx,
-@@ -XXX,XX +XXX,XX @@ NetClientState *net_hub_add_port(int hub_id, const char *name)
+@@ -XXX,XX +XXX,XX @@ static bool vhost_vdpa_svq_setup(struct vhost_dev *dev,
-         hub = net_hub_new(hub_id);
+     r = vhost_vdpa_set_vring_dev_kick(dev, &file);
      if (unlikely(r != 0)) {
          error_setg_errno(errp, -r, "Can't set device kick fd");
 -        return false;
 +        return r;
      }
--    port = net_hub_port_new(hub, name);
+     event_notifier = &svq->hdev_call;
-+    port = net_hub_port_new(hub, name, hubpeer);
+@@ -XXX,XX +XXX,XX @@ static bool vhost_vdpa_svq_setup(struct vhost_dev *dev,
-     return &port->nc;
+         error_setg_errno(errp, -r, "Can't set device call fd");
      }
 +    return r;
 +}
 +
 +/**
 + * Unmap a SVQ area in the device
 + */
 +static bool vhost_vdpa_svq_unmap_ring(struct vhost_vdpa *v, hwaddr iova,
 +                                      hwaddr size)
 +{
 +    int r;
 +
 +    size = ROUND_UP(size, qemu_real_host_page_size);
 +    r = vhost_vdpa_dma_unmap(v, iova, size);
 +    return r == 0;
 +}
 +
 +static bool vhost_vdpa_svq_unmap_rings(struct vhost_dev *dev,
 +                                       const VhostShadowVirtqueue *svq)
 +{
 +    struct vhost_vdpa *v = dev->opaque;
 +    struct vhost_vring_addr svq_addr;
 +    size_t device_size = vhost_svq_device_area_size(svq);
 +    size_t driver_size = vhost_svq_driver_area_size(svq);
 +    bool ok;
 +
 +    vhost_svq_get_vring_addr(svq, &svq_addr);
 +
 +    ok = vhost_vdpa_svq_unmap_ring(v, svq_addr.desc_user_addr, driver_size);
 +    if (unlikely(!ok)) {
 +        return false;
 +    }
 +
 +    return vhost_vdpa_svq_unmap_ring(v, svq_addr.used_user_addr, device_size);
 +}
 +
 +/**
 + * Map the shadow virtqueue rings in the device
 + *
 + * @dev: The vhost device
 + * @svq: The shadow virtqueue
 + * @addr: Assigned IOVA addresses
 + * @errp: Error pointer
 + */
 +static bool vhost_vdpa_svq_map_rings(struct vhost_dev *dev,
 +                                     const VhostShadowVirtqueue *svq,
 +                                     struct vhost_vring_addr *addr,
 +                                     Error **errp)
 +{
 +    struct vhost_vdpa *v = dev->opaque;
 +    size_t device_size = vhost_svq_device_area_size(svq);
 +    size_t driver_size = vhost_svq_driver_area_size(svq);
 +    int r;
 +
 +    ERRP_GUARD();
 +    vhost_svq_get_vring_addr(svq, addr);
 +
 +    r = vhost_vdpa_dma_map(v, addr->desc_user_addr, driver_size,
 +                           (void *)addr->desc_user_addr, true);
 +    if (unlikely(r != 0)) {
 +        error_setg_errno(errp, -r, "Cannot create vq driver region: ");
 +        return false;
 +    }
 +
 +    r = vhost_vdpa_dma_map(v, addr->used_user_addr, device_size,
 +                           (void *)addr->used_user_addr, false);
 +    if (unlikely(r != 0)) {
 +        error_setg_errno(errp, -r, "Cannot create vq device region: ");
 +    }
 +
 +    return r == 0;
 +}
 +
 +static bool vhost_vdpa_svq_setup(struct vhost_dev *dev,
 +                                 VhostShadowVirtqueue *svq,
 +                                 unsigned idx,
 +                                 Error **errp)
 +{
 +    uint16_t vq_index = dev->vq_index + idx;
 +    struct vhost_vring_state s = {
 +        .index = vq_index,
 +    };
 +    int r;
 +
 +    r = vhost_vdpa_set_dev_vring_base(dev, &s);
 +    if (unlikely(r)) {
 +        error_setg_errno(errp, -r, "Cannot set vring base");
 +        return false;
 +    }
 +
 +    r = vhost_vdpa_svq_set_fds(dev, svq, idx, errp);
      return r == 0;
  }
-@@ -XXX,XX +XXX,XX @@ NetClientState *net_hub_port_find(int hub_id)
+@@ -XXX,XX +XXX,XX @@ static bool vhost_vdpa_svqs_start(struct vhost_dev *dev)
      }
      for (i = 0; i < v->shadow_vqs->len; ++i) {
 +        VirtQueue *vq = virtio_get_queue(dev->vdev, dev->vq_index + i);
          VhostShadowVirtqueue *svq = g_ptr_array_index(v->shadow_vqs, i);
 +        struct vhost_vring_addr addr = {
 +            .index = i,
 +        };
 +        int r;
          bool ok = vhost_vdpa_svq_setup(dev, svq, i, &err);
          if (unlikely(!ok)) {
 -            error_reportf_err(err, "Cannot setup SVQ %u: ", i);
 +            goto err;
 +        }
 +
 +        vhost_svq_start(svq, dev->vdev, vq);
 +        ok = vhost_vdpa_svq_map_rings(dev, svq, &addr, &err);
 +        if (unlikely(!ok)) {
 +            goto err_map;
 +        }
 +
 +        /* Override vring GPA set by vhost subsystem */
 +        r = vhost_vdpa_set_vring_dev_addr(dev, &addr);
 +        if (unlikely(r != 0)) {
 +            error_setg_errno(&err, -r, "Cannot set device address");
 +            goto err_set_addr;
 +        }
 +    }
 +
 +    return true;
 +
 +err_set_addr:
 +    vhost_vdpa_svq_unmap_rings(dev, g_ptr_array_index(v->shadow_vqs, i));
 +
 +err_map:
 +    vhost_svq_stop(g_ptr_array_index(v->shadow_vqs, i));
 +
 +err:
 +    error_reportf_err(err, "Cannot setup SVQ %u: ", i);
 +    for (unsigned j = 0; j < i; ++j) {
 +        VhostShadowVirtqueue *svq = g_ptr_array_index(v->shadow_vqs, j);
 +        vhost_vdpa_svq_unmap_rings(dev, svq);
 +        vhost_svq_stop(svq);
 +    }
 +
 +    return false;
 +}
 +
 +static bool vhost_vdpa_svqs_stop(struct vhost_dev *dev)
 +{
 +    struct vhost_vdpa *v = dev->opaque;
 +
 +    if (!v->shadow_vqs) {
 +        return true;
 +    }
 +
 +    for (unsigned i = 0; i < v->shadow_vqs->len; ++i) {
 +        VhostShadowVirtqueue *svq = g_ptr_array_index(v->shadow_vqs,
 +                                                      i);
 +        bool ok = vhost_vdpa_svq_unmap_rings(dev, svq);
 +        if (unlikely(!ok)) {
              return false;
          }
      }
+@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_dev_start(struct vhost_dev *dev, bool started)
--    nc = net_hub_add_port(hub_id, NULL);
+         }
-+    nc = net_hub_add_port(hub_id, NULL, NULL);
+         vhost_vdpa_set_vring_ready(dev);
-     return nc;
+     } else {
- }
++        ok = vhost_vdpa_svqs_stop(dev);
++        if (unlikely(!ok)) {
@@ -XXX,XX +XXX,XX @@ int net_init_hubport(const Netdev *netdev, const char *name,
                       NetClientState *peer, Error **errp)
  {
      const NetdevHubPortOptions *hubport;
 +    NetClientState *hubpeer = NULL;
      assert(netdev->type == NET_CLIENT_DRIVER_HUBPORT);
      assert(!peer);
      hubport = &netdev->u.hubport;
 -    net_hub_add_port(hubport->hubid, name);
 +    if (hubport->has_netdev) {
 +        hubpeer = qemu_find_netdev(hubport->netdev);
 +        if (!hubpeer) {
 +            error_setg(errp, "netdev '%s' not found", hubport->netdev);
 +            return -1;
 +        }
-+    }
+         vhost_vdpa_host_notifiers_uninit(dev, dev->nvqs);
-+
+     }
 +    net_hub_add_port(hubport->hubid, name, hubpeer);
 +
      return 0;
  }
 diff --git a/net/hub.h b/net/hub.h
 index XXXXXXX..XXXXXXX 100644
 --- a/net/hub.h
 +++ b/net/hub.h
@@ -XXX,XX +XXX,XX @@
  #include "qemu-common.h"
 -NetClientState *net_hub_add_port(int hub_id, const char *name);
 +NetClientState *net_hub_add_port(int hub_id, const char *name,
 +                                 NetClientState *hubpeer);
  NetClientState *net_hub_find_client_by_name(int hub_id, const char *name);
  void net_hub_info(Monitor *mon);
  void net_hub_check_clients(void);
 diff --git a/net/net.c b/net/net.c
 index XXXXXXX..XXXXXXX 100644
 --- a/net/net.c
 +++ b/net/net.c
@@ -XXX,XX +XXX,XX @@ static int net_client_init1(const void *object, bool is_netdev, Error **errp)
          /* Do not add to a vlan if it's a nic with a netdev= parameter. */
          if (netdev->type != NET_CLIENT_DRIVER_NIC ||
              !opts->u.nic.has_netdev) {
 -            peer = net_hub_add_port(net->has_vlan ? net->vlan : 0, NULL);
 +            peer = net_hub_add_port(net->has_vlan ? net->vlan : 0, NULL, NULL);
          }
          if (net->has_vlan && !vlan_warned) {
 diff --git a/qapi/net.json b/qapi/net.json
 index XXXXXXX..XXXXXXX 100644
 --- a/qapi/net.json
 +++ b/qapi/net.json
@@ -XXX,XX +XXX,XX @@
  # Connect two or more net clients through a software hub.
  #
  # @hubid: hub identifier number
 +# @netdev: used to connect hub to a netdev instead of a device (since 2.12)
  #
  # Since: 1.2
  ##
  { 'struct': 'NetdevHubPortOptions',
    'data': {
 -    'hubid':     'int32' } }
 +    'hubid':     'int32',
 +    '*netdev':    'str' } }
  ##
  # @NetdevNetmapOptions:
 diff --git a/qemu-options.hx b/qemu-options.hx
 index XXXXXXX..XXXXXXX 100644
 --- a/qemu-options.hx
 +++ b/qemu-options.hx
@@ -XXX,XX +XXX,XX @@ DEF("netdev", HAS_ARG, QEMU_OPTION_netdev,
  #endif
      "-netdev vhost-user,id=str,chardev=dev[,vhostforce=on|off]\n"
      "                configure a vhost-user network, backed by a chardev 'dev'\n"
 -    "-netdev hubport,id=str,hubid=n\n"
 +    "-netdev hubport,id=str,hubid=n[,netdev=nd]\n"
      "                configure a hub port on QEMU VLAN 'n'\n", QEMU_ARCH_ALL)
  DEF("net", HAS_ARG, QEMU_OPTION_net,
      "-net nic[,vlan=n][,netdev=nd][,macaddr=mac][,model=type][,name=str][,addr=str][,vectors=v]\n"
@@ -XXX,XX +XXX,XX @@ vde_switch -F -sock /tmp/myswitch
  qemu-system-i386 linux.img -net nic -net vde,sock=/tmp/myswitch
  @end example
 -@item -netdev hubport,id=@var{id},hubid=@var{hubid}
 +@item -netdev hubport,id=@var{id},hubid=@var{hubid}[,netdev=@var{nd}]
  Create a hub port on QEMU "vlan" @var{hubid}.
  The hubport netdev lets you connect a NIC to a QEMU "vlan" instead of a single
  netdev.  @code{-net} and @code{-device} with parameter @option{vlan} create the
 -required hub automatically.
 +required hub automatically. Alternatively, you can also connect the hubport
 +to another netdev with ID @var{nd} by using the @option{netdev=@var{nd}}
 +option.
  @item -netdev vhost-user,chardev=@var{id}[,vhostforce=on|off][,queues=n]
 --
 .7.4

-New patch
+[PULL 09/15] util: Add iova_tree_alloc_map
+From: Eugenio Pérez <eperezma@redhat.com>
 This iova tree function allows it to look for a hole in allocated
 regions and return a totally new translation for a given translated
 address.
 It's usage is mainly to allow devices to access qemu address space,
 remapping guest's one into a new iova space where qemu can add chunks of
 addresses.
 Acked-by: Michael S. Tsirkin <mst@redhat.com>
 Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
 Reviewed-by: Peter Xu <peterx@redhat.com>
 Signed-off-by: Jason Wang <jasowang@redhat.com>
 ---
  include/qemu/iova-tree.h |  18 +++++++
  util/iova-tree.c         | 135 +++++++++++++++++++++++++++++++++++++++++++++++
 files changed, 153 insertions(+)
 diff --git a/include/qemu/iova-tree.h b/include/qemu/iova-tree.h
 index XXXXXXX..XXXXXXX 100644
 --- a/include/qemu/iova-tree.h
 +++ b/include/qemu/iova-tree.h
@@ -XXX,XX +XXX,XX @@
  #define  IOVA_OK           (0)
  #define  IOVA_ERR_INVALID  (-1) /* Invalid parameters */
  #define  IOVA_ERR_OVERLAP  (-2) /* IOVA range overlapped */
 +#define  IOVA_ERR_NOMEM    (-3) /* Cannot allocate */
  typedef struct IOVATree IOVATree;
  typedef struct DMAMap {
@@ -XXX,XX +XXX,XX @@ const DMAMap *iova_tree_find_address(const IOVATree *tree, hwaddr iova);
  void iova_tree_foreach(IOVATree *tree, iova_tree_iterator iterator);
  /**
 + * iova_tree_alloc_map:
 + *
 + * @tree: the iova tree to allocate from
 + * @map: the new map (as translated addr & size) to allocate in the iova region
 + * @iova_begin: the minimum address of the allocation
 + * @iova_end: the maximum addressable direction of the allocation
 + *
 + * Allocates a new region of a given size, between iova_min and iova_max.
 + *
 + * Return: Same as iova_tree_insert, but cannot overlap and can return error if
 + * iova tree is out of free contiguous range. The caller gets the assigned iova
 + * in map->iova.
 + */
 +int iova_tree_alloc_map(IOVATree *tree, DMAMap *map, hwaddr iova_begin,
 +                        hwaddr iova_end);
 +
 +/**
   * iova_tree_destroy:
   *
   * @tree: the iova tree to destroy
 diff --git a/util/iova-tree.c b/util/iova-tree.c
 index XXXXXXX..XXXXXXX 100644
 --- a/util/iova-tree.c
 +++ b/util/iova-tree.c
@@ -XXX,XX +XXX,XX @@ struct IOVATree {
      GTree *tree;
  };
 +/* Args to pass to iova_tree_alloc foreach function. */
 +struct IOVATreeAllocArgs {
 +    /* Size of the desired allocation */
 +    size_t new_size;
 +
 +    /* The minimum address allowed in the allocation */
 +    hwaddr iova_begin;
 +
 +    /* Map at the left of the hole, can be NULL if "this" is first one */
 +    const DMAMap *prev;
 +
 +    /* Map at the right of the hole, can be NULL if "prev" is the last one */
 +    const DMAMap *this;
 +
 +    /* If found, we fill in the IOVA here */
 +    hwaddr iova_result;
 +
 +    /* Whether have we found a valid IOVA */
 +    bool iova_found;
 +};
 +
 +/**
 + * Iterate args to the next hole
 + *
 + * @args: The alloc arguments
 + * @next: The next mapping in the tree. Can be NULL to signal the last one
 + */
 +static void iova_tree_alloc_args_iterate(struct IOVATreeAllocArgs *args,
 +                                         const DMAMap *next) {
 +    args->prev = args->this;
 +    args->this = next;
 +}
 +
  static int iova_tree_compare(gconstpointer a, gconstpointer b, gpointer data)
  {
      const DMAMap *m1 = a, *m2 = b;
@@ -XXX,XX +XXX,XX @@ int iova_tree_remove(IOVATree *tree, const DMAMap *map)
      return IOVA_OK;
  }
 +/**
 + * Try to find an unallocated IOVA range between prev and this elements.
 + *
 + * @args: Arguments to allocation
 + *
 + * Cases:
 + *
 + * (1) !prev, !this: No entries allocated, always succeed
 + *
 + * (2) !prev, this: We're iterating at the 1st element.
 + *
 + * (3) prev, !this: We're iterating at the last element.
 + *
 + * (4) prev, this: this is the most common case, we'll try to find a hole
 + * between "prev" and "this" mapping.
 + *
 + * Note that this function assumes the last valid iova is HWADDR_MAX, but it
 + * searches linearly so it's easy to discard the result if it's not the case.
 + */
 +static void iova_tree_alloc_map_in_hole(struct IOVATreeAllocArgs *args)
 +{
 +    const DMAMap *prev = args->prev, *this = args->this;
 +    uint64_t hole_start, hole_last;
 +
 +    if (this && this->iova + this->size < args->iova_begin) {
 +        return;
 +    }
 +
 +    hole_start = MAX(prev ? prev->iova + prev->size + 1 : 0, args->iova_begin);
 +    hole_last = this ? this->iova : HWADDR_MAX;
 +
 +    if (hole_last - hole_start > args->new_size) {
 +        args->iova_result = hole_start;
 +        args->iova_found = true;
 +    }
 +}
 +
 +/**
 + * Foreach dma node in the tree, compare if there is a hole with its previous
 + * node (or minimum iova address allowed) and the node.
 + *
 + * @key: Node iterating
 + * @value: Node iterating
 + * @pargs: Struct to communicate with the outside world
 + *
 + * Return: false to keep iterating, true if needs break.
 + */
 +static gboolean iova_tree_alloc_traverse(gpointer key, gpointer value,
 +                                         gpointer pargs)
 +{
 +    struct IOVATreeAllocArgs *args = pargs;
 +    DMAMap *node = value;
 +
 +    assert(key == value);
 +
 +    iova_tree_alloc_args_iterate(args, node);
 +    iova_tree_alloc_map_in_hole(args);
 +    return args->iova_found;
 +}
 +
 +int iova_tree_alloc_map(IOVATree *tree, DMAMap *map, hwaddr iova_begin,
 +                        hwaddr iova_last)
 +{
 +    struct IOVATreeAllocArgs args = {
 +        .new_size = map->size,
 +        .iova_begin = iova_begin,
 +    };
 +
 +    if (unlikely(iova_last < iova_begin)) {
 +        return IOVA_ERR_INVALID;
 +    }
 +
 +    /*
 +     * Find a valid hole for the mapping
 +     *
 +     * Assuming low iova_begin, so no need to do a binary search to
 +     * locate the first node.
 +     *
 +     * TODO: Replace all this with g_tree_node_first/next/last when available
 +     * (from glib since 2.68). To do it with g_tree_foreach complicates the
 +     * code a lot.
 +     *
 +     */
 +    g_tree_foreach(tree->tree, iova_tree_alloc_traverse, &args);
 +    if (!args.iova_found) {
 +        /*
 +         * Either tree is empty or the last hole is still not checked.
 +         * g_tree_foreach does not compare (last, iova_last] range, so we check
 +         * it here.
 +         */
 +        iova_tree_alloc_args_iterate(&args, NULL);
 +        iova_tree_alloc_map_in_hole(&args);
 +    }
 +
 +    if (!args.iova_found || args.iova_result + map->size > iova_last) {
 +        return IOVA_ERR_NOMEM;
 +    }
 +
 +    map->iova = args.iova_result;
 +    return iova_tree_insert(tree, map);
 +}
 +
  void iova_tree_destroy(IOVATree *tree)
  {
      g_tree_destroy(tree->tree);
 --
 .7.4

-New patch
+[PULL 10/15] util: add iova_tree_find_iova
+From: Eugenio Pérez <eperezma@redhat.com>
+This function does the reverse operation of iova_tree_find: To look for
+a mapping that match a translated address so we can do the reverse.
+This have linear complexity instead of logarithmic, but it supports
+overlapping HVA. Future developments could reduce it.
+Acked-by: Michael S. Tsirkin <mst@redhat.com>
+Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
+Signed-off-by: Jason Wang <jasowang@redhat.com>
+---
+ include/qemu/iova-tree.h | 20 +++++++++++++++++++-
+ util/iova-tree.c         | 34 ++++++++++++++++++++++++++++++++++
+files changed, 53 insertions(+), 1 deletion(-)
+diff --git a/include/qemu/iova-tree.h b/include/qemu/iova-tree.h
+index XXXXXXX..XXXXXXX 100644
+--- a/include/qemu/iova-tree.h
++++ b/include/qemu/iova-tree.h
+@@ -XXX,XX +XXX,XX @@ int iova_tree_remove(IOVATree *tree, const DMAMap *map);
+  * @tree: the iova tree to search from
+  * @map: the mapping to search
+  *
+- * Search for a mapping in the iova tree that overlaps with the
++ * Search for a mapping in the iova tree that iova overlaps with the
+  * mapping range specified.  Only the first found mapping will be
+  * returned.
+  *
+@@ -XXX,XX +XXX,XX @@ int iova_tree_remove(IOVATree *tree, const DMAMap *map);
+ const DMAMap *iova_tree_find(const IOVATree *tree, const DMAMap *map);
+ /**
++ * iova_tree_find_iova:
++ *
++ * @tree: the iova tree to search from
++ * @map: the mapping to search
++ *
++ * Search for a mapping in the iova tree that translated_addr overlaps with the
++ * mapping range specified.  Only the first found mapping will be
++ * returned.
++ *
++ * Return: DMAMap pointer if found, or NULL if not found.  Note that
++ * the returned DMAMap pointer is maintained internally.  User should
++ * only read the content but never modify or free the content.  Also,
++ * user is responsible to make sure the pointer is valid (say, no
++ * concurrent deletion in progress).
++ */
++const DMAMap *iova_tree_find_iova(const IOVATree *tree, const DMAMap *map);
++
++/**
+  * iova_tree_find_address:
+  *
+  * @tree: the iova tree to search from
+diff --git a/util/iova-tree.c b/util/iova-tree.c
+index XXXXXXX..XXXXXXX 100644
+--- a/util/iova-tree.c
++++ b/util/iova-tree.c
+@@ -XXX,XX +XXX,XX @@ struct IOVATreeAllocArgs {
+     bool iova_found;
+ };
++typedef struct IOVATreeFindIOVAArgs {
++    const DMAMap *needle;
++    const DMAMap *result;
++} IOVATreeFindIOVAArgs;
++
+ /**
+  * Iterate args to the next hole
+  *
+@@ -XXX,XX +XXX,XX @@ const DMAMap *iova_tree_find(const IOVATree *tree, const DMAMap *map)
+     return g_tree_lookup(tree->tree, map);
+ }
++static gboolean iova_tree_find_address_iterator(gpointer key, gpointer value,
++                                                gpointer data)
++{
++    const DMAMap *map = key;
++    IOVATreeFindIOVAArgs *args = data;
++    const DMAMap *needle;
++
++    g_assert(key == value);
++
++    needle = args->needle;
++    if (map->translated_addr + map->size < needle->translated_addr ||
++        needle->translated_addr + needle->size < map->translated_addr) {
++        return false;
++    }
++
++    args->result = map;
++    return true;
++}
++
++const DMAMap *iova_tree_find_iova(const IOVATree *tree, const DMAMap *map)
++{
++    IOVATreeFindIOVAArgs args = {
++        .needle = map,
++    };
++
++    g_tree_foreach(tree->tree, iova_tree_find_address_iterator, &args);
++    return args.result;
++}
++
+ const DMAMap *iova_tree_find_address(const IOVATree *tree, hwaddr iova)
+ {
+     const DMAMap map = { .iova = iova, .size = 0 };
+--
+.7.4

-[Qemu-devel] [PULL 2/6] colo: compare the packet based on the tcp sequence number
+[PULL 11/15] vhost: Add VhostIOVATree
-From: Mao Zhongyi <maozy.fnst@cn.fujitsu.com>
+From: Eugenio Pérez <eperezma@redhat.com>
-Packet size some time different or when network is busy.
+This tree is able to look for a translated address from an IOVA address.
 Based on same payload size, but TCP protocol can not
 guarantee send the same one packet in the same way,
-like that:
+At first glance it is similar to util/iova-tree. However, SVQ working on
-We send this payload:
+devices with limited IOVA space need more capabilities, like allocating
-------------------------------
+IOVA chunks or performing reverse translations (qemu addresses to iova).
 | header |1|2|3|4|5|6|7|8|9|0|
 ------------------------------
-primary:
+The allocation capability, as "assign a free IOVA address to this chunk
-ppkt1:
+of memory in qemu's address space" allows shadow virtqueue to create a
-----------------
+new address space that is not restricted by guest's addressable one, so
-| header |1|2|3|
+we can allocate shadow vqs vrings outside of it.
 ----------------
 ppkt2:
 ------------------------
 | header |4|5|6|7|8|9|0|
 ------------------------
-secondary:
+It duplicates the tree so it can search efficiently in both directions,
-spkt1:
+and it will signal overlap if iova or the translated address is present
-------------------------------
+in any tree.
 | header |1|2|3|4|5|6|7|8|9|0|
 ------------------------------
-In the original method, ppkt1 and ppkt2 are different in size and
+Acked-by: Michael S. Tsirkin <mst@redhat.com>
-spkt1, so they can't compare and trigger the checkpoint.
+Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
 I have tested FTP get 200M and 1G file many times, I found that
 the performance was less than 1% of the native.
 Now I reconstructed the comparison of TCP packets based on the
 TCP sequence number. first of all, ppkt1 and spkt1 have the same
 starting sequence number, so they can compare, even though their
 length is different. And then ppkt1 with a smaller payload length
 is used as the comparison length, if the payload is same, send
 out the ppkt1 and record the offset(the length of ppkt1 payload)
 in spkt1. The next comparison, ppkt2 and spkt1 can be compared
 from the recorded position of spkt1.
 like that:
 ----------------
 | header |1|2|3| ppkt1
 ---------|-----|
          |     |
 ---------v-----v--------------
 | header |1|2|3|4|5|6|7|8|9|0| spkt1
 ---------------|\------------|
                | \offset     |
       ---------v-------------v
       | header |4|5|6|7|8|9|0| ppkt2
       ------------------------
 In this way, the performance can reach native 20% in my multiple
 tests.
 Cc: Zhang Chen <zhangckid@gmail.com>
 Cc: Li Zhijian <lizhijian@cn.fujitsu.com>
 Cc: Jason Wang <jasowang@redhat.com>
 Signed-off-by: Mao Zhongyi <maozy.fnst@cn.fujitsu.com>
 Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com>
 Signed-off-by: Zhang Chen <zhangckid@gmail.com>
 Reviewed-by: Zhang Chen <zhangckid@gmail.com>
 Tested-by: Zhang Chen <zhangckid@gmail.com>
 Signed-off-by: Jason Wang <jasowang@redhat.com>
 ---
- net/colo-compare.c | 343 +++++++++++++++++++++++++++++++++++------------------
+ hw/virtio/meson.build       |   2 +-
- net/colo.c         |   9 ++
+ hw/virtio/vhost-iova-tree.c | 110 ++++++++++++++++++++++++++++++++++++++++++++
- net/colo.h         |  15 +++
+ hw/virtio/vhost-iova-tree.h |  27 +++++++++++
- net/trace-events   |   2 +-
+files changed, 138 insertions(+), 1 deletion(-)
-files changed, 250 insertions(+), 119 deletions(-)
+ create mode 100644 hw/virtio/vhost-iova-tree.c
  create mode 100644 hw/virtio/vhost-iova-tree.h
-diff --git a/net/colo-compare.c b/net/colo-compare.c
+diff --git a/hw/virtio/meson.build b/hw/virtio/meson.build
 index XXXXXXX..XXXXXXX 100644
---- a/net/colo-compare.c
+--- a/hw/virtio/meson.build
-+++ b/net/colo-compare.c
++++ b/hw/virtio/meson.build
@@ -XXX,XX +XXX,XX @@ softmmu_ss.add(when: 'CONFIG_ALL', if_true: files('vhost-stub.c'))
  virtio_ss = ss.source_set()
  virtio_ss.add(files('virtio.c'))
 -virtio_ss.add(when: 'CONFIG_VHOST', if_true: files('vhost.c', 'vhost-backend.c', 'vhost-shadow-virtqueue.c'))
 +virtio_ss.add(when: 'CONFIG_VHOST', if_true: files('vhost.c', 'vhost-backend.c', 'vhost-shadow-virtqueue.c', 'vhost-iova-tree.c'))
  virtio_ss.add(when: 'CONFIG_VHOST_USER', if_true: files('vhost-user.c'))
  virtio_ss.add(when: 'CONFIG_VHOST_VDPA', if_true: files('vhost-vdpa.c'))
  virtio_ss.add(when: 'CONFIG_VIRTIO_BALLOON', if_true: files('virtio-balloon.c'))
 diff --git a/hw/virtio/vhost-iova-tree.c b/hw/virtio/vhost-iova-tree.c
 new file mode 100644
 index XXXXXXX..XXXXXXX
 --- /dev/null
 +++ b/hw/virtio/vhost-iova-tree.c
 @@ -XXX,XX +XXX,XX @@
- #define COMPARE_READ_LEN_MAX NET_BUFSIZE
++/*
- #define MAX_QUEUE_SIZE 1024
++ * vhost software live migration iova tree
++ *
-+#define COLO_COMPARE_FREE_PRIMARY     0x01
++ * SPDX-FileCopyrightText: Red Hat, Inc. 2021
-+#define COLO_COMPARE_FREE_SECONDARY   0x02
++ * SPDX-FileContributor: Author: Eugenio Pérez <eperezma@redhat.com>
 + *
 + * SPDX-License-Identifier: GPL-2.0-or-later
 + */
 +
- /* TODO: Should be configurable */
++#include "qemu/osdep.h"
- #define REGULAR_PACKET_CHECK_MS 3000
++#include "qemu/iova-tree.h"
++#include "vhost-iova-tree.h"
-@@ -XXX,XX +XXX,XX @@ static gint seq_sorter(Packet *a, Packet *b, gpointer data)
++
-     return ntohl(atcp->th_seq) - ntohl(btcp->th_seq);
++#define iova_min_addr qemu_real_host_page_size
- }
++
++/**
-+static void fill_pkt_tcp_info(void *data, uint32_t *max_ack)
++ * VhostIOVATree, able to:
 + * - Translate iova address
 + * - Reverse translate iova address (from translated to iova)
 + * - Allocate IOVA regions for translated range (linear operation)
 + */
 +struct VhostIOVATree {
 +    /* First addressable iova address in the device */
 +    uint64_t iova_first;
 +
 +    /* Last addressable iova address in the device */
 +    uint64_t iova_last;
 +
 +    /* IOVA address to qemu memory maps. */
 +    IOVATree *iova_taddr_map;
 +};
 +
 +/**
 + * Create a new IOVA tree
 + *
 + * Returns the new IOVA tree
 + */
 +VhostIOVATree *vhost_iova_tree_new(hwaddr iova_first, hwaddr iova_last)
 +{
-+    Packet *pkt = data;
++    VhostIOVATree *tree = g_new(VhostIOVATree, 1);
 +    struct tcphdr *tcphd;
 +
-+    tcphd = (struct tcphdr *)pkt->transport_header;
++    /* Some devices do not like 0 addresses */
 +    tree->iova_first = MAX(iova_first, iova_min_addr);
 +    tree->iova_last = iova_last;
 +
-+    pkt->tcp_seq = ntohl(tcphd->th_seq);
++    tree->iova_taddr_map = iova_tree_new();
-+    pkt->tcp_ack = ntohl(tcphd->th_ack);
++    return tree;
 +    *max_ack = *max_ack > pkt->tcp_ack ? *max_ack : pkt->tcp_ack;
 +    pkt->header_size = pkt->transport_header - (uint8_t *)pkt->data
 +                       + (tcphd->th_off << 2) - pkt->vnet_hdr_len;
 +    pkt->payload_size = pkt->size - pkt->header_size;
 +    pkt->seq_end = pkt->tcp_seq + pkt->payload_size;
 +    pkt->flags = tcphd->th_flags;
 +}
 +
- /*
++/**
-  * Return 1 on success, if return 0 means the
++ * Delete an iova tree
-  * packet will be dropped
++ */
-  */
++void vhost_iova_tree_delete(VhostIOVATree *iova_tree)
 -static int colo_insert_packet(GQueue *queue, Packet *pkt)
 +static int colo_insert_packet(GQueue *queue, Packet *pkt, uint32_t *max_ack)
  {
      if (g_queue_get_length(queue) <= MAX_QUEUE_SIZE) {
          if (pkt->ip->ip_p == IPPROTO_TCP) {
 +            fill_pkt_tcp_info(pkt, max_ack);
              g_queue_insert_sorted(queue,
                                    pkt,
                                    (GCompareDataFunc)seq_sorter,
@@ -XXX,XX +XXX,XX @@ static int packet_enqueue(CompareState *s, int mode, Connection **con)
      }
      if (mode == PRIMARY_IN) {
 -        if (!colo_insert_packet(&conn->primary_list, pkt)) {
 +        if (!colo_insert_packet(&conn->primary_list, pkt, &conn->pack)) {
              error_report("colo compare primary queue size too big,"
                           "drop packet");
          }
      } else {
 -        if (!colo_insert_packet(&conn->secondary_list, pkt)) {
 +        if (!colo_insert_packet(&conn->secondary_list, pkt, &conn->sack)) {
              error_report("colo compare secondary queue size too big,"
                           "drop packet");
          }
@@ -XXX,XX +XXX,XX @@ static int packet_enqueue(CompareState *s, int mode, Connection **con)
      return 0;
  }
 +static inline bool after(uint32_t seq1, uint32_t seq2)
 +{
-+        return (int32_t)(seq1 - seq2) > 0;
++    iova_tree_destroy(iova_tree->iova_taddr_map);
 +    g_free(iova_tree);
 +}
 +
-+static void colo_release_primary_pkt(CompareState *s, Packet *pkt)
++/**
 + * Find the IOVA address stored from a memory address
 + *
 + * @tree: The iova tree
 + * @map: The map with the memory address
 + *
 + * Return the stored mapping, or NULL if not found.
 + */
 +const DMAMap *vhost_iova_tree_find_iova(const VhostIOVATree *tree,
 +                                        const DMAMap *map)
 +{
-+    int ret;
++    return iova_tree_find_iova(tree->iova_taddr_map, map);
 +    ret = compare_chr_send(s,
 +                           pkt->data,
 +                           pkt->size,
 +                           pkt->vnet_hdr_len);
 +    if (ret < 0) {
 +        error_report("colo send primary packet failed");
 +    }
 +    trace_colo_compare_main("packet same and release packet");
 +    packet_destroy(pkt, NULL);
 +}
 +
- /*
++/**
-  * The IP packets sent by primary and secondary
++ * Allocate a new mapping
-  * will be compared in here
++ *
-@@ -XXX,XX +XXX,XX @@ static int colo_compare_packet_payload(Packet *ppkt,
++ * @tree: The iova tree
- }
++ * @map: The iova map
++ *
- /*
++ * Returns:
-- * Called from the compare thread on the primary
++ * - IOVA_OK if the map fits in the container
-- * for compare tcp packet
++ * - IOVA_ERR_INVALID if the map does not make sense (like size overflow)
-- * compare_tcp copied from Dr. David Alan Gilbert's branch
++ * - IOVA_ERR_NOMEM if tree cannot allocate more space.
-- */
++ *
--static int colo_packet_compare_tcp(Packet *spkt, Packet *ppkt)
++ * It returns assignated iova in map->iova if return value is VHOST_DMA_MAP_OK.
-+ * return true means that the payload is consist and
++ */
-+ * need to make the next comparison, false means do
++int vhost_iova_tree_map_alloc(VhostIOVATree *tree, DMAMap *map)
-+ * the checkpoint
++{
-+*/
++    /* Some vhost devices do not like addr 0. Skip first page */
-+static bool colo_mark_tcp_pkt(Packet *ppkt, Packet *spkt,
++    hwaddr iova_first = tree->iova_first ?: qemu_real_host_page_size;
 +                              int8_t *mark, uint32_t max_ack)
  {
 -    struct tcphdr *ptcp, *stcp;
 -    int res;
 +    *mark = 0;
 +
-+    if (ppkt->tcp_seq == spkt->tcp_seq && ppkt->seq_end == spkt->seq_end) {
++    if (map->translated_addr + map->size < map->translated_addr ||
-+        if (colo_compare_packet_payload(ppkt, spkt,
++        map->perm == IOMMU_NONE) {
-+                                        ppkt->header_size, spkt->header_size,
++        return IOVA_ERR_INVALID;
 +                                        ppkt->payload_size)) {
 +            *mark = COLO_COMPARE_FREE_SECONDARY | COLO_COMPARE_FREE_PRIMARY;
 +            return true;
 +        }
 +    }
 +    if (ppkt->tcp_seq == spkt->tcp_seq && ppkt->seq_end == spkt->seq_end) {
 +        if (colo_compare_packet_payload(ppkt, spkt,
 +                                        ppkt->header_size, spkt->header_size,
 +                                        ppkt->payload_size)) {
 +            *mark = COLO_COMPARE_FREE_SECONDARY | COLO_COMPARE_FREE_PRIMARY;
 +            return true;
 +        }
 +    }
 +
-+    /* one part of secondary packet payload still need to be compared */
++    /* Allocate a node in IOVA address */
-+    if (!after(ppkt->seq_end, spkt->seq_end)) {
++    return iova_tree_alloc_map(tree->iova_taddr_map, map, iova_first,
-+        if (colo_compare_packet_payload(ppkt, spkt,
++                               tree->iova_last);
 +                                        ppkt->header_size + ppkt->offset,
 +                                        spkt->header_size + spkt->offset,
 +                                        ppkt->payload_size - ppkt->offset)) {
 +            if (!after(ppkt->tcp_ack, max_ack)) {
 +                *mark = COLO_COMPARE_FREE_PRIMARY;
 +                spkt->offset += ppkt->payload_size - ppkt->offset;
 +                return true;
 +            } else {
 +                /* secondary guest hasn't ack the data, don't send
 +                 * out this packet
 +                 */
 +                return false;
 +            }
 +        }
 +    } else {
 +        /* primary packet is longer than secondary packet, compare
 +         * the same part and mark the primary packet offset
 +         */
 +        if (colo_compare_packet_payload(ppkt, spkt,
 +                                        ppkt->header_size + ppkt->offset,
 +                                        spkt->header_size + spkt->offset,
 +                                        spkt->payload_size - spkt->offset)) {
 +            *mark = COLO_COMPARE_FREE_SECONDARY;
 +            ppkt->offset += spkt->payload_size - spkt->offset;
 +            return true;
 +        }
 +    }
 -    trace_colo_compare_main("compare tcp");
 +    return false;
 +}
 -    ptcp = (struct tcphdr *)ppkt->transport_header;
 -    stcp = (struct tcphdr *)spkt->transport_header;
 +static void colo_compare_tcp(CompareState *s, Connection *conn)
 +{
 +    Packet *ppkt = NULL, *spkt = NULL;
 +    int8_t mark;
      /*
 -     * The 'identification' field in the IP header is *very* random
 -     * it almost never matches.  Fudge this by ignoring differences in
 -     * unfragmented packets; they'll normally sort themselves out if different
 -     * anyway, and it should recover at the TCP level.
 -     * An alternative would be to get both the primary and secondary to rewrite
 -     * somehow; but that would need some sync traffic to sync the state
 -     */
 -    if (ntohs(ppkt->ip->ip_off) & IP_DF) {
 -        spkt->ip->ip_id = ppkt->ip->ip_id;
 -        /* and the sum will be different if the IDs were different */
 -        spkt->ip->ip_sum = ppkt->ip->ip_sum;
 +     * If ppkt and spkt have the same payload, but ppkt's ACK
 +     * is greater than spkt's ACK, in this case we can not
 +     * send the ppkt because it will cause the secondary guest
 +     * to miss sending some data in the next. Therefore, we
 +     * record the maximum ACK in the current queue at both
 +     * primary side and secondary side. Only when the ack is
 +     * less than the smaller of the two maximum ack, then we
 +     * can ensure that the packet's payload is acknowledged by
 +     * primary and secondary.
 +    */
 +    uint32_t min_ack = conn->pack > conn->sack ? conn->sack : conn->pack;
 +
 +pri:
 +    if (g_queue_is_empty(&conn->primary_list)) {
 +        return;
      }
 +    ppkt = g_queue_pop_head(&conn->primary_list);
 +sec:
 +    if (g_queue_is_empty(&conn->secondary_list)) {
 +        g_queue_push_head(&conn->primary_list, ppkt);
 +        return;
 +    }
 +    spkt = g_queue_pop_head(&conn->secondary_list);
 -    /*
 -     * Check tcp header length for tcp option field.
 -     * th_off > 5 means this tcp packet have options field.
 -     * The tcp options maybe always different.
 -     * for example:
 -     * From RFC 7323.
 -     * TCP Timestamps option (TSopt):
 -     * Kind: 8
 -     *
 -     * Length: 10 bytes
 -     *
 -     *    +-------+-------+---------------------+---------------------+
 -     *    |Kind=8 |  10   |   TS Value (TSval)  |TS Echo Reply (TSecr)|
 -     *    +-------+-------+---------------------+---------------------+
 -     *       1       1              4                     4
 -     *
 -     * In this case the primary guest's timestamp always different with
 -     * the secondary guest's timestamp. COLO just focus on payload,
 -     * so we just need skip this field.
 -     */
 +    if (ppkt->tcp_seq == ppkt->seq_end) {
 +        colo_release_primary_pkt(s, ppkt);
 +        ppkt = NULL;
 +    }
 -    ptrdiff_t ptcp_offset, stcp_offset;
 +    if (ppkt && conn->compare_seq && !after(ppkt->seq_end, conn->compare_seq)) {
 +        trace_colo_compare_main("pri: this packet has compared");
 +        colo_release_primary_pkt(s, ppkt);
 +        ppkt = NULL;
 +    }
 -    ptcp_offset = ppkt->transport_header - (uint8_t *)ppkt->data
 -                  + (ptcp->th_off << 2) - ppkt->vnet_hdr_len;
 -    stcp_offset = spkt->transport_header - (uint8_t *)spkt->data
 -                  + (stcp->th_off << 2) - spkt->vnet_hdr_len;
 -    if (ppkt->size - ptcp_offset == spkt->size - stcp_offset) {
 -        res = colo_compare_packet_payload(ppkt, spkt,
 -                                          ptcp_offset, stcp_offset,
 -                                          ppkt->size - ptcp_offset);
 +    if (spkt->tcp_seq == spkt->seq_end) {
 +        packet_destroy(spkt, NULL);
 +        if (!ppkt) {
 +            goto pri;
 +        } else {
 +            goto sec;
 +        }
      } else {
 -        trace_colo_compare_main("TCP: payload size of packets are different");
 -        res = -1;
 +        if (conn->compare_seq && !after(spkt->seq_end, conn->compare_seq)) {
 +            trace_colo_compare_main("sec: this packet has compared");
 +            packet_destroy(spkt, NULL);
 +            if (!ppkt) {
 +                goto pri;
 +            } else {
 +                goto sec;
 +            }
 +        }
 +        if (!ppkt) {
 +            g_queue_push_head(&conn->secondary_list, spkt);
 +            goto pri;
 +        }
      }
 -    if (res != 0 &&
 -        trace_event_get_state_backends(TRACE_COLO_COMPARE_MISCOMPARE)) {
 -        char pri_ip_src[20], pri_ip_dst[20], sec_ip_src[20], sec_ip_dst[20];
 -
 -        strcpy(pri_ip_src, inet_ntoa(ppkt->ip->ip_src));
 -        strcpy(pri_ip_dst, inet_ntoa(ppkt->ip->ip_dst));
 -        strcpy(sec_ip_src, inet_ntoa(spkt->ip->ip_src));
 -        strcpy(sec_ip_dst, inet_ntoa(spkt->ip->ip_dst));
 -
 -        trace_colo_compare_ip_info(ppkt->size, pri_ip_src,
 -                                   pri_ip_dst, spkt->size,
 -                                   sec_ip_src, sec_ip_dst);
 -
 -        trace_colo_compare_tcp_info("pri tcp packet",
 -                                    ntohl(ptcp->th_seq),
 -                                    ntohl(ptcp->th_ack),
 -                                    res, ptcp->th_flags,
 -                                    ppkt->size);
 -
 -        trace_colo_compare_tcp_info("sec tcp packet",
 -                                    ntohl(stcp->th_seq),
 -                                    ntohl(stcp->th_ack),
 -                                    res, stcp->th_flags,
 -                                    spkt->size);
 +    if (colo_mark_tcp_pkt(ppkt, spkt, &mark, min_ack)) {
 +        trace_colo_compare_tcp_info("pri",
 +                                    ppkt->tcp_seq, ppkt->tcp_ack,
 +                                    ppkt->header_size, ppkt->payload_size,
 +                                    ppkt->offset, ppkt->flags);
 +
 +        trace_colo_compare_tcp_info("sec",
 +                                    spkt->tcp_seq, spkt->tcp_ack,
 +                                    spkt->header_size, spkt->payload_size,
 +                                    spkt->offset, spkt->flags);
 +
 +        if (mark == COLO_COMPARE_FREE_PRIMARY) {
 +            conn->compare_seq = ppkt->seq_end;
 +            colo_release_primary_pkt(s, ppkt);
 +            g_queue_push_head(&conn->secondary_list, spkt);
 +            goto pri;
 +        }
 +        if (mark == COLO_COMPARE_FREE_SECONDARY) {
 +            conn->compare_seq = spkt->seq_end;
 +            packet_destroy(spkt, NULL);
 +            goto sec;
 +        }
 +        if (mark == (COLO_COMPARE_FREE_PRIMARY | COLO_COMPARE_FREE_SECONDARY)) {
 +            conn->compare_seq = ppkt->seq_end;
 +            colo_release_primary_pkt(s, ppkt);
 +            packet_destroy(spkt, NULL);
 +            goto pri;
 +        }
 +    } else {
 +        g_queue_push_head(&conn->primary_list, ppkt);
 +        g_queue_push_head(&conn->secondary_list, spkt);
          qemu_hexdump((char *)ppkt->data, stderr,
                       "colo-compare ppkt", ppkt->size);
          qemu_hexdump((char *)spkt->data, stderr,
                       "colo-compare spkt", spkt->size);
 -    }
 -    return res;
 +        /*
 +         * colo_compare_inconsistent_notify();
 +         * TODO: notice to checkpoint();
 +         */
 +    }
  }
 +
  /*
   * Called from the compare thread on the primary
   * for compare udp packet
@@ -XXX,XX +XXX,XX @@ static void colo_old_packet_check(void *opaque)
                          (GCompareFunc)colo_old_packet_check_one_conn);
  }
 -/*
 - * Called from the compare thread on the primary
 - * for compare packet with secondary list of the
 - * specified connection when a new packet was
 - * queued to it.
 - */
 -static void colo_compare_connection(void *opaque, void *user_data)
 +static void colo_compare_packet(CompareState *s, Connection *conn,
 +                                int (*HandlePacket)(Packet *spkt,
 +                                Packet *ppkt))
  {
 -    CompareState *s = user_data;
 -    Connection *conn = opaque;
      Packet *pkt = NULL;
      GList *result = NULL;
 -    int ret;
      while (!g_queue_is_empty(&conn->primary_list) &&
             !g_queue_is_empty(&conn->secondary_list)) {
          pkt = g_queue_pop_head(&conn->primary_list);
 -        switch (conn->ip_proto) {
 -        case IPPROTO_TCP:
 -            result = g_queue_find_custom(&conn->secondary_list,
 -                     pkt, (GCompareFunc)colo_packet_compare_tcp);
 -            break;
 -        case IPPROTO_UDP:
 -            result = g_queue_find_custom(&conn->secondary_list,
 -                     pkt, (GCompareFunc)colo_packet_compare_udp);
 -            break;
 -        case IPPROTO_ICMP:
 -            result = g_queue_find_custom(&conn->secondary_list,
 -                     pkt, (GCompareFunc)colo_packet_compare_icmp);
 -            break;
 -        default:
 -            result = g_queue_find_custom(&conn->secondary_list,
 -                     pkt, (GCompareFunc)colo_packet_compare_other);
 -            break;
 -        }
 +        result = g_queue_find_custom(&conn->secondary_list,
 +                 pkt, (GCompareFunc)HandlePacket);
          if (result) {
 -            ret = compare_chr_send(s,
 -                                   pkt->data,
 -                                   pkt->size,
 -                                   pkt->vnet_hdr_len);
 -            if (ret < 0) {
 -                error_report("colo_send_primary_packet failed");
 -            }
 -            trace_colo_compare_main("packet same and release packet");
 +            colo_release_primary_pkt(s, pkt);
              g_queue_remove(&conn->secondary_list, result->data);
 -            packet_destroy(pkt, NULL);
          } else {
              /*
               * If one packet arrive late, the secondary_list or
@@ -XXX,XX +XXX,XX @@ static void colo_compare_connection(void *opaque, void *user_data)
      }
  }
 +/*
 + * Called from the compare thread on the primary
 + * for compare packet with secondary list of the
 + * specified connection when a new packet was
 + * queued to it.
 + */
 +static void colo_compare_connection(void *opaque, void *user_data)
 +{
 +    CompareState *s = user_data;
 +    Connection *conn = opaque;
 +
 +    switch (conn->ip_proto) {
 +    case IPPROTO_TCP:
 +        colo_compare_tcp(s, conn);
 +        break;
 +    case IPPROTO_UDP:
 +        colo_compare_packet(s, conn, colo_packet_compare_udp);
 +        break;
 +    case IPPROTO_ICMP:
 +        colo_compare_packet(s, conn, colo_packet_compare_icmp);
 +        break;
 +    default:
 +        colo_compare_packet(s, conn, colo_packet_compare_other);
 +        break;
 +    }
 +}
 +
- static int compare_chr_send(CompareState *s,
++/**
-                             const uint8_t *buf,
++ * Remove existing mappings from iova tree
-                             uint32_t size,
++ *
-diff --git a/net/colo.c b/net/colo.c
++ * @iova_tree: The vhost iova tree
-index XXXXXXX..XXXXXXX 100644
++ * @map: The map to remove
---- a/net/colo.c
++ */
-+++ b/net/colo.c
++void vhost_iova_tree_remove(VhostIOVATree *iova_tree, const DMAMap *map)
-@@ -XXX,XX +XXX,XX @@ Connection *connection_new(ConnectionKey *key)
++{
-     conn->processing = false;
++    iova_tree_remove(iova_tree->iova_taddr_map, map);
-     conn->offset = 0;
++}
-     conn->syn_flag = 0;
+diff --git a/hw/virtio/vhost-iova-tree.h b/hw/virtio/vhost-iova-tree.h
-+    conn->pack = 0;
+new file mode 100644
-+    conn->sack = 0;
+index XXXXXXX..XXXXXXX
-     g_queue_init(&conn->primary_list);
+--- /dev/null
-     g_queue_init(&conn->secondary_list);
++++ b/hw/virtio/vhost-iova-tree.h
+@@ -XXX,XX +XXX,XX @@
-@@ -XXX,XX +XXX,XX @@ Packet *packet_new(const void *data, int size, int vnet_hdr_len)
++/*
-     pkt->size = size;
++ * vhost software live migration iova tree
-     pkt->creation_ms = qemu_clock_get_ms(QEMU_CLOCK_HOST);
++ *
-     pkt->vnet_hdr_len = vnet_hdr_len;
++ * SPDX-FileCopyrightText: Red Hat, Inc. 2021
-+    pkt->tcp_seq = 0;
++ * SPDX-FileContributor: Author: Eugenio Pérez <eperezma@redhat.com>
-+    pkt->tcp_ack = 0;
++ *
-+    pkt->seq_end = 0;
++ * SPDX-License-Identifier: GPL-2.0-or-later
-+    pkt->header_size = 0;
++ */
-+    pkt->payload_size = 0;
++
-+    pkt->offset = 0;
++#ifndef HW_VIRTIO_VHOST_IOVA_TREE_H
-+    pkt->flags = 0;
++#define HW_VIRTIO_VHOST_IOVA_TREE_H
++
-     return pkt;
++#include "qemu/iova-tree.h"
- }
++#include "exec/memory.h"
-diff --git a/net/colo.h b/net/colo.h
++
-index XXXXXXX..XXXXXXX 100644
++typedef struct VhostIOVATree VhostIOVATree;
---- a/net/colo.h
++
-+++ b/net/colo.h
++VhostIOVATree *vhost_iova_tree_new(uint64_t iova_first, uint64_t iova_last);
-@@ -XXX,XX +XXX,XX @@ typedef struct Packet {
++void vhost_iova_tree_delete(VhostIOVATree *iova_tree);
-     int64_t creation_ms;
++G_DEFINE_AUTOPTR_CLEANUP_FUNC(VhostIOVATree, vhost_iova_tree_delete);
-     /* Get vnet_hdr_len from filter */
++
-     uint32_t vnet_hdr_len;
++const DMAMap *vhost_iova_tree_find_iova(const VhostIOVATree *iova_tree,
-+    uint32_t tcp_seq; /* sequence number */
++                                        const DMAMap *map);
-+    uint32_t tcp_ack; /* acknowledgement number */
++int vhost_iova_tree_map_alloc(VhostIOVATree *iova_tree, DMAMap *map);
-+    /* the sequence number of the last byte of the packet */
++void vhost_iova_tree_remove(VhostIOVATree *iova_tree, const DMAMap *map);
-+    uint32_t seq_end;
++
-+    uint8_t header_size;  /* the header length */
++#endif
 +    uint16_t payload_size; /* the payload length */
 +    /* record the payload offset(the length that has been compared) */
 +    uint16_t offset;
 +    uint8_t flags; /* Flags(aka Control bits) */
  } Packet;
  typedef struct ConnectionKey {
@@ -XXX,XX +XXX,XX @@ typedef struct Connection {
      /* flag to enqueue unprocessed_connections */
      bool processing;
      uint8_t ip_proto;
 +    /* record the sequence number that has been compared */
 +    uint32_t compare_seq;
 +    /* the maximum of acknowledgement number in primary_list queue */
 +    uint32_t pack;
 +    /* the maximum of acknowledgement number in secondary_list queue */
 +    uint32_t sack;
      /* offset = secondary_seq - primary_seq */
      tcp_seq  offset;
      /*
 diff --git a/net/trace-events b/net/trace-events
 index XXXXXXX..XXXXXXX 100644
 --- a/net/trace-events
 +++ b/net/trace-events
@@ -XXX,XX +XXX,XX @@ colo_compare_icmp_miscompare(const char *sta, int size) ": %s = %d"
  colo_compare_ip_info(int psize, const char *sta, const char *stb, int ssize, const char *stc, const char *std) "ppkt size = %d, ip_src = %s, ip_dst = %s, spkt size = %d, ip_src = %s, ip_dst = %s"
  colo_old_packet_check_found(int64_t old_time) "%" PRId64
  colo_compare_miscompare(void) ""
 -colo_compare_tcp_info(const char *pkt, uint32_t seq, uint32_t ack, int res, uint32_t flag, int size) "side: %s seq/ack= %u/%u res= %d flags= 0x%x pkt_size: %d\n"
 +colo_compare_tcp_info(const char *pkt, uint32_t seq, uint32_t ack, int hdlen, int pdlen, int offset, int flags) "%s: seq/ack= %u/%u hdlen= %d pdlen= %d offset= %d flags=%d\n"
  # net/filter-rewriter.c
  colo_filter_rewriter_debug(void) ""
 --
 .7.4

-[Qemu-devel] [PULL 4/6] net: Allow netdevs to be used with 'hostfwd_add' and 'hostfwd_remove'
+[PULL 12/15] vdpa: Add custom IOTLB translations to SVQ
-From: Thomas Huth <thuth@redhat.com>
+From: Eugenio Pérez <eperezma@redhat.com>
-It does not make much sense to limit these commands to the legacy 'vlan'
+Use translations added in VhostIOVATree in SVQ.
-concept only, they should work with the modern netdevs, too. So now
-it is possible to use this command with one, two or three parameters.
+Only introduce usage here, not allocation and deallocation. As with
+previous patches, we use the dead code paths of shadow_vqs_enabled to
-With one parameter, the command installs a hostfwd rule on the default
+avoid commiting too many changes at once. These are impossible to take
-"user" network:
+at the moment.
-    hostfwd_add tcp:...
+Acked-by: Michael S. Tsirkin <mst@redhat.com>
-With two parameters, the command installs a hostfwd rule on a netdev
+Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
 (that's the new way of using this command):
     hostfwd_add netdev_id tcp:...
 With three parameters, the command installs a rule on a 'vlan' (aka hub):
     hostfwd_add hub_id name tcp:...
 Same applies to the hostfwd_remove command now.
 Signed-off-by: Thomas Huth <thuth@redhat.com>
 Signed-off-by: Jason Wang <jasowang@redhat.com>
 ---
- hmp-commands.hx |  4 ++--
+ hw/virtio/vhost-shadow-virtqueue.c |  75 +++++++++++++++++++++--
- net/slirp.c     | 33 +++++++++++++++++++++++----------
+ hw/virtio/vhost-shadow-virtqueue.h |   6 +-
-files changed, 25 insertions(+), 12 deletions(-)
+ hw/virtio/vhost-vdpa.c             | 122 +++++++++++++++++++++++++++++++------
+ include/hw/virtio/vhost-vdpa.h     |   3 +
-diff --git a/hmp-commands.hx b/hmp-commands.hx
+files changed, 181 insertions(+), 25 deletions(-)
 diff --git a/hw/virtio/vhost-shadow-virtqueue.c b/hw/virtio/vhost-shadow-virtqueue.c
 index XXXXXXX..XXXXXXX 100644
---- a/hmp-commands.hx
+--- a/hw/virtio/vhost-shadow-virtqueue.c
-+++ b/hmp-commands.hx
++++ b/hw/virtio/vhost-shadow-virtqueue.c
-@@ -XXX,XX +XXX,XX @@ ETEXI
+@@ -XXX,XX +XXX,XX @@ static uint16_t vhost_svq_available_slots(const VhostShadowVirtqueue *svq)
-     {
+     return svq->vring.num - (svq->shadow_avail_idx - svq->shadow_used_idx);
-         .name       = "hostfwd_add",
+ }
-         .args_type  = "arg1:s,arg2:s?,arg3:s?",
--        .params     = "[vlan_id name] [tcp|udp]:[hostaddr]:hostport-[guestaddr]:guestport",
++/**
-+        .params     = "[hub_id name]|[netdev_id] [tcp|udp]:[hostaddr]:hostport-[guestaddr]:guestport",
++ * Translate addresses between the qemu's virtual address and the SVQ IOVA
-         .help       = "redirect TCP or UDP connections from host to guest (requires -net user)",
++ *
-         .cmd        = hmp_hostfwd_add,
++ * @svq: Shadow VirtQueue
-     },
++ * @vaddr: Translated IOVA addresses
-@@ -XXX,XX +XXX,XX @@ ETEXI
++ * @iovec: Source qemu's VA addresses
-     {
++ * @num: Length of iovec and minimum length of vaddr
-         .name       = "hostfwd_remove",
++ */
-         .args_type  = "arg1:s,arg2:s?,arg3:s?",
++static bool vhost_svq_translate_addr(const VhostShadowVirtqueue *svq,
--        .params     = "[vlan_id name] [tcp|udp]:[hostaddr]:hostport",
++                                     void **addrs, const struct iovec *iovec,
-+        .params     = "[hub_id name]|[netdev_id] [tcp|udp]:[hostaddr]:hostport",
++                                     size_t num)
-         .help       = "remove host-to-guest TCP or UDP redirection",
++{
-         .cmd        = hmp_hostfwd_remove,
++    if (num == 0) {
-     },
++        return true;
-diff --git a/net/slirp.c b/net/slirp.c
++    }
 +
 +    for (size_t i = 0; i < num; ++i) {
 +        DMAMap needle = {
 +            .translated_addr = (hwaddr)iovec[i].iov_base,
 +            .size = iovec[i].iov_len,
 +        };
 +        size_t off;
 +
 +        const DMAMap *map = vhost_iova_tree_find_iova(svq->iova_tree, &needle);
 +        /*
 +         * Map cannot be NULL since iova map contains all guest space and
 +         * qemu already has a physical address mapped
 +         */
 +        if (unlikely(!map)) {
 +            qemu_log_mask(LOG_GUEST_ERROR,
 +                          "Invalid address 0x%"HWADDR_PRIx" given by guest",
 +                          needle.translated_addr);
 +            return false;
 +        }
 +
 +        off = needle.translated_addr - map->translated_addr;
 +        addrs[i] = (void *)(map->iova + off);
 +
 +        if (unlikely(int128_gt(int128_add(needle.translated_addr,
 +                                          iovec[i].iov_len),
 +                               map->translated_addr + map->size))) {
 +            qemu_log_mask(LOG_GUEST_ERROR,
 +                          "Guest buffer expands over iova range");
 +            return false;
 +        }
 +    }
 +
 +    return true;
 +}
 +
  static void vhost_vring_write_descs(VhostShadowVirtqueue *svq,
 +                                    void * const *sg,
                                      const struct iovec *iovec,
                                      size_t num, bool more_descs, bool write)
  {
@@ -XXX,XX +XXX,XX @@ static void vhost_vring_write_descs(VhostShadowVirtqueue *svq,
          } else {
              descs[i].flags = flags;
          }
 -        descs[i].addr = cpu_to_le64((hwaddr)iovec[n].iov_base);
 +        descs[i].addr = cpu_to_le64((hwaddr)sg[n]);
          descs[i].len = cpu_to_le32(iovec[n].iov_len);
          last = i;
@@ -XXX,XX +XXX,XX @@ static bool vhost_svq_add_split(VhostShadowVirtqueue *svq,
  {
      unsigned avail_idx;
      vring_avail_t *avail = svq->vring.avail;
 +    bool ok;
 +    g_autofree void **sgs = g_new(void *, MAX(elem->out_num, elem->in_num));
      *head = svq->free_head;
@@ -XXX,XX +XXX,XX @@ static bool vhost_svq_add_split(VhostShadowVirtqueue *svq,
          return false;
      }
 -    vhost_vring_write_descs(svq, elem->out_sg, elem->out_num,
 +    ok = vhost_svq_translate_addr(svq, sgs, elem->out_sg, elem->out_num);
 +    if (unlikely(!ok)) {
 +        return false;
 +    }
 +    vhost_vring_write_descs(svq, sgs, elem->out_sg, elem->out_num,
                              elem->in_num > 0, false);
 -    vhost_vring_write_descs(svq, elem->in_sg, elem->in_num, false, true);
 +
 +
 +    ok = vhost_svq_translate_addr(svq, sgs, elem->in_sg, elem->in_num);
 +    if (unlikely(!ok)) {
 +        return false;
 +    }
 +
 +    vhost_vring_write_descs(svq, sgs, elem->in_sg, elem->in_num, false, true);
      /*
       * Put the entry in the available array (but don't update avail->idx until
@@ -XXX,XX +XXX,XX @@ void vhost_svq_stop(VhostShadowVirtqueue *svq)
   * Creates vhost shadow virtqueue, and instructs the vhost device to use the
   * shadow methods and file descriptors.
   *
 + * @iova_tree: Tree to perform descriptors translations
 + *
   * Returns the new virtqueue or NULL.
   *
   * In case of error, reason is reported through error_report.
   */
 -VhostShadowVirtqueue *vhost_svq_new(void)
 +VhostShadowVirtqueue *vhost_svq_new(VhostIOVATree *iova_tree)
  {
      g_autofree VhostShadowVirtqueue *svq = g_new0(VhostShadowVirtqueue, 1);
      int r;
@@ -XXX,XX +XXX,XX @@ VhostShadowVirtqueue *vhost_svq_new(void)
      event_notifier_init_fd(&svq->svq_kick, VHOST_FILE_UNBIND);
      event_notifier_set_handler(&svq->hdev_call, vhost_svq_handle_call);
 +    svq->iova_tree = iova_tree;
      return g_steal_pointer(&svq);
  err_init_hdev_call:
 diff --git a/hw/virtio/vhost-shadow-virtqueue.h b/hw/virtio/vhost-shadow-virtqueue.h
 index XXXXXXX..XXXXXXX 100644
---- a/net/slirp.c
+--- a/hw/virtio/vhost-shadow-virtqueue.h
-+++ b/net/slirp.c
++++ b/hw/virtio/vhost-shadow-virtqueue.h
-@@ -XXX,XX +XXX,XX @@ error:
+@@ -XXX,XX +XXX,XX @@
-     return -1;
+ #include "qemu/event_notifier.h"
  #include "hw/virtio/virtio.h"
  #include "standard-headers/linux/vhost_types.h"
 +#include "hw/virtio/vhost-iova-tree.h"
  /* Shadow virtqueue to relay notifications */
  typedef struct VhostShadowVirtqueue {
@@ -XXX,XX +XXX,XX @@ typedef struct VhostShadowVirtqueue {
      /* Virtio device */
      VirtIODevice *vdev;
 +    /* IOVA mapping */
 +    VhostIOVATree *iova_tree;
 +
      /* Map for use the guest's descriptors */
      VirtQueueElement **ring_id_maps;
@@ -XXX,XX +XXX,XX @@ void vhost_svq_start(VhostShadowVirtqueue *svq, VirtIODevice *vdev,
                       VirtQueue *vq);
  void vhost_svq_stop(VhostShadowVirtqueue *svq);
 -VhostShadowVirtqueue *vhost_svq_new(void);
 +VhostShadowVirtqueue *vhost_svq_new(VhostIOVATree *iova_tree);
  void vhost_svq_free(gpointer vq);
  G_DEFINE_AUTOPTR_CLEANUP_FUNC(VhostShadowVirtqueue, vhost_svq_free);
 diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c
 index XXXXXXX..XXXXXXX 100644
 --- a/hw/virtio/vhost-vdpa.c
 +++ b/hw/virtio/vhost-vdpa.c
@@ -XXX,XX +XXX,XX @@ static void vhost_vdpa_listener_region_add(MemoryListener *listener,
                                           vaddr, section->readonly);
      llsize = int128_sub(llend, int128_make64(iova));
 +    if (v->shadow_vqs_enabled) {
 +        DMAMap mem_region = {
 +            .translated_addr = (hwaddr)vaddr,
 +            .size = int128_get64(llsize) - 1,
 +            .perm = IOMMU_ACCESS_FLAG(true, section->readonly),
 +        };
 +
 +        int r = vhost_iova_tree_map_alloc(v->iova_tree, &mem_region);
 +        if (unlikely(r != IOVA_OK)) {
 +            error_report("Can't allocate a mapping (%d)", r);
 +            goto fail;
 +        }
 +
 +        iova = mem_region.iova;
 +    }
      vhost_vdpa_iotlb_batch_begin_once(v);
      ret = vhost_vdpa_dma_map(v, iova, int128_get64(llsize),
@@ -XXX,XX +XXX,XX @@ static void vhost_vdpa_listener_region_del(MemoryListener *listener,
      llsize = int128_sub(llend, int128_make64(iova));
 +    if (v->shadow_vqs_enabled) {
 +        const DMAMap *result;
 +        const void *vaddr = memory_region_get_ram_ptr(section->mr) +
 +            section->offset_within_region +
 +            (iova - section->offset_within_address_space);
 +        DMAMap mem_region = {
 +            .translated_addr = (hwaddr)vaddr,
 +            .size = int128_get64(llsize) - 1,
 +        };
 +
 +        result = vhost_iova_tree_find_iova(v->iova_tree, &mem_region);
 +        iova = result->iova;
 +        vhost_iova_tree_remove(v->iova_tree, &mem_region);
 +    }
      vhost_vdpa_iotlb_batch_begin_once(v);
      ret = vhost_vdpa_dma_unmap(v, iova, int128_get64(llsize));
      if (ret) {
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_init_svq(struct vhost_dev *hdev, struct vhost_vdpa *v,
      shadow_vqs = g_ptr_array_new_full(hdev->nvqs, vhost_svq_free);
      for (unsigned n = 0; n < hdev->nvqs; ++n) {
 -        g_autoptr(VhostShadowVirtqueue) svq = vhost_svq_new();
 +        g_autoptr(VhostShadowVirtqueue) svq = vhost_svq_new(v->iova_tree);
          if (unlikely(!svq)) {
              error_setg(errp, "Cannot create svq %u", n);
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_svq_set_fds(struct vhost_dev *dev,
  /**
   * Unmap a SVQ area in the device
   */
 -static bool vhost_vdpa_svq_unmap_ring(struct vhost_vdpa *v, hwaddr iova,
 -                                      hwaddr size)
 +static bool vhost_vdpa_svq_unmap_ring(struct vhost_vdpa *v,
 +                                      const DMAMap *needle)
  {
 +    const DMAMap *result = vhost_iova_tree_find_iova(v->iova_tree, needle);
 +    hwaddr size;
      int r;
 -    size = ROUND_UP(size, qemu_real_host_page_size);
 -    r = vhost_vdpa_dma_unmap(v, iova, size);
 +    if (unlikely(!result)) {
 +        error_report("Unable to find SVQ address to unmap");
 +        return false;
 +    }
 +
 +    size = ROUND_UP(result->size, qemu_real_host_page_size);
 +    r = vhost_vdpa_dma_unmap(v, result->iova, size);
      return r == 0;
  }
--static SlirpState *slirp_lookup(Monitor *mon, const char *vlan,
+ static bool vhost_vdpa_svq_unmap_rings(struct vhost_dev *dev,
--                                const char *stack)
+                                        const VhostShadowVirtqueue *svq)
-+static SlirpState *slirp_lookup(Monitor *mon, const char *hub_id,
+ {
-+                                const char *name)
++    DMAMap needle = {};
- {
+     struct vhost_vdpa *v = dev->opaque;
--
+     struct vhost_vring_addr svq_addr;
--    if (vlan) {
+-    size_t device_size = vhost_svq_device_area_size(svq);
-+    if (name) {
+-    size_t driver_size = vhost_svq_driver_area_size(svq);
-         NetClientState *nc;
+     bool ok;
--        nc = net_hub_find_client_by_name(strtol(vlan, NULL, 0), stack);
--        if (!nc) {
+     vhost_svq_get_vring_addr(svq, &svq_addr);
--            monitor_printf(mon, "unrecognized (vlan-id, stackname) pair\n");
--            return NULL;
+-    ok = vhost_vdpa_svq_unmap_ring(v, svq_addr.desc_user_addr, driver_size);
-+        if (hub_id) {
++    needle.translated_addr = svq_addr.desc_user_addr;
-+            nc = net_hub_find_client_by_name(strtol(hub_id, NULL, 0), name);
++    ok = vhost_vdpa_svq_unmap_ring(v, &needle);
-+            if (!nc) {
+     if (unlikely(!ok)) {
-+                monitor_printf(mon, "unrecognized (vlan-id, stackname) pair\n");
+         return false;
-+                return NULL;
+     }
-+            }
-+        } else {
+-    return vhost_vdpa_svq_unmap_ring(v, svq_addr.used_user_addr, device_size);
-+            nc = qemu_find_netdev(name);
++    needle.translated_addr = svq_addr.used_user_addr;
-+            if (!nc) {
++    return vhost_vdpa_svq_unmap_ring(v, &needle);
-+                monitor_printf(mon, "unrecognized netdev id '%s'\n", name);
++}
-+                return NULL;
++
-+            }
++/**
-         }
++ * Map the SVQ area in the device
-         if (strcmp(nc->model, "user")) {
++ *
-             monitor_printf(mon, "invalid device specified\n");
++ * @v: Vhost-vdpa device
-@@ -XXX,XX +XXX,XX @@ void hmp_hostfwd_remove(Monitor *mon, const QDict *qdict)
++ * @needle: The area to search iova
-     const char *arg2 = qdict_get_try_str(qdict, "arg2");
++ * @errorp: Error pointer
-     const char *arg3 = qdict_get_try_str(qdict, "arg3");
++ */
++static bool vhost_vdpa_svq_map_ring(struct vhost_vdpa *v, DMAMap *needle,
--    if (arg2) {
++                                    Error **errp)
-+    if (arg3) {
++{
-         s = slirp_lookup(mon, arg1, arg2);
++    int r;
-         src_str = arg3;
++
-+    } else if (arg2) {
++    r = vhost_iova_tree_map_alloc(v->iova_tree, needle);
-+        s = slirp_lookup(mon, NULL, arg1);
++    if (unlikely(r != IOVA_OK)) {
-+        src_str = arg2;
++        error_setg(errp, "Cannot allocate iova (%d)", r);
-     } else {
++        return false;
-         s = slirp_lookup(mon, NULL, NULL);
++    }
-         src_str = arg1;
++
-@@ -XXX,XX +XXX,XX @@ void hmp_hostfwd_add(Monitor *mon, const QDict *qdict)
++    r = vhost_vdpa_dma_map(v, needle->iova, needle->size + 1,
-     const char *arg2 = qdict_get_try_str(qdict, "arg2");
++                           (void *)needle->translated_addr,
-     const char *arg3 = qdict_get_try_str(qdict, "arg3");
++                           needle->perm == IOMMU_RO);
++    if (unlikely(r != 0)) {
--    if (arg2) {
++        error_setg_errno(errp, -r, "Cannot map region to device");
-+    if (arg3) {
++        vhost_iova_tree_remove(v->iova_tree, needle);
-         s = slirp_lookup(mon, arg1, arg2);
++    }
-         redir_str = arg3;
++
-+    } else if (arg2) {
++    return r == 0;
-+        s = slirp_lookup(mon, NULL, arg1);
+ }
-+        redir_str = arg2;
-     } else {
+ /**
-         s = slirp_lookup(mon, NULL, NULL);
+@@ -XXX,XX +XXX,XX @@ static bool vhost_vdpa_svq_map_rings(struct vhost_dev *dev,
-         redir_str = arg1;
+                                      struct vhost_vring_addr *addr,
                                       Error **errp)
  {
 +    DMAMap device_region, driver_region;
 +    struct vhost_vring_addr svq_addr;
      struct vhost_vdpa *v = dev->opaque;
      size_t device_size = vhost_svq_device_area_size(svq);
      size_t driver_size = vhost_svq_driver_area_size(svq);
 -    int r;
 +    size_t avail_offset;
 +    bool ok;
      ERRP_GUARD();
 -    vhost_svq_get_vring_addr(svq, addr);
 +    vhost_svq_get_vring_addr(svq, &svq_addr);
 -    r = vhost_vdpa_dma_map(v, addr->desc_user_addr, driver_size,
 -                           (void *)addr->desc_user_addr, true);
 -    if (unlikely(r != 0)) {
 -        error_setg_errno(errp, -r, "Cannot create vq driver region: ");
 +    driver_region = (DMAMap) {
 +        .translated_addr = svq_addr.desc_user_addr,
 +        .size = driver_size - 1,
 +        .perm = IOMMU_RO,
 +    };
 +    ok = vhost_vdpa_svq_map_ring(v, &driver_region, errp);
 +    if (unlikely(!ok)) {
 +        error_prepend(errp, "Cannot create vq driver region: ");
          return false;
      }
 +    addr->desc_user_addr = driver_region.iova;
 +    avail_offset = svq_addr.avail_user_addr - svq_addr.desc_user_addr;
 +    addr->avail_user_addr = driver_region.iova + avail_offset;
 -    r = vhost_vdpa_dma_map(v, addr->used_user_addr, device_size,
 -                           (void *)addr->used_user_addr, false);
 -    if (unlikely(r != 0)) {
 -        error_setg_errno(errp, -r, "Cannot create vq device region: ");
 +    device_region = (DMAMap) {
 +        .translated_addr = svq_addr.used_user_addr,
 +        .size = device_size - 1,
 +        .perm = IOMMU_RW,
 +    };
 +    ok = vhost_vdpa_svq_map_ring(v, &device_region, errp);
 +    if (unlikely(!ok)) {
 +        error_prepend(errp, "Cannot create vq device region: ");
 +        vhost_vdpa_svq_unmap_ring(v, &driver_region);
      }
 +    addr->used_user_addr = device_region.iova;
 -    return r == 0;
 +    return ok;
  }
  static bool vhost_vdpa_svq_setup(struct vhost_dev *dev,
 diff --git a/include/hw/virtio/vhost-vdpa.h b/include/hw/virtio/vhost-vdpa.h
 index XXXXXXX..XXXXXXX 100644
 --- a/include/hw/virtio/vhost-vdpa.h
 +++ b/include/hw/virtio/vhost-vdpa.h
@@ -XXX,XX +XXX,XX @@
  #include <gmodule.h>
 +#include "hw/virtio/vhost-iova-tree.h"
  #include "hw/virtio/virtio.h"
  #include "standard-headers/linux/vhost_types.h"
@@ -XXX,XX +XXX,XX @@ typedef struct vhost_vdpa {
      MemoryListener listener;
      struct vhost_vdpa_iova_range iova_range;
      bool shadow_vqs_enabled;
 +    /* IOVA mapping used by the Shadow Virtqueue */
 +    VhostIOVATree *iova_tree;
      GPtrArray *shadow_vqs;
      struct vhost_dev *dev;
      VhostVDPAHostNotifier notifier[VIRTIO_QUEUE_MAX];
 --
 .7.4

-[Qemu-devel] [PULL 5/6] qemu-doc: Get rid of "vlan=X" example in the documentation
+[PULL 13/15] vdpa: Adapt vhost_vdpa_get_vring_base to SVQ
-From: Thomas Huth <thuth@redhat.com>
+From: Eugenio Pérez <eperezma@redhat.com>
-The vlan concept is marked as deprecated, so we should not use
+This is needed to achieve migration, so the destination can restore its
-this for examples in the documentation anymore.
+index.
-Signed-off-by: Thomas Huth <thuth@redhat.com>
+Setting base as last used idx, so destination will see as available all
 the entries that the device did not use, including the in-flight
 processing ones.
 This is ok for networking, but other kinds of devices might have
 problems with these retransmissions.
 Acked-by: Michael S. Tsirkin <mst@redhat.com>
 Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
 Signed-off-by: Jason Wang <jasowang@redhat.com>
 ---
- qemu-options.hx | 4 ++--
+ hw/virtio/vhost-vdpa.c | 17 +++++++++++++++++
-file changed, 2 insertions(+), 2 deletions(-)
+file changed, 17 insertions(+)
-diff --git a/qemu-options.hx b/qemu-options.hx
+diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c
 index XXXXXXX..XXXXXXX 100644
---- a/qemu-options.hx
+--- a/hw/virtio/vhost-vdpa.c
-+++ b/qemu-options.hx
++++ b/hw/virtio/vhost-vdpa.c
-@@ -XXX,XX +XXX,XX @@ qemu-system-i386 linux.img -net nic -net tap
+@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_vring_base(struct vhost_dev *dev,
- #launch a QEMU instance with two NICs, each one connected
+ static int vhost_vdpa_get_vring_base(struct vhost_dev *dev,
- #to a TAP device
+                                        struct vhost_vring_state *ring)
- qemu-system-i386 linux.img \
+ {
--                 -net nic,vlan=0 -net tap,vlan=0,ifname=tap0 \
++    struct vhost_vdpa *v = dev->opaque;
--                 -net nic,vlan=1 -net tap,vlan=1,ifname=tap1
+     int ret;
-+        -netdev tap,id=nd0,ifname=tap0 -device e1000,netdev=nd0 \
-+        -netdev tap,id=nd1,ifname=tap1 -device rtl8139,netdev=nd1
++    if (v->shadow_vqs_enabled) {
- @end example
++        VhostShadowVirtqueue *svq = g_ptr_array_index(v->shadow_vqs,
++                                                      ring->index);
- @example
++
 +        /*
 +         * Setting base as last used idx, so destination will see as available
 +         * all the entries that the device did not use, including the in-flight
 +         * processing ones.
 +         *
 +         * TODO: This is ok for networking, but other kinds of devices might
 +         * have problems with these retransmissions.
 +         */
 +        ring->num = svq->last_used_idx;
 +        return 0;
 +    }
 +
      ret = vhost_vdpa_call(dev, VHOST_GET_VRING_BASE, ring);
      trace_vhost_vdpa_get_vring_base(dev, ring->index, ring->num);
      return ret;
 --
 .7.4

-[Qemu-devel] [PULL 6/6] MAINTAINERS: update Dmitry Fleytman email
+[PULL 14/15] vdpa: Never set log_base addr if SVQ is enabled
-From: Philippe Mathieu-Daudé <f4bug@amsat.org>
+From: Eugenio Pérez <eperezma@redhat.com>
-gently asked by his automatic reply :)
+Setting the log address would make the device start reporting invalid
 dirty memory because the SVQ vrings are located in qemu's memory.
-Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
+Acked-by: Michael S. Tsirkin <mst@redhat.com>
 Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
 Signed-off-by: Jason Wang <jasowang@redhat.com>
 ---
- MAINTAINERS | 8 ++++----
+ hw/virtio/vhost-vdpa.c | 3 ++-
-file changed, 4 insertions(+), 4 deletions(-)
+file changed, 2 insertions(+), 1 deletion(-)
-diff --git a/MAINTAINERS b/MAINTAINERS
+diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c
 index XXXXXXX..XXXXXXX 100644
---- a/MAINTAINERS
+--- a/hw/virtio/vhost-vdpa.c
-+++ b/MAINTAINERS
++++ b/hw/virtio/vhost-vdpa.c
-@@ -XXX,XX +XXX,XX @@ F: hw/scsi/mfi.h
+@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_dev_start(struct vhost_dev *dev, bool started)
- F: tests/megasas-test.c
+ static int vhost_vdpa_set_log_base(struct vhost_dev *dev, uint64_t base,
+                                      struct vhost_log *log)
- Network packet abstractions
+ {
--M: Dmitry Fleytman <dmitry@daynix.com>
+-    if (vhost_vdpa_one_time_request(dev)) {
-+M: Dmitry Fleytman <dmitry.fleytman@gmail.com>
++    struct vhost_vdpa *v = dev->opaque;
- S: Maintained
++    if (v->shadow_vqs_enabled || vhost_vdpa_one_time_request(dev)) {
- F: include/net/eth.h
+         return 0;
- F: net/eth.c
+     }
@@ -XXX,XX +XXX,XX @@ F: hw/net/net_rx_pkt*
  F: hw/net/net_tx_pkt*
  Vmware
 -M: Dmitry Fleytman <dmitry@daynix.com>
 +M: Dmitry Fleytman <dmitry.fleytman@gmail.com>
  S: Maintained
  F: hw/net/vmxnet*
  F: hw/scsi/vmw_pvscsi*
@@ -XXX,XX +XXX,XX @@ F: hw/mem/nvdimm.c
  F: include/hw/mem/nvdimm.h
  e1000x
 -M: Dmitry Fleytman <dmitry@daynix.com>
 +M: Dmitry Fleytman <dmitry.fleytman@gmail.com>
  S: Maintained
  F: hw/net/e1000x*
  e1000e
 -M: Dmitry Fleytman <dmitry@daynix.com>
 +M: Dmitry Fleytman <dmitry.fleytman@gmail.com>
  S: Maintained
  F: hw/net/e1000e*
 --
 .7.4

-[Qemu-devel] [PULL 1/6] colo: modified the payload compare function
+[PULL 15/15] vdpa: Expose VHOST_F_LOG_ALL on SVQ
-From: Mao Zhongyi <maozy.fnst@cn.fujitsu.com>
+From: Eugenio Pérez <eperezma@redhat.com>
-Modified the function colo_packet_compare_common to prepare for the
+SVQ is able to log the dirty bits by itself, so let's use it to not
-tcp packet comparison in the next patch.
+block migration.
-Cc: Zhang Chen <zhangckid@gmail.com>
+Also, ignore set and clear of VHOST_F_LOG_ALL on set_features if SVQ is
-Cc: Li Zhijian <lizhijian@cn.fujitsu.com>
+enabled. Even if the device supports it, the reports would be nonsense
-Cc: Jason Wang <jasowang@redhat.com>
+because SVQ memory is in the qemu region.
-Signed-off-by: Mao Zhongyi <maozy.fnst@cn.fujitsu.com>
+The log region is still allocated. Future changes might skip that, but
-Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com>
+this series is already long enough.
-Signed-off-by: Zhang Chen <zhangckid@gmail.com>
-Reviewed-by: Zhang Chen <zhangckid@gmail.com>
+Acked-by: Michael S. Tsirkin <mst@redhat.com>
 Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
 Signed-off-by: Jason Wang <jasowang@redhat.com>
 ---
- net/colo-compare.c | 88 +++++++++++++++++++++++++++---------------------------
+ hw/virtio/vhost-vdpa.c         | 39 +++++++++++++++++++++++++++++++++++----
-file changed, 44 insertions(+), 44 deletions(-)
+ include/hw/virtio/vhost-vdpa.h |  1 +
 files changed, 36 insertions(+), 4 deletions(-)
-diff --git a/net/colo-compare.c b/net/colo-compare.c
+diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c
 index XXXXXXX..XXXXXXX 100644
---- a/net/colo-compare.c
+--- a/hw/virtio/vhost-vdpa.c
-+++ b/net/colo-compare.c
++++ b/hw/virtio/vhost-vdpa.c
-@@ -XXX,XX +XXX,XX @@ static int packet_enqueue(CompareState *s, int mode, Connection **con)
+@@ -XXX,XX +XXX,XX @@ static bool vhost_vdpa_one_time_request(struct vhost_dev *dev)
-  * return:    0  means packet same
+     return v->index != 0;
-  *            > 0 || < 0 means packet different
+ }
-  */
--static int colo_packet_compare_common(Packet *ppkt,
++static int vhost_vdpa_get_dev_features(struct vhost_dev *dev,
--                                      Packet *spkt,
++                                       uint64_t *features)
--                                      int poffset,
++{
--                                      int soffset)
++    int ret;
 +static int colo_compare_packet_payload(Packet *ppkt,
 +                                       Packet *spkt,
 +                                       uint16_t poffset,
 +                                       uint16_t soffset,
 +                                       uint16_t len)
 +
++    ret = vhost_vdpa_call(dev, VHOST_GET_FEATURES, features);
++    trace_vhost_vdpa_get_features(dev, *features);
++    return ret;
++}
++
+ static int vhost_vdpa_init_svq(struct vhost_dev *hdev, struct vhost_vdpa *v,
+                                Error **errp)
  {
-     if (trace_event_get_state_backends(TRACE_COLO_COMPARE_MISCOMPARE)) {
+@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_init_svq(struct vhost_dev *hdev, struct vhost_vdpa *v,
-         char pri_ip_src[20], pri_ip_dst[20], sec_ip_src[20], sec_ip_dst[20];
+         return 0;
@@ -XXX,XX +XXX,XX @@ static int colo_packet_compare_common(Packet *ppkt,
                                     sec_ip_src, sec_ip_dst);
      }
--    poffset = ppkt->vnet_hdr_len + poffset;
+-    r = hdev->vhost_ops->vhost_get_features(hdev, &dev_features);
--    soffset = ppkt->vnet_hdr_len + soffset;
++    r = vhost_vdpa_get_dev_features(hdev, &dev_features);
--
+     if (r != 0) {
--    if (ppkt->size - poffset == spkt->size - soffset) {
+         error_setg_errno(errp, -r, "Can't get vdpa device features");
--        return memcmp(ppkt->data + poffset,
+         return r;
--                      spkt->data + soffset,
+@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_mem_table(struct vhost_dev *dev,
--                      spkt->size - soffset);
+ static int vhost_vdpa_set_features(struct vhost_dev *dev,
--    } else {
+                                    uint64_t features)
--        trace_colo_compare_main("Net packet size are not the same");
+ {
--        return -1;
++    struct vhost_vdpa *v = dev->opaque;
--    }
+     int ret;
-+    return memcmp(ppkt->data + poffset, spkt->data + soffset, len);
- }
+     if (vhost_vdpa_one_time_request(dev)) {
+         return 0;
  /*
@@ -XXX,XX +XXX,XX @@ static int colo_packet_compare_tcp(Packet *spkt, Packet *ppkt)
       * the secondary guest's timestamp. COLO just focus on payload,
       * so we just need skip this field.
       */
 -    if (ptcp->th_off > 5) {
 -        ptrdiff_t ptcp_offset, stcp_offset;
 -        ptcp_offset = ppkt->transport_header - (uint8_t *)ppkt->data
 -                      + (ptcp->th_off * 4) - ppkt->vnet_hdr_len;
 -        stcp_offset = spkt->transport_header - (uint8_t *)spkt->data
 -                      + (stcp->th_off * 4) - spkt->vnet_hdr_len;
 +    ptrdiff_t ptcp_offset, stcp_offset;
 -        /*
 -         * When network is busy, some tcp options(like sack) will unpredictable
 -         * occur in primary side or secondary side. it will make packet size
 -         * not same, but the two packet's payload is identical. colo just
 -         * care about packet payload, so we skip the option field.
 -         */
 -        res = colo_packet_compare_common(ppkt, spkt, ptcp_offset, stcp_offset);
 -    } else if (ptcp->th_sum == stcp->th_sum) {
 -        res = colo_packet_compare_common(ppkt, spkt, ETH_HLEN, ETH_HLEN);
 +    ptcp_offset = ppkt->transport_header - (uint8_t *)ppkt->data
 +                  + (ptcp->th_off << 2) - ppkt->vnet_hdr_len;
 +    stcp_offset = spkt->transport_header - (uint8_t *)spkt->data
 +                  + (stcp->th_off << 2) - spkt->vnet_hdr_len;
 +    if (ppkt->size - ptcp_offset == spkt->size - stcp_offset) {
 +        res = colo_compare_packet_payload(ppkt, spkt,
 +                                          ptcp_offset, stcp_offset,
 +                                          ppkt->size - ptcp_offset);
      } else {
 +        trace_colo_compare_main("TCP: payload size of packets are different");
          res = -1;
      }
-@@ -XXX,XX +XXX,XX @@ static int colo_packet_compare_tcp(Packet *spkt, Packet *ppkt)
++    if (v->shadow_vqs_enabled) {
-  */
++        if ((v->acked_features ^ features) == BIT_ULL(VHOST_F_LOG_ALL)) {
- static int colo_packet_compare_udp(Packet *spkt, Packet *ppkt)
++            /*
 +             * QEMU is just trying to enable or disable logging. SVQ handles
 +             * this sepparately, so no need to forward this.
 +             */
 +            v->acked_features = features;
 +            return 0;
 +        }
 +
 +        v->acked_features = features;
 +
 +        /* We must not ack _F_LOG if SVQ is enabled */
 +        features &= ~BIT_ULL(VHOST_F_LOG_ALL);
 +    }
 +
      trace_vhost_vdpa_set_features(dev, features);
      ret = vhost_vdpa_call(dev, VHOST_SET_FEATURES, &features);
      if (ret) {
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_vring_call(struct vhost_dev *dev,
  static int vhost_vdpa_get_features(struct vhost_dev *dev,
                                       uint64_t *features)
  {
 -    int ret;
--    int network_header_length = ppkt->ip->ip_hl * 4;
++    struct vhost_vdpa *v = dev->opaque;
-+    uint16_t network_header_length = ppkt->ip->ip_hl << 2;
++    int ret = vhost_vdpa_get_dev_features(dev, features);
-+    uint16_t offset = network_header_length + ETH_HLEN + ppkt->vnet_hdr_len;
++
++    if (ret == 0 && v->shadow_vqs_enabled) {
-     trace_colo_compare_main("compare udp");
++        /* Add SVQ logging capabilities */
++        *features |= BIT_ULL(VHOST_F_LOG_ALL);
@@ -XXX,XX +XXX,XX @@ static int colo_packet_compare_udp(Packet *spkt, Packet *ppkt)
       * other field like TOS,TTL,IP Checksum. we only need to compare
       * the ip payload here.
       */
 -    ret = colo_packet_compare_common(ppkt, spkt,
 -                                     network_header_length + ETH_HLEN,
 -                                     network_header_length + ETH_HLEN);
 -
 -    if (ret) {
 +    if (ppkt->size != spkt->size) {
 +        trace_colo_compare_main("UDP: payload size of packets are different");
 +        return -1;
 +    }
-+    if (colo_compare_packet_payload(ppkt, spkt, offset, offset,
-+                                    ppkt->size - offset)) {
+-    ret = vhost_vdpa_call(dev, VHOST_GET_FEATURES, features);
-         trace_colo_compare_udp_miscompare("primary pkt size", ppkt->size);
+-    trace_vhost_vdpa_get_features(dev, *features);
-         trace_colo_compare_udp_miscompare("Secondary pkt size", spkt->size);
+     return ret;
          if (trace_event_get_state_backends(TRACE_COLO_COMPARE_MISCOMPARE)) {
@@ -XXX,XX +XXX,XX @@ static int colo_packet_compare_udp(Packet *spkt, Packet *ppkt)
              qemu_hexdump((char *)spkt->data, stderr, "colo-compare sec pkt",
                           spkt->size);
          }
 +        return -1;
 +    } else {
 +        return 0;
      }
 -
 -    return ret;
  }
- /*
+diff --git a/include/hw/virtio/vhost-vdpa.h b/include/hw/virtio/vhost-vdpa.h
-@@ -XXX,XX +XXX,XX @@ static int colo_packet_compare_udp(Packet *spkt, Packet *ppkt)
+index XXXXXXX..XXXXXXX 100644
-  */
+--- a/include/hw/virtio/vhost-vdpa.h
- static int colo_packet_compare_icmp(Packet *spkt, Packet *ppkt)
++++ b/include/hw/virtio/vhost-vdpa.h
- {
+@@ -XXX,XX +XXX,XX @@ typedef struct vhost_vdpa {
--    int network_header_length = ppkt->ip->ip_hl * 4;
+     bool iotlb_batch_begin_sent;
-+    uint16_t network_header_length = ppkt->ip->ip_hl << 2;
+     MemoryListener listener;
-+    uint16_t offset = network_header_length + ETH_HLEN + ppkt->vnet_hdr_len;
+     struct vhost_vdpa_iova_range iova_range;
++    uint64_t acked_features;
-     trace_colo_compare_main("compare icmp");
+     bool shadow_vqs_enabled;
+     /* IOVA mapping used by the Shadow Virtqueue */
-@@ -XXX,XX +XXX,XX @@ static int colo_packet_compare_icmp(Packet *spkt, Packet *ppkt)
+     VhostIOVATree *iova_tree;
       * other field like TOS,TTL,IP Checksum. we only need to compare
       * the ip payload here.
       */
 -    if (colo_packet_compare_common(ppkt, spkt,
 -                                   network_header_length + ETH_HLEN,
 -                                   network_header_length + ETH_HLEN)) {
 +    if (ppkt->size != spkt->size) {
 +        trace_colo_compare_main("ICMP: payload size of packets are different");
 +        return -1;
 +    }
 +    if (colo_compare_packet_payload(ppkt, spkt, offset, offset,
 +                                    ppkt->size - offset)) {
          trace_colo_compare_icmp_miscompare("primary pkt size",
                                             ppkt->size);
          trace_colo_compare_icmp_miscompare("Secondary pkt size",
@@ -XXX,XX +XXX,XX @@ static int colo_packet_compare_icmp(Packet *spkt, Packet *ppkt)
   */
  static int colo_packet_compare_other(Packet *spkt, Packet *ppkt)
  {
 +    uint16_t offset = ppkt->vnet_hdr_len;
 +
      trace_colo_compare_main("compare other");
      if (trace_event_get_state_backends(TRACE_COLO_COMPARE_MISCOMPARE)) {
          char pri_ip_src[20], pri_ip_dst[20], sec_ip_src[20], sec_ip_dst[20];
@@ -XXX,XX +XXX,XX @@ static int colo_packet_compare_other(Packet *spkt, Packet *ppkt)
                                     sec_ip_src, sec_ip_dst);
      }
 -    return colo_packet_compare_common(ppkt, spkt, 0, 0);
 +    if (ppkt->size != spkt->size) {
 +        trace_colo_compare_main("Other: payload size of packets are different");
 +        return -1;
 +    }
 +    return colo_compare_packet_payload(ppkt, spkt, offset, offset,
 +                                       ppkt->size - offset);
  }
  static int colo_old_packet_check_one(Packet *pkt, int64_t *check_time)
 --
 .7.4

The following changes since commit e607bbee553cfe73072870cef458cfa4e78133e2:

Merge remote-tracking branch 'remotes/edgar/tags/edgar/xilinx-next-2018-01-26.for-upstream' into staging (2018-01-26 14:24:25 +0000)

are available in the git repository at:

https://github.com/jasowang/qemu.git tags/net-pull-request

for you to fetch changes up to bf4835a4d5338bb7424827715df22570a8adc67c:

MAINTAINERS: update Dmitry Fleytman email (2018-01-29 16:05:38 +0800)

----------------------------------------------------------------

----------------------------------------------------------------
Mao Zhongyi (2):
      colo: modified the payload compare function
      colo: compare the packet based on the tcp sequence number

Philippe Mathieu-Daudé (1):
      MAINTAINERS: update Dmitry Fleytman email

Thomas Huth (3):
      net: Allow hubports to connect to other netdevs
      net: Allow netdevs to be used with 'hostfwd_add' and 'hostfwd_remove'
      qemu-doc: Get rid of "vlan=X" example in the documentation

From: Mao Zhongyi <maozy.fnst@cn.fujitsu.com>

Modified the function colo_packet_compare_common to prepare for the
tcp packet comparison in the next patch.

Cc: Zhang Chen <zhangckid@gmail.com>
Cc: Li Zhijian <lizhijian@cn.fujitsu.com>
Cc: Jason Wang <jasowang@redhat.com>

Signed-off-by: Mao Zhongyi <maozy.fnst@cn.fujitsu.com>
Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com>
Signed-off-by: Zhang Chen <zhangckid@gmail.com>
Reviewed-by: Zhang Chen <zhangckid@gmail.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 net/colo-compare.c | 88 +++++++++++++++++++++++++++---------------------------
 1 file changed, 44 insertions(+), 44 deletions(-)

diff --git a/net/colo-compare.c b/net/colo-compare.c
index XXXXXXX..XXXXXXX 100644
--- a/net/colo-compare.c
+++ b/net/colo-compare.c
@@ -XXX,XX +XXX,XX @@ static int packet_enqueue(CompareState *s, int mode, Connection **con)
  * return:    0  means packet same
  *            > 0 || < 0 means packet different
  */
-static int colo_packet_compare_common(Packet *ppkt,
-                                      Packet *spkt,
-                                      int poffset,
-                                      int soffset)
+static int colo_compare_packet_payload(Packet *ppkt,
+                                       Packet *spkt,
+                                       uint16_t poffset,
+                                       uint16_t soffset,
+                                       uint16_t len)
+
 {
     if (trace_event_get_state_backends(TRACE_COLO_COMPARE_MISCOMPARE)) {
         char pri_ip_src[20], pri_ip_dst[20], sec_ip_src[20], sec_ip_dst[20];
@@ -XXX,XX +XXX,XX @@ static int colo_packet_compare_common(Packet *ppkt,
                                    sec_ip_src, sec_ip_dst);
     }
 
-    poffset = ppkt->vnet_hdr_len + poffset;
-    soffset = ppkt->vnet_hdr_len + soffset;
-
-    if (ppkt->size - poffset == spkt->size - soffset) {
-        return memcmp(ppkt->data + poffset,
-                      spkt->data + soffset,
-                      spkt->size - soffset);
-    } else {
-        trace_colo_compare_main("Net packet size are not the same");
-        return -1;
-    }
+    return memcmp(ppkt->data + poffset, spkt->data + soffset, len);
 }
 
 /*
@@ -XXX,XX +XXX,XX @@ static int colo_packet_compare_tcp(Packet *spkt, Packet *ppkt)
      * the secondary guest's timestamp. COLO just focus on payload,
      * so we just need skip this field.
      */
-    if (ptcp->th_off > 5) {
-        ptrdiff_t ptcp_offset, stcp_offset;
 
-        ptcp_offset = ppkt->transport_header - (uint8_t *)ppkt->data
-                      + (ptcp->th_off * 4) - ppkt->vnet_hdr_len;
-        stcp_offset = spkt->transport_header - (uint8_t *)spkt->data
-                      + (stcp->th_off * 4) - spkt->vnet_hdr_len;
+    ptrdiff_t ptcp_offset, stcp_offset;
 
-        /*
-         * When network is busy, some tcp options(like sack) will unpredictable
-         * occur in primary side or secondary side. it will make packet size
-         * not same, but the two packet's payload is identical. colo just
-         * care about packet payload, so we skip the option field.
-         */
-        res = colo_packet_compare_common(ppkt, spkt, ptcp_offset, stcp_offset);
-    } else if (ptcp->th_sum == stcp->th_sum) {
-        res = colo_packet_compare_common(ppkt, spkt, ETH_HLEN, ETH_HLEN);
+    ptcp_offset = ppkt->transport_header - (uint8_t *)ppkt->data
+                  + (ptcp->th_off << 2) - ppkt->vnet_hdr_len;
+    stcp_offset = spkt->transport_header - (uint8_t *)spkt->data
+                  + (stcp->th_off << 2) - spkt->vnet_hdr_len;
+    if (ppkt->size - ptcp_offset == spkt->size - stcp_offset) {
+        res = colo_compare_packet_payload(ppkt, spkt,
+                                          ptcp_offset, stcp_offset,
+                                          ppkt->size - ptcp_offset);
     } else {
+        trace_colo_compare_main("TCP: payload size of packets are different");
         res = -1;
     }
 
@@ -XXX,XX +XXX,XX @@ static int colo_packet_compare_tcp(Packet *spkt, Packet *ppkt)
  */
 static int colo_packet_compare_udp(Packet *spkt, Packet *ppkt)
 {
-    int ret;
-    int network_header_length = ppkt->ip->ip_hl * 4;
+    uint16_t network_header_length = ppkt->ip->ip_hl << 2;
+    uint16_t offset = network_header_length + ETH_HLEN + ppkt->vnet_hdr_len;
 
     trace_colo_compare_main("compare udp");
 
@@ -XXX,XX +XXX,XX @@ static int colo_packet_compare_udp(Packet *spkt, Packet *ppkt)
      * other field like TOS,TTL,IP Checksum. we only need to compare
      * the ip payload here.
      */
-    ret = colo_packet_compare_common(ppkt, spkt,
-                                     network_header_length + ETH_HLEN,
-                                     network_header_length + ETH_HLEN);
-
-    if (ret) {
+    if (ppkt->size != spkt->size) {
+        trace_colo_compare_main("UDP: payload size of packets are different");
+        return -1;
+    }
+    if (colo_compare_packet_payload(ppkt, spkt, offset, offset,
+                                    ppkt->size - offset)) {
         trace_colo_compare_udp_miscompare("primary pkt size", ppkt->size);
         trace_colo_compare_udp_miscompare("Secondary pkt size", spkt->size);
         if (trace_event_get_state_backends(TRACE_COLO_COMPARE_MISCOMPARE)) {
@@ -XXX,XX +XXX,XX @@ static int colo_packet_compare_udp(Packet *spkt, Packet *ppkt)
             qemu_hexdump((char *)spkt->data, stderr, "colo-compare sec pkt",
                          spkt->size);
         }
+        return -1;
+    } else {
+        return 0;
     }
-
-    return ret;
 }
 
 /*
@@ -XXX,XX +XXX,XX @@ static int colo_packet_compare_udp(Packet *spkt, Packet *ppkt)
  */
 static int colo_packet_compare_icmp(Packet *spkt, Packet *ppkt)
 {
-    int network_header_length = ppkt->ip->ip_hl * 4;
+    uint16_t network_header_length = ppkt->ip->ip_hl << 2;
+    uint16_t offset = network_header_length + ETH_HLEN + ppkt->vnet_hdr_len;
 
     trace_colo_compare_main("compare icmp");
 
@@ -XXX,XX +XXX,XX @@ static int colo_packet_compare_icmp(Packet *spkt, Packet *ppkt)
      * other field like TOS,TTL,IP Checksum. we only need to compare
      * the ip payload here.
      */
-    if (colo_packet_compare_common(ppkt, spkt,
-                                   network_header_length + ETH_HLEN,
-                                   network_header_length + ETH_HLEN)) {
+    if (ppkt->size != spkt->size) {
+        trace_colo_compare_main("ICMP: payload size of packets are different");
+        return -1;
+    }
+    if (colo_compare_packet_payload(ppkt, spkt, offset, offset,
+                                    ppkt->size - offset)) {
         trace_colo_compare_icmp_miscompare("primary pkt size",
                                            ppkt->size);
         trace_colo_compare_icmp_miscompare("Secondary pkt size",
@@ -XXX,XX +XXX,XX @@ static int colo_packet_compare_icmp(Packet *spkt, Packet *ppkt)
  */
 static int colo_packet_compare_other(Packet *spkt, Packet *ppkt)
 {
+    uint16_t offset = ppkt->vnet_hdr_len;
+
     trace_colo_compare_main("compare other");
     if (trace_event_get_state_backends(TRACE_COLO_COMPARE_MISCOMPARE)) {
         char pri_ip_src[20], pri_ip_dst[20], sec_ip_src[20], sec_ip_dst[20];
@@ -XXX,XX +XXX,XX @@ static int colo_packet_compare_other(Packet *spkt, Packet *ppkt)
                                    sec_ip_src, sec_ip_dst);
     }
 
-    return colo_packet_compare_common(ppkt, spkt, 0, 0);
+    if (ppkt->size != spkt->size) {
+        trace_colo_compare_main("Other: payload size of packets are different");
+        return -1;
+    }
+    return colo_compare_packet_payload(ppkt, spkt, offset, offset,
+                                       ppkt->size - offset);
 }
 
 static int colo_old_packet_check_one(Packet *pkt, int64_t *check_time)
-- 
2.7.4

From: Mao Zhongyi <maozy.fnst@cn.fujitsu.com>

Packet size some time different or when network is busy.
Based on same payload size, but TCP protocol can not
guarantee send the same one packet in the same way,

like that:
We send this payload:
------------------------------
| header |1|2|3|4|5|6|7|8|9|0|
------------------------------

primary:
ppkt1:
----------------
| header |1|2|3|
----------------
ppkt2:
------------------------
| header |4|5|6|7|8|9|0|
------------------------

secondary:
spkt1:
------------------------------
| header |1|2|3|4|5|6|7|8|9|0|
------------------------------

In the original method, ppkt1 and ppkt2 are different in size and
spkt1, so they can't compare and trigger the checkpoint.

I have tested FTP get 200M and 1G file many times, I found that
the performance was less than 1% of the native.

Now I reconstructed the comparison of TCP packets based on the
TCP sequence number. first of all, ppkt1 and spkt1 have the same
starting sequence number, so they can compare, even though their
length is different. And then ppkt1 with a smaller payload length
is used as the comparison length, if the payload is same, send
out the ppkt1 and record the offset(the length of ppkt1 payload)
in spkt1. The next comparison, ppkt2 and spkt1 can be compared
from the recorded position of spkt1.

like that:
----------------
| header |1|2|3| ppkt1
---------|-----|
         |     |
---------v-----v--------------
| header |1|2|3|4|5|6|7|8|9|0| spkt1
---------------|\------------|
               | \offset     |
      ---------v-------------v
      | header |4|5|6|7|8|9|0| ppkt2
      ------------------------

In this way, the performance can reach native 20% in my multiple
tests.

Cc: Zhang Chen <zhangckid@gmail.com>
Cc: Li Zhijian <lizhijian@cn.fujitsu.com>
Cc: Jason Wang <jasowang@redhat.com>

Signed-off-by: Mao Zhongyi <maozy.fnst@cn.fujitsu.com>
Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com>
Signed-off-by: Zhang Chen <zhangckid@gmail.com>
Reviewed-by: Zhang Chen <zhangckid@gmail.com>
Tested-by: Zhang Chen <zhangckid@gmail.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 net/colo-compare.c | 343 +++++++++++++++++++++++++++++++++++------------------
 net/colo.c         |   9 ++
 net/colo.h         |  15 +++
 net/trace-events   |   2 +-
 4 files changed, 250 insertions(+), 119 deletions(-)

diff --git a/net/colo-compare.c b/net/colo-compare.c
index XXXXXXX..XXXXXXX 100644
--- a/net/colo-compare.c
+++ b/net/colo-compare.c
@@ -XXX,XX +XXX,XX @@
 #define COMPARE_READ_LEN_MAX NET_BUFSIZE
 #define MAX_QUEUE_SIZE 1024
 
+#define COLO_COMPARE_FREE_PRIMARY     0x01
+#define COLO_COMPARE_FREE_SECONDARY   0x02
+
 /* TODO: Should be configurable */
 #define REGULAR_PACKET_CHECK_MS 3000
 
@@ -XXX,XX +XXX,XX @@ static gint seq_sorter(Packet *a, Packet *b, gpointer data)
     return ntohl(atcp->th_seq) - ntohl(btcp->th_seq);
 }
 
+static void fill_pkt_tcp_info(void *data, uint32_t *max_ack)
+{
+    Packet *pkt = data;
+    struct tcphdr *tcphd;
+
+    tcphd = (struct tcphdr *)pkt->transport_header;
+
+    pkt->tcp_seq = ntohl(tcphd->th_seq);
+    pkt->tcp_ack = ntohl(tcphd->th_ack);
+    *max_ack = *max_ack > pkt->tcp_ack ? *max_ack : pkt->tcp_ack;
+    pkt->header_size = pkt->transport_header - (uint8_t *)pkt->data
+                       + (tcphd->th_off << 2) - pkt->vnet_hdr_len;
+    pkt->payload_size = pkt->size - pkt->header_size;
+    pkt->seq_end = pkt->tcp_seq + pkt->payload_size;
+    pkt->flags = tcphd->th_flags;
+}
+
 /*
  * Return 1 on success, if return 0 means the
  * packet will be dropped
  */
-static int colo_insert_packet(GQueue *queue, Packet *pkt)
+static int colo_insert_packet(GQueue *queue, Packet *pkt, uint32_t *max_ack)
 {
     if (g_queue_get_length(queue) <= MAX_QUEUE_SIZE) {
         if (pkt->ip->ip_p == IPPROTO_TCP) {
+            fill_pkt_tcp_info(pkt, max_ack);
             g_queue_insert_sorted(queue,
                                   pkt,
                                   (GCompareDataFunc)seq_sorter,
@@ -XXX,XX +XXX,XX @@ static int packet_enqueue(CompareState *s, int mode, Connection **con)
     }
 
     if (mode == PRIMARY_IN) {
-        if (!colo_insert_packet(&conn->primary_list, pkt)) {
+        if (!colo_insert_packet(&conn->primary_list, pkt, &conn->pack)) {
             error_report("colo compare primary queue size too big,"
                          "drop packet");
         }
     } else {
-        if (!colo_insert_packet(&conn->secondary_list, pkt)) {
+        if (!colo_insert_packet(&conn->secondary_list, pkt, &conn->sack)) {
             error_report("colo compare secondary queue size too big,"
                          "drop packet");
         }
@@ -XXX,XX +XXX,XX @@ static int packet_enqueue(CompareState *s, int mode, Connection **con)
     return 0;
 }
 
+static inline bool after(uint32_t seq1, uint32_t seq2)
+{
+        return (int32_t)(seq1 - seq2) > 0;
+}
+
+static void colo_release_primary_pkt(CompareState *s, Packet *pkt)
+{
+    int ret;
+    ret = compare_chr_send(s,
+                           pkt->data,
+                           pkt->size,
+                           pkt->vnet_hdr_len);
+    if (ret < 0) {
+        error_report("colo send primary packet failed");
+    }
+    trace_colo_compare_main("packet same and release packet");
+    packet_destroy(pkt, NULL);
+}
+
 /*
  * The IP packets sent by primary and secondary
  * will be compared in here
@@ -XXX,XX +XXX,XX @@ static int colo_compare_packet_payload(Packet *ppkt,
 }
 
 /*
- * Called from the compare thread on the primary
- * for compare tcp packet
- * compare_tcp copied from Dr. David Alan Gilbert's branch
- */
-static int colo_packet_compare_tcp(Packet *spkt, Packet *ppkt)
+ * return true means that the payload is consist and
+ * need to make the next comparison, false means do
+ * the checkpoint
+*/
+static bool colo_mark_tcp_pkt(Packet *ppkt, Packet *spkt,
+                              int8_t *mark, uint32_t max_ack)
 {
-    struct tcphdr *ptcp, *stcp;
-    int res;
+    *mark = 0;
+
+    if (ppkt->tcp_seq == spkt->tcp_seq && ppkt->seq_end == spkt->seq_end) {
+        if (colo_compare_packet_payload(ppkt, spkt,
+                                        ppkt->header_size, spkt->header_size,
+                                        ppkt->payload_size)) {
+            *mark = COLO_COMPARE_FREE_SECONDARY | COLO_COMPARE_FREE_PRIMARY;
+            return true;
+        }
+    }
+    if (ppkt->tcp_seq == spkt->tcp_seq && ppkt->seq_end == spkt->seq_end) {
+        if (colo_compare_packet_payload(ppkt, spkt,
+                                        ppkt->header_size, spkt->header_size,
+                                        ppkt->payload_size)) {
+            *mark = COLO_COMPARE_FREE_SECONDARY | COLO_COMPARE_FREE_PRIMARY;
+            return true;
+        }
+    }
+
+    /* one part of secondary packet payload still need to be compared */
+    if (!after(ppkt->seq_end, spkt->seq_end)) {
+        if (colo_compare_packet_payload(ppkt, spkt,
+                                        ppkt->header_size + ppkt->offset,
+                                        spkt->header_size + spkt->offset,
+                                        ppkt->payload_size - ppkt->offset)) {
+            if (!after(ppkt->tcp_ack, max_ack)) {
+                *mark = COLO_COMPARE_FREE_PRIMARY;
+                spkt->offset += ppkt->payload_size - ppkt->offset;
+                return true;
+            } else {
+                /* secondary guest hasn't ack the data, don't send
+                 * out this packet
+                 */
+                return false;
+            }
+        }
+    } else {
+        /* primary packet is longer than secondary packet, compare
+         * the same part and mark the primary packet offset
+         */
+        if (colo_compare_packet_payload(ppkt, spkt,
+                                        ppkt->header_size + ppkt->offset,
+                                        spkt->header_size + spkt->offset,
+                                        spkt->payload_size - spkt->offset)) {
+            *mark = COLO_COMPARE_FREE_SECONDARY;
+            ppkt->offset += spkt->payload_size - spkt->offset;
+            return true;
+        }
+    }
 
-    trace_colo_compare_main("compare tcp");
+    return false;
+}
 
-    ptcp = (struct tcphdr *)ppkt->transport_header;
-    stcp = (struct tcphdr *)spkt->transport_header;
+static void colo_compare_tcp(CompareState *s, Connection *conn)
+{
+    Packet *ppkt = NULL, *spkt = NULL;
+    int8_t mark;
 
     /*
-     * The 'identification' field in the IP header is *very* random
-     * it almost never matches.  Fudge this by ignoring differences in
-     * unfragmented packets; they'll normally sort themselves out if different
-     * anyway, and it should recover at the TCP level.
-     * An alternative would be to get both the primary and secondary to rewrite
-     * somehow; but that would need some sync traffic to sync the state
-     */
-    if (ntohs(ppkt->ip->ip_off) & IP_DF) {
-        spkt->ip->ip_id = ppkt->ip->ip_id;
-        /* and the sum will be different if the IDs were different */
-        spkt->ip->ip_sum = ppkt->ip->ip_sum;
+     * If ppkt and spkt have the same payload, but ppkt's ACK
+     * is greater than spkt's ACK, in this case we can not
+     * send the ppkt because it will cause the secondary guest
+     * to miss sending some data in the next. Therefore, we
+     * record the maximum ACK in the current queue at both
+     * primary side and secondary side. Only when the ack is
+     * less than the smaller of the two maximum ack, then we
+     * can ensure that the packet's payload is acknowledged by
+     * primary and secondary.
+    */
+    uint32_t min_ack = conn->pack > conn->sack ? conn->sack : conn->pack;
+
+pri:
+    if (g_queue_is_empty(&conn->primary_list)) {
+        return;
     }
+    ppkt = g_queue_pop_head(&conn->primary_list);
+sec:
+    if (g_queue_is_empty(&conn->secondary_list)) {
+        g_queue_push_head(&conn->primary_list, ppkt);
+        return;
+    }
+    spkt = g_queue_pop_head(&conn->secondary_list);
 
-    /*
-     * Check tcp header length for tcp option field.
-     * th_off > 5 means this tcp packet have options field.
-     * The tcp options maybe always different.
-     * for example:
-     * From RFC 7323.
-     * TCP Timestamps option (TSopt):
-     * Kind: 8
-     *
-     * Length: 10 bytes
-     *
-     *    +-------+-------+---------------------+---------------------+
-     *    |Kind=8 |  10   |   TS Value (TSval)  |TS Echo Reply (TSecr)|
-     *    +-------+-------+---------------------+---------------------+
-     *       1       1              4                     4
-     *
-     * In this case the primary guest's timestamp always different with
-     * the secondary guest's timestamp. COLO just focus on payload,
-     * so we just need skip this field.
-     */
+    if (ppkt->tcp_seq == ppkt->seq_end) {
+        colo_release_primary_pkt(s, ppkt);
+        ppkt = NULL;
+    }
 
-    ptrdiff_t ptcp_offset, stcp_offset;
+    if (ppkt && conn->compare_seq && !after(ppkt->seq_end, conn->compare_seq)) {
+        trace_colo_compare_main("pri: this packet has compared");
+        colo_release_primary_pkt(s, ppkt);
+        ppkt = NULL;
+    }
 
-    ptcp_offset = ppkt->transport_header - (uint8_t *)ppkt->data
-                  + (ptcp->th_off << 2) - ppkt->vnet_hdr_len;
-    stcp_offset = spkt->transport_header - (uint8_t *)spkt->data
-                  + (stcp->th_off << 2) - spkt->vnet_hdr_len;
-    if (ppkt->size - ptcp_offset == spkt->size - stcp_offset) {
-        res = colo_compare_packet_payload(ppkt, spkt,
-                                          ptcp_offset, stcp_offset,
-                                          ppkt->size - ptcp_offset);
+    if (spkt->tcp_seq == spkt->seq_end) {
+        packet_destroy(spkt, NULL);
+        if (!ppkt) {
+            goto pri;
+        } else {
+            goto sec;
+        }
     } else {
-        trace_colo_compare_main("TCP: payload size of packets are different");
-        res = -1;
+        if (conn->compare_seq && !after(spkt->seq_end, conn->compare_seq)) {
+            trace_colo_compare_main("sec: this packet has compared");
+            packet_destroy(spkt, NULL);
+            if (!ppkt) {
+                goto pri;
+            } else {
+                goto sec;
+            }
+        }
+        if (!ppkt) {
+            g_queue_push_head(&conn->secondary_list, spkt);
+            goto pri;
+        }
     }
 
-    if (res != 0 &&
-        trace_event_get_state_backends(TRACE_COLO_COMPARE_MISCOMPARE)) {
-        char pri_ip_src[20], pri_ip_dst[20], sec_ip_src[20], sec_ip_dst[20];
-
-        strcpy(pri_ip_src, inet_ntoa(ppkt->ip->ip_src));
-        strcpy(pri_ip_dst, inet_ntoa(ppkt->ip->ip_dst));
-        strcpy(sec_ip_src, inet_ntoa(spkt->ip->ip_src));
-        strcpy(sec_ip_dst, inet_ntoa(spkt->ip->ip_dst));
-
-        trace_colo_compare_ip_info(ppkt->size, pri_ip_src,
-                                   pri_ip_dst, spkt->size,
-                                   sec_ip_src, sec_ip_dst);
-
-        trace_colo_compare_tcp_info("pri tcp packet",
-                                    ntohl(ptcp->th_seq),
-                                    ntohl(ptcp->th_ack),
-                                    res, ptcp->th_flags,
-                                    ppkt->size);
-
-        trace_colo_compare_tcp_info("sec tcp packet",
-                                    ntohl(stcp->th_seq),
-                                    ntohl(stcp->th_ack),
-                                    res, stcp->th_flags,
-                                    spkt->size);
+    if (colo_mark_tcp_pkt(ppkt, spkt, &mark, min_ack)) {
+        trace_colo_compare_tcp_info("pri",
+                                    ppkt->tcp_seq, ppkt->tcp_ack,
+                                    ppkt->header_size, ppkt->payload_size,
+                                    ppkt->offset, ppkt->flags);
+
+        trace_colo_compare_tcp_info("sec",
+                                    spkt->tcp_seq, spkt->tcp_ack,
+                                    spkt->header_size, spkt->payload_size,
+                                    spkt->offset, spkt->flags);
+
+        if (mark == COLO_COMPARE_FREE_PRIMARY) {
+            conn->compare_seq = ppkt->seq_end;
+            colo_release_primary_pkt(s, ppkt);
+            g_queue_push_head(&conn->secondary_list, spkt);
+            goto pri;
+        }
+        if (mark == COLO_COMPARE_FREE_SECONDARY) {
+            conn->compare_seq = spkt->seq_end;
+            packet_destroy(spkt, NULL);
+            goto sec;
+        }
+        if (mark == (COLO_COMPARE_FREE_PRIMARY | COLO_COMPARE_FREE_SECONDARY)) {
+            conn->compare_seq = ppkt->seq_end;
+            colo_release_primary_pkt(s, ppkt);
+            packet_destroy(spkt, NULL);
+            goto pri;
+        }
+    } else {
+        g_queue_push_head(&conn->primary_list, ppkt);
+        g_queue_push_head(&conn->secondary_list, spkt);
 
         qemu_hexdump((char *)ppkt->data, stderr,
                      "colo-compare ppkt", ppkt->size);
         qemu_hexdump((char *)spkt->data, stderr,
                      "colo-compare spkt", spkt->size);
-    }
 
-    return res;
+        /*
+         * colo_compare_inconsistent_notify();
+         * TODO: notice to checkpoint();
+         */
+    }
 }
 
+
 /*
  * Called from the compare thread on the primary
  * for compare udp packet
@@ -XXX,XX +XXX,XX @@ static void colo_old_packet_check(void *opaque)
                         (GCompareFunc)colo_old_packet_check_one_conn);
 }
 
-/*
- * Called from the compare thread on the primary
- * for compare packet with secondary list of the
- * specified connection when a new packet was
- * queued to it.
- */
-static void colo_compare_connection(void *opaque, void *user_data)
+static void colo_compare_packet(CompareState *s, Connection *conn,
+                                int (*HandlePacket)(Packet *spkt,
+                                Packet *ppkt))
 {
-    CompareState *s = user_data;
-    Connection *conn = opaque;
     Packet *pkt = NULL;
     GList *result = NULL;
-    int ret;
 
     while (!g_queue_is_empty(&conn->primary_list) &&
            !g_queue_is_empty(&conn->secondary_list)) {
         pkt = g_queue_pop_head(&conn->primary_list);
-        switch (conn->ip_proto) {
-        case IPPROTO_TCP:
-            result = g_queue_find_custom(&conn->secondary_list,
-                     pkt, (GCompareFunc)colo_packet_compare_tcp);
-            break;
-        case IPPROTO_UDP:
-            result = g_queue_find_custom(&conn->secondary_list,
-                     pkt, (GCompareFunc)colo_packet_compare_udp);
-            break;
-        case IPPROTO_ICMP:
-            result = g_queue_find_custom(&conn->secondary_list,
-                     pkt, (GCompareFunc)colo_packet_compare_icmp);
-            break;
-        default:
-            result = g_queue_find_custom(&conn->secondary_list,
-                     pkt, (GCompareFunc)colo_packet_compare_other);
-            break;
-        }
+        result = g_queue_find_custom(&conn->secondary_list,
+                 pkt, (GCompareFunc)HandlePacket);
 
         if (result) {
-            ret = compare_chr_send(s,
-                                   pkt->data,
-                                   pkt->size,
-                                   pkt->vnet_hdr_len);
-            if (ret < 0) {
-                error_report("colo_send_primary_packet failed");
-            }
-            trace_colo_compare_main("packet same and release packet");
+            colo_release_primary_pkt(s, pkt);
             g_queue_remove(&conn->secondary_list, result->data);
-            packet_destroy(pkt, NULL);
         } else {
             /*
              * If one packet arrive late, the secondary_list or
@@ -XXX,XX +XXX,XX @@ static void colo_compare_connection(void *opaque, void *user_data)
     }
 }
 
+/*
+ * Called from the compare thread on the primary
+ * for compare packet with secondary list of the
+ * specified connection when a new packet was
+ * queued to it.
+ */
+static void colo_compare_connection(void *opaque, void *user_data)
+{
+    CompareState *s = user_data;
+    Connection *conn = opaque;
+
+    switch (conn->ip_proto) {
+    case IPPROTO_TCP:
+        colo_compare_tcp(s, conn);
+        break;
+    case IPPROTO_UDP:
+        colo_compare_packet(s, conn, colo_packet_compare_udp);
+        break;
+    case IPPROTO_ICMP:
+        colo_compare_packet(s, conn, colo_packet_compare_icmp);
+        break;
+    default:
+        colo_compare_packet(s, conn, colo_packet_compare_other);
+        break;
+    }
+}
+
 static int compare_chr_send(CompareState *s,
                             const uint8_t *buf,
                             uint32_t size,
diff --git a/net/colo.c b/net/colo.c
index XXXXXXX..XXXXXXX 100644
--- a/net/colo.c
+++ b/net/colo.c
@@ -XXX,XX +XXX,XX @@ Connection *connection_new(ConnectionKey *key)
     conn->processing = false;
     conn->offset = 0;
     conn->syn_flag = 0;
+    conn->pack = 0;
+    conn->sack = 0;
     g_queue_init(&conn->primary_list);
     g_queue_init(&conn->secondary_list);
 
@@ -XXX,XX +XXX,XX @@ Packet *packet_new(const void *data, int size, int vnet_hdr_len)
     pkt->size = size;
     pkt->creation_ms = qemu_clock_get_ms(QEMU_CLOCK_HOST);
     pkt->vnet_hdr_len = vnet_hdr_len;
+    pkt->tcp_seq = 0;
+    pkt->tcp_ack = 0;
+    pkt->seq_end = 0;
+    pkt->header_size = 0;
+    pkt->payload_size = 0;
+    pkt->offset = 0;
+    pkt->flags = 0;
 
     return pkt;
 }
diff --git a/net/colo.h b/net/colo.h
index XXXXXXX..XXXXXXX 100644
--- a/net/colo.h
+++ b/net/colo.h
@@ -XXX,XX +XXX,XX @@ typedef struct Packet {
     int64_t creation_ms;
     /* Get vnet_hdr_len from filter */
     uint32_t vnet_hdr_len;
+    uint32_t tcp_seq; /* sequence number */
+    uint32_t tcp_ack; /* acknowledgement number */
+    /* the sequence number of the last byte of the packet */
+    uint32_t seq_end;
+    uint8_t header_size;  /* the header length */
+    uint16_t payload_size; /* the payload length */
+    /* record the payload offset(the length that has been compared) */
+    uint16_t offset;
+    uint8_t flags; /* Flags(aka Control bits) */
 } Packet;
 
 typedef struct ConnectionKey {
@@ -XXX,XX +XXX,XX @@ typedef struct Connection {
     /* flag to enqueue unprocessed_connections */
     bool processing;
     uint8_t ip_proto;
+    /* record the sequence number that has been compared */
+    uint32_t compare_seq;
+    /* the maximum of acknowledgement number in primary_list queue */
+    uint32_t pack;
+    /* the maximum of acknowledgement number in secondary_list queue */
+    uint32_t sack;
     /* offset = secondary_seq - primary_seq */
     tcp_seq  offset;
     /*
diff --git a/net/trace-events b/net/trace-events
index XXXXXXX..XXXXXXX 100644
--- a/net/trace-events
+++ b/net/trace-events
@@ -XXX,XX +XXX,XX @@ colo_compare_icmp_miscompare(const char *sta, int size) ": %s = %d"
 colo_compare_ip_info(int psize, const char *sta, const char *stb, int ssize, const char *stc, const char *std) "ppkt size = %d, ip_src = %s, ip_dst = %s, spkt size = %d, ip_src = %s, ip_dst = %s"
 colo_old_packet_check_found(int64_t old_time) "%" PRId64
 colo_compare_miscompare(void) ""
-colo_compare_tcp_info(const char *pkt, uint32_t seq, uint32_t ack, int res, uint32_t flag, int size) "side: %s seq/ack= %u/%u res= %d flags= 0x%x pkt_size: %d\n"
+colo_compare_tcp_info(const char *pkt, uint32_t seq, uint32_t ack, int hdlen, int pdlen, int offset, int flags) "%s: seq/ack= %u/%u hdlen= %d pdlen= %d offset= %d flags=%d\n"
 
 # net/filter-rewriter.c
 colo_filter_rewriter_debug(void) ""
-- 
2.7.4

From: Thomas Huth <thuth@redhat.com>

QEMU can emulate hubs to connect NICs and netdevs. This is currently
primarily used for the mis-named 'vlan' feature of the networking
subsystem. Now the 'vlan' feature has been marked as deprecated, since
its name is rather confusing and the users often rather mis-configure
their network when trying to use it. But while the 'vlan' parameter
should be removed at one point in time, the basic idea of emulating
a hub in QEMU is still good: It's useful for bundling up the output of
multiple NICs into one single l2tp netdev for example.

Now to be able to use the hubport feature without 'vlan's, there is one
missing piece: The possibility to connect a hubport to a netdev, too.
This patch adds this possibility by introducing a new "netdev=..."
parameter to the hubports.

To bundle up the output of multiple NICs into one socket netdev, you can
now run QEMU with these parameters for example:

qemu-system-ppc64 ... -netdev socket,id=s1,connect=:11122 \
    -netdev hubport,hubid=1,id=h1,netdev=s1 \
    -netdev hubport,hubid=1,id=h2 -device e1000,netdev=h2 \
    -netdev hubport,hubid=1,id=h3 -device virtio-net-pci,netdev=h3

For using the socket netdev, you have got to start another QEMU as the
receiving side first, for example with network dumping enabled:

qemu-system-x86_64 -M isapc -netdev socket,id=s0,listen=:11122 \
    -device ne2k_isa,netdev=s0 \
    -object filter-dump,id=f1,netdev=s0,file=/tmp/dump.dat

After the ppc64 guest tried to boot from both NICs, you can see in the
dump file (using Wireshark, for example), that the output of both NICs
(the e1000 and the virtio-net-pci) has been successfully transfered
via the socket netdev in this case.

Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 net/hub.c       | 27 +++++++++++++++++++++------
 net/hub.h       |  3 ++-
 net/net.c       |  2 +-
 qapi/net.json   |  4 +++-
 qemu-options.hx |  8 +++++---
 5 files changed, 32 insertions(+), 12 deletions(-)

diff --git a/net/hub.c b/net/hub.c
index XXXXXXX..XXXXXXX 100644
--- a/net/hub.c
+++ b/net/hub.c
@@ -XXX,XX +XXX,XX @@
  */
 
 #include "qemu/osdep.h"
+#include "qapi/error.h"
 #include "monitor/monitor.h"
 #include "net/net.h"
 #include "clients.h"
@@ -XXX,XX +XXX,XX @@ static NetClientInfo net_hub_port_info = {
     .cleanup = net_hub_port_cleanup,
 };
 
-static NetHubPort *net_hub_port_new(NetHub *hub, const char *name)
+static NetHubPort *net_hub_port_new(NetHub *hub, const char *name,
+                                    NetClientState *hubpeer)
 {
     NetClientState *nc;
     NetHubPort *port;
@@ -XXX,XX +XXX,XX @@ static NetHubPort *net_hub_port_new(NetHub *hub, const char *name)
         name = default_name;
     }
 
-    nc = qemu_new_net_client(&net_hub_port_info, NULL, "hub", name);
+    nc = qemu_new_net_client(&net_hub_port_info, hubpeer, "hub", name);
     port = DO_UPCAST(NetHubPort, nc, nc);
     port->id = id;
     port->hub = hub;
@@ -XXX,XX +XXX,XX @@ static NetHubPort *net_hub_port_new(NetHub *hub, const char *name)
 
 /**
  * Create a port on a given hub
+ * @hub_id: Number of the hub
  * @name: Net client name or NULL for default name.
+ * @hubpeer: Peer to use (if "netdev=id" has been specified)
  *
  * If there is no existing hub with the given id then a new hub is created.
  */
-NetClientState *net_hub_add_port(int hub_id, const char *name)
+NetClientState *net_hub_add_port(int hub_id, const char *name,
+                                 NetClientState *hubpeer)
 {
     NetHub *hub;
     NetHubPort *port;
@@ -XXX,XX +XXX,XX @@ NetClientState *net_hub_add_port(int hub_id, const char *name)
         hub = net_hub_new(hub_id);
     }
 
-    port = net_hub_port_new(hub, name);
+    port = net_hub_port_new(hub, name, hubpeer);
     return &port->nc;
 }
 
@@ -XXX,XX +XXX,XX @@ NetClientState *net_hub_port_find(int hub_id)
         }
     }
 
-    nc = net_hub_add_port(hub_id, NULL);
+    nc = net_hub_add_port(hub_id, NULL, NULL);
     return nc;
 }
 
@@ -XXX,XX +XXX,XX @@ int net_init_hubport(const Netdev *netdev, const char *name,
                      NetClientState *peer, Error **errp)
 {
     const NetdevHubPortOptions *hubport;
+    NetClientState *hubpeer = NULL;
 
     assert(netdev->type == NET_CLIENT_DRIVER_HUBPORT);
     assert(!peer);
     hubport = &netdev->u.hubport;
 
-    net_hub_add_port(hubport->hubid, name);
+    if (hubport->has_netdev) {
+        hubpeer = qemu_find_netdev(hubport->netdev);
+        if (!hubpeer) {
+            error_setg(errp, "netdev '%s' not found", hubport->netdev);
+            return -1;
+        }
+    }
+
+    net_hub_add_port(hubport->hubid, name, hubpeer);
+
     return 0;
 }
 
diff --git a/net/hub.h b/net/hub.h
index XXXXXXX..XXXXXXX 100644
--- a/net/hub.h
+++ b/net/hub.h
@@ -XXX,XX +XXX,XX @@
 
 #include "qemu-common.h"
 
-NetClientState *net_hub_add_port(int hub_id, const char *name);
+NetClientState *net_hub_add_port(int hub_id, const char *name,
+                                 NetClientState *hubpeer);
 NetClientState *net_hub_find_client_by_name(int hub_id, const char *name);
 void net_hub_info(Monitor *mon);
 void net_hub_check_clients(void);
diff --git a/net/net.c b/net/net.c
index XXXXXXX..XXXXXXX 100644
--- a/net/net.c
+++ b/net/net.c
@@ -XXX,XX +XXX,XX @@ static int net_client_init1(const void *object, bool is_netdev, Error **errp)
         /* Do not add to a vlan if it's a nic with a netdev= parameter. */
         if (netdev->type != NET_CLIENT_DRIVER_NIC ||
             !opts->u.nic.has_netdev) {
-            peer = net_hub_add_port(net->has_vlan ? net->vlan : 0, NULL);
+            peer = net_hub_add_port(net->has_vlan ? net->vlan : 0, NULL, NULL);
         }
 
         if (net->has_vlan && !vlan_warned) {
diff --git a/qapi/net.json b/qapi/net.json
index XXXXXXX..XXXXXXX 100644
--- a/qapi/net.json
+++ b/qapi/net.json
@@ -XXX,XX +XXX,XX @@
 # Connect two or more net clients through a software hub.
 #
 # @hubid: hub identifier number
+# @netdev: used to connect hub to a netdev instead of a device (since 2.12)
 #
 # Since: 1.2
 ##
 { 'struct': 'NetdevHubPortOptions',
   'data': {
-    'hubid':     'int32' } }
+    'hubid':     'int32',
+    '*netdev':    'str' } }
 
 ##
 # @NetdevNetmapOptions:
diff --git a/qemu-options.hx b/qemu-options.hx
index XXXXXXX..XXXXXXX 100644
--- a/qemu-options.hx
+++ b/qemu-options.hx
@@ -XXX,XX +XXX,XX @@ DEF("netdev", HAS_ARG, QEMU_OPTION_netdev,
 #endif
     "-netdev vhost-user,id=str,chardev=dev[,vhostforce=on|off]\n"
     "                configure a vhost-user network, backed by a chardev 'dev'\n"
-    "-netdev hubport,id=str,hubid=n\n"
+    "-netdev hubport,id=str,hubid=n[,netdev=nd]\n"
     "                configure a hub port on QEMU VLAN 'n'\n", QEMU_ARCH_ALL)
 DEF("net", HAS_ARG, QEMU_OPTION_net,
     "-net nic[,vlan=n][,netdev=nd][,macaddr=mac][,model=type][,name=str][,addr=str][,vectors=v]\n"
@@ -XXX,XX +XXX,XX @@ vde_switch -F -sock /tmp/myswitch
 qemu-system-i386 linux.img -net nic -net vde,sock=/tmp/myswitch
 @end example
 
-@item -netdev hubport,id=@var{id},hubid=@var{hubid}
+@item -netdev hubport,id=@var{id},hubid=@var{hubid}[,netdev=@var{nd}]
 
 Create a hub port on QEMU "vlan" @var{hubid}.
 
 The hubport netdev lets you connect a NIC to a QEMU "vlan" instead of a single
 netdev.  @code{-net} and @code{-device} with parameter @option{vlan} create the
-required hub automatically.
+required hub automatically. Alternatively, you can also connect the hubport
+to another netdev with ID @var{nd} by using the @option{netdev=@var{nd}}
+option.
 
 @item -netdev vhost-user,chardev=@var{id}[,vhostforce=on|off][,queues=n]
 
-- 
2.7.4

From: Thomas Huth <thuth@redhat.com>

It does not make much sense to limit these commands to the legacy 'vlan'
concept only, they should work with the modern netdevs, too. So now
it is possible to use this command with one, two or three parameters.

With one parameter, the command installs a hostfwd rule on the default
"user" network:
    hostfwd_add tcp:...

With two parameters, the command installs a hostfwd rule on a netdev
(that's the new way of using this command):
    hostfwd_add netdev_id tcp:...

With three parameters, the command installs a rule on a 'vlan' (aka hub):
    hostfwd_add hub_id name tcp:...

Same applies to the hostfwd_remove command now.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 hmp-commands.hx |  4 ++--
 net/slirp.c     | 33 +++++++++++++++++++++++----------
 2 files changed, 25 insertions(+), 12 deletions(-)

diff --git a/hmp-commands.hx b/hmp-commands.hx
index XXXXXXX..XXXXXXX 100644
--- a/hmp-commands.hx
+++ b/hmp-commands.hx
@@ -XXX,XX +XXX,XX @@ ETEXI
     {
         .name       = "hostfwd_add",
         .args_type  = "arg1:s,arg2:s?,arg3:s?",
-        .params     = "[vlan_id name] [tcp|udp]:[hostaddr]:hostport-[guestaddr]:guestport",
+        .params     = "[hub_id name]|[netdev_id] [tcp|udp]:[hostaddr]:hostport-[guestaddr]:guestport",
         .help       = "redirect TCP or UDP connections from host to guest (requires -net user)",
         .cmd        = hmp_hostfwd_add,
     },
@@ -XXX,XX +XXX,XX @@ ETEXI
     {
         .name       = "hostfwd_remove",
         .args_type  = "arg1:s,arg2:s?,arg3:s?",
-        .params     = "[vlan_id name] [tcp|udp]:[hostaddr]:hostport",
+        .params     = "[hub_id name]|[netdev_id] [tcp|udp]:[hostaddr]:hostport",
         .help       = "remove host-to-guest TCP or UDP redirection",
         .cmd        = hmp_hostfwd_remove,
     },
diff --git a/net/slirp.c b/net/slirp.c
index XXXXXXX..XXXXXXX 100644
--- a/net/slirp.c
+++ b/net/slirp.c
@@ -XXX,XX +XXX,XX @@ error:
     return -1;
 }
 
-static SlirpState *slirp_lookup(Monitor *mon, const char *vlan,
-                                const char *stack)
+static SlirpState *slirp_lookup(Monitor *mon, const char *hub_id,
+                                const char *name)
 {
-
-    if (vlan) {
+    if (name) {
         NetClientState *nc;
-        nc = net_hub_find_client_by_name(strtol(vlan, NULL, 0), stack);
-        if (!nc) {
-            monitor_printf(mon, "unrecognized (vlan-id, stackname) pair\n");
-            return NULL;
+        if (hub_id) {
+            nc = net_hub_find_client_by_name(strtol(hub_id, NULL, 0), name);
+            if (!nc) {
+                monitor_printf(mon, "unrecognized (vlan-id, stackname) pair\n");
+                return NULL;
+            }
+        } else {
+            nc = qemu_find_netdev(name);
+            if (!nc) {
+                monitor_printf(mon, "unrecognized netdev id '%s'\n", name);
+                return NULL;
+            }
         }
         if (strcmp(nc->model, "user")) {
             monitor_printf(mon, "invalid device specified\n");
@@ -XXX,XX +XXX,XX @@ void hmp_hostfwd_remove(Monitor *mon, const QDict *qdict)
     const char *arg2 = qdict_get_try_str(qdict, "arg2");
     const char *arg3 = qdict_get_try_str(qdict, "arg3");
 
-    if (arg2) {
+    if (arg3) {
         s = slirp_lookup(mon, arg1, arg2);
         src_str = arg3;
+    } else if (arg2) {
+        s = slirp_lookup(mon, NULL, arg1);
+        src_str = arg2;
     } else {
         s = slirp_lookup(mon, NULL, NULL);
         src_str = arg1;
@@ -XXX,XX +XXX,XX @@ void hmp_hostfwd_add(Monitor *mon, const QDict *qdict)
     const char *arg2 = qdict_get_try_str(qdict, "arg2");
     const char *arg3 = qdict_get_try_str(qdict, "arg3");
 
-    if (arg2) {
+    if (arg3) {
         s = slirp_lookup(mon, arg1, arg2);
         redir_str = arg3;
+    } else if (arg2) {
+        s = slirp_lookup(mon, NULL, arg1);
+        redir_str = arg2;
     } else {
         s = slirp_lookup(mon, NULL, NULL);
         redir_str = arg1;
-- 
2.7.4

From: Philippe Mathieu-Daudé <f4bug@amsat.org>

gently asked by his automatic reply :)

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 MAINTAINERS | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/MAINTAINERS b/MAINTAINERS
index XXXXXXX..XXXXXXX 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -XXX,XX +XXX,XX @@ F: hw/scsi/mfi.h
 F: tests/megasas-test.c
 
 Network packet abstractions
-M: Dmitry Fleytman <dmitry@daynix.com>
+M: Dmitry Fleytman <dmitry.fleytman@gmail.com>
 S: Maintained
 F: include/net/eth.h
 F: net/eth.c
@@ -XXX,XX +XXX,XX @@ F: hw/net/net_rx_pkt*
 F: hw/net/net_tx_pkt*
 
 Vmware
-M: Dmitry Fleytman <dmitry@daynix.com>
+M: Dmitry Fleytman <dmitry.fleytman@gmail.com>
 S: Maintained
 F: hw/net/vmxnet*
 F: hw/scsi/vmw_pvscsi*
@@ -XXX,XX +XXX,XX @@ F: hw/mem/nvdimm.c
 F: include/hw/mem/nvdimm.h
 
 e1000x
-M: Dmitry Fleytman <dmitry@daynix.com>
+M: Dmitry Fleytman <dmitry.fleytman@gmail.com>
 S: Maintained
 F: hw/net/e1000x*
 
 e1000e
-M: Dmitry Fleytman <dmitry@daynix.com>
+M: Dmitry Fleytman <dmitry.fleytman@gmail.com>
 S: Maintained
 F: hw/net/e1000e*
 
-- 
2.7.4

The following changes since commit f45cc81911adc7726e8a2801986b6998b91b816e:

Merge remote-tracking branch 'remotes/cschoenebeck/tags/pull-9p-20220307' into staging (2022-03-08 09:06:57 +0000)

are available in the git repository at:

https://github.com/jasowang/qemu.git tags/net-pull-request

for you to fetch changes up to a10dd1e279fc56cebc7e738925e0db85d0cea089:

vdpa: Expose VHOST_F_LOG_ALL on SVQ (2022-03-08 21:18:41 +0800)

----------------------------------------------------------------

----------------------------------------------------------------
Eugenio Pérez (14):
      vhost: Add VhostShadowVirtqueue
      vhost: Add Shadow VirtQueue kick forwarding capabilities
      vhost: Add Shadow VirtQueue call forwarding capabilities
      vhost: Add vhost_svq_valid_features to shadow vq
      virtio: Add vhost_svq_get_vring_addr
      vdpa: adapt vhost_ops callbacks to svq
      vhost: Shadow virtqueue buffers forwarding
      util: Add iova_tree_alloc_map
      util: add iova_tree_find_iova
      vhost: Add VhostIOVATree
      vdpa: Add custom IOTLB translations to SVQ
      vdpa: Adapt vhost_vdpa_get_vring_base to SVQ
      vdpa: Never set log_base addr if SVQ is enabled
      vdpa: Expose VHOST_F_LOG_ALL on SVQ

Jason Wang (1):
      virtio-net: fix map leaking on error during receive

hw/net/virtio-net.c                |   1 +
 hw/virtio/meson.build              |   2 +-
 hw/virtio/vhost-iova-tree.c        | 110 +++++++
 hw/virtio/vhost-iova-tree.h        |  27 ++
 hw/virtio/vhost-shadow-virtqueue.c | 637 +++++++++++++++++++++++++++++++++++++
 hw/virtio/vhost-shadow-virtqueue.h |  87 +++++
 hw/virtio/vhost-vdpa.c             | 525 +++++++++++++++++++++++++++++-
 include/hw/virtio/vhost-vdpa.h     |   8 +
 include/qemu/iova-tree.h           |  38 ++-
 util/iova-tree.c                   | 169 ++++++++++
 10 files changed, 1587 insertions(+), 17 deletions(-)
 create mode 100644 hw/virtio/vhost-iova-tree.c
 create mode 100644 hw/virtio/vhost-iova-tree.h
 create mode 100644 hw/virtio/vhost-shadow-virtqueue.c
 create mode 100644 hw/virtio/vhost-shadow-virtqueue.h

Commit bedd7e93d0196 ("virtio-net: fix use after unmap/free for sg")
tries to fix the use after free of the sg by caching the virtqueue
elements in an array and unmap them at once after receiving the
packets, But it forgot to unmap the cached elements on error which
will lead to leaking of mapping and other unexpected results.

Fixing this by detaching the cached elements on error. This addresses
CVE-2022-26353.

Reported-by: Victor Tom <vv474172261@gmail.com>
Cc: qemu-stable@nongnu.org
Fixes: CVE-2022-26353
Fixes: bedd7e93d0196 ("virtio-net: fix use after unmap/free for sg")
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 hw/net/virtio-net.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/net/virtio-net.c
+++ b/hw/net/virtio-net.c
@@ -XXX,XX +XXX,XX @@ static ssize_t virtio_net_receive_rcu(NetClientState *nc, const uint8_t *buf,
 
 err:
     for (j = 0; j < i; j++) {
+        virtqueue_detach_element(q->rx_vq, elems[j], lens[j]);
         g_free(elems[j]);
     }
 
-- 
2.7.4

From: Eugenio Pérez <eperezma@redhat.com>

Vhost shadow virtqueue (SVQ) is an intermediate jump for virtqueue
notifications and buffers, allowing qemu to track them. While qemu is
forwarding the buffers and virtqueue changes, it is able to commit the
memory it's being dirtied, the same way regular qemu's VirtIO devices
do.

This commit only exposes basic SVQ allocation and free. Next patches of
the series add functionality like notifications and buffers forwarding.

Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 hw/virtio/meson.build              |  2 +-
 hw/virtio/vhost-shadow-virtqueue.c | 62 ++++++++++++++++++++++++++++++++++++++
 hw/virtio/vhost-shadow-virtqueue.h | 28 +++++++++++++++++
 3 files changed, 91 insertions(+), 1 deletion(-)
 create mode 100644 hw/virtio/vhost-shadow-virtqueue.c
 create mode 100644 hw/virtio/vhost-shadow-virtqueue.h

diff --git a/hw/virtio/meson.build b/hw/virtio/meson.build
index XXXXXXX..XXXXXXX 100644
--- a/hw/virtio/meson.build
+++ b/hw/virtio/meson.build
@@ -XXX,XX +XXX,XX @@ softmmu_ss.add(when: 'CONFIG_ALL', if_true: files('vhost-stub.c'))
 
 virtio_ss = ss.source_set()
 virtio_ss.add(files('virtio.c'))
-virtio_ss.add(when: 'CONFIG_VHOST', if_true: files('vhost.c', 'vhost-backend.c'))
+virtio_ss.add(when: 'CONFIG_VHOST', if_true: files('vhost.c', 'vhost-backend.c', 'vhost-shadow-virtqueue.c'))
 virtio_ss.add(when: 'CONFIG_VHOST_USER', if_true: files('vhost-user.c'))
 virtio_ss.add(when: 'CONFIG_VHOST_VDPA', if_true: files('vhost-vdpa.c'))
 virtio_ss.add(when: 'CONFIG_VIRTIO_BALLOON', if_true: files('virtio-balloon.c'))
diff --git a/hw/virtio/vhost-shadow-virtqueue.c b/hw/virtio/vhost-shadow-virtqueue.c
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/hw/virtio/vhost-shadow-virtqueue.c
@@ -XXX,XX +XXX,XX @@
+/*
+ * vhost shadow virtqueue
+ *
+ * SPDX-FileCopyrightText: Red Hat, Inc. 2021
+ * SPDX-FileContributor: Author: Eugenio Pérez <eperezma@redhat.com>
+ *
+ * SPDX-License-Identifier: GPL-2.0-or-later
+ */
+
+#include "qemu/osdep.h"
+#include "hw/virtio/vhost-shadow-virtqueue.h"
+
+#include "qemu/error-report.h"
+
+/**
+ * Creates vhost shadow virtqueue, and instructs the vhost device to use the
+ * shadow methods and file descriptors.
+ *
+ * Returns the new virtqueue or NULL.
+ *
+ * In case of error, reason is reported through error_report.
+ */
+VhostShadowVirtqueue *vhost_svq_new(void)
+{
+    g_autofree VhostShadowVirtqueue *svq = g_new0(VhostShadowVirtqueue, 1);
+    int r;
+
+    r = event_notifier_init(&svq->hdev_kick, 0);
+    if (r != 0) {
+        error_report("Couldn't create kick event notifier: %s (%d)",
+                     g_strerror(errno), errno);
+        goto err_init_hdev_kick;
+    }
+
+    r = event_notifier_init(&svq->hdev_call, 0);
+    if (r != 0) {
+        error_report("Couldn't create call event notifier: %s (%d)",
+                     g_strerror(errno), errno);
+        goto err_init_hdev_call;
+    }
+
+    return g_steal_pointer(&svq);
+
+err_init_hdev_call:
+    event_notifier_cleanup(&svq->hdev_kick);
+
+err_init_hdev_kick:
+    return NULL;
+}
+
+/**
+ * Free the resources of the shadow virtqueue.
+ *
+ * @pvq: gpointer to SVQ so it can be used by autofree functions.
+ */
+void vhost_svq_free(gpointer pvq)
+{
+    VhostShadowVirtqueue *vq = pvq;
+    event_notifier_cleanup(&vq->hdev_kick);
+    event_notifier_cleanup(&vq->hdev_call);
+    g_free(vq);
+}
diff --git a/hw/virtio/vhost-shadow-virtqueue.h b/hw/virtio/vhost-shadow-virtqueue.h
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/hw/virtio/vhost-shadow-virtqueue.h
@@ -XXX,XX +XXX,XX @@
+/*
+ * vhost shadow virtqueue
+ *
+ * SPDX-FileCopyrightText: Red Hat, Inc. 2021
+ * SPDX-FileContributor: Author: Eugenio Pérez <eperezma@redhat.com>
+ *
+ * SPDX-License-Identifier: GPL-2.0-or-later
+ */
+
+#ifndef VHOST_SHADOW_VIRTQUEUE_H
+#define VHOST_SHADOW_VIRTQUEUE_H
+
+#include "qemu/event_notifier.h"
+
+/* Shadow virtqueue to relay notifications */
+typedef struct VhostShadowVirtqueue {
+    /* Shadow kick notifier, sent to vhost */
+    EventNotifier hdev_kick;
+    /* Shadow call notifier, sent to vhost */
+    EventNotifier hdev_call;
+} VhostShadowVirtqueue;
+
+VhostShadowVirtqueue *vhost_svq_new(void);
+
+void vhost_svq_free(gpointer vq);
+G_DEFINE_AUTOPTR_CLEANUP_FUNC(VhostShadowVirtqueue, vhost_svq_free);
+
+#endif
-- 
2.7.4

From: Eugenio Pérez <eperezma@redhat.com>

At this mode no buffer forwarding will be performed in SVQ mode: Qemu
will just forward the guest's kicks to the device.

Host memory notifiers regions are left out for simplicity, and they will
not be addressed in this series.

Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 hw/virtio/vhost-shadow-virtqueue.c |  56 ++++++++++++++
 hw/virtio/vhost-shadow-virtqueue.h |  14 ++++
 hw/virtio/vhost-vdpa.c             | 145 ++++++++++++++++++++++++++++++++++++-
 include/hw/virtio/vhost-vdpa.h     |   4 +
 4 files changed, 217 insertions(+), 2 deletions(-)

From: Eugenio Pérez <eperezma@redhat.com>

This will make qemu aware of the device used buffers, allowing it to
write the guest memory with its contents if needed.

Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 hw/virtio/vhost-shadow-virtqueue.c | 38 ++++++++++++++++++++++++++++++++++++++
 hw/virtio/vhost-shadow-virtqueue.h |  4 ++++
 hw/virtio/vhost-vdpa.c             | 31 +++++++++++++++++++++++++++++--
 3 files changed, 71 insertions(+), 2 deletions(-)

diff --git a/hw/virtio/vhost-shadow-virtqueue.c b/hw/virtio/vhost-shadow-virtqueue.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/virtio/vhost-shadow-virtqueue.c
+++ b/hw/virtio/vhost-shadow-virtqueue.c
@@ -XXX,XX +XXX,XX @@ static void vhost_handle_guest_kick(EventNotifier *n)
 }
 
 /**
+ * Forward vhost notifications
+ *
+ * @n: hdev call event notifier, the one that device set to notify svq.
+ */
+static void vhost_svq_handle_call(EventNotifier *n)
+{
+    VhostShadowVirtqueue *svq = container_of(n, VhostShadowVirtqueue,
+                                             hdev_call);
+    event_notifier_test_and_clear(n);
+    event_notifier_set(&svq->svq_call);
+}
+
+/**
+ * Set the call notifier for the SVQ to call the guest
+ *
+ * @svq: Shadow virtqueue
+ * @call_fd: call notifier
+ *
+ * Called on BQL context.
+ */
+void vhost_svq_set_svq_call_fd(VhostShadowVirtqueue *svq, int call_fd)
+{
+    if (call_fd == VHOST_FILE_UNBIND) {
+        /*
+         * Fail event_notifier_set if called handling device call.
+         *
+         * SVQ still needs device notifications, since it needs to keep
+         * forwarding used buffers even with the unbind.
+         */
+        memset(&svq->svq_call, 0, sizeof(svq->svq_call));
+    } else {
+        event_notifier_init_fd(&svq->svq_call, call_fd);
+    }
+}
+
+/**
  * Set a new file descriptor for the guest to kick the SVQ and notify for avail
  *
  * @svq: The svq
@@ -XXX,XX +XXX,XX @@ VhostShadowVirtqueue *vhost_svq_new(void)
     }
 
     event_notifier_init_fd(&svq->svq_kick, VHOST_FILE_UNBIND);
+    event_notifier_set_handler(&svq->hdev_call, vhost_svq_handle_call);
     return g_steal_pointer(&svq);
 
 err_init_hdev_call:
@@ -XXX,XX +XXX,XX @@ void vhost_svq_free(gpointer pvq)
     VhostShadowVirtqueue *vq = pvq;
     vhost_svq_stop(vq);
     event_notifier_cleanup(&vq->hdev_kick);
+    event_notifier_set_handler(&vq->hdev_call, NULL);
     event_notifier_cleanup(&vq->hdev_call);
     g_free(vq);
 }
diff --git a/hw/virtio/vhost-shadow-virtqueue.h b/hw/virtio/vhost-shadow-virtqueue.h
index XXXXXXX..XXXXXXX 100644
--- a/hw/virtio/vhost-shadow-virtqueue.h
+++ b/hw/virtio/vhost-shadow-virtqueue.h
@@ -XXX,XX +XXX,XX @@ typedef struct VhostShadowVirtqueue {
      * So shadow virtqueue must not clean it, or we would lose VirtQueue one.
      */
     EventNotifier svq_kick;
+
+    /* Guest's call notifier, where the SVQ calls guest. */
+    EventNotifier svq_call;
 } VhostShadowVirtqueue;
 
 void vhost_svq_set_svq_kick_fd(VhostShadowVirtqueue *svq, int svq_kick_fd);
+void vhost_svq_set_svq_call_fd(VhostShadowVirtqueue *svq, int call_fd);
 
 void vhost_svq_stop(VhostShadowVirtqueue *svq);
 
diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/virtio/vhost-vdpa.c
+++ b/hw/virtio/vhost-vdpa.c
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_vring_dev_kick(struct vhost_dev *dev,
     return vhost_vdpa_call(dev, VHOST_SET_VRING_KICK, file);
 }
 
+static int vhost_vdpa_set_vring_dev_call(struct vhost_dev *dev,
+                                         struct vhost_vring_file *file)
+{
+    trace_vhost_vdpa_set_vring_call(dev, file->index, file->fd);
+    return vhost_vdpa_call(dev, VHOST_SET_VRING_CALL, file);
+}
+
 /**
  * Set the shadow virtqueue descriptors to the device
  *
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_vring_dev_kick(struct vhost_dev *dev,
  * @svq: The shadow virtqueue
  * @idx: The index of the virtqueue in the vhost device
  * @errp: Error
+ *
+ * Note that this function does not rewind kick file descriptor if cannot set
+ * call one.
  */
 static bool vhost_vdpa_svq_setup(struct vhost_dev *dev,
                                  VhostShadowVirtqueue *svq,
@@ -XXX,XX +XXX,XX @@ static bool vhost_vdpa_svq_setup(struct vhost_dev *dev,
     r = vhost_vdpa_set_vring_dev_kick(dev, &file);
     if (unlikely(r != 0)) {
         error_setg_errno(errp, -r, "Can't set device kick fd");
+        return false;
+    }
+
+    event_notifier = &svq->hdev_call;
+    file.fd = event_notifier_get_fd(event_notifier);
+    r = vhost_vdpa_set_vring_dev_call(dev, &file);
+    if (unlikely(r != 0)) {
+        error_setg_errno(errp, -r, "Can't set device call fd");
     }
 
     return r == 0;
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_vring_kick(struct vhost_dev *dev,
 static int vhost_vdpa_set_vring_call(struct vhost_dev *dev,
                                        struct vhost_vring_file *file)
 {
-    trace_vhost_vdpa_set_vring_call(dev, file->index, file->fd);
-    return vhost_vdpa_call(dev, VHOST_SET_VRING_CALL, file);
+    struct vhost_vdpa *v = dev->opaque;
+
+    if (v->shadow_vqs_enabled) {
+        int vdpa_idx = file->index - dev->vq_index;
+        VhostShadowVirtqueue *svq = g_ptr_array_index(v->shadow_vqs, vdpa_idx);
+
+        vhost_svq_set_svq_call_fd(svq, file->fd);
+        return 0;
+    } else {
+        return vhost_vdpa_set_vring_dev_call(dev, file);
+    }
 }
 
 static int vhost_vdpa_get_features(struct vhost_dev *dev,
-- 
2.7.4

From: Eugenio Pérez <eperezma@redhat.com>

This allows SVQ to negotiate features with the guest and the device. For
the device, SVQ is a driver. While this function bypasses all
non-transport features, it needs to disable the features that SVQ does
not support when forwarding buffers. This includes packed vq layout,
indirect descriptors or event idx.

Future changes can add support to offer more features to the guest,
since the use of VirtQueue gives this for free. This is left out at the
moment for simplicity.

Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 hw/virtio/vhost-shadow-virtqueue.c | 44 ++++++++++++++++++++++++++++++++++++++
 hw/virtio/vhost-shadow-virtqueue.h |  2 ++
 hw/virtio/vhost-vdpa.c             | 15 +++++++++++++
 3 files changed, 61 insertions(+)

From: Eugenio Pérez <eperezma@redhat.com>

It reports the shadow virtqueue address from qemu virtual address space.

Since this will be different from the guest's vaddr, but the device can
access it, SVQ takes special care about its alignment & lack of garbage
data. It assumes that IOMMU will work in host_page_size ranges for that.

Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 hw/virtio/vhost-shadow-virtqueue.c | 29 +++++++++++++++++++++++++++++
 hw/virtio/vhost-shadow-virtqueue.h |  9 +++++++++
 2 files changed, 38 insertions(+)

diff --git a/hw/virtio/vhost-shadow-virtqueue.c b/hw/virtio/vhost-shadow-virtqueue.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/virtio/vhost-shadow-virtqueue.c
+++ b/hw/virtio/vhost-shadow-virtqueue.c
@@ -XXX,XX +XXX,XX @@ void vhost_svq_set_svq_call_fd(VhostShadowVirtqueue *svq, int call_fd)
 }
 
 /**
+ * Get the shadow vq vring address.
+ * @svq: Shadow virtqueue
+ * @addr: Destination to store address
+ */
+void vhost_svq_get_vring_addr(const VhostShadowVirtqueue *svq,
+                              struct vhost_vring_addr *addr)
+{
+    addr->desc_user_addr = (uint64_t)svq->vring.desc;
+    addr->avail_user_addr = (uint64_t)svq->vring.avail;
+    addr->used_user_addr = (uint64_t)svq->vring.used;
+}
+
+size_t vhost_svq_driver_area_size(const VhostShadowVirtqueue *svq)
+{
+    size_t desc_size = sizeof(vring_desc_t) * svq->vring.num;
+    size_t avail_size = offsetof(vring_avail_t, ring) +
+                                             sizeof(uint16_t) * svq->vring.num;
+
+    return ROUND_UP(desc_size + avail_size, qemu_real_host_page_size);
+}
+
+size_t vhost_svq_device_area_size(const VhostShadowVirtqueue *svq)
+{
+    size_t used_size = offsetof(vring_used_t, ring) +
+                                    sizeof(vring_used_elem_t) * svq->vring.num;
+    return ROUND_UP(used_size, qemu_real_host_page_size);
+}
+
+/**
  * Set a new file descriptor for the guest to kick the SVQ and notify for avail
  *
  * @svq: The svq
diff --git a/hw/virtio/vhost-shadow-virtqueue.h b/hw/virtio/vhost-shadow-virtqueue.h
index XXXXXXX..XXXXXXX 100644
--- a/hw/virtio/vhost-shadow-virtqueue.h
+++ b/hw/virtio/vhost-shadow-virtqueue.h
@@ -XXX,XX +XXX,XX @@
 #define VHOST_SHADOW_VIRTQUEUE_H
 
 #include "qemu/event_notifier.h"
+#include "hw/virtio/virtio.h"
+#include "standard-headers/linux/vhost_types.h"
 
 /* Shadow virtqueue to relay notifications */
 typedef struct VhostShadowVirtqueue {
+    /* Shadow vring */
+    struct vring vring;
+
     /* Shadow kick notifier, sent to vhost */
     EventNotifier hdev_kick;
     /* Shadow call notifier, sent to vhost */
@@ -XXX,XX +XXX,XX @@ bool vhost_svq_valid_features(uint64_t features, Error **errp);
 
 void vhost_svq_set_svq_kick_fd(VhostShadowVirtqueue *svq, int svq_kick_fd);
 void vhost_svq_set_svq_call_fd(VhostShadowVirtqueue *svq, int call_fd);
+void vhost_svq_get_vring_addr(const VhostShadowVirtqueue *svq,
+                              struct vhost_vring_addr *addr);
+size_t vhost_svq_driver_area_size(const VhostShadowVirtqueue *svq);
+size_t vhost_svq_device_area_size(const VhostShadowVirtqueue *svq);
 
 void vhost_svq_stop(VhostShadowVirtqueue *svq);
 
-- 
2.7.4

From: Eugenio Pérez <eperezma@redhat.com>

First half of the buffers forwarding part, preparing vhost-vdpa
callbacks to SVQ to offer it. QEMU cannot enable it at this moment, so
this is effectively dead code at the moment, but it helps to reduce
patch size.

diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/virtio/vhost-vdpa.c
+++ b/hw/virtio/vhost-vdpa.c
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_get_config(struct vhost_dev *dev, uint8_t *config,
     return ret;
  }
 
+static int vhost_vdpa_set_dev_vring_base(struct vhost_dev *dev,
+                                         struct vhost_vring_state *ring)
+{
+    trace_vhost_vdpa_set_vring_base(dev, ring->index, ring->num);
+    return vhost_vdpa_call(dev, VHOST_SET_VRING_BASE, ring);
+}
+
 static int vhost_vdpa_set_vring_dev_kick(struct vhost_dev *dev,
                                          struct vhost_vring_file *file)
 {
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_vring_dev_call(struct vhost_dev *dev,
     return vhost_vdpa_call(dev, VHOST_SET_VRING_CALL, file);
 }
 
+static int vhost_vdpa_set_vring_dev_addr(struct vhost_dev *dev,
+                                         struct vhost_vring_addr *addr)
+{
+    trace_vhost_vdpa_set_vring_addr(dev, addr->index, addr->flags,
+                                addr->desc_user_addr, addr->used_user_addr,
+                                addr->avail_user_addr,
+                                addr->log_guest_addr);
+
+    return vhost_vdpa_call(dev, VHOST_SET_VRING_ADDR, addr);
+
+}
+
 /**
  * Set the shadow virtqueue descriptors to the device
  *
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_log_base(struct vhost_dev *dev, uint64_t base,
 static int vhost_vdpa_set_vring_addr(struct vhost_dev *dev,
                                        struct vhost_vring_addr *addr)
 {
-    trace_vhost_vdpa_set_vring_addr(dev, addr->index, addr->flags,
-                                    addr->desc_user_addr, addr->used_user_addr,
-                                    addr->avail_user_addr,
-                                    addr->log_guest_addr);
-    return vhost_vdpa_call(dev, VHOST_SET_VRING_ADDR, addr);
+    struct vhost_vdpa *v = dev->opaque;
+
+    if (v->shadow_vqs_enabled) {
+        /*
+         * Device vring addr was set at device start. SVQ base is handled by
+         * VirtQueue code.
+         */
+        return 0;
+    }
+
+    return vhost_vdpa_set_vring_dev_addr(dev, addr);
 }
 
 static int vhost_vdpa_set_vring_num(struct vhost_dev *dev,
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_vring_num(struct vhost_dev *dev,
 static int vhost_vdpa_set_vring_base(struct vhost_dev *dev,
                                        struct vhost_vring_state *ring)
 {
-    trace_vhost_vdpa_set_vring_base(dev, ring->index, ring->num);
-    return vhost_vdpa_call(dev, VHOST_SET_VRING_BASE, ring);
+    struct vhost_vdpa *v = dev->opaque;
+
+    if (v->shadow_vqs_enabled) {
+        /*
+         * Device vring base was set at device start. SVQ base is handled by
+         * VirtQueue code.
+         */
+        return 0;
+    }
+
+    return vhost_vdpa_set_dev_vring_base(dev, ring);
 }
 
 static int vhost_vdpa_get_vring_base(struct vhost_dev *dev,
-- 
2.7.4

From: Eugenio Pérez <eperezma@redhat.com>

Initial version of shadow virtqueue that actually forward buffers. There
is no iommu support at the moment, and that will be addressed in future
patches of this series. Since all vhost-vdpa devices use forced IOMMU,
this means that SVQ is not usable at this point of the series on any
device.

For simplicity it only supports modern devices, that expects vring
in little endian, with split ring and no event idx or indirect
descriptors. Support for them will not be added in this series.

It reuses the VirtQueue code for the device part. The driver part is
based on Linux's virtio_ring driver, but with stripped functionality
and optimizations so it's easier to review.

However, forwarding buffers have some particular pieces: One of the most
unexpected ones is that a guest's buffer can expand through more than
one descriptor in SVQ. While this is handled gracefully by qemu's
emulated virtio devices, it may cause unexpected SVQ queue full. This
patch also solves it by checking for this condition at both guest's
kicks and device's calls. The code may be more elegant in the future if
SVQ code runs in its own iocontext.

Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 hw/virtio/vhost-shadow-virtqueue.c | 353 ++++++++++++++++++++++++++++++++++++-
 hw/virtio/vhost-shadow-virtqueue.h |  26 +++
 hw/virtio/vhost-vdpa.c             | 159 ++++++++++++++++-
 3 files changed, 526 insertions(+), 12 deletions(-)

diff --git a/hw/virtio/vhost-shadow-virtqueue.c b/hw/virtio/vhost-shadow-virtqueue.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/virtio/vhost-shadow-virtqueue.c
+++ b/hw/virtio/vhost-shadow-virtqueue.c
@@ -XXX,XX +XXX,XX @@
 #include "qemu/error-report.h"
 #include "qapi/error.h"
 #include "qemu/main-loop.h"
+#include "qemu/log.h"
 #include "linux-headers/linux/vhost.h"
 
 /**
@@ -XXX,XX +XXX,XX @@ bool vhost_svq_valid_features(uint64_t features, Error **errp)
 }
 
 /**
- * Forward guest notifications.
+ * Number of descriptors that the SVQ can make available from the guest.
+ *
+ * @svq: The svq
+ */
+static uint16_t vhost_svq_available_slots(const VhostShadowVirtqueue *svq)
+{
+    return svq->vring.num - (svq->shadow_avail_idx - svq->shadow_used_idx);
+}
+
+static void vhost_vring_write_descs(VhostShadowVirtqueue *svq,
+                                    const struct iovec *iovec,
+                                    size_t num, bool more_descs, bool write)
+{
+    uint16_t i = svq->free_head, last = svq->free_head;
+    unsigned n;
+    uint16_t flags = write ? cpu_to_le16(VRING_DESC_F_WRITE) : 0;
+    vring_desc_t *descs = svq->vring.desc;
+
+    if (num == 0) {
+        return;
+    }
+
+    for (n = 0; n < num; n++) {
+        if (more_descs || (n + 1 < num)) {
+            descs[i].flags = flags | cpu_to_le16(VRING_DESC_F_NEXT);
+        } else {
+            descs[i].flags = flags;
+        }
+        descs[i].addr = cpu_to_le64((hwaddr)iovec[n].iov_base);
+        descs[i].len = cpu_to_le32(iovec[n].iov_len);
+
+        last = i;
+        i = cpu_to_le16(descs[i].next);
+    }
+
+    svq->free_head = le16_to_cpu(descs[last].next);
+}
+
+static bool vhost_svq_add_split(VhostShadowVirtqueue *svq,
+                                VirtQueueElement *elem,
+                                unsigned *head)
+{
+    unsigned avail_idx;
+    vring_avail_t *avail = svq->vring.avail;
+
+    *head = svq->free_head;
+
+    /* We need some descriptors here */
+    if (unlikely(!elem->out_num && !elem->in_num)) {
+        qemu_log_mask(LOG_GUEST_ERROR,
+            "Guest provided element with no descriptors");
+        return false;
+    }
+
+    vhost_vring_write_descs(svq, elem->out_sg, elem->out_num,
+                            elem->in_num > 0, false);
+    vhost_vring_write_descs(svq, elem->in_sg, elem->in_num, false, true);
+
+    /*
+     * Put the entry in the available array (but don't update avail->idx until
+     * they do sync).
+     */
+    avail_idx = svq->shadow_avail_idx & (svq->vring.num - 1);
+    avail->ring[avail_idx] = cpu_to_le16(*head);
+    svq->shadow_avail_idx++;
+
+    /* Update the avail index after write the descriptor */
+    smp_wmb();
+    avail->idx = cpu_to_le16(svq->shadow_avail_idx);
+
+    return true;
+}
+
+static bool vhost_svq_add(VhostShadowVirtqueue *svq, VirtQueueElement *elem)
+{
+    unsigned qemu_head;
+    bool ok = vhost_svq_add_split(svq, elem, &qemu_head);
+    if (unlikely(!ok)) {
+        return false;
+    }
+
+    svq->ring_id_maps[qemu_head] = elem;
+    return true;
+}
+
+static void vhost_svq_kick(VhostShadowVirtqueue *svq)
+{
+    /*
+     * We need to expose the available array entries before checking the used
+     * flags
+     */
+    smp_mb();
+    if (svq->vring.used->flags & VRING_USED_F_NO_NOTIFY) {
+        return;
+    }
+
+    event_notifier_set(&svq->hdev_kick);
+}
+
+/**
+ * Forward available buffers.
+ *
+ * @svq: Shadow VirtQueue
+ *
+ * Note that this function does not guarantee that all guest's available
+ * buffers are available to the device in SVQ avail ring. The guest may have
+ * exposed a GPA / GIOVA contiguous buffer, but it may not be contiguous in
+ * qemu vaddr.
+ *
+ * If that happens, guest's kick notifications will be disabled until the
+ * device uses some buffers.
+ */
+static void vhost_handle_guest_kick(VhostShadowVirtqueue *svq)
+{
+    /* Clear event notifier */
+    event_notifier_test_and_clear(&svq->svq_kick);
+
+    /* Forward to the device as many available buffers as possible */
+    do {
+        virtio_queue_set_notification(svq->vq, false);
+
+        while (true) {
+            VirtQueueElement *elem;
+            bool ok;
+
+            if (svq->next_guest_avail_elem) {
+                elem = g_steal_pointer(&svq->next_guest_avail_elem);
+            } else {
+                elem = virtqueue_pop(svq->vq, sizeof(*elem));
+            }
+
+            if (!elem) {
+                break;
+            }
+
+            if (elem->out_num + elem->in_num >
+                vhost_svq_available_slots(svq)) {
+                /*
+                 * This condition is possible since a contiguous buffer in GPA
+                 * does not imply a contiguous buffer in qemu's VA
+                 * scatter-gather segments. If that happens, the buffer exposed
+                 * to the device needs to be a chain of descriptors at this
+                 * moment.
+                 *
+                 * SVQ cannot hold more available buffers if we are here:
+                 * queue the current guest descriptor and ignore further kicks
+                 * until some elements are used.
+                 */
+                svq->next_guest_avail_elem = elem;
+                return;
+            }
+
+            ok = vhost_svq_add(svq, elem);
+            if (unlikely(!ok)) {
+                /* VQ is broken, just return and ignore any other kicks */
+                return;
+            }
+            vhost_svq_kick(svq);
+        }
+
+        virtio_queue_set_notification(svq->vq, true);
+    } while (!virtio_queue_empty(svq->vq));
+}
+
+/**
+ * Handle guest's kick.
  *
  * @n: guest kick event notifier, the one that guest set to notify svq.
  */
-static void vhost_handle_guest_kick(EventNotifier *n)
+static void vhost_handle_guest_kick_notifier(EventNotifier *n)
 {
     VhostShadowVirtqueue *svq = container_of(n, VhostShadowVirtqueue,
                                              svq_kick);
     event_notifier_test_and_clear(n);
-    event_notifier_set(&svq->hdev_kick);
+    vhost_handle_guest_kick(svq);
+}
+
+static bool vhost_svq_more_used(VhostShadowVirtqueue *svq)
+{
+    if (svq->last_used_idx != svq->shadow_used_idx) {
+        return true;
+    }
+
+    svq->shadow_used_idx = cpu_to_le16(svq->vring.used->idx);
+
+    return svq->last_used_idx != svq->shadow_used_idx;
 }
 
 /**
- * Forward vhost notifications
+ * Enable vhost device calls after disable them.
+ *
+ * @svq: The svq
+ *
+ * It returns false if there are pending used buffers from the vhost device,
+ * avoiding the possible races between SVQ checking for more work and enabling
+ * callbacks. True if SVQ used vring has no more pending buffers.
+ */
+static bool vhost_svq_enable_notification(VhostShadowVirtqueue *svq)
+{
+    svq->vring.avail->flags &= ~cpu_to_le16(VRING_AVAIL_F_NO_INTERRUPT);
+    /* Make sure the flag is written before the read of used_idx */
+    smp_mb();
+    return !vhost_svq_more_used(svq);
+}
+
+static void vhost_svq_disable_notification(VhostShadowVirtqueue *svq)
+{
+    svq->vring.avail->flags |= cpu_to_le16(VRING_AVAIL_F_NO_INTERRUPT);
+}
+
+static VirtQueueElement *vhost_svq_get_buf(VhostShadowVirtqueue *svq,
+                                           uint32_t *len)
+{
+    vring_desc_t *descs = svq->vring.desc;
+    const vring_used_t *used = svq->vring.used;
+    vring_used_elem_t used_elem;
+    uint16_t last_used;
+
+    if (!vhost_svq_more_used(svq)) {
+        return NULL;
+    }
+
+    /* Only get used array entries after they have been exposed by dev */
+    smp_rmb();
+    last_used = svq->last_used_idx & (svq->vring.num - 1);
+    used_elem.id = le32_to_cpu(used->ring[last_used].id);
+    used_elem.len = le32_to_cpu(used->ring[last_used].len);
+
+    svq->last_used_idx++;
+    if (unlikely(used_elem.id >= svq->vring.num)) {
+        qemu_log_mask(LOG_GUEST_ERROR, "Device %s says index %u is used",
+                      svq->vdev->name, used_elem.id);
+        return NULL;
+    }
+
+    if (unlikely(!svq->ring_id_maps[used_elem.id])) {
+        qemu_log_mask(LOG_GUEST_ERROR,
+            "Device %s says index %u is used, but it was not available",
+            svq->vdev->name, used_elem.id);
+        return NULL;
+    }
+
+    descs[used_elem.id].next = svq->free_head;
+    svq->free_head = used_elem.id;
+
+    *len = used_elem.len;
+    return g_steal_pointer(&svq->ring_id_maps[used_elem.id]);
+}
+
+static void vhost_svq_flush(VhostShadowVirtqueue *svq,
+                            bool check_for_avail_queue)
+{
+    VirtQueue *vq = svq->vq;
+
+    /* Forward as many used buffers as possible. */
+    do {
+        unsigned i = 0;
+
+        vhost_svq_disable_notification(svq);
+        while (true) {
+            uint32_t len;
+            g_autofree VirtQueueElement *elem = vhost_svq_get_buf(svq, &len);
+            if (!elem) {
+                break;
+            }
+
+            if (unlikely(i >= svq->vring.num)) {
+                qemu_log_mask(LOG_GUEST_ERROR,
+                         "More than %u used buffers obtained in a %u size SVQ",
+                         i, svq->vring.num);
+                virtqueue_fill(vq, elem, len, i);
+                virtqueue_flush(vq, i);
+                return;
+            }
+            virtqueue_fill(vq, elem, len, i++);
+        }
+
+        virtqueue_flush(vq, i);
+        event_notifier_set(&svq->svq_call);
+
+        if (check_for_avail_queue && svq->next_guest_avail_elem) {
+            /*
+             * Avail ring was full when vhost_svq_flush was called, so it's a
+             * good moment to make more descriptors available if possible.
+             */
+            vhost_handle_guest_kick(svq);
+        }
+    } while (!vhost_svq_enable_notification(svq));
+}
+
+/**
+ * Forward used buffers.
  *
  * @n: hdev call event notifier, the one that device set to notify svq.
+ *
+ * Note that we are not making any buffers available in the loop, there is no
+ * way that it runs more than virtqueue size times.
  */
 static void vhost_svq_handle_call(EventNotifier *n)
 {
     VhostShadowVirtqueue *svq = container_of(n, VhostShadowVirtqueue,
                                              hdev_call);
     event_notifier_test_and_clear(n);
-    event_notifier_set(&svq->svq_call);
+    vhost_svq_flush(svq, true);
 }
 
 /**
@@ -XXX,XX +XXX,XX @@ void vhost_svq_set_svq_kick_fd(VhostShadowVirtqueue *svq, int svq_kick_fd)
     if (poll_start) {
         event_notifier_init_fd(svq_kick, svq_kick_fd);
         event_notifier_set(svq_kick);
-        event_notifier_set_handler(svq_kick, vhost_handle_guest_kick);
+        event_notifier_set_handler(svq_kick, vhost_handle_guest_kick_notifier);
+    }
+}
+
+/**
+ * Start the shadow virtqueue operation.
+ *
+ * @svq: Shadow Virtqueue
+ * @vdev: VirtIO device
+ * @vq: Virtqueue to shadow
+ */
+void vhost_svq_start(VhostShadowVirtqueue *svq, VirtIODevice *vdev,
+                     VirtQueue *vq)
+{
+    size_t desc_size, driver_size, device_size;
+
+    svq->next_guest_avail_elem = NULL;
+    svq->shadow_avail_idx = 0;
+    svq->shadow_used_idx = 0;
+    svq->last_used_idx = 0;
+    svq->vdev = vdev;
+    svq->vq = vq;
+
+    svq->vring.num = virtio_queue_get_num(vdev, virtio_get_queue_index(vq));
+    driver_size = vhost_svq_driver_area_size(svq);
+    device_size = vhost_svq_device_area_size(svq);
+    svq->vring.desc = qemu_memalign(qemu_real_host_page_size, driver_size);
+    desc_size = sizeof(vring_desc_t) * svq->vring.num;
+    svq->vring.avail = (void *)((char *)svq->vring.desc + desc_size);
+    memset(svq->vring.desc, 0, driver_size);
+    svq->vring.used = qemu_memalign(qemu_real_host_page_size, device_size);
+    memset(svq->vring.used, 0, device_size);
+    svq->ring_id_maps = g_new0(VirtQueueElement *, svq->vring.num);
+    for (unsigned i = 0; i < svq->vring.num - 1; i++) {
+        svq->vring.desc[i].next = cpu_to_le16(i + 1);
     }
 }
 
@@ -XXX,XX +XXX,XX @@ void vhost_svq_set_svq_kick_fd(VhostShadowVirtqueue *svq, int svq_kick_fd)
 void vhost_svq_stop(VhostShadowVirtqueue *svq)
 {
     event_notifier_set_handler(&svq->svq_kick, NULL);
+    g_autofree VirtQueueElement *next_avail_elem = NULL;
+
+    if (!svq->vq) {
+        return;
+    }
+
+    /* Send all pending used descriptors to guest */
+    vhost_svq_flush(svq, false);
+
+    for (unsigned i = 0; i < svq->vring.num; ++i) {
+        g_autofree VirtQueueElement *elem = NULL;
+        elem = g_steal_pointer(&svq->ring_id_maps[i]);
+        if (elem) {
+            virtqueue_detach_element(svq->vq, elem, 0);
+        }
+    }
+
+    next_avail_elem = g_steal_pointer(&svq->next_guest_avail_elem);
+    if (next_avail_elem) {
+        virtqueue_detach_element(svq->vq, next_avail_elem, 0);
+    }
+    svq->vq = NULL;
+    g_free(svq->ring_id_maps);
+    qemu_vfree(svq->vring.desc);
+    qemu_vfree(svq->vring.used);
 }
 
 /**
diff --git a/hw/virtio/vhost-shadow-virtqueue.h b/hw/virtio/vhost-shadow-virtqueue.h
index XXXXXXX..XXXXXXX 100644
--- a/hw/virtio/vhost-shadow-virtqueue.h
+++ b/hw/virtio/vhost-shadow-virtqueue.h
@@ -XXX,XX +XXX,XX @@ typedef struct VhostShadowVirtqueue {
 
     /* Guest's call notifier, where the SVQ calls guest. */
     EventNotifier svq_call;
+
+    /* Virtio queue shadowing */
+    VirtQueue *vq;
+
+    /* Virtio device */
+    VirtIODevice *vdev;
+
+    /* Map for use the guest's descriptors */
+    VirtQueueElement **ring_id_maps;
+
+    /* Next VirtQueue element that guest made available */
+    VirtQueueElement *next_guest_avail_elem;
+
+    /* Next head to expose to the device */
+    uint16_t shadow_avail_idx;
+
+    /* Next free descriptor */
+    uint16_t free_head;
+
+    /* Last seen used idx */
+    uint16_t shadow_used_idx;
+
+    /* Next head to consume from the device */
+    uint16_t last_used_idx;
 } VhostShadowVirtqueue;
 
 bool vhost_svq_valid_features(uint64_t features, Error **errp);
@@ -XXX,XX +XXX,XX @@ void vhost_svq_get_vring_addr(const VhostShadowVirtqueue *svq,
 size_t vhost_svq_driver_area_size(const VhostShadowVirtqueue *svq);
 size_t vhost_svq_device_area_size(const VhostShadowVirtqueue *svq);
 
+void vhost_svq_start(VhostShadowVirtqueue *svq, VirtIODevice *vdev,
+                     VirtQueue *vq);
 void vhost_svq_stop(VhostShadowVirtqueue *svq);
 
 VhostShadowVirtqueue *vhost_svq_new(void);
diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/virtio/vhost-vdpa.c
+++ b/hw/virtio/vhost-vdpa.c
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_vring_dev_addr(struct vhost_dev *dev,
  * Note that this function does not rewind kick file descriptor if cannot set
  * call one.
  */
-static bool vhost_vdpa_svq_setup(struct vhost_dev *dev,
-                                 VhostShadowVirtqueue *svq,
-                                 unsigned idx,
-                                 Error **errp)
+static int vhost_vdpa_svq_set_fds(struct vhost_dev *dev,
+                                  VhostShadowVirtqueue *svq,
+                                  unsigned idx,
+                                  Error **errp)
 {
     struct vhost_vring_file file = {
         .index = dev->vq_index + idx,
@@ -XXX,XX +XXX,XX @@ static bool vhost_vdpa_svq_setup(struct vhost_dev *dev,
     r = vhost_vdpa_set_vring_dev_kick(dev, &file);
     if (unlikely(r != 0)) {
         error_setg_errno(errp, -r, "Can't set device kick fd");
-        return false;
+        return r;
     }
 
     event_notifier = &svq->hdev_call;
@@ -XXX,XX +XXX,XX @@ static bool vhost_vdpa_svq_setup(struct vhost_dev *dev,
         error_setg_errno(errp, -r, "Can't set device call fd");
     }
 
+    return r;
+}
+
+/**
+ * Unmap a SVQ area in the device
+ */
+static bool vhost_vdpa_svq_unmap_ring(struct vhost_vdpa *v, hwaddr iova,
+                                      hwaddr size)
+{
+    int r;
+
+    size = ROUND_UP(size, qemu_real_host_page_size);
+    r = vhost_vdpa_dma_unmap(v, iova, size);
+    return r == 0;
+}
+
+static bool vhost_vdpa_svq_unmap_rings(struct vhost_dev *dev,
+                                       const VhostShadowVirtqueue *svq)
+{
+    struct vhost_vdpa *v = dev->opaque;
+    struct vhost_vring_addr svq_addr;
+    size_t device_size = vhost_svq_device_area_size(svq);
+    size_t driver_size = vhost_svq_driver_area_size(svq);
+    bool ok;
+
+    vhost_svq_get_vring_addr(svq, &svq_addr);
+
+    ok = vhost_vdpa_svq_unmap_ring(v, svq_addr.desc_user_addr, driver_size);
+    if (unlikely(!ok)) {
+        return false;
+    }
+
+    return vhost_vdpa_svq_unmap_ring(v, svq_addr.used_user_addr, device_size);
+}
+
+/**
+ * Map the shadow virtqueue rings in the device
+ *
+ * @dev: The vhost device
+ * @svq: The shadow virtqueue
+ * @addr: Assigned IOVA addresses
+ * @errp: Error pointer
+ */
+static bool vhost_vdpa_svq_map_rings(struct vhost_dev *dev,
+                                     const VhostShadowVirtqueue *svq,
+                                     struct vhost_vring_addr *addr,
+                                     Error **errp)
+{
+    struct vhost_vdpa *v = dev->opaque;
+    size_t device_size = vhost_svq_device_area_size(svq);
+    size_t driver_size = vhost_svq_driver_area_size(svq);
+    int r;
+
+    ERRP_GUARD();
+    vhost_svq_get_vring_addr(svq, addr);
+
+    r = vhost_vdpa_dma_map(v, addr->desc_user_addr, driver_size,
+                           (void *)addr->desc_user_addr, true);
+    if (unlikely(r != 0)) {
+        error_setg_errno(errp, -r, "Cannot create vq driver region: ");
+        return false;
+    }
+
+    r = vhost_vdpa_dma_map(v, addr->used_user_addr, device_size,
+                           (void *)addr->used_user_addr, false);
+    if (unlikely(r != 0)) {
+        error_setg_errno(errp, -r, "Cannot create vq device region: ");
+    }
+
+    return r == 0;
+}
+
+static bool vhost_vdpa_svq_setup(struct vhost_dev *dev,
+                                 VhostShadowVirtqueue *svq,
+                                 unsigned idx,
+                                 Error **errp)
+{
+    uint16_t vq_index = dev->vq_index + idx;
+    struct vhost_vring_state s = {
+        .index = vq_index,
+    };
+    int r;
+
+    r = vhost_vdpa_set_dev_vring_base(dev, &s);
+    if (unlikely(r)) {
+        error_setg_errno(errp, -r, "Cannot set vring base");
+        return false;
+    }
+
+    r = vhost_vdpa_svq_set_fds(dev, svq, idx, errp);
     return r == 0;
 }
 
@@ -XXX,XX +XXX,XX @@ static bool vhost_vdpa_svqs_start(struct vhost_dev *dev)
     }
 
     for (i = 0; i < v->shadow_vqs->len; ++i) {
+        VirtQueue *vq = virtio_get_queue(dev->vdev, dev->vq_index + i);
         VhostShadowVirtqueue *svq = g_ptr_array_index(v->shadow_vqs, i);
+        struct vhost_vring_addr addr = {
+            .index = i,
+        };
+        int r;
         bool ok = vhost_vdpa_svq_setup(dev, svq, i, &err);
         if (unlikely(!ok)) {
-            error_reportf_err(err, "Cannot setup SVQ %u: ", i);
+            goto err;
+        }
+
+        vhost_svq_start(svq, dev->vdev, vq);
+        ok = vhost_vdpa_svq_map_rings(dev, svq, &addr, &err);
+        if (unlikely(!ok)) {
+            goto err_map;
+        }
+
+        /* Override vring GPA set by vhost subsystem */
+        r = vhost_vdpa_set_vring_dev_addr(dev, &addr);
+        if (unlikely(r != 0)) {
+            error_setg_errno(&err, -r, "Cannot set device address");
+            goto err_set_addr;
+        }
+    }
+
+    return true;
+
+err_set_addr:
+    vhost_vdpa_svq_unmap_rings(dev, g_ptr_array_index(v->shadow_vqs, i));
+
+err_map:
+    vhost_svq_stop(g_ptr_array_index(v->shadow_vqs, i));
+
+err:
+    error_reportf_err(err, "Cannot setup SVQ %u: ", i);
+    for (unsigned j = 0; j < i; ++j) {
+        VhostShadowVirtqueue *svq = g_ptr_array_index(v->shadow_vqs, j);
+        vhost_vdpa_svq_unmap_rings(dev, svq);
+        vhost_svq_stop(svq);
+    }
+
+    return false;
+}
+
+static bool vhost_vdpa_svqs_stop(struct vhost_dev *dev)
+{
+    struct vhost_vdpa *v = dev->opaque;
+
+    if (!v->shadow_vqs) {
+        return true;
+    }
+
+    for (unsigned i = 0; i < v->shadow_vqs->len; ++i) {
+        VhostShadowVirtqueue *svq = g_ptr_array_index(v->shadow_vqs,
+                                                      i);
+        bool ok = vhost_vdpa_svq_unmap_rings(dev, svq);
+        if (unlikely(!ok)) {
             return false;
         }
     }
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_dev_start(struct vhost_dev *dev, bool started)
         }
         vhost_vdpa_set_vring_ready(dev);
     } else {
+        ok = vhost_vdpa_svqs_stop(dev);
+        if (unlikely(!ok)) {
+            return -1;
+        }
         vhost_vdpa_host_notifiers_uninit(dev, dev->nvqs);
     }
 
-- 
2.7.4

From: Eugenio Pérez <eperezma@redhat.com>

This iova tree function allows it to look for a hole in allocated
regions and return a totally new translation for a given translated
address.

It's usage is mainly to allow devices to access qemu address space,
remapping guest's one into a new iova space where qemu can add chunks of
addresses.

Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 include/qemu/iova-tree.h |  18 +++++++
 util/iova-tree.c         | 135 +++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 153 insertions(+)

diff --git a/include/qemu/iova-tree.h b/include/qemu/iova-tree.h
index XXXXXXX..XXXXXXX 100644
--- a/include/qemu/iova-tree.h
+++ b/include/qemu/iova-tree.h
@@ -XXX,XX +XXX,XX @@
 #define  IOVA_OK           (0)
 #define  IOVA_ERR_INVALID  (-1) /* Invalid parameters */
 #define  IOVA_ERR_OVERLAP  (-2) /* IOVA range overlapped */
+#define  IOVA_ERR_NOMEM    (-3) /* Cannot allocate */
 
 typedef struct IOVATree IOVATree;
 typedef struct DMAMap {
@@ -XXX,XX +XXX,XX @@ const DMAMap *iova_tree_find_address(const IOVATree *tree, hwaddr iova);
 void iova_tree_foreach(IOVATree *tree, iova_tree_iterator iterator);
 
 /**
+ * iova_tree_alloc_map:
+ *
+ * @tree: the iova tree to allocate from
+ * @map: the new map (as translated addr & size) to allocate in the iova region
+ * @iova_begin: the minimum address of the allocation
+ * @iova_end: the maximum addressable direction of the allocation
+ *
+ * Allocates a new region of a given size, between iova_min and iova_max.
+ *
+ * Return: Same as iova_tree_insert, but cannot overlap and can return error if
+ * iova tree is out of free contiguous range. The caller gets the assigned iova
+ * in map->iova.
+ */
+int iova_tree_alloc_map(IOVATree *tree, DMAMap *map, hwaddr iova_begin,
+                        hwaddr iova_end);
+
+/**
  * iova_tree_destroy:
  *
  * @tree: the iova tree to destroy
diff --git a/util/iova-tree.c b/util/iova-tree.c
index XXXXXXX..XXXXXXX 100644
--- a/util/iova-tree.c
+++ b/util/iova-tree.c
@@ -XXX,XX +XXX,XX @@ struct IOVATree {
     GTree *tree;
 };
 
+/* Args to pass to iova_tree_alloc foreach function. */
+struct IOVATreeAllocArgs {
+    /* Size of the desired allocation */
+    size_t new_size;
+
+    /* The minimum address allowed in the allocation */
+    hwaddr iova_begin;
+
+    /* Map at the left of the hole, can be NULL if "this" is first one */
+    const DMAMap *prev;
+
+    /* Map at the right of the hole, can be NULL if "prev" is the last one */
+    const DMAMap *this;
+
+    /* If found, we fill in the IOVA here */
+    hwaddr iova_result;
+
+    /* Whether have we found a valid IOVA */
+    bool iova_found;
+};
+
+/**
+ * Iterate args to the next hole
+ *
+ * @args: The alloc arguments
+ * @next: The next mapping in the tree. Can be NULL to signal the last one
+ */
+static void iova_tree_alloc_args_iterate(struct IOVATreeAllocArgs *args,
+                                         const DMAMap *next) {
+    args->prev = args->this;
+    args->this = next;
+}
+
 static int iova_tree_compare(gconstpointer a, gconstpointer b, gpointer data)
 {
     const DMAMap *m1 = a, *m2 = b;
@@ -XXX,XX +XXX,XX @@ int iova_tree_remove(IOVATree *tree, const DMAMap *map)
     return IOVA_OK;
 }
 
+/**
+ * Try to find an unallocated IOVA range between prev and this elements.
+ *
+ * @args: Arguments to allocation
+ *
+ * Cases:
+ *
+ * (1) !prev, !this: No entries allocated, always succeed
+ *
+ * (2) !prev, this: We're iterating at the 1st element.
+ *
+ * (3) prev, !this: We're iterating at the last element.
+ *
+ * (4) prev, this: this is the most common case, we'll try to find a hole
+ * between "prev" and "this" mapping.
+ *
+ * Note that this function assumes the last valid iova is HWADDR_MAX, but it
+ * searches linearly so it's easy to discard the result if it's not the case.
+ */
+static void iova_tree_alloc_map_in_hole(struct IOVATreeAllocArgs *args)
+{
+    const DMAMap *prev = args->prev, *this = args->this;
+    uint64_t hole_start, hole_last;
+
+    if (this && this->iova + this->size < args->iova_begin) {
+        return;
+    }
+
+    hole_start = MAX(prev ? prev->iova + prev->size + 1 : 0, args->iova_begin);
+    hole_last = this ? this->iova : HWADDR_MAX;
+
+    if (hole_last - hole_start > args->new_size) {
+        args->iova_result = hole_start;
+        args->iova_found = true;
+    }
+}
+
+/**
+ * Foreach dma node in the tree, compare if there is a hole with its previous
+ * node (or minimum iova address allowed) and the node.
+ *
+ * @key: Node iterating
+ * @value: Node iterating
+ * @pargs: Struct to communicate with the outside world
+ *
+ * Return: false to keep iterating, true if needs break.
+ */
+static gboolean iova_tree_alloc_traverse(gpointer key, gpointer value,
+                                         gpointer pargs)
+{
+    struct IOVATreeAllocArgs *args = pargs;
+    DMAMap *node = value;
+
+    assert(key == value);
+
+    iova_tree_alloc_args_iterate(args, node);
+    iova_tree_alloc_map_in_hole(args);
+    return args->iova_found;
+}
+
+int iova_tree_alloc_map(IOVATree *tree, DMAMap *map, hwaddr iova_begin,
+                        hwaddr iova_last)
+{
+    struct IOVATreeAllocArgs args = {
+        .new_size = map->size,
+        .iova_begin = iova_begin,
+    };
+
+    if (unlikely(iova_last < iova_begin)) {
+        return IOVA_ERR_INVALID;
+    }
+
+    /*
+     * Find a valid hole for the mapping
+     *
+     * Assuming low iova_begin, so no need to do a binary search to
+     * locate the first node.
+     *
+     * TODO: Replace all this with g_tree_node_first/next/last when available
+     * (from glib since 2.68). To do it with g_tree_foreach complicates the
+     * code a lot.
+     *
+     */
+    g_tree_foreach(tree->tree, iova_tree_alloc_traverse, &args);
+    if (!args.iova_found) {
+        /*
+         * Either tree is empty or the last hole is still not checked.
+         * g_tree_foreach does not compare (last, iova_last] range, so we check
+         * it here.
+         */
+        iova_tree_alloc_args_iterate(&args, NULL);
+        iova_tree_alloc_map_in_hole(&args);
+    }
+
+    if (!args.iova_found || args.iova_result + map->size > iova_last) {
+        return IOVA_ERR_NOMEM;
+    }
+
+    map->iova = args.iova_result;
+    return iova_tree_insert(tree, map);
+}
+
 void iova_tree_destroy(IOVATree *tree)
 {
     g_tree_destroy(tree->tree);
-- 
2.7.4

From: Eugenio Pérez <eperezma@redhat.com>

This function does the reverse operation of iova_tree_find: To look for
a mapping that match a translated address so we can do the reverse.

This have linear complexity instead of logarithmic, but it supports
overlapping HVA. Future developments could reduce it.

Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 include/qemu/iova-tree.h | 20 +++++++++++++++++++-
 util/iova-tree.c         | 34 ++++++++++++++++++++++++++++++++++
 2 files changed, 53 insertions(+), 1 deletion(-)

diff --git a/include/qemu/iova-tree.h b/include/qemu/iova-tree.h
index XXXXXXX..XXXXXXX 100644
--- a/include/qemu/iova-tree.h
+++ b/include/qemu/iova-tree.h
@@ -XXX,XX +XXX,XX @@ int iova_tree_remove(IOVATree *tree, const DMAMap *map);
  * @tree: the iova tree to search from
  * @map: the mapping to search
  *
- * Search for a mapping in the iova tree that overlaps with the
+ * Search for a mapping in the iova tree that iova overlaps with the
  * mapping range specified.  Only the first found mapping will be
  * returned.
  *
@@ -XXX,XX +XXX,XX @@ int iova_tree_remove(IOVATree *tree, const DMAMap *map);
 const DMAMap *iova_tree_find(const IOVATree *tree, const DMAMap *map);
 
 /**
+ * iova_tree_find_iova:
+ *
+ * @tree: the iova tree to search from
+ * @map: the mapping to search
+ *
+ * Search for a mapping in the iova tree that translated_addr overlaps with the
+ * mapping range specified.  Only the first found mapping will be
+ * returned.
+ *
+ * Return: DMAMap pointer if found, or NULL if not found.  Note that
+ * the returned DMAMap pointer is maintained internally.  User should
+ * only read the content but never modify or free the content.  Also,
+ * user is responsible to make sure the pointer is valid (say, no
+ * concurrent deletion in progress).
+ */
+const DMAMap *iova_tree_find_iova(const IOVATree *tree, const DMAMap *map);
+
+/**
  * iova_tree_find_address:
  *
  * @tree: the iova tree to search from
diff --git a/util/iova-tree.c b/util/iova-tree.c
index XXXXXXX..XXXXXXX 100644
--- a/util/iova-tree.c
+++ b/util/iova-tree.c
@@ -XXX,XX +XXX,XX @@ struct IOVATreeAllocArgs {
     bool iova_found;
 };
 
+typedef struct IOVATreeFindIOVAArgs {
+    const DMAMap *needle;
+    const DMAMap *result;
+} IOVATreeFindIOVAArgs;
+
 /**
  * Iterate args to the next hole
  *
@@ -XXX,XX +XXX,XX @@ const DMAMap *iova_tree_find(const IOVATree *tree, const DMAMap *map)
     return g_tree_lookup(tree->tree, map);
 }
 
+static gboolean iova_tree_find_address_iterator(gpointer key, gpointer value,
+                                                gpointer data)
+{
+    const DMAMap *map = key;
+    IOVATreeFindIOVAArgs *args = data;
+    const DMAMap *needle;
+
+    g_assert(key == value);
+
+    needle = args->needle;
+    if (map->translated_addr + map->size < needle->translated_addr ||
+        needle->translated_addr + needle->size < map->translated_addr) {
+        return false;
+    }
+
+    args->result = map;
+    return true;
+}
+
+const DMAMap *iova_tree_find_iova(const IOVATree *tree, const DMAMap *map)
+{
+    IOVATreeFindIOVAArgs args = {
+        .needle = map,
+    };
+
+    g_tree_foreach(tree->tree, iova_tree_find_address_iterator, &args);
+    return args.result;
+}
+
 const DMAMap *iova_tree_find_address(const IOVATree *tree, hwaddr iova)
 {
     const DMAMap map = { .iova = iova, .size = 0 };
-- 
2.7.4

From: Eugenio Pérez <eperezma@redhat.com>

This tree is able to look for a translated address from an IOVA address.

At first glance it is similar to util/iova-tree. However, SVQ working on
devices with limited IOVA space need more capabilities, like allocating
IOVA chunks or performing reverse translations (qemu addresses to iova).

The allocation capability, as "assign a free IOVA address to this chunk
of memory in qemu's address space" allows shadow virtqueue to create a
new address space that is not restricted by guest's addressable one, so
we can allocate shadow vqs vrings outside of it.

It duplicates the tree so it can search efficiently in both directions,
and it will signal overlap if iova or the translated address is present
in any tree.

Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 hw/virtio/meson.build       |   2 +-
 hw/virtio/vhost-iova-tree.c | 110 ++++++++++++++++++++++++++++++++++++++++++++
 hw/virtio/vhost-iova-tree.h |  27 +++++++++++
 3 files changed, 138 insertions(+), 1 deletion(-)
 create mode 100644 hw/virtio/vhost-iova-tree.c
 create mode 100644 hw/virtio/vhost-iova-tree.h

diff --git a/hw/virtio/meson.build b/hw/virtio/meson.build
index XXXXXXX..XXXXXXX 100644
--- a/hw/virtio/meson.build
+++ b/hw/virtio/meson.build
@@ -XXX,XX +XXX,XX @@ softmmu_ss.add(when: 'CONFIG_ALL', if_true: files('vhost-stub.c'))
 
 virtio_ss = ss.source_set()
 virtio_ss.add(files('virtio.c'))
-virtio_ss.add(when: 'CONFIG_VHOST', if_true: files('vhost.c', 'vhost-backend.c', 'vhost-shadow-virtqueue.c'))
+virtio_ss.add(when: 'CONFIG_VHOST', if_true: files('vhost.c', 'vhost-backend.c', 'vhost-shadow-virtqueue.c', 'vhost-iova-tree.c'))
 virtio_ss.add(when: 'CONFIG_VHOST_USER', if_true: files('vhost-user.c'))
 virtio_ss.add(when: 'CONFIG_VHOST_VDPA', if_true: files('vhost-vdpa.c'))
 virtio_ss.add(when: 'CONFIG_VIRTIO_BALLOON', if_true: files('virtio-balloon.c'))
diff --git a/hw/virtio/vhost-iova-tree.c b/hw/virtio/vhost-iova-tree.c
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/hw/virtio/vhost-iova-tree.c
@@ -XXX,XX +XXX,XX @@
+/*
+ * vhost software live migration iova tree
+ *
+ * SPDX-FileCopyrightText: Red Hat, Inc. 2021
+ * SPDX-FileContributor: Author: Eugenio Pérez <eperezma@redhat.com>
+ *
+ * SPDX-License-Identifier: GPL-2.0-or-later
+ */
+
+#include "qemu/osdep.h"
+#include "qemu/iova-tree.h"
+#include "vhost-iova-tree.h"
+
+#define iova_min_addr qemu_real_host_page_size
+
+/**
+ * VhostIOVATree, able to:
+ * - Translate iova address
+ * - Reverse translate iova address (from translated to iova)
+ * - Allocate IOVA regions for translated range (linear operation)
+ */
+struct VhostIOVATree {
+    /* First addressable iova address in the device */
+    uint64_t iova_first;
+
+    /* Last addressable iova address in the device */
+    uint64_t iova_last;
+
+    /* IOVA address to qemu memory maps. */
+    IOVATree *iova_taddr_map;
+};
+
+/**
+ * Create a new IOVA tree
+ *
+ * Returns the new IOVA tree
+ */
+VhostIOVATree *vhost_iova_tree_new(hwaddr iova_first, hwaddr iova_last)
+{
+    VhostIOVATree *tree = g_new(VhostIOVATree, 1);
+
+    /* Some devices do not like 0 addresses */
+    tree->iova_first = MAX(iova_first, iova_min_addr);
+    tree->iova_last = iova_last;
+
+    tree->iova_taddr_map = iova_tree_new();
+    return tree;
+}
+
+/**
+ * Delete an iova tree
+ */
+void vhost_iova_tree_delete(VhostIOVATree *iova_tree)
+{
+    iova_tree_destroy(iova_tree->iova_taddr_map);
+    g_free(iova_tree);
+}
+
+/**
+ * Find the IOVA address stored from a memory address
+ *
+ * @tree: The iova tree
+ * @map: The map with the memory address
+ *
+ * Return the stored mapping, or NULL if not found.
+ */
+const DMAMap *vhost_iova_tree_find_iova(const VhostIOVATree *tree,
+                                        const DMAMap *map)
+{
+    return iova_tree_find_iova(tree->iova_taddr_map, map);
+}
+
+/**
+ * Allocate a new mapping
+ *
+ * @tree: The iova tree
+ * @map: The iova map
+ *
+ * Returns:
+ * - IOVA_OK if the map fits in the container
+ * - IOVA_ERR_INVALID if the map does not make sense (like size overflow)
+ * - IOVA_ERR_NOMEM if tree cannot allocate more space.
+ *
+ * It returns assignated iova in map->iova if return value is VHOST_DMA_MAP_OK.
+ */
+int vhost_iova_tree_map_alloc(VhostIOVATree *tree, DMAMap *map)
+{
+    /* Some vhost devices do not like addr 0. Skip first page */
+    hwaddr iova_first = tree->iova_first ?: qemu_real_host_page_size;
+
+    if (map->translated_addr + map->size < map->translated_addr ||
+        map->perm == IOMMU_NONE) {
+        return IOVA_ERR_INVALID;
+    }
+
+    /* Allocate a node in IOVA address */
+    return iova_tree_alloc_map(tree->iova_taddr_map, map, iova_first,
+                               tree->iova_last);
+}
+
+/**
+ * Remove existing mappings from iova tree
+ *
+ * @iova_tree: The vhost iova tree
+ * @map: The map to remove
+ */
+void vhost_iova_tree_remove(VhostIOVATree *iova_tree, const DMAMap *map)
+{
+    iova_tree_remove(iova_tree->iova_taddr_map, map);
+}
diff --git a/hw/virtio/vhost-iova-tree.h b/hw/virtio/vhost-iova-tree.h
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/hw/virtio/vhost-iova-tree.h
@@ -XXX,XX +XXX,XX @@
+/*
+ * vhost software live migration iova tree
+ *
+ * SPDX-FileCopyrightText: Red Hat, Inc. 2021
+ * SPDX-FileContributor: Author: Eugenio Pérez <eperezma@redhat.com>
+ *
+ * SPDX-License-Identifier: GPL-2.0-or-later
+ */
+
+#ifndef HW_VIRTIO_VHOST_IOVA_TREE_H
+#define HW_VIRTIO_VHOST_IOVA_TREE_H
+
+#include "qemu/iova-tree.h"
+#include "exec/memory.h"
+
+typedef struct VhostIOVATree VhostIOVATree;
+
+VhostIOVATree *vhost_iova_tree_new(uint64_t iova_first, uint64_t iova_last);
+void vhost_iova_tree_delete(VhostIOVATree *iova_tree);
+G_DEFINE_AUTOPTR_CLEANUP_FUNC(VhostIOVATree, vhost_iova_tree_delete);
+
+const DMAMap *vhost_iova_tree_find_iova(const VhostIOVATree *iova_tree,
+                                        const DMAMap *map);
+int vhost_iova_tree_map_alloc(VhostIOVATree *iova_tree, DMAMap *map);
+void vhost_iova_tree_remove(VhostIOVATree *iova_tree, const DMAMap *map);
+
+#endif
-- 
2.7.4

From: Eugenio Pérez <eperezma@redhat.com>

Use translations added in VhostIOVATree in SVQ.

Only introduce usage here, not allocation and deallocation. As with
previous patches, we use the dead code paths of shadow_vqs_enabled to
avoid commiting too many changes at once. These are impossible to take
at the moment.

Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 hw/virtio/vhost-shadow-virtqueue.c |  75 +++++++++++++++++++++--
 hw/virtio/vhost-shadow-virtqueue.h |   6 +-
 hw/virtio/vhost-vdpa.c             | 122 +++++++++++++++++++++++++++++++------
 include/hw/virtio/vhost-vdpa.h     |   3 +
 4 files changed, 181 insertions(+), 25 deletions(-)

diff --git a/hw/virtio/vhost-shadow-virtqueue.c b/hw/virtio/vhost-shadow-virtqueue.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/virtio/vhost-shadow-virtqueue.c
+++ b/hw/virtio/vhost-shadow-virtqueue.c
@@ -XXX,XX +XXX,XX @@ static uint16_t vhost_svq_available_slots(const VhostShadowVirtqueue *svq)
     return svq->vring.num - (svq->shadow_avail_idx - svq->shadow_used_idx);
 }
 
+/**
+ * Translate addresses between the qemu's virtual address and the SVQ IOVA
+ *
+ * @svq: Shadow VirtQueue
+ * @vaddr: Translated IOVA addresses
+ * @iovec: Source qemu's VA addresses
+ * @num: Length of iovec and minimum length of vaddr
+ */
+static bool vhost_svq_translate_addr(const VhostShadowVirtqueue *svq,
+                                     void **addrs, const struct iovec *iovec,
+                                     size_t num)
+{
+    if (num == 0) {
+        return true;
+    }
+
+    for (size_t i = 0; i < num; ++i) {
+        DMAMap needle = {
+            .translated_addr = (hwaddr)iovec[i].iov_base,
+            .size = iovec[i].iov_len,
+        };
+        size_t off;
+
+        const DMAMap *map = vhost_iova_tree_find_iova(svq->iova_tree, &needle);
+        /*
+         * Map cannot be NULL since iova map contains all guest space and
+         * qemu already has a physical address mapped
+         */
+        if (unlikely(!map)) {
+            qemu_log_mask(LOG_GUEST_ERROR,
+                          "Invalid address 0x%"HWADDR_PRIx" given by guest",
+                          needle.translated_addr);
+            return false;
+        }
+
+        off = needle.translated_addr - map->translated_addr;
+        addrs[i] = (void *)(map->iova + off);
+
+        if (unlikely(int128_gt(int128_add(needle.translated_addr,
+                                          iovec[i].iov_len),
+                               map->translated_addr + map->size))) {
+            qemu_log_mask(LOG_GUEST_ERROR,
+                          "Guest buffer expands over iova range");
+            return false;
+        }
+    }
+
+    return true;
+}
+
 static void vhost_vring_write_descs(VhostShadowVirtqueue *svq,
+                                    void * const *sg,
                                     const struct iovec *iovec,
                                     size_t num, bool more_descs, bool write)
 {
@@ -XXX,XX +XXX,XX @@ static void vhost_vring_write_descs(VhostShadowVirtqueue *svq,
         } else {
             descs[i].flags = flags;
         }
-        descs[i].addr = cpu_to_le64((hwaddr)iovec[n].iov_base);
+        descs[i].addr = cpu_to_le64((hwaddr)sg[n]);
         descs[i].len = cpu_to_le32(iovec[n].iov_len);
 
         last = i;
@@ -XXX,XX +XXX,XX @@ static bool vhost_svq_add_split(VhostShadowVirtqueue *svq,
 {
     unsigned avail_idx;
     vring_avail_t *avail = svq->vring.avail;
+    bool ok;
+    g_autofree void **sgs = g_new(void *, MAX(elem->out_num, elem->in_num));
 
     *head = svq->free_head;
 
@@ -XXX,XX +XXX,XX @@ static bool vhost_svq_add_split(VhostShadowVirtqueue *svq,
         return false;
     }
 
-    vhost_vring_write_descs(svq, elem->out_sg, elem->out_num,
+    ok = vhost_svq_translate_addr(svq, sgs, elem->out_sg, elem->out_num);
+    if (unlikely(!ok)) {
+        return false;
+    }
+    vhost_vring_write_descs(svq, sgs, elem->out_sg, elem->out_num,
                             elem->in_num > 0, false);
-    vhost_vring_write_descs(svq, elem->in_sg, elem->in_num, false, true);
+
+
+    ok = vhost_svq_translate_addr(svq, sgs, elem->in_sg, elem->in_num);
+    if (unlikely(!ok)) {
+        return false;
+    }
+
+    vhost_vring_write_descs(svq, sgs, elem->in_sg, elem->in_num, false, true);
 
     /*
      * Put the entry in the available array (but don't update avail->idx until
@@ -XXX,XX +XXX,XX @@ void vhost_svq_stop(VhostShadowVirtqueue *svq)
  * Creates vhost shadow virtqueue, and instructs the vhost device to use the
  * shadow methods and file descriptors.
  *
+ * @iova_tree: Tree to perform descriptors translations
+ *
  * Returns the new virtqueue or NULL.
  *
  * In case of error, reason is reported through error_report.
  */
-VhostShadowVirtqueue *vhost_svq_new(void)
+VhostShadowVirtqueue *vhost_svq_new(VhostIOVATree *iova_tree)
 {
     g_autofree VhostShadowVirtqueue *svq = g_new0(VhostShadowVirtqueue, 1);
     int r;
@@ -XXX,XX +XXX,XX @@ VhostShadowVirtqueue *vhost_svq_new(void)
 
     event_notifier_init_fd(&svq->svq_kick, VHOST_FILE_UNBIND);
     event_notifier_set_handler(&svq->hdev_call, vhost_svq_handle_call);
+    svq->iova_tree = iova_tree;
     return g_steal_pointer(&svq);
 
 err_init_hdev_call:
diff --git a/hw/virtio/vhost-shadow-virtqueue.h b/hw/virtio/vhost-shadow-virtqueue.h
index XXXXXXX..XXXXXXX 100644
--- a/hw/virtio/vhost-shadow-virtqueue.h
+++ b/hw/virtio/vhost-shadow-virtqueue.h
@@ -XXX,XX +XXX,XX @@
 #include "qemu/event_notifier.h"
 #include "hw/virtio/virtio.h"
 #include "standard-headers/linux/vhost_types.h"
+#include "hw/virtio/vhost-iova-tree.h"
 
 /* Shadow virtqueue to relay notifications */
 typedef struct VhostShadowVirtqueue {
@@ -XXX,XX +XXX,XX @@ typedef struct VhostShadowVirtqueue {
     /* Virtio device */
     VirtIODevice *vdev;
 
+    /* IOVA mapping */
+    VhostIOVATree *iova_tree;
+
     /* Map for use the guest's descriptors */
     VirtQueueElement **ring_id_maps;
 
@@ -XXX,XX +XXX,XX @@ void vhost_svq_start(VhostShadowVirtqueue *svq, VirtIODevice *vdev,
                      VirtQueue *vq);
 void vhost_svq_stop(VhostShadowVirtqueue *svq);
 
-VhostShadowVirtqueue *vhost_svq_new(void);
+VhostShadowVirtqueue *vhost_svq_new(VhostIOVATree *iova_tree);
 
 void vhost_svq_free(gpointer vq);
 G_DEFINE_AUTOPTR_CLEANUP_FUNC(VhostShadowVirtqueue, vhost_svq_free);
diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/virtio/vhost-vdpa.c
+++ b/hw/virtio/vhost-vdpa.c
@@ -XXX,XX +XXX,XX @@ static void vhost_vdpa_listener_region_add(MemoryListener *listener,
                                          vaddr, section->readonly);
 
     llsize = int128_sub(llend, int128_make64(iova));
+    if (v->shadow_vqs_enabled) {
+        DMAMap mem_region = {
+            .translated_addr = (hwaddr)vaddr,
+            .size = int128_get64(llsize) - 1,
+            .perm = IOMMU_ACCESS_FLAG(true, section->readonly),
+        };
+
+        int r = vhost_iova_tree_map_alloc(v->iova_tree, &mem_region);
+        if (unlikely(r != IOVA_OK)) {
+            error_report("Can't allocate a mapping (%d)", r);
+            goto fail;
+        }
+
+        iova = mem_region.iova;
+    }
 
     vhost_vdpa_iotlb_batch_begin_once(v);
     ret = vhost_vdpa_dma_map(v, iova, int128_get64(llsize),
@@ -XXX,XX +XXX,XX @@ static void vhost_vdpa_listener_region_del(MemoryListener *listener,
 
     llsize = int128_sub(llend, int128_make64(iova));
 
+    if (v->shadow_vqs_enabled) {
+        const DMAMap *result;
+        const void *vaddr = memory_region_get_ram_ptr(section->mr) +
+            section->offset_within_region +
+            (iova - section->offset_within_address_space);
+        DMAMap mem_region = {
+            .translated_addr = (hwaddr)vaddr,
+            .size = int128_get64(llsize) - 1,
+        };
+
+        result = vhost_iova_tree_find_iova(v->iova_tree, &mem_region);
+        iova = result->iova;
+        vhost_iova_tree_remove(v->iova_tree, &mem_region);
+    }
     vhost_vdpa_iotlb_batch_begin_once(v);
     ret = vhost_vdpa_dma_unmap(v, iova, int128_get64(llsize));
     if (ret) {
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_init_svq(struct vhost_dev *hdev, struct vhost_vdpa *v,
 
     shadow_vqs = g_ptr_array_new_full(hdev->nvqs, vhost_svq_free);
     for (unsigned n = 0; n < hdev->nvqs; ++n) {
-        g_autoptr(VhostShadowVirtqueue) svq = vhost_svq_new();
+        g_autoptr(VhostShadowVirtqueue) svq = vhost_svq_new(v->iova_tree);
 
         if (unlikely(!svq)) {
             error_setg(errp, "Cannot create svq %u", n);
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_svq_set_fds(struct vhost_dev *dev,
 /**
  * Unmap a SVQ area in the device
  */
-static bool vhost_vdpa_svq_unmap_ring(struct vhost_vdpa *v, hwaddr iova,
-                                      hwaddr size)
+static bool vhost_vdpa_svq_unmap_ring(struct vhost_vdpa *v,
+                                      const DMAMap *needle)
 {
+    const DMAMap *result = vhost_iova_tree_find_iova(v->iova_tree, needle);
+    hwaddr size;
     int r;
 
-    size = ROUND_UP(size, qemu_real_host_page_size);
-    r = vhost_vdpa_dma_unmap(v, iova, size);
+    if (unlikely(!result)) {
+        error_report("Unable to find SVQ address to unmap");
+        return false;
+    }
+
+    size = ROUND_UP(result->size, qemu_real_host_page_size);
+    r = vhost_vdpa_dma_unmap(v, result->iova, size);
     return r == 0;
 }
 
 static bool vhost_vdpa_svq_unmap_rings(struct vhost_dev *dev,
                                        const VhostShadowVirtqueue *svq)
 {
+    DMAMap needle = {};
     struct vhost_vdpa *v = dev->opaque;
     struct vhost_vring_addr svq_addr;
-    size_t device_size = vhost_svq_device_area_size(svq);
-    size_t driver_size = vhost_svq_driver_area_size(svq);
     bool ok;
 
     vhost_svq_get_vring_addr(svq, &svq_addr);
 
-    ok = vhost_vdpa_svq_unmap_ring(v, svq_addr.desc_user_addr, driver_size);
+    needle.translated_addr = svq_addr.desc_user_addr;
+    ok = vhost_vdpa_svq_unmap_ring(v, &needle);
     if (unlikely(!ok)) {
         return false;
     }
 
-    return vhost_vdpa_svq_unmap_ring(v, svq_addr.used_user_addr, device_size);
+    needle.translated_addr = svq_addr.used_user_addr;
+    return vhost_vdpa_svq_unmap_ring(v, &needle);
+}
+
+/**
+ * Map the SVQ area in the device
+ *
+ * @v: Vhost-vdpa device
+ * @needle: The area to search iova
+ * @errorp: Error pointer
+ */
+static bool vhost_vdpa_svq_map_ring(struct vhost_vdpa *v, DMAMap *needle,
+                                    Error **errp)
+{
+    int r;
+
+    r = vhost_iova_tree_map_alloc(v->iova_tree, needle);
+    if (unlikely(r != IOVA_OK)) {
+        error_setg(errp, "Cannot allocate iova (%d)", r);
+        return false;
+    }
+
+    r = vhost_vdpa_dma_map(v, needle->iova, needle->size + 1,
+                           (void *)needle->translated_addr,
+                           needle->perm == IOMMU_RO);
+    if (unlikely(r != 0)) {
+        error_setg_errno(errp, -r, "Cannot map region to device");
+        vhost_iova_tree_remove(v->iova_tree, needle);
+    }
+
+    return r == 0;
 }
 
 /**
@@ -XXX,XX +XXX,XX @@ static bool vhost_vdpa_svq_map_rings(struct vhost_dev *dev,
                                      struct vhost_vring_addr *addr,
                                      Error **errp)
 {
+    DMAMap device_region, driver_region;
+    struct vhost_vring_addr svq_addr;
     struct vhost_vdpa *v = dev->opaque;
     size_t device_size = vhost_svq_device_area_size(svq);
     size_t driver_size = vhost_svq_driver_area_size(svq);
-    int r;
+    size_t avail_offset;
+    bool ok;
 
     ERRP_GUARD();
-    vhost_svq_get_vring_addr(svq, addr);
+    vhost_svq_get_vring_addr(svq, &svq_addr);
 
-    r = vhost_vdpa_dma_map(v, addr->desc_user_addr, driver_size,
-                           (void *)addr->desc_user_addr, true);
-    if (unlikely(r != 0)) {
-        error_setg_errno(errp, -r, "Cannot create vq driver region: ");
+    driver_region = (DMAMap) {
+        .translated_addr = svq_addr.desc_user_addr,
+        .size = driver_size - 1,
+        .perm = IOMMU_RO,
+    };
+    ok = vhost_vdpa_svq_map_ring(v, &driver_region, errp);
+    if (unlikely(!ok)) {
+        error_prepend(errp, "Cannot create vq driver region: ");
         return false;
     }
+    addr->desc_user_addr = driver_region.iova;
+    avail_offset = svq_addr.avail_user_addr - svq_addr.desc_user_addr;
+    addr->avail_user_addr = driver_region.iova + avail_offset;
 
-    r = vhost_vdpa_dma_map(v, addr->used_user_addr, device_size,
-                           (void *)addr->used_user_addr, false);
-    if (unlikely(r != 0)) {
-        error_setg_errno(errp, -r, "Cannot create vq device region: ");
+    device_region = (DMAMap) {
+        .translated_addr = svq_addr.used_user_addr,
+        .size = device_size - 1,
+        .perm = IOMMU_RW,
+    };
+    ok = vhost_vdpa_svq_map_ring(v, &device_region, errp);
+    if (unlikely(!ok)) {
+        error_prepend(errp, "Cannot create vq device region: ");
+        vhost_vdpa_svq_unmap_ring(v, &driver_region);
     }
+    addr->used_user_addr = device_region.iova;
 
-    return r == 0;
+    return ok;
 }
 
 static bool vhost_vdpa_svq_setup(struct vhost_dev *dev,
diff --git a/include/hw/virtio/vhost-vdpa.h b/include/hw/virtio/vhost-vdpa.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/virtio/vhost-vdpa.h
+++ b/include/hw/virtio/vhost-vdpa.h
@@ -XXX,XX +XXX,XX @@
 
 #include <gmodule.h>
 
+#include "hw/virtio/vhost-iova-tree.h"
 #include "hw/virtio/virtio.h"
 #include "standard-headers/linux/vhost_types.h"
 
@@ -XXX,XX +XXX,XX @@ typedef struct vhost_vdpa {
     MemoryListener listener;
     struct vhost_vdpa_iova_range iova_range;
     bool shadow_vqs_enabled;
+    /* IOVA mapping used by the Shadow Virtqueue */
+    VhostIOVATree *iova_tree;
     GPtrArray *shadow_vqs;
     struct vhost_dev *dev;
     VhostVDPAHostNotifier notifier[VIRTIO_QUEUE_MAX];
-- 
2.7.4

From: Eugenio Pérez <eperezma@redhat.com>

This is needed to achieve migration, so the destination can restore its
index.

Setting base as last used idx, so destination will see as available all
the entries that the device did not use, including the in-flight
processing ones.

This is ok for networking, but other kinds of devices might have
problems with these retransmissions.

diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/virtio/vhost-vdpa.c
+++ b/hw/virtio/vhost-vdpa.c
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_vring_base(struct vhost_dev *dev,
 static int vhost_vdpa_get_vring_base(struct vhost_dev *dev,
                                        struct vhost_vring_state *ring)
 {
+    struct vhost_vdpa *v = dev->opaque;
     int ret;
 
+    if (v->shadow_vqs_enabled) {
+        VhostShadowVirtqueue *svq = g_ptr_array_index(v->shadow_vqs,
+                                                      ring->index);
+
+        /*
+         * Setting base as last used idx, so destination will see as available
+         * all the entries that the device did not use, including the in-flight
+         * processing ones.
+         *
+         * TODO: This is ok for networking, but other kinds of devices might
+         * have problems with these retransmissions.
+         */
+        ring->num = svq->last_used_idx;
+        return 0;
+    }
+
     ret = vhost_vdpa_call(dev, VHOST_GET_VRING_BASE, ring);
     trace_vhost_vdpa_get_vring_base(dev, ring->index, ring->num);
     return ret;
-- 
2.7.4

From: Eugenio Pérez <eperezma@redhat.com>

SVQ is able to log the dirty bits by itself, so let's use it to not
block migration.

Also, ignore set and clear of VHOST_F_LOG_ALL on set_features if SVQ is
enabled. Even if the device supports it, the reports would be nonsense
because SVQ memory is in the qemu region.

The log region is still allocated. Future changes might skip that, but
this series is already long enough.

Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 hw/virtio/vhost-vdpa.c         | 39 +++++++++++++++++++++++++++++++++++----
 include/hw/virtio/vhost-vdpa.h |  1 +
 2 files changed, 36 insertions(+), 4 deletions(-)

diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/virtio/vhost-vdpa.c
+++ b/hw/virtio/vhost-vdpa.c
@@ -XXX,XX +XXX,XX @@ static bool vhost_vdpa_one_time_request(struct vhost_dev *dev)
     return v->index != 0;
 }
 
+static int vhost_vdpa_get_dev_features(struct vhost_dev *dev,
+                                       uint64_t *features)
+{
+    int ret;
+
+    ret = vhost_vdpa_call(dev, VHOST_GET_FEATURES, features);
+    trace_vhost_vdpa_get_features(dev, *features);
+    return ret;
+}
+
 static int vhost_vdpa_init_svq(struct vhost_dev *hdev, struct vhost_vdpa *v,
                                Error **errp)
 {
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_init_svq(struct vhost_dev *hdev, struct vhost_vdpa *v,
         return 0;
     }
 
-    r = hdev->vhost_ops->vhost_get_features(hdev, &dev_features);
+    r = vhost_vdpa_get_dev_features(hdev, &dev_features);
     if (r != 0) {
         error_setg_errno(errp, -r, "Can't get vdpa device features");
         return r;
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_mem_table(struct vhost_dev *dev,
 static int vhost_vdpa_set_features(struct vhost_dev *dev,
                                    uint64_t features)
 {
+    struct vhost_vdpa *v = dev->opaque;
     int ret;
 
     if (vhost_vdpa_one_time_request(dev)) {
         return 0;
     }
 
+    if (v->shadow_vqs_enabled) {
+        if ((v->acked_features ^ features) == BIT_ULL(VHOST_F_LOG_ALL)) {
+            /*
+             * QEMU is just trying to enable or disable logging. SVQ handles
+             * this sepparately, so no need to forward this.
+             */
+            v->acked_features = features;
+            return 0;
+        }
+
+        v->acked_features = features;
+
+        /* We must not ack _F_LOG if SVQ is enabled */
+        features &= ~BIT_ULL(VHOST_F_LOG_ALL);
+    }
+
     trace_vhost_vdpa_set_features(dev, features);
     ret = vhost_vdpa_call(dev, VHOST_SET_FEATURES, &features);
     if (ret) {
@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_vring_call(struct vhost_dev *dev,
 static int vhost_vdpa_get_features(struct vhost_dev *dev,
                                      uint64_t *features)
 {
-    int ret;
+    struct vhost_vdpa *v = dev->opaque;
+    int ret = vhost_vdpa_get_dev_features(dev, features);
+
+    if (ret == 0 && v->shadow_vqs_enabled) {
+        /* Add SVQ logging capabilities */
+        *features |= BIT_ULL(VHOST_F_LOG_ALL);
+    }
 
-    ret = vhost_vdpa_call(dev, VHOST_GET_FEATURES, features);
-    trace_vhost_vdpa_get_features(dev, *features);
     return ret;
 }
 
diff --git a/include/hw/virtio/vhost-vdpa.h b/include/hw/virtio/vhost-vdpa.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/virtio/vhost-vdpa.h
+++ b/include/hw/virtio/vhost-vdpa.h
@@ -XXX,XX +XXX,XX @@ typedef struct vhost_vdpa {
     bool iotlb_batch_begin_sent;
     MemoryListener listener;
     struct vhost_vdpa_iova_range iova_range;
+    uint64_t acked_features;
     bool shadow_vqs_enabled;
     /* IOVA mapping used by the Shadow Virtqueue */
     VhostIOVATree *iova_tree;
-- 
2.7.4