Series comparison

-[Qemu-devel] [PULL 0/6] Net patches
+[PULL 00/15] Net patches
-The following changes since commit 2e02083438962d26ef9dcc7100f3b378104183db:
+The following changes since commit f45cc81911adc7726e8a2801986b6998b91b816e:
-  Merge remote-tracking branch 'remotes/kevin/tags/for-upstream' into staging (2017-11-17 19:08:07 +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 c527e0afcd7d719abc3a5ca5e4c8ac2fe48b999f:
+for you to fetch changes up to a10dd1e279fc56cebc7e738925e0db85d0cea089:
-  hw/net/vmxnet3: Fix code to work on big endian hosts, too (2017-11-20 11:08:00 +0800)
+  vdpa: Expose VHOST_F_LOG_ALL on SVQ (2022-03-08 21:18:41 +0800)
 ----------------------------------------------------------------
 ----------------------------------------------------------------
-Ed Swierk (1):
+Eugenio Pérez (14):
-      net: Transmit zero UDP checksum as 0xFFFF
+      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):
-      Revert "Add new PCI ID for i82559a"
+      virtio-net: fix map leaking on error during receive
-Mao Zhongyi (1):
+ hw/net/virtio-net.c                |   1 +
-      colo-compare: fix the dangerous assignment
+ hw/virtio/meson.build              |   2 +-
+ hw/virtio/vhost-iova-tree.c        | 110 +++++++
-Stefan Weil (1):
+ hw/virtio/vhost-iova-tree.h        |  27 ++
-      MAINTAINERS: Add missing entry for eepro100 emulation
+ hw/virtio/vhost-shadow-virtqueue.c | 637 +++++++++++++++++++++++++++++++++++++
+ hw/virtio/vhost-shadow-virtqueue.h |  87 +++++
-Thomas Huth (2):
+ hw/virtio/vhost-vdpa.c             | 525 +++++++++++++++++++++++++++++-
-      hw/net/eepro100: Fix endianness problem on big endian hosts
+ include/hw/virtio/vhost-vdpa.h     |   8 +
-      hw/net/vmxnet3: Fix code to work on big endian hosts, too
+ include/qemu/iova-tree.h           |  38 ++-
+ util/iova-tree.c                   | 169 ++++++++++
- MAINTAINERS            |   5 ++
+files changed, 1587 insertions(+), 17 deletions(-)
- hw/net/e1000.c         |   2 +-
+ create mode 100644 hw/virtio/vhost-iova-tree.c
- hw/net/eepro100.c      |  17 +---
+ create mode 100644 hw/virtio/vhost-iova-tree.h
- hw/net/net_rx_pkt.c    |   2 +-
+ create mode 100644 hw/virtio/vhost-shadow-virtqueue.c
- hw/net/net_tx_pkt.c    |   2 +-
+ create mode 100644 hw/virtio/vhost-shadow-virtqueue.h
  hw/net/vmware_utils.h  |   6 ++
  hw/net/vmxnet3.c       |  49 ++++++++---
  hw/net/vmxnet3.h       | 230 ++++++++++++++++++++++++++++++-------------------
  include/hw/compat.h    |   4 -
  include/hw/pci/pci.h   |   1 -
  include/net/checksum.h |   6 ++
  net/colo-compare.c     |   2 +-
  qemu-options.hx        |   2 +-
 files changed, 201 insertions(+), 127 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

-[Qemu-devel] [PULL 1/6] colo-compare: fix the dangerous assignment
+[PULL 03/15] vhost: Add Shadow VirtQueue kick forwarding capabilities
-From: Mao Zhongyi <maozy.fnst@cn.fujitsu.com>
+From: Eugenio Pérez <eperezma@redhat.com>
-Cc: Peter Maydell <peter.maydell@linaro.org>
+At this mode no buffer forwarding will be performed in SVQ mode: Qemu
-Cc: Jason Wang <jasowang@redhat.com>
+will just forward the guest's kicks to the device.
-Cc: Zhang Chen <zhangckid@gmail.com>
-Cc: Li Zhijian <lizhijian@cn.fujitsu.com>
+Host memory notifiers regions are left out for simplicity, and they will
-Cc: Paolo Bonzini <pbonzini@redhat.com>
+not be addressed in this series.
-Fixes: 8ec14402029d783720f4312ed8a925548e1dad61
-Reported-by: Peter Maydell <peter.maydell@linaro.org>
+Acked-by: Michael S. Tsirkin <mst@redhat.com>
-Reported-by: Paolo Bonzini <pbonzini@redhat.com>
+Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
 Signed-off-by: Mao Zhongyi <maozy.fnst@cn.fujitsu.com>
 Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
 Signed-off-by: Jason Wang <jasowang@redhat.com>
 ---
- net/colo-compare.c | 2 +-
+ hw/virtio/vhost-shadow-virtqueue.c |  56 ++++++++++++++
-file changed, 1 insertion(+), 1 deletion(-)
+ hw/virtio/vhost-shadow-virtqueue.h |  14 ++++
+ hw/virtio/vhost-vdpa.c             | 145 ++++++++++++++++++++++++++++++++++++-
-diff --git a/net/colo-compare.c b/net/colo-compare.c
+ 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/net/colo-compare.c
+--- a/hw/virtio/vhost-shadow-virtqueue.c
-+++ b/net/colo-compare.c
++++ b/hw/virtio/vhost-shadow-virtqueue.c
-@@ -XXX,XX +XXX,XX @@ static int packet_enqueue(CompareState *s, int mode, Connection **con)
+@@ -XXX,XX +XXX,XX @@
-                          "drop packet");
+ #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;
      }
--    con = &conn;
-+    *con = conn;
++    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 6/6] hw/net/vmxnet3: Fix code to work on big endian hosts, too
+[PULL 08/15] vhost: Shadow virtqueue buffers forwarding
-From: Thomas Huth <thuth@redhat.com>
+From: Eugenio Pérez <eperezma@redhat.com>
-Since commit ab06ec43577177a442e8 we test the vmxnet3 device in the
+Initial version of shadow virtqueue that actually forward buffers. There
-pxe-tester, too (when running "make check SPEED=slow"). This now
+is no iommu support at the moment, and that will be addressed in future
-revealed that the code is not working there if the host is a big
+patches of this series. Since all vhost-vdpa devices use forced IOMMU,
-endian machine (for example ppc64 or s390x) - "make check SPEED=slow"
+this means that SVQ is not usable at this point of the series on any
-is now failing on such hosts.
+device.
-The vmxnet3 code lacks endianness conversions in a couple of places.
+For simplicity it only supports modern devices, that expects vring
-Interestingly, the bitfields in the structs in vmxnet3.h already tried to
+in little endian, with split ring and no event idx or indirect
-take care of the *bit* endianness of the C compilers - but the code missed
+descriptors. Support for them will not be added in this series.
-to change the *byte* endianness when reading or writing the corresponding
-structs. So the bitfields are now wrapped into unions which allow to change
+It reuses the VirtQueue code for the device part. The driver part is
-the byte endianness during runtime with the non-bitfield member of the union.
+based on Linux's virtio_ring driver, but with stripped functionality
-With these changes, "make check SPEED=slow" now properly works on big endian
+and optimizations so it's easier to review.
-hosts, too.
+However, forwarding buffers have some particular pieces: One of the most
-Reported-by: David Gibson <dgibson@redhat.com>
+unexpected ones is that a guest's buffer can expand through more than
-Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
+one descriptor in SVQ. While this is handled gracefully by qemu's
-Reviewed-by: David Gibson <dgibson@redhat.com>
+emulated virtio devices, it may cause unexpected SVQ queue full. This
-Signed-off-by: Thomas Huth <thuth@redhat.com>
+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/net/vmware_utils.h |   6 ++
+ hw/virtio/vhost-shadow-virtqueue.c | 353 ++++++++++++++++++++++++++++++++++++-
- hw/net/vmxnet3.c      |  46 +++++++---
+ hw/virtio/vhost-shadow-virtqueue.h |  26 +++
- hw/net/vmxnet3.h      | 230 ++++++++++++++++++++++++++++++--------------------
+ hw/virtio/vhost-vdpa.c             | 159 ++++++++++++++++-
-files changed, 181 insertions(+), 101 deletions(-)
+files changed, 526 insertions(+), 12 deletions(-)
-diff --git a/hw/net/vmware_utils.h b/hw/net/vmware_utils.h
+diff --git a/hw/virtio/vhost-shadow-virtqueue.c b/hw/virtio/vhost-shadow-virtqueue.c
 index XXXXXXX..XXXXXXX 100644
---- a/hw/net/vmware_utils.h
+--- a/hw/virtio/vhost-shadow-virtqueue.c
-+++ b/hw/net/vmware_utils.h
++++ b/hw/virtio/vhost-shadow-virtqueue.c
-@@ -XXX,XX +XXX,XX @@ vmw_shmem_ld16(PCIDevice *d, hwaddr addr)
+@@ -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)
  {
-     uint16_t res;
+     VhostShadowVirtqueue *svq = container_of(n, VhostShadowVirtqueue,
-     pci_dma_read(d, addr, &res, 2);
+                                              svq_kick);
-+    res = le16_to_cpu(res);
+     event_notifier_test_and_clear(n);
-     VMW_SHPRN("SHMEM load16: %" PRIx64 " (value 0x%X)", addr, res);
+-    event_notifier_set(&svq->hdev_kick);
-     return res;
++    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;
  }
-@@ -XXX,XX +XXX,XX @@ static inline void
- vmw_shmem_st16(PCIDevice *d, hwaddr addr, uint16_t value)
+ /**
 - * 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)
  {
-     VMW_SHPRN("SHMEM store16: %" PRIx64 " (value 0x%X)", addr, value);
+     VhostShadowVirtqueue *svq = container_of(n, VhostShadowVirtqueue,
-+    value = cpu_to_le16(value);
+                                              hdev_call);
-     pci_dma_write(d, addr, &value, 2);
+     event_notifier_test_and_clear(n);
 -    event_notifier_set(&svq->svq_call);
 +    vhost_svq_flush(svq, true);
  }
-@@ -XXX,XX +XXX,XX @@ vmw_shmem_ld32(PCIDevice *d, hwaddr addr)
+ /**
- {
+@@ -XXX,XX +XXX,XX @@ void vhost_svq_set_svq_kick_fd(VhostShadowVirtqueue *svq, int svq_kick_fd)
-     uint32_t res;
+     if (poll_start) {
-     pci_dma_read(d, addr, &res, 4);
+         event_notifier_init_fd(svq_kick, svq_kick_fd);
-+    res = le32_to_cpu(res);
+         event_notifier_set(svq_kick);
-     VMW_SHPRN("SHMEM load32: %" PRIx64 " (value 0x%X)", addr, res);
+-        event_notifier_set_handler(svq_kick, vhost_handle_guest_kick);
-     return res;
++        event_notifier_set_handler(svq_kick, vhost_handle_guest_kick_notifier);
- }
++    }
-@@ -XXX,XX +XXX,XX @@ static inline void
++}
- vmw_shmem_st32(PCIDevice *d, hwaddr addr, uint32_t value)
++
- {
++/**
-     VMW_SHPRN("SHMEM store32: %" PRIx64 " (value 0x%X)", addr, value);
++ * Start the shadow virtqueue operation.
-+    value = cpu_to_le32(value);
++ *
-     pci_dma_write(d, addr, &value, 4);
++ * @svq: Shadow Virtqueue
- }
++ * @vdev: VirtIO device
++ * @vq: Virtqueue to shadow
-@@ -XXX,XX +XXX,XX @@ vmw_shmem_ld64(PCIDevice *d, hwaddr addr)
++ */
- {
++void vhost_svq_start(VhostShadowVirtqueue *svq, VirtIODevice *vdev,
-     uint64_t res;
++                     VirtQueue *vq)
-     pci_dma_read(d, addr, &res, 8);
++{
-+    res = le64_to_cpu(res);
++    size_t desc_size, driver_size, device_size;
-     VMW_SHPRN("SHMEM load64: %" PRIx64 " (value %" PRIx64 ")", addr, res);
++
-     return res;
++    svq->next_guest_avail_elem = NULL;
- }
++    svq->shadow_avail_idx = 0;
-@@ -XXX,XX +XXX,XX @@ static inline void
++    svq->shadow_used_idx = 0;
- vmw_shmem_st64(PCIDevice *d, hwaddr addr, uint64_t value)
++    svq->last_used_idx = 0;
- {
++    svq->vdev = vdev;
-     VMW_SHPRN("SHMEM store64: %" PRIx64 " (value %" PRIx64 ")", addr, value);
++    svq->vq = vq;
-+    value = cpu_to_le64(value);
++
-     pci_dma_write(d, addr, &value, 8);
++    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);
-diff --git a/hw/net/vmxnet3.c b/hw/net/vmxnet3.c
++    svq->vring.desc = qemu_memalign(qemu_real_host_page_size, driver_size);
-index XXXXXXX..XXXXXXX 100644
++    desc_size = sizeof(vring_desc_t) * svq->vring.num;
---- a/hw/net/vmxnet3.c
++    svq->vring.avail = (void *)((char *)svq->vring.desc + desc_size);
-+++ b/hw/net/vmxnet3.c
++    memset(svq->vring.desc, 0, driver_size);
-@@ -XXX,XX +XXX,XX @@ vmxnet3_dump_tx_descr(struct Vmxnet3_TxDesc *descr)
++    svq->vring.used = qemu_memalign(qemu_real_host_page_size, device_size);
-               "addr %" PRIx64 ", len: %d, gen: %d, rsvd: %d, "
++    memset(svq->vring.used, 0, device_size);
-               "dtype: %d, ext1: %d, msscof: %d, hlen: %d, om: %d, "
++    svq->ring_id_maps = g_new0(VirtQueueElement *, svq->vring.num);
-               "eop: %d, cq: %d, ext2: %d, ti: %d, tci: %d",
++    for (unsigned i = 0; i < svq->vring.num - 1; i++) {
--              le64_to_cpu(descr->addr), descr->len, descr->gen, descr->rsvd,
++        svq->vring.desc[i].next = cpu_to_le16(i + 1);
 +              descr->addr, descr->len, descr->gen, descr->rsvd,
                descr->dtype, descr->ext1, descr->msscof, descr->hlen, descr->om,
                descr->eop, descr->cq, descr->ext2, descr->ti, descr->tci);
  }
@@ -XXX,XX +XXX,XX @@ vmxnet3_dump_rx_descr(struct Vmxnet3_RxDesc *descr)
  {
      VMW_PKPRN("RX DESCR: addr %" PRIx64 ", len: %d, gen: %d, rsvd: %d, "
                "dtype: %d, ext1: %d, btype: %d",
 -              le64_to_cpu(descr->addr), descr->len, descr->gen,
 +              descr->addr, descr->len, descr->gen,
                descr->rsvd, descr->dtype, descr->ext1, descr->btype);
  }
@@ -XXX,XX +XXX,XX @@ static void vmxnet3_complete_packet(VMXNET3State *s, int qidx, uint32_t tx_ridx)
      memset(&txcq_descr, 0, sizeof(txcq_descr));
      txcq_descr.txdIdx = tx_ridx;
      txcq_descr.gen = vmxnet3_ring_curr_gen(&s->txq_descr[qidx].comp_ring);
 -
 +    txcq_descr.val1 = cpu_to_le32(txcq_descr.val1);
 +    txcq_descr.val2 = cpu_to_le32(txcq_descr.val2);
      vmxnet3_ring_write_curr_cell(d, &s->txq_descr[qidx].comp_ring, &txcq_descr);
      /* Flush changes in TX descriptor before changing the counter value */
@@ -XXX,XX +XXX,XX @@ vmxnet3_on_rx_done_update_stats(VMXNET3State *s,
      }
  }
-+static inline void
+@@ -XXX,XX +XXX,XX @@ void vhost_svq_set_svq_kick_fd(VhostShadowVirtqueue *svq, int svq_kick_fd)
-+vmxnet3_ring_read_curr_txdesc(PCIDevice *pcidev, Vmxnet3Ring *ring,
+ void vhost_svq_stop(VhostShadowVirtqueue *svq)
-+                              struct Vmxnet3_TxDesc *txd)
+ {
-+{
+     event_notifier_set_handler(&svq->svq_kick, NULL);
-+    vmxnet3_ring_read_curr_cell(pcidev, ring, txd);
++    g_autofree VirtQueueElement *next_avail_elem = NULL;
-+    txd->addr = le64_to_cpu(txd->addr);
++
-+    txd->val1 = le32_to_cpu(txd->val1);
++    if (!svq->vq) {
-+    txd->val2 = le32_to_cpu(txd->val2);
++        return;
-+}
++    }
 +
- static inline bool
++    /* Send all pending used descriptors to guest */
- vmxnet3_pop_next_tx_descr(VMXNET3State *s,
++    vhost_svq_flush(svq, false);
-                           int qidx,
++
-@@ -XXX,XX +XXX,XX @@ vmxnet3_pop_next_tx_descr(VMXNET3State *s,
++    for (unsigned i = 0; i < svq->vring.num; ++i) {
-     Vmxnet3Ring *ring = &s->txq_descr[qidx].tx_ring;
++        g_autofree VirtQueueElement *elem = NULL;
-     PCIDevice *d = PCI_DEVICE(s);
++        elem = g_steal_pointer(&svq->ring_id_maps[i]);
++        if (elem) {
--    vmxnet3_ring_read_curr_cell(d, ring, txd);
++            virtqueue_detach_element(svq->vq, elem, 0);
-+    vmxnet3_ring_read_curr_txdesc(d, ring, txd);
++        }
-     if (txd->gen == vmxnet3_ring_curr_gen(ring)) {
++    }
-         /* Only read after generation field verification */
++
-         smp_rmb();
++    next_avail_elem = g_steal_pointer(&svq->next_guest_avail_elem);
-         /* Re-read to be sure we got the latest version */
++    if (next_avail_elem) {
--        vmxnet3_ring_read_curr_cell(d, ring, txd);
++        virtqueue_detach_element(svq->vq, next_avail_elem, 0);
-+        vmxnet3_ring_read_curr_txdesc(d, ring, txd);
++    }
-         VMXNET3_RING_DUMP(VMW_RIPRN, "TX", qidx, ring);
++    svq->vq = NULL;
-         *descr_idx = vmxnet3_ring_curr_cell_idx(ring);
++    g_free(svq->ring_id_maps);
-         vmxnet3_inc_tx_consumption_counter(s, qidx);
++    qemu_vfree(svq->vring.desc);
-@@ -XXX,XX +XXX,XX @@ static void vmxnet3_process_tx_queue(VMXNET3State *s, int qidx)
++    qemu_vfree(svq->vring.used);
          if (!s->skip_current_tx_pkt) {
              data_len = (txd.len > 0) ? txd.len : VMXNET3_MAX_TX_BUF_SIZE;
 -            data_pa = le64_to_cpu(txd.addr);
 +            data_pa = txd.addr;
              if (!net_tx_pkt_add_raw_fragment(s->tx_pkt,
                                                  data_pa,
@@ -XXX,XX +XXX,XX @@ vmxnet3_read_next_rx_descr(VMXNET3State *s, int qidx, int ridx,
      Vmxnet3Ring *ring = &s->rxq_descr[qidx].rx_ring[ridx];
      *didx = vmxnet3_ring_curr_cell_idx(ring);
      vmxnet3_ring_read_curr_cell(d, ring, dbuf);
 +    dbuf->addr = le64_to_cpu(dbuf->addr);
 +    dbuf->val1 = le32_to_cpu(dbuf->val1);
 +    dbuf->ext1 = le32_to_cpu(dbuf->ext1);
  }
- static inline uint8_t
+ /**
-@@ -XXX,XX +XXX,XX @@ vmxnet3_pop_rxc_descr(VMXNET3State *s, int qidx, uint32_t *descr_gen)
+diff --git a/hw/virtio/vhost-shadow-virtqueue.h b/hw/virtio/vhost-shadow-virtqueue.h
+index XXXXXXX..XXXXXXX 100644
-     pci_dma_read(PCI_DEVICE(s),
+--- a/hw/virtio/vhost-shadow-virtqueue.h
-                  daddr, &rxcd, sizeof(struct Vmxnet3_RxCompDesc));
++++ b/hw/virtio/vhost-shadow-virtqueue.h
-+    rxcd.val1 = le32_to_cpu(rxcd.val1);
+@@ -XXX,XX +XXX,XX @@ typedef struct VhostShadowVirtqueue {
-+    rxcd.val2 = le32_to_cpu(rxcd.val2);
-+    rxcd.val3 = le32_to_cpu(rxcd.val3);
+     /* Guest's call notifier, where the SVQ calls guest. */
-     ring_gen = vmxnet3_ring_curr_gen(&s->rxq_descr[qidx].comp_ring);
+     EventNotifier svq_call;
++
-     if (rxcd.gen != ring_gen) {
++    /* Virtio queue shadowing */
-@@ -XXX,XX +XXX,XX @@ vmxnet3_pci_dma_writev(PCIDevice *pci_dev,
++    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;
  }
-+static void
+@@ -XXX,XX +XXX,XX @@ static bool vhost_vdpa_svqs_start(struct vhost_dev *dev)
-+vmxnet3_pci_dma_write_rxcd(PCIDevice *pcidev, dma_addr_t pa,
+     }
-+                           struct Vmxnet3_RxCompDesc *rxcd)
-+{
+     for (i = 0; i < v->shadow_vqs->len; ++i) {
-+    rxcd->val1 = cpu_to_le32(rxcd->val1);
++        VirtQueue *vq = virtio_get_queue(dev->vdev, dev->vq_index + i);
-+    rxcd->val2 = cpu_to_le32(rxcd->val2);
+         VhostShadowVirtqueue *svq = g_ptr_array_index(v->shadow_vqs, i);
-+    rxcd->val3 = cpu_to_le32(rxcd->val3);
++        struct vhost_vring_addr addr = {
-+    pci_dma_write(pcidev, pa, rxcd, sizeof(*rxcd));
++            .index = i,
-+}
++        };
-+
++        int r;
- static bool
+         bool ok = vhost_vdpa_svq_setup(dev, svq, i, &err);
- vmxnet3_indicate_packet(VMXNET3State *s)
+         if (unlikely(!ok)) {
- {
+-            error_reportf_err(err, "Cannot setup SVQ %u: ", i);
-@@ -XXX,XX +XXX,XX @@ vmxnet3_indicate_packet(VMXNET3State *s)
++            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;
          }
+     }
-         chunk_size = MIN(bytes_left, rxd.len);
+@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_dev_start(struct vhost_dev *dev, bool started)
 -        vmxnet3_pci_dma_writev(d, data, bytes_copied,
 -                               le64_to_cpu(rxd.addr), chunk_size);
 +        vmxnet3_pci_dma_writev(d, data, bytes_copied, rxd.addr, chunk_size);
          bytes_copied += chunk_size;
          bytes_left -= chunk_size;
          vmxnet3_dump_rx_descr(&rxd);
          if (ready_rxcd_pa != 0) {
 -            pci_dma_write(d, ready_rxcd_pa, &rxcd, sizeof(rxcd));
 +            vmxnet3_pci_dma_write_rxcd(d, ready_rxcd_pa, &rxcd);
          }
+         vhost_vdpa_set_vring_ready(dev);
-         memset(&rxcd, 0, sizeof(struct Vmxnet3_RxCompDesc));
+     } else {
-@@ -XXX,XX +XXX,XX @@ vmxnet3_indicate_packet(VMXNET3State *s)
++        ok = vhost_vdpa_svqs_stop(dev);
-         rxcd.eop = 1;
++        if (unlikely(!ok)) {
-         rxcd.err = (bytes_left != 0);
++            return -1;
++        }
--        pci_dma_write(d, ready_rxcd_pa, &rxcd, sizeof(rxcd));
+         vhost_vdpa_host_notifiers_uninit(dev, dev->nvqs);
-+        vmxnet3_pci_dma_write_rxcd(d, ready_rxcd_pa, &rxcd);
+     }
          /* Flush RX descriptor changes */
          smp_wmb();
 diff --git a/hw/net/vmxnet3.h b/hw/net/vmxnet3.h
 index XXXXXXX..XXXXXXX 100644
 --- a/hw/net/vmxnet3.h
 +++ b/hw/net/vmxnet3.h
@@ -XXX,XX +XXX,XX @@ enum {
  struct Vmxnet3_TxDesc {
      __le64 addr;
 +    union {
 +        struct {
  #ifdef __BIG_ENDIAN_BITFIELD
 -    u32 msscof:14;  /* MSS, checksum offset, flags */
 -    u32 ext1:1;
 -    u32 dtype:1;    /* descriptor type */
 -    u32 rsvd:1;
 -    u32 gen:1;      /* generation bit */
 -    u32 len:14;
 +            u32 msscof:14;  /* MSS, checksum offset, flags */
 +            u32 ext1:1;
 +            u32 dtype:1;    /* descriptor type */
 +            u32 rsvd:1;
 +            u32 gen:1;      /* generation bit */
 +            u32 len:14;
  #else
 -    u32 len:14;
 -    u32 gen:1;      /* generation bit */
 -    u32 rsvd:1;
 -    u32 dtype:1;    /* descriptor type */
 -    u32 ext1:1;
 -    u32 msscof:14;  /* MSS, checksum offset, flags */
 +            u32 len:14;
 +            u32 gen:1;      /* generation bit */
 +            u32 rsvd:1;
 +            u32 dtype:1;    /* descriptor type */
 +            u32 ext1:1;
 +            u32 msscof:14;  /* MSS, checksum offset, flags */
  #endif  /* __BIG_ENDIAN_BITFIELD */
 -
 +        };
 +        u32 val1;
 +    };
 +
 +    union {
 +        struct {
  #ifdef __BIG_ENDIAN_BITFIELD
 -    u32 tci:16;     /* Tag to Insert */
 -    u32 ti:1;       /* VLAN Tag Insertion */
 -    u32 ext2:1;
 -    u32 cq:1;       /* completion request */
 -    u32 eop:1;      /* End Of Packet */
 -    u32 om:2;       /* offload mode */
 -    u32 hlen:10;    /* header len */
 +            u32 tci:16;     /* Tag to Insert */
 +            u32 ti:1;       /* VLAN Tag Insertion */
 +            u32 ext2:1;
 +            u32 cq:1;       /* completion request */
 +            u32 eop:1;      /* End Of Packet */
 +            u32 om:2;       /* offload mode */
 +            u32 hlen:10;    /* header len */
  #else
 -    u32 hlen:10;    /* header len */
 -    u32 om:2;       /* offload mode */
 -    u32 eop:1;      /* End Of Packet */
 -    u32 cq:1;       /* completion request */
 -    u32 ext2:1;
 -    u32 ti:1;       /* VLAN Tag Insertion */
 -    u32 tci:16;     /* Tag to Insert */
 +            u32 hlen:10;    /* header len */
 +            u32 om:2;       /* offload mode */
 +            u32 eop:1;      /* End Of Packet */
 +            u32 cq:1;       /* completion request */
 +            u32 ext2:1;
 +            u32 ti:1;       /* VLAN Tag Insertion */
 +            u32 tci:16;     /* Tag to Insert */
  #endif  /* __BIG_ENDIAN_BITFIELD */
 +        };
 +        u32 val2;
 +    };
  };
  /* TxDesc.OM values */
@@ -XXX,XX +XXX,XX @@ struct Vmxnet3_TxDataDesc {
  #define VMXNET3_TCD_GEN_DWORD_SHIFT    3
  struct Vmxnet3_TxCompDesc {
 -    u32        txdIdx:12;    /* Index of the EOP TxDesc */
 -    u32        ext1:20;
 -
 +    union {
 +        struct {
 +#ifdef __BIG_ENDIAN_BITFIELD
 +            u32 ext1:20;
 +            u32 txdIdx:12;    /* Index of the EOP TxDesc */
 +#else
 +            u32 txdIdx:12;    /* Index of the EOP TxDesc */
 +            u32 ext1:20;
 +#endif
 +        };
 +        u32 val1;
 +    };
      __le32        ext2;
      __le32        ext3;
 -    u32        rsvd:24;
 -    u32        type:7;       /* completion type */
 -    u32        gen:1;        /* generation bit */
 +    union {
 +        struct {
 +#ifdef __BIG_ENDIAN_BITFIELD
 +            u32 gen:1;        /* generation bit */
 +            u32 type:7;       /* completion type */
 +            u32 rsvd:24;
 +#else
 +            u32 rsvd:24;
 +            u32 type:7;       /* completion type */
 +            u32 gen:1;        /* generation bit */
 +#endif
 +        };
 +        u32 val2;
 +    };
  };
  struct Vmxnet3_RxDesc {
      __le64        addr;
 -
 +    union {
 +        struct {
  #ifdef __BIG_ENDIAN_BITFIELD
 -    u32        gen:1;        /* Generation bit */
 -    u32        rsvd:15;
 -    u32        dtype:1;      /* Descriptor type */
 -    u32        btype:1;      /* Buffer Type */
 -    u32        len:14;
 +            u32 gen:1;        /* Generation bit */
 +            u32 rsvd:15;
 +            u32 dtype:1;      /* Descriptor type */
 +            u32 btype:1;      /* Buffer Type */
 +            u32 len:14;
  #else
 -    u32        len:14;
 -    u32        btype:1;      /* Buffer Type */
 -    u32        dtype:1;      /* Descriptor type */
 -    u32        rsvd:15;
 -    u32        gen:1;        /* Generation bit */
 +            u32 len:14;
 +            u32 btype:1;      /* Buffer Type */
 +            u32 dtype:1;      /* Descriptor type */
 +            u32 rsvd:15;
 +            u32 gen:1;        /* Generation bit */
  #endif
 +        };
 +        u32 val1;
 +    };
      u32        ext1;
  };
@@ -XXX,XX +XXX,XX @@ struct Vmxnet3_RxDesc {
  #define VMXNET3_RXD_GEN_SHIFT    31
  struct Vmxnet3_RxCompDesc {
 +    union {
 +        struct {
  #ifdef __BIG_ENDIAN_BITFIELD
 -    u32        ext2:1;
 -    u32        cnc:1;        /* Checksum Not Calculated */
 -    u32        rssType:4;    /* RSS hash type used */
 -    u32        rqID:10;      /* rx queue/ring ID */
 -    u32        sop:1;        /* Start of Packet */
 -    u32        eop:1;        /* End of Packet */
 -    u32        ext1:2;
 -    u32        rxdIdx:12;    /* Index of the RxDesc */
 +            u32 ext2:1;
 +            u32 cnc:1;        /* Checksum Not Calculated */
 +            u32 rssType:4;    /* RSS hash type used */
 +            u32 rqID:10;      /* rx queue/ring ID */
 +            u32 sop:1;        /* Start of Packet */
 +            u32 eop:1;        /* End of Packet */
 +            u32 ext1:2;
 +            u32 rxdIdx:12;    /* Index of the RxDesc */
  #else
 -    u32        rxdIdx:12;    /* Index of the RxDesc */
 -    u32        ext1:2;
 -    u32        eop:1;        /* End of Packet */
 -    u32        sop:1;        /* Start of Packet */
 -    u32        rqID:10;      /* rx queue/ring ID */
 -    u32        rssType:4;    /* RSS hash type used */
 -    u32        cnc:1;        /* Checksum Not Calculated */
 -    u32        ext2:1;
 +            u32 rxdIdx:12;    /* Index of the RxDesc */
 +            u32 ext1:2;
 +            u32 eop:1;        /* End of Packet */
 +            u32 sop:1;        /* Start of Packet */
 +            u32 rqID:10;      /* rx queue/ring ID */
 +            u32 rssType:4;    /* RSS hash type used */
 +            u32 cnc:1;        /* Checksum Not Calculated */
 +            u32 ext2:1;
  #endif  /* __BIG_ENDIAN_BITFIELD */
 +        };
 +        u32 val1;
 +    };
      __le32        rssHash;      /* RSS hash value */
 +    union {
 +        struct {
  #ifdef __BIG_ENDIAN_BITFIELD
 -    u32        tci:16;       /* Tag stripped */
 -    u32        ts:1;         /* Tag is stripped */
 -    u32        err:1;        /* Error */
 -    u32        len:14;       /* data length */
 +            u32 tci:16;       /* Tag stripped */
 +            u32 ts:1;         /* Tag is stripped */
 +            u32 err:1;        /* Error */
 +            u32 len:14;       /* data length */
  #else
 -    u32        len:14;       /* data length */
 -    u32        err:1;        /* Error */
 -    u32        ts:1;         /* Tag is stripped */
 -    u32        tci:16;       /* Tag stripped */
 +            u32 len:14;       /* data length */
 +            u32 err:1;        /* Error */
 +            u32 ts:1;         /* Tag is stripped */
 +            u32 tci:16;       /* Tag stripped */
  #endif  /* __BIG_ENDIAN_BITFIELD */
 +        };
 +        u32 val2;
 +    };
 -
 +    union {
 +        struct {
  #ifdef __BIG_ENDIAN_BITFIELD
 -    u32        gen:1;        /* generation bit */
 -    u32        type:7;       /* completion type */
 -    u32        fcs:1;        /* Frame CRC correct */
 -    u32        frg:1;        /* IP Fragment */
 -    u32        v4:1;         /* IPv4 */
 -    u32        v6:1;         /* IPv6 */
 -    u32        ipc:1;        /* IP Checksum Correct */
 -    u32        tcp:1;        /* TCP packet */
 -    u32        udp:1;        /* UDP packet */
 -    u32        tuc:1;        /* TCP/UDP Checksum Correct */
 -    u32        csum:16;
 +            u32 gen:1;        /* generation bit */
 +            u32 type:7;       /* completion type */
 +            u32 fcs:1;        /* Frame CRC correct */
 +            u32 frg:1;        /* IP Fragment */
 +            u32 v4:1;         /* IPv4 */
 +            u32 v6:1;         /* IPv6 */
 +            u32 ipc:1;        /* IP Checksum Correct */
 +            u32 tcp:1;        /* TCP packet */
 +            u32 udp:1;        /* UDP packet */
 +            u32 tuc:1;        /* TCP/UDP Checksum Correct */
 +            u32 csum:16;
  #else
 -    u32        csum:16;
 -    u32        tuc:1;        /* TCP/UDP Checksum Correct */
 -    u32        udp:1;        /* UDP packet */
 -    u32        tcp:1;        /* TCP packet */
 -    u32        ipc:1;        /* IP Checksum Correct */
 -    u32        v6:1;         /* IPv6 */
 -    u32        v4:1;         /* IPv4 */
 -    u32        frg:1;        /* IP Fragment */
 -    u32        fcs:1;        /* Frame CRC correct */
 -    u32        type:7;       /* completion type */
 -    u32        gen:1;        /* generation bit */
 +            u32 csum:16;
 +            u32 tuc:1;        /* TCP/UDP Checksum Correct */
 +            u32 udp:1;        /* UDP packet */
 +            u32 tcp:1;        /* TCP packet */
 +            u32 ipc:1;        /* IP Checksum Correct */
 +            u32 v6:1;         /* IPv6 */
 +            u32 v4:1;         /* IPv4 */
 +            u32 frg:1;        /* IP Fragment */
 +            u32 fcs:1;        /* Frame CRC correct */
 +            u32 type:7;       /* completion type */
 +            u32 gen:1;        /* generation bit */
  #endif  /* __BIG_ENDIAN_BITFIELD */
 +        };
 +        u32 val3;
 +    };
  };
  /* fields in RxCompDesc we access via Vmxnet3_GenericDesc.dword[3] */
 --
 .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

-[Qemu-devel] [PULL 2/6] Revert "Add new PCI ID for i82559a"
+[PULL 10/15] util: add iova_tree_find_iova
-This reverts commit 5e89dc01133f8f5e621f6b66b356c6f37d31dafb since:
+From: Eugenio Pérez <eperezma@redhat.com>
-- we should use ID in the spec instead the one used by OEM
+This function does the reverse operation of iova_tree_find: To look for
-- in the future, we should allow changing id through either property
+a mapping that match a translated address so we can do the reverse.
   or EEPROM file.
-Cc: Stefan Weil <sw@weilnetz.de>
+This have linear complexity instead of logarithmic, but it supports
-Cc: Michael Nawrocki <michael.nawrocki@gtri.gatech.edu>
+overlapping HVA. Future developments could reduce it.
-Cc: Peter Maydell <peter.maydell@linaro.org>
-Cc: Michael S. Tsirkin <mst@redhat.com>
+Acked-by: Michael S. Tsirkin <mst@redhat.com>
-Reviewed-by: Stefan Weil <sw@weilnetz.de>
+Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
 Signed-off-by: Jason Wang <jasowang@redhat.com>
 ---
- hw/net/eepro100.c    | 13 -------------
+ include/qemu/iova-tree.h | 20 +++++++++++++++++++-
- include/hw/compat.h  |  4 ----
+ util/iova-tree.c         | 34 ++++++++++++++++++++++++++++++++++
- include/hw/pci/pci.h |  1 -
+files changed, 53 insertions(+), 1 deletion(-)
  qemu-options.hx      |  2 +-
 files changed, 1 insertion(+), 19 deletions(-)
-diff --git a/hw/net/eepro100.c b/hw/net/eepro100.c
+diff --git a/include/qemu/iova-tree.h b/include/qemu/iova-tree.h
 index XXXXXXX..XXXXXXX 100644
---- a/hw/net/eepro100.c
+--- a/include/qemu/iova-tree.h
-+++ b/hw/net/eepro100.c
++++ b/include/qemu/iova-tree.h
-@@ -XXX,XX +XXX,XX @@ typedef struct {
+@@ -XXX,XX +XXX,XX @@ int iova_tree_remove(IOVATree *tree, const DMAMap *map);
-     const char *name;
+  * @tree: the iova tree to search from
-     const char *desc;
+  * @map: the mapping to search
-     uint16_t device_id;
+  *
--    uint16_t alt_device_id;
+- * Search for a mapping in the iova tree that overlaps with the
-     uint8_t revision;
++ * Search for a mapping in the iova tree that iova overlaps with the
-     uint16_t subsystem_vendor_id;
+  * mapping range specified.  Only the first found mapping will be
-     uint16_t subsystem_id;
+  * returned.
-@@ -XXX,XX +XXX,XX @@ typedef struct {
+  *
-     /* Quasi static device properties (no need to save them). */
+@@ -XXX,XX +XXX,XX @@ int iova_tree_remove(IOVATree *tree, const DMAMap *map);
-     uint16_t stats_size;
+ const DMAMap *iova_tree_find(const IOVATree *tree, const DMAMap *map);
-     bool has_extended_tcb_support;
--    bool use_alt_device_id;
+ /**
- } EEPRO100State;
++ * iova_tree_find_iova:
++ *
- /* Word indices in EEPROM. */
++ * @tree: the iova tree to search from
-@@ -XXX,XX +XXX,XX @@ static void e100_nic_realize(PCIDevice *pci_dev, Error **errp)
++ * @map: the mapping to search
++ *
-     TRACE(OTHER, logout("\n"));
++ * Search for a mapping in the iova tree that translated_addr overlaps with the
++ * mapping range specified.  Only the first found mapping will be
--    /* By default, the i82559a adapter uses the legacy PCI ID (for the
++ * returned.
--     * i82557). This allows the PCI ID to be changed to the alternate
++ *
--     * i82559 ID if needed.
++ * Return: DMAMap pointer if found, or NULL if not found.  Note that
--     */
++ * the returned DMAMap pointer is maintained internally.  User should
--    if (s->use_alt_device_id && strcmp(info->name, "i82559a") == 0) {
++ * only read the content but never modify or free the content.  Also,
--        pci_config_set_device_id(s->dev.config, info->alt_device_id);
++ * user is responsible to make sure the pointer is valid (say, no
--    }
++ * concurrent deletion in progress).
--
++ */
-     s->device = info->device;
++const DMAMap *iova_tree_find_iova(const IOVATree *tree, const DMAMap *map);
++
-     e100_pci_reset(s, &local_err);
++/**
-@@ -XXX,XX +XXX,XX @@ static E100PCIDeviceInfo e100_devices[] = {
+  * iova_tree_find_address:
-         .desc = "Intel i82559A Ethernet",
+  *
-         .device = i82559A,
+  * @tree: the iova tree to search from
-         .device_id = PCI_DEVICE_ID_INTEL_82557,
+diff --git a/util/iova-tree.c b/util/iova-tree.c
--        .alt_device_id = PCI_DEVICE_ID_INTEL_82559,
+index XXXXXXX..XXXXXXX 100644
-         .revision = 0x06,
+--- a/util/iova-tree.c
-         .stats_size = 80,
++++ b/util/iova-tree.c
-         .has_extended_tcb_support = true,
+@@ -XXX,XX +XXX,XX @@ struct IOVATreeAllocArgs {
-@@ -XXX,XX +XXX,XX @@ static E100PCIDeviceInfo *eepro100_get_class(EEPRO100State *s)
+     bool iova_found;
  static Property e100_properties[] = {
      DEFINE_NIC_PROPERTIES(EEPRO100State, conf),
 -    DEFINE_PROP_BOOL("x-use-alt-device-id", EEPRO100State, use_alt_device_id,
 -                     true),
      DEFINE_PROP_END_OF_LIST(),
  };
-diff --git a/include/hw/compat.h b/include/hw/compat.h
++typedef struct IOVATreeFindIOVAArgs {
-index XXXXXXX..XXXXXXX 100644
++    const DMAMap *needle;
---- a/include/hw/compat.h
++    const DMAMap *result;
-+++ b/include/hw/compat.h
++} IOVATreeFindIOVAArgs;
-@@ -XXX,XX +XXX,XX @@
++
-         .driver   = "virtio-tablet-device",\
+ /**
-         .property = "wheel-axis",\
+  * Iterate args to the next hole
-         .value    = "false",\
+  *
--    },{\
+@@ -XXX,XX +XXX,XX @@ const DMAMap *iova_tree_find(const IOVATree *tree, const DMAMap *map)
--        .driver   = "i82559a",\
+     return g_tree_lookup(tree->tree, map);
--        .property = "x-use-alt-device-id",\
+ }
--        .value    = "false",\
-     },
++static gboolean iova_tree_find_address_iterator(gpointer key, gpointer value,
++                                                gpointer data)
- #define HW_COMPAT_2_9 \
++{
-diff --git a/include/hw/pci/pci.h b/include/hw/pci/pci.h
++    const DMAMap *map = key;
-index XXXXXXX..XXXXXXX 100644
++    IOVATreeFindIOVAArgs *args = data;
---- a/include/hw/pci/pci.h
++    const DMAMap *needle;
-+++ b/include/hw/pci/pci.h
++
-@@ -XXX,XX +XXX,XX @@ extern bool pci_available;
++    g_assert(key == value);
- /* Intel (0x8086) */
++
- #define PCI_DEVICE_ID_INTEL_82551IT      0x1209
++    needle = args->needle;
- #define PCI_DEVICE_ID_INTEL_82557        0x1229
++    if (map->translated_addr + map->size < needle->translated_addr ||
--#define PCI_DEVICE_ID_INTEL_82559        0x1030
++        needle->translated_addr + needle->size < map->translated_addr) {
- #define PCI_DEVICE_ID_INTEL_82801IR      0x2922
++        return false;
++    }
- /* Red Hat / Qumranet (for QEMU) -- see pci-ids.txt */
++
-diff --git a/qemu-options.hx b/qemu-options.hx
++    args->result = map;
-index XXXXXXX..XXXXXXX 100644
++    return true;
---- a/qemu-options.hx
++}
-+++ b/qemu-options.hx
++
-@@ -XXX,XX +XXX,XX @@ that the card should have; this option currently only affects virtio cards; set
++const DMAMap *iova_tree_find_iova(const IOVATree *tree, const DMAMap *map)
- @var{v} = 0 to disable MSI-X. If no @option{-net} option is specified, a single
++{
- NIC is created.  QEMU can emulate several different models of network card.
++    IOVATreeFindIOVAArgs args = {
- Valid values for @var{type} are
++        .needle = map,
--@code{virtio}, @code{i82551}, @code{i82557b}, @code{i82559a}, @code{i82559er},
++    };
-+@code{virtio}, @code{i82551}, @code{i82557b}, @code{i82559er},
++
- @code{ne2k_pci}, @code{ne2k_isa}, @code{pcnet}, @code{rtl8139},
++    g_tree_foreach(tree->tree, iova_tree_find_address_iterator, &args);
- @code{e1000}, @code{smc91c111}, @code{lance} and @code{mcf_fec}.
++    return args.result;
- Not all devices are supported on all targets.  Use @code{-net nic,model=help}
++}
 +
  const DMAMap *iova_tree_find_address(const IOVATree *tree, hwaddr iova)
  {
      const DMAMap map = { .iova = iova, .size = 0 };
 --
 .7.4

-New patch
+[PULL 11/15] vhost: Add VhostIOVATree
+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 +++++++++++
+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
+--
+.7.4

-New patch
+[PULL 12/15] vdpa: Add custom IOTLB translations to SVQ
+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 +
 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];
 --
 .7.4

-[Qemu-devel] [PULL 4/6] MAINTAINERS: Add missing entry for eepro100 emulation
+[PULL 13/15] vdpa: Adapt vhost_vdpa_get_vring_base to SVQ
-From: Stefan Weil <sw@weilnetz.de>
+From: Eugenio Pérez <eperezma@redhat.com>
-Signed-off-by: Stefan Weil <sw@weilnetz.de>
+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.
 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 | 5 +++++
+ hw/virtio/vhost-vdpa.c | 17 +++++++++++++++++
-file changed, 5 insertions(+)
+file changed, 17 insertions(+)
-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 @@ M: Dmitry Fleytman <dmitry@daynix.com>
+@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_set_vring_base(struct vhost_dev *dev,
- S: Maintained
+ static int vhost_vdpa_get_vring_base(struct vhost_dev *dev,
- F: hw/net/e1000e*
+                                        struct vhost_vring_state *ring)
+ {
-+eepro100
++    struct vhost_vdpa *v = dev->opaque;
-+M: Stefan Weil <sw@weilnetz.de>
+     int ret;
-+S: Maintained
-+F: hw/net/eepro100.c
++    if (v->shadow_vqs_enabled) {
 +        VhostShadowVirtqueue *svq = g_ptr_array_index(v->shadow_vqs,
 +                                                      ring->index);
 +
- Generic Loader
++        /*
- M: Alistair Francis <alistair.francis@xilinx.com>
++         * Setting base as last used idx, so destination will see as available
- S: Maintained
++         * 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 3/6] hw/net/eepro100: Fix endianness problem on big endian hosts
+[PULL 14/15] vdpa: Never set log_base addr if SVQ is enabled
-From: Thomas Huth <thuth@redhat.com>
+From: Eugenio Pérez <eperezma@redhat.com>
-Since commit 1865e288a823c764cd4344d ("Fix eepro100 simple transmission
+Setting the log address would make the device start reporting invalid
-mode"), the test/pxe-test is broken for the eepro100 device on big
+dirty memory because the SVQ vrings are located in qemu's memory.
 endian hosts. However, it seems like that commit did not introduce the
 problem, but just uncovered it: The EEPRO100State->tx.tbd_array_addr and
 EEPRO100State->tx.tcb_bytes fields are already in host byte order, since
 they have already been byte-swapped in the read_cb() function.
 Thus byte-swapping them in tx_command() again results in the wrong
 endianness. Removing the byte-swapping here fixes the pxe-test.
-Signed-off-by: Thomas Huth <thuth@redhat.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>
 ---
- hw/net/eepro100.c | 4 ++--
+ hw/virtio/vhost-vdpa.c | 3 ++-
-file changed, 2 insertions(+), 2 deletions(-)
+file changed, 2 insertions(+), 1 deletion(-)
-diff --git a/hw/net/eepro100.c b/hw/net/eepro100.c
+diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c
 index XXXXXXX..XXXXXXX 100644
---- a/hw/net/eepro100.c
+--- a/hw/virtio/vhost-vdpa.c
-+++ b/hw/net/eepro100.c
++++ b/hw/virtio/vhost-vdpa.c
-@@ -XXX,XX +XXX,XX @@ static void read_cb(EEPRO100State *s)
+@@ -XXX,XX +XXX,XX @@ static int vhost_vdpa_dev_start(struct vhost_dev *dev, bool started)
+ static int vhost_vdpa_set_log_base(struct vhost_dev *dev, uint64_t base,
- static void tx_command(EEPRO100State *s)
+                                      struct vhost_log *log)
  {
--    uint32_t tbd_array = le32_to_cpu(s->tx.tbd_array_addr);
+-    if (vhost_vdpa_one_time_request(dev)) {
--    uint16_t tcb_bytes = (le16_to_cpu(s->tx.tcb_bytes) & 0x3fff);
++    struct vhost_vdpa *v = dev->opaque;
-+    uint32_t tbd_array = s->tx.tbd_array_addr;
++    if (v->shadow_vqs_enabled || vhost_vdpa_one_time_request(dev)) {
-+    uint16_t tcb_bytes = s->tx.tcb_bytes & 0x3fff;
+         return 0;
-     /* Sends larger than MAX_ETH_FRAME_SIZE are allowed, up to 2600 bytes. */
+     }
-     uint8_t buf[2600];
      uint16_t size = 0;
 --
 .7.4

-[Qemu-devel] [PULL 5/6] net: Transmit zero UDP checksum as 0xFFFF
+[PULL 15/15] vdpa: Expose VHOST_F_LOG_ALL on SVQ
-From: Ed Swierk <eswierk@skyportsystems.com>
+From: Eugenio Pérez <eperezma@redhat.com>
-The checksum algorithm used by IPv4, TCP and UDP allows a zero value
+SVQ is able to log the dirty bits by itself, so let's use it to not
-to be represented by either 0x0000 and 0xFFFF. But per RFC 768, a zero
+block migration.
 UDP checksum must be transmitted as 0xFFFF because 0x0000 is a special
 value meaning no checksum.
-Substitute 0xFFFF whenever a checksum is computed as zero when
+Also, ignore set and clear of VHOST_F_LOG_ALL on set_features if SVQ is
-modifying a UDP datagram header. Doing this on IPv4 and TCP checksums
+enabled. Even if the device supports it, the reports would be nonsense
-is unnecessary but legal. Add a wrapper for net_checksum_finish() that
+because SVQ memory is in the qemu region.
 makes the substitution.
-(We can't just change net_checksum_finish(), as that function is also
+The log region is still allocated. Future changes might skip that, but
-used by receivers to verify checksums, and in that case the expected
+this series is already long enough.
 value is always 0x0000.)
-Signed-off-by: Ed Swierk <eswierk@skyportsystems.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>
 ---
- hw/net/e1000.c         | 2 +-
+ hw/virtio/vhost-vdpa.c         | 39 +++++++++++++++++++++++++++++++++++----
- hw/net/net_rx_pkt.c    | 2 +-
+ include/hw/virtio/vhost-vdpa.h |  1 +
- hw/net/net_tx_pkt.c    | 2 +-
+files changed, 36 insertions(+), 4 deletions(-)
  hw/net/vmxnet3.c       | 3 ++-
  include/net/checksum.h | 6 ++++++
 files changed, 11 insertions(+), 4 deletions(-)
-diff --git a/hw/net/e1000.c b/hw/net/e1000.c
+diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c
 index XXXXXXX..XXXXXXX 100644
---- a/hw/net/e1000.c
+--- a/hw/virtio/vhost-vdpa.c
-+++ b/hw/net/e1000.c
++++ b/hw/virtio/vhost-vdpa.c
-@@ -XXX,XX +XXX,XX @@ putsum(uint8_t *data, uint32_t n, uint32_t sloc, uint32_t css, uint32_t cse)
+@@ -XXX,XX +XXX,XX @@ static bool vhost_vdpa_one_time_request(struct vhost_dev *dev)
-         n = cse + 1;
+     return v->index != 0;
      if (sloc < n-1) {
          sum = net_checksum_add(n-css, data+css);
 -        stw_be_p(data + sloc, net_checksum_finish(sum));
 +        stw_be_p(data + sloc, net_checksum_finish_nozero(sum));
      }
  }
-diff --git a/hw/net/net_rx_pkt.c b/hw/net/net_rx_pkt.c
++static int vhost_vdpa_get_dev_features(struct vhost_dev *dev,
-index XXXXXXX..XXXXXXX 100644
++                                       uint64_t *features)
 --- a/hw/net/net_rx_pkt.c
 +++ b/hw/net/net_rx_pkt.c
@@ -XXX,XX +XXX,XX @@ _net_rx_pkt_calc_l4_csum(struct NetRxPkt *pkt)
      cntr += net_checksum_add_iov(pkt->vec, pkt->vec_len,
                                   pkt->l4hdr_off, csl, cso);
 -    csum = net_checksum_finish(cntr);
 +    csum = net_checksum_finish_nozero(cntr);
      trace_net_rx_pkt_l4_csum_calc_csum(pkt->l4hdr_off, csl, cntr, csum);
 diff --git a/hw/net/net_tx_pkt.c b/hw/net/net_tx_pkt.c
 index XXXXXXX..XXXXXXX 100644
 --- a/hw/net/net_tx_pkt.c
 +++ b/hw/net/net_tx_pkt.c
@@ -XXX,XX +XXX,XX @@ static void net_tx_pkt_do_sw_csum(struct NetTxPkt *pkt)
          net_checksum_add_iov(iov, iov_len, pkt->virt_hdr.csum_start, csl, cso);
      /* Put the checksum obtained into the packet */
 -    csum = cpu_to_be16(net_checksum_finish(csum_cntr));
 +    csum = cpu_to_be16(net_checksum_finish_nozero(csum_cntr));
      iov_from_buf(iov, iov_len, csum_offset, &csum, sizeof csum);
  }
 diff --git a/hw/net/vmxnet3.c b/hw/net/vmxnet3.c
 index XXXXXXX..XXXXXXX 100644
 --- a/hw/net/vmxnet3.c
 +++ b/hw/net/vmxnet3.c
@@ -XXX,XX +XXX,XX @@ static void vmxnet3_rx_need_csum_calculate(struct NetRxPkt *pkt,
      data = (uint8_t *)pkt_data + vhdr->csum_start;
      len = pkt_len - vhdr->csum_start;
      /* Put the checksum obtained into the packet */
 -    stw_be_p(data + vhdr->csum_offset, net_raw_checksum(data, len));
 +    stw_be_p(data + vhdr->csum_offset,
 +             net_checksum_finish_nozero(net_checksum_add(len, data)));
      vhdr->flags &= ~VIRTIO_NET_HDR_F_NEEDS_CSUM;
      vhdr->flags |= VIRTIO_NET_HDR_F_DATA_VALID;
 diff --git a/include/net/checksum.h b/include/net/checksum.h
 index XXXXXXX..XXXXXXX 100644
 --- a/include/net/checksum.h
 +++ b/include/net/checksum.h
@@ -XXX,XX +XXX,XX @@ net_checksum_add(int len, uint8_t *buf)
  }
  static inline uint16_t
 +net_checksum_finish_nozero(uint32_t sum)
 +{
-+    return net_checksum_finish(sum) ?: 0xFFFF;
++    int ret;
 +
 +    ret = vhost_vdpa_call(dev, VHOST_GET_FEATURES, features);
 +    trace_vhost_vdpa_get_features(dev, *features);
 +    return ret;
 +}
 +
-+static inline uint16_t
+ static int vhost_vdpa_init_svq(struct vhost_dev *hdev, struct vhost_vdpa *v,
- net_raw_checksum(uint8_t *data, int length)
+                                Error **errp)
  {
-     return net_checksum_finish(net_checksum_add(length, data));
+@@ -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;
 --
 .7.4

The following changes since commit 2e02083438962d26ef9dcc7100f3b378104183db:

Merge remote-tracking branch 'remotes/kevin/tags/for-upstream' into staging (2017-11-17 19:08:07 +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 c527e0afcd7d719abc3a5ca5e4c8ac2fe48b999f:

hw/net/vmxnet3: Fix code to work on big endian hosts, too (2017-11-20 11:08:00 +0800)

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

----------------------------------------------------------------
Ed Swierk (1):
      net: Transmit zero UDP checksum as 0xFFFF

Jason Wang (1):
      Revert "Add new PCI ID for i82559a"

Mao Zhongyi (1):
      colo-compare: fix the dangerous assignment

Stefan Weil (1):
      MAINTAINERS: Add missing entry for eepro100 emulation

Thomas Huth (2):
      hw/net/eepro100: Fix endianness problem on big endian hosts
      hw/net/vmxnet3: Fix code to work on big endian hosts, too

This reverts commit 5e89dc01133f8f5e621f6b66b356c6f37d31dafb since:

- we should use ID in the spec instead the one used by OEM
- in the future, we should allow changing id through either property
  or EEPROM file.

Cc: Stefan Weil <sw@weilnetz.de>
Cc: Michael Nawrocki <michael.nawrocki@gtri.gatech.edu>
Cc: Peter Maydell <peter.maydell@linaro.org>
Cc: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 hw/net/eepro100.c    | 13 -------------
 include/hw/compat.h  |  4 ----
 include/hw/pci/pci.h |  1 -
 qemu-options.hx      |  2 +-
 4 files changed, 1 insertion(+), 19 deletions(-)

diff --git a/hw/net/eepro100.c b/hw/net/eepro100.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/net/eepro100.c
+++ b/hw/net/eepro100.c
@@ -XXX,XX +XXX,XX @@ typedef struct {
     const char *name;
     const char *desc;
     uint16_t device_id;
-    uint16_t alt_device_id;
     uint8_t revision;
     uint16_t subsystem_vendor_id;
     uint16_t subsystem_id;
@@ -XXX,XX +XXX,XX @@ typedef struct {
     /* Quasi static device properties (no need to save them). */
     uint16_t stats_size;
     bool has_extended_tcb_support;
-    bool use_alt_device_id;
 } EEPRO100State;
 
 /* Word indices in EEPROM. */
@@ -XXX,XX +XXX,XX @@ static void e100_nic_realize(PCIDevice *pci_dev, Error **errp)
 
     TRACE(OTHER, logout("\n"));
 
-    /* By default, the i82559a adapter uses the legacy PCI ID (for the
-     * i82557). This allows the PCI ID to be changed to the alternate
-     * i82559 ID if needed.
-     */
-    if (s->use_alt_device_id && strcmp(info->name, "i82559a") == 0) {
-        pci_config_set_device_id(s->dev.config, info->alt_device_id);
-    }
-
     s->device = info->device;
 
     e100_pci_reset(s, &local_err);
@@ -XXX,XX +XXX,XX @@ static E100PCIDeviceInfo e100_devices[] = {
         .desc = "Intel i82559A Ethernet",
         .device = i82559A,
         .device_id = PCI_DEVICE_ID_INTEL_82557,
-        .alt_device_id = PCI_DEVICE_ID_INTEL_82559,
         .revision = 0x06,
         .stats_size = 80,
         .has_extended_tcb_support = true,
@@ -XXX,XX +XXX,XX @@ static E100PCIDeviceInfo *eepro100_get_class(EEPRO100State *s)
 
 static Property e100_properties[] = {
     DEFINE_NIC_PROPERTIES(EEPRO100State, conf),
-    DEFINE_PROP_BOOL("x-use-alt-device-id", EEPRO100State, use_alt_device_id,
-                     true),
     DEFINE_PROP_END_OF_LIST(),
 };
 
diff --git a/include/hw/compat.h b/include/hw/compat.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/compat.h
+++ b/include/hw/compat.h
@@ -XXX,XX +XXX,XX @@
         .driver   = "virtio-tablet-device",\
         .property = "wheel-axis",\
         .value    = "false",\
-    },{\
-        .driver   = "i82559a",\
-        .property = "x-use-alt-device-id",\
-        .value    = "false",\
     },
 
 #define HW_COMPAT_2_9 \
diff --git a/include/hw/pci/pci.h b/include/hw/pci/pci.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/pci/pci.h
+++ b/include/hw/pci/pci.h
@@ -XXX,XX +XXX,XX @@ extern bool pci_available;
 /* Intel (0x8086) */
 #define PCI_DEVICE_ID_INTEL_82551IT      0x1209
 #define PCI_DEVICE_ID_INTEL_82557        0x1229
-#define PCI_DEVICE_ID_INTEL_82559        0x1030
 #define PCI_DEVICE_ID_INTEL_82801IR      0x2922
 
 /* Red Hat / Qumranet (for QEMU) -- see pci-ids.txt */
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 @@ that the card should have; this option currently only affects virtio cards; set
 @var{v} = 0 to disable MSI-X. If no @option{-net} option is specified, a single
 NIC is created.  QEMU can emulate several different models of network card.
 Valid values for @var{type} are
-@code{virtio}, @code{i82551}, @code{i82557b}, @code{i82559a}, @code{i82559er},
+@code{virtio}, @code{i82551}, @code{i82557b}, @code{i82559er},
 @code{ne2k_pci}, @code{ne2k_isa}, @code{pcnet}, @code{rtl8139},
 @code{e1000}, @code{smc91c111}, @code{lance} and @code{mcf_fec}.
 Not all devices are supported on all targets.  Use @code{-net nic,model=help}
-- 
2.7.4

From: Thomas Huth <thuth@redhat.com>

Since commit 1865e288a823c764cd4344d ("Fix eepro100 simple transmission
mode"), the test/pxe-test is broken for the eepro100 device on big
endian hosts. However, it seems like that commit did not introduce the
problem, but just uncovered it: The EEPRO100State->tx.tbd_array_addr and
EEPRO100State->tx.tcb_bytes fields are already in host byte order, since
they have already been byte-swapped in the read_cb() function.
Thus byte-swapping them in tx_command() again results in the wrong
endianness. Removing the byte-swapping here fixes the pxe-test.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 hw/net/eepro100.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/hw/net/eepro100.c b/hw/net/eepro100.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/net/eepro100.c
+++ b/hw/net/eepro100.c
@@ -XXX,XX +XXX,XX @@ static void read_cb(EEPRO100State *s)
 
 static void tx_command(EEPRO100State *s)
 {
-    uint32_t tbd_array = le32_to_cpu(s->tx.tbd_array_addr);
-    uint16_t tcb_bytes = (le16_to_cpu(s->tx.tcb_bytes) & 0x3fff);
+    uint32_t tbd_array = s->tx.tbd_array_addr;
+    uint16_t tcb_bytes = s->tx.tcb_bytes & 0x3fff;
     /* Sends larger than MAX_ETH_FRAME_SIZE are allowed, up to 2600 bytes. */
     uint8_t buf[2600];
     uint16_t size = 0;
-- 
2.7.4

From: Ed Swierk <eswierk@skyportsystems.com>

The checksum algorithm used by IPv4, TCP and UDP allows a zero value
to be represented by either 0x0000 and 0xFFFF. But per RFC 768, a zero
UDP checksum must be transmitted as 0xFFFF because 0x0000 is a special
value meaning no checksum.

Substitute 0xFFFF whenever a checksum is computed as zero when
modifying a UDP datagram header. Doing this on IPv4 and TCP checksums
is unnecessary but legal. Add a wrapper for net_checksum_finish() that
makes the substitution.

(We can't just change net_checksum_finish(), as that function is also
used by receivers to verify checksums, and in that case the expected
value is always 0x0000.)

Signed-off-by: Ed Swierk <eswierk@skyportsystems.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 hw/net/e1000.c         | 2 +-
 hw/net/net_rx_pkt.c    | 2 +-
 hw/net/net_tx_pkt.c    | 2 +-
 hw/net/vmxnet3.c       | 3 ++-
 include/net/checksum.h | 6 ++++++
 5 files changed, 11 insertions(+), 4 deletions(-)

diff --git a/hw/net/e1000.c b/hw/net/e1000.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/net/e1000.c
+++ b/hw/net/e1000.c
@@ -XXX,XX +XXX,XX @@ putsum(uint8_t *data, uint32_t n, uint32_t sloc, uint32_t css, uint32_t cse)
         n = cse + 1;
     if (sloc < n-1) {
         sum = net_checksum_add(n-css, data+css);
-        stw_be_p(data + sloc, net_checksum_finish(sum));
+        stw_be_p(data + sloc, net_checksum_finish_nozero(sum));
     }
 }
 
diff --git a/hw/net/net_rx_pkt.c b/hw/net/net_rx_pkt.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/net/net_rx_pkt.c
+++ b/hw/net/net_rx_pkt.c
@@ -XXX,XX +XXX,XX @@ _net_rx_pkt_calc_l4_csum(struct NetRxPkt *pkt)
     cntr += net_checksum_add_iov(pkt->vec, pkt->vec_len,
                                  pkt->l4hdr_off, csl, cso);
 
-    csum = net_checksum_finish(cntr);
+    csum = net_checksum_finish_nozero(cntr);
 
     trace_net_rx_pkt_l4_csum_calc_csum(pkt->l4hdr_off, csl, cntr, csum);
 
diff --git a/hw/net/net_tx_pkt.c b/hw/net/net_tx_pkt.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/net/net_tx_pkt.c
+++ b/hw/net/net_tx_pkt.c
@@ -XXX,XX +XXX,XX @@ static void net_tx_pkt_do_sw_csum(struct NetTxPkt *pkt)
         net_checksum_add_iov(iov, iov_len, pkt->virt_hdr.csum_start, csl, cso);
 
     /* Put the checksum obtained into the packet */
-    csum = cpu_to_be16(net_checksum_finish(csum_cntr));
+    csum = cpu_to_be16(net_checksum_finish_nozero(csum_cntr));
     iov_from_buf(iov, iov_len, csum_offset, &csum, sizeof csum);
 }
 
diff --git a/hw/net/vmxnet3.c b/hw/net/vmxnet3.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/net/vmxnet3.c
+++ b/hw/net/vmxnet3.c
@@ -XXX,XX +XXX,XX @@ static void vmxnet3_rx_need_csum_calculate(struct NetRxPkt *pkt,
     data = (uint8_t *)pkt_data + vhdr->csum_start;
     len = pkt_len - vhdr->csum_start;
     /* Put the checksum obtained into the packet */
-    stw_be_p(data + vhdr->csum_offset, net_raw_checksum(data, len));
+    stw_be_p(data + vhdr->csum_offset,
+             net_checksum_finish_nozero(net_checksum_add(len, data)));
 
     vhdr->flags &= ~VIRTIO_NET_HDR_F_NEEDS_CSUM;
     vhdr->flags |= VIRTIO_NET_HDR_F_DATA_VALID;
diff --git a/include/net/checksum.h b/include/net/checksum.h
index XXXXXXX..XXXXXXX 100644
--- a/include/net/checksum.h
+++ b/include/net/checksum.h
@@ -XXX,XX +XXX,XX @@ net_checksum_add(int len, uint8_t *buf)
 }
 
 static inline uint16_t
+net_checksum_finish_nozero(uint32_t sum)
+{
+    return net_checksum_finish(sum) ?: 0xFFFF;
+}
+
+static inline uint16_t
 net_raw_checksum(uint8_t *data, int length)
 {
     return net_checksum_finish(net_checksum_add(length, data));
-- 
2.7.4

From: Thomas Huth <thuth@redhat.com>

Since commit ab06ec43577177a442e8 we test the vmxnet3 device in the
pxe-tester, too (when running "make check SPEED=slow"). This now
revealed that the code is not working there if the host is a big
endian machine (for example ppc64 or s390x) - "make check SPEED=slow"
is now failing on such hosts.

The vmxnet3 code lacks endianness conversions in a couple of places.
Interestingly, the bitfields in the structs in vmxnet3.h already tried to
take care of the *bit* endianness of the C compilers - but the code missed
to change the *byte* endianness when reading or writing the corresponding
structs. So the bitfields are now wrapped into unions which allow to change
the byte endianness during runtime with the non-bitfield member of the union.
With these changes, "make check SPEED=slow" now properly works on big endian
hosts, too.

Reported-by: David Gibson <dgibson@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: David Gibson <dgibson@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 hw/net/vmware_utils.h |   6 ++
 hw/net/vmxnet3.c      |  46 +++++++---
 hw/net/vmxnet3.h      | 230 ++++++++++++++++++++++++++++++--------------------
 3 files changed, 181 insertions(+), 101 deletions(-)

diff --git a/hw/net/vmware_utils.h b/hw/net/vmware_utils.h
index XXXXXXX..XXXXXXX 100644
--- a/hw/net/vmware_utils.h
+++ b/hw/net/vmware_utils.h
@@ -XXX,XX +XXX,XX @@ vmw_shmem_ld16(PCIDevice *d, hwaddr addr)
 {
     uint16_t res;
     pci_dma_read(d, addr, &res, 2);
+    res = le16_to_cpu(res);
     VMW_SHPRN("SHMEM load16: %" PRIx64 " (value 0x%X)", addr, res);
     return res;
 }
@@ -XXX,XX +XXX,XX @@ static inline void
 vmw_shmem_st16(PCIDevice *d, hwaddr addr, uint16_t value)
 {
     VMW_SHPRN("SHMEM store16: %" PRIx64 " (value 0x%X)", addr, value);
+    value = cpu_to_le16(value);
     pci_dma_write(d, addr, &value, 2);
 }
 
@@ -XXX,XX +XXX,XX @@ vmw_shmem_ld32(PCIDevice *d, hwaddr addr)
 {
     uint32_t res;
     pci_dma_read(d, addr, &res, 4);
+    res = le32_to_cpu(res);
     VMW_SHPRN("SHMEM load32: %" PRIx64 " (value 0x%X)", addr, res);
     return res;
 }
@@ -XXX,XX +XXX,XX @@ static inline void
 vmw_shmem_st32(PCIDevice *d, hwaddr addr, uint32_t value)
 {
     VMW_SHPRN("SHMEM store32: %" PRIx64 " (value 0x%X)", addr, value);
+    value = cpu_to_le32(value);
     pci_dma_write(d, addr, &value, 4);
 }
 
@@ -XXX,XX +XXX,XX @@ vmw_shmem_ld64(PCIDevice *d, hwaddr addr)
 {
     uint64_t res;
     pci_dma_read(d, addr, &res, 8);
+    res = le64_to_cpu(res);
     VMW_SHPRN("SHMEM load64: %" PRIx64 " (value %" PRIx64 ")", addr, res);
     return res;
 }
@@ -XXX,XX +XXX,XX @@ static inline void
 vmw_shmem_st64(PCIDevice *d, hwaddr addr, uint64_t value)
 {
     VMW_SHPRN("SHMEM store64: %" PRIx64 " (value %" PRIx64 ")", addr, value);
+    value = cpu_to_le64(value);
     pci_dma_write(d, addr, &value, 8);
 }
 
diff --git a/hw/net/vmxnet3.c b/hw/net/vmxnet3.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/net/vmxnet3.c
+++ b/hw/net/vmxnet3.c
@@ -XXX,XX +XXX,XX @@ vmxnet3_dump_tx_descr(struct Vmxnet3_TxDesc *descr)
               "addr %" PRIx64 ", len: %d, gen: %d, rsvd: %d, "
               "dtype: %d, ext1: %d, msscof: %d, hlen: %d, om: %d, "
               "eop: %d, cq: %d, ext2: %d, ti: %d, tci: %d",
-              le64_to_cpu(descr->addr), descr->len, descr->gen, descr->rsvd,
+              descr->addr, descr->len, descr->gen, descr->rsvd,
               descr->dtype, descr->ext1, descr->msscof, descr->hlen, descr->om,
               descr->eop, descr->cq, descr->ext2, descr->ti, descr->tci);
 }
@@ -XXX,XX +XXX,XX @@ vmxnet3_dump_rx_descr(struct Vmxnet3_RxDesc *descr)
 {
     VMW_PKPRN("RX DESCR: addr %" PRIx64 ", len: %d, gen: %d, rsvd: %d, "
               "dtype: %d, ext1: %d, btype: %d",
-              le64_to_cpu(descr->addr), descr->len, descr->gen,
+              descr->addr, descr->len, descr->gen,
               descr->rsvd, descr->dtype, descr->ext1, descr->btype);
 }
 
@@ -XXX,XX +XXX,XX @@ static void vmxnet3_complete_packet(VMXNET3State *s, int qidx, uint32_t tx_ridx)
     memset(&txcq_descr, 0, sizeof(txcq_descr));
     txcq_descr.txdIdx = tx_ridx;
     txcq_descr.gen = vmxnet3_ring_curr_gen(&s->txq_descr[qidx].comp_ring);
-
+    txcq_descr.val1 = cpu_to_le32(txcq_descr.val1);
+    txcq_descr.val2 = cpu_to_le32(txcq_descr.val2);
     vmxnet3_ring_write_curr_cell(d, &s->txq_descr[qidx].comp_ring, &txcq_descr);
 
     /* Flush changes in TX descriptor before changing the counter value */
@@ -XXX,XX +XXX,XX @@ vmxnet3_on_rx_done_update_stats(VMXNET3State *s,
     }
 }
 
+static inline void
+vmxnet3_ring_read_curr_txdesc(PCIDevice *pcidev, Vmxnet3Ring *ring,
+                              struct Vmxnet3_TxDesc *txd)
+{
+    vmxnet3_ring_read_curr_cell(pcidev, ring, txd);
+    txd->addr = le64_to_cpu(txd->addr);
+    txd->val1 = le32_to_cpu(txd->val1);
+    txd->val2 = le32_to_cpu(txd->val2);
+}
+
 static inline bool
 vmxnet3_pop_next_tx_descr(VMXNET3State *s,
                           int qidx,
@@ -XXX,XX +XXX,XX @@ vmxnet3_pop_next_tx_descr(VMXNET3State *s,
     Vmxnet3Ring *ring = &s->txq_descr[qidx].tx_ring;
     PCIDevice *d = PCI_DEVICE(s);
 
-    vmxnet3_ring_read_curr_cell(d, ring, txd);
+    vmxnet3_ring_read_curr_txdesc(d, ring, txd);
     if (txd->gen == vmxnet3_ring_curr_gen(ring)) {
         /* Only read after generation field verification */
         smp_rmb();
         /* Re-read to be sure we got the latest version */
-        vmxnet3_ring_read_curr_cell(d, ring, txd);
+        vmxnet3_ring_read_curr_txdesc(d, ring, txd);
         VMXNET3_RING_DUMP(VMW_RIPRN, "TX", qidx, ring);
         *descr_idx = vmxnet3_ring_curr_cell_idx(ring);
         vmxnet3_inc_tx_consumption_counter(s, qidx);
@@ -XXX,XX +XXX,XX @@ static void vmxnet3_process_tx_queue(VMXNET3State *s, int qidx)
 
         if (!s->skip_current_tx_pkt) {
             data_len = (txd.len > 0) ? txd.len : VMXNET3_MAX_TX_BUF_SIZE;
-            data_pa = le64_to_cpu(txd.addr);
+            data_pa = txd.addr;
 
             if (!net_tx_pkt_add_raw_fragment(s->tx_pkt,
                                                 data_pa,
@@ -XXX,XX +XXX,XX @@ vmxnet3_read_next_rx_descr(VMXNET3State *s, int qidx, int ridx,
     Vmxnet3Ring *ring = &s->rxq_descr[qidx].rx_ring[ridx];
     *didx = vmxnet3_ring_curr_cell_idx(ring);
     vmxnet3_ring_read_curr_cell(d, ring, dbuf);
+    dbuf->addr = le64_to_cpu(dbuf->addr);
+    dbuf->val1 = le32_to_cpu(dbuf->val1);
+    dbuf->ext1 = le32_to_cpu(dbuf->ext1);
 }
 
 static inline uint8_t
@@ -XXX,XX +XXX,XX @@ vmxnet3_pop_rxc_descr(VMXNET3State *s, int qidx, uint32_t *descr_gen)
 
     pci_dma_read(PCI_DEVICE(s),
                  daddr, &rxcd, sizeof(struct Vmxnet3_RxCompDesc));
+    rxcd.val1 = le32_to_cpu(rxcd.val1);
+    rxcd.val2 = le32_to_cpu(rxcd.val2);
+    rxcd.val3 = le32_to_cpu(rxcd.val3);
     ring_gen = vmxnet3_ring_curr_gen(&s->rxq_descr[qidx].comp_ring);
 
     if (rxcd.gen != ring_gen) {
@@ -XXX,XX +XXX,XX @@ vmxnet3_pci_dma_writev(PCIDevice *pci_dev,
     }
 }
 
+static void
+vmxnet3_pci_dma_write_rxcd(PCIDevice *pcidev, dma_addr_t pa,
+                           struct Vmxnet3_RxCompDesc *rxcd)
+{
+    rxcd->val1 = cpu_to_le32(rxcd->val1);
+    rxcd->val2 = cpu_to_le32(rxcd->val2);
+    rxcd->val3 = cpu_to_le32(rxcd->val3);
+    pci_dma_write(pcidev, pa, rxcd, sizeof(*rxcd));
+}
+
 static bool
 vmxnet3_indicate_packet(VMXNET3State *s)
 {
@@ -XXX,XX +XXX,XX @@ vmxnet3_indicate_packet(VMXNET3State *s)
         }
 
         chunk_size = MIN(bytes_left, rxd.len);
-        vmxnet3_pci_dma_writev(d, data, bytes_copied,
-                               le64_to_cpu(rxd.addr), chunk_size);
+        vmxnet3_pci_dma_writev(d, data, bytes_copied, rxd.addr, chunk_size);
         bytes_copied += chunk_size;
         bytes_left -= chunk_size;
 
         vmxnet3_dump_rx_descr(&rxd);
 
         if (ready_rxcd_pa != 0) {
-            pci_dma_write(d, ready_rxcd_pa, &rxcd, sizeof(rxcd));
+            vmxnet3_pci_dma_write_rxcd(d, ready_rxcd_pa, &rxcd);
         }
 
         memset(&rxcd, 0, sizeof(struct Vmxnet3_RxCompDesc));
@@ -XXX,XX +XXX,XX @@ vmxnet3_indicate_packet(VMXNET3State *s)
         rxcd.eop = 1;
         rxcd.err = (bytes_left != 0);
 
-        pci_dma_write(d, ready_rxcd_pa, &rxcd, sizeof(rxcd));
+        vmxnet3_pci_dma_write_rxcd(d, ready_rxcd_pa, &rxcd);
 
         /* Flush RX descriptor changes */
         smp_wmb();
diff --git a/hw/net/vmxnet3.h b/hw/net/vmxnet3.h
index XXXXXXX..XXXXXXX 100644
--- a/hw/net/vmxnet3.h
+++ b/hw/net/vmxnet3.h
@@ -XXX,XX +XXX,XX @@ enum {
 struct Vmxnet3_TxDesc {
     __le64 addr;
 
+    union {
+        struct {
 #ifdef __BIG_ENDIAN_BITFIELD
-    u32 msscof:14;  /* MSS, checksum offset, flags */
-    u32 ext1:1;
-    u32 dtype:1;    /* descriptor type */
-    u32 rsvd:1;
-    u32 gen:1;      /* generation bit */
-    u32 len:14;
+            u32 msscof:14;  /* MSS, checksum offset, flags */
+            u32 ext1:1;
+            u32 dtype:1;    /* descriptor type */
+            u32 rsvd:1;
+            u32 gen:1;      /* generation bit */
+            u32 len:14;
 #else
-    u32 len:14;
-    u32 gen:1;      /* generation bit */
-    u32 rsvd:1;
-    u32 dtype:1;    /* descriptor type */
-    u32 ext1:1;
-    u32 msscof:14;  /* MSS, checksum offset, flags */
+            u32 len:14;
+            u32 gen:1;      /* generation bit */
+            u32 rsvd:1;
+            u32 dtype:1;    /* descriptor type */
+            u32 ext1:1;
+            u32 msscof:14;  /* MSS, checksum offset, flags */
 #endif  /* __BIG_ENDIAN_BITFIELD */
-
+        };
+        u32 val1;
+    };
+    
+    union {
+        struct {
 #ifdef __BIG_ENDIAN_BITFIELD
-    u32 tci:16;     /* Tag to Insert */
-    u32 ti:1;       /* VLAN Tag Insertion */
-    u32 ext2:1;
-    u32 cq:1;       /* completion request */
-    u32 eop:1;      /* End Of Packet */
-    u32 om:2;       /* offload mode */
-    u32 hlen:10;    /* header len */
+            u32 tci:16;     /* Tag to Insert */
+            u32 ti:1;       /* VLAN Tag Insertion */
+            u32 ext2:1;
+            u32 cq:1;       /* completion request */
+            u32 eop:1;      /* End Of Packet */
+            u32 om:2;       /* offload mode */
+            u32 hlen:10;    /* header len */
 #else
-    u32 hlen:10;    /* header len */
-    u32 om:2;       /* offload mode */
-    u32 eop:1;      /* End Of Packet */
-    u32 cq:1;       /* completion request */
-    u32 ext2:1;
-    u32 ti:1;       /* VLAN Tag Insertion */
-    u32 tci:16;     /* Tag to Insert */
+            u32 hlen:10;    /* header len */
+            u32 om:2;       /* offload mode */
+            u32 eop:1;      /* End Of Packet */
+            u32 cq:1;       /* completion request */
+            u32 ext2:1;
+            u32 ti:1;       /* VLAN Tag Insertion */
+            u32 tci:16;     /* Tag to Insert */
 #endif  /* __BIG_ENDIAN_BITFIELD */
+        };
+        u32 val2;
+    };
 };
 
 /* TxDesc.OM values */
@@ -XXX,XX +XXX,XX @@ struct Vmxnet3_TxDataDesc {
 #define VMXNET3_TCD_GEN_DWORD_SHIFT    3
 
 struct Vmxnet3_TxCompDesc {
-    u32        txdIdx:12;    /* Index of the EOP TxDesc */
-    u32        ext1:20;
-
+    union {
+        struct {
+#ifdef __BIG_ENDIAN_BITFIELD
+            u32 ext1:20;
+            u32 txdIdx:12;    /* Index of the EOP TxDesc */
+#else
+            u32 txdIdx:12;    /* Index of the EOP TxDesc */
+            u32 ext1:20;
+#endif
+        };
+        u32 val1;
+    };
     __le32        ext2;
     __le32        ext3;
 
-    u32        rsvd:24;
-    u32        type:7;       /* completion type */
-    u32        gen:1;        /* generation bit */
+    union {
+        struct {
+#ifdef __BIG_ENDIAN_BITFIELD
+            u32 gen:1;        /* generation bit */
+            u32 type:7;       /* completion type */
+            u32 rsvd:24;
+#else
+            u32 rsvd:24;
+            u32 type:7;       /* completion type */
+            u32 gen:1;        /* generation bit */
+#endif
+        };
+        u32 val2;
+    };
 };
 
 struct Vmxnet3_RxDesc {
     __le64        addr;
-
+    union {
+        struct {
 #ifdef __BIG_ENDIAN_BITFIELD
-    u32        gen:1;        /* Generation bit */
-    u32        rsvd:15;
-    u32        dtype:1;      /* Descriptor type */
-    u32        btype:1;      /* Buffer Type */
-    u32        len:14;
+            u32 gen:1;        /* Generation bit */
+            u32 rsvd:15;
+            u32 dtype:1;      /* Descriptor type */
+            u32 btype:1;      /* Buffer Type */
+            u32 len:14;
 #else
-    u32        len:14;
-    u32        btype:1;      /* Buffer Type */
-    u32        dtype:1;      /* Descriptor type */
-    u32        rsvd:15;
-    u32        gen:1;        /* Generation bit */
+            u32 len:14;
+            u32 btype:1;      /* Buffer Type */
+            u32 dtype:1;      /* Descriptor type */
+            u32 rsvd:15;
+            u32 gen:1;        /* Generation bit */
 #endif
+        };
+        u32 val1;
+    };
     u32        ext1;
 };
 
@@ -XXX,XX +XXX,XX @@ struct Vmxnet3_RxDesc {
 #define VMXNET3_RXD_GEN_SHIFT    31
 
 struct Vmxnet3_RxCompDesc {
+    union {
+        struct {
 #ifdef __BIG_ENDIAN_BITFIELD
-    u32        ext2:1;
-    u32        cnc:1;        /* Checksum Not Calculated */
-    u32        rssType:4;    /* RSS hash type used */
-    u32        rqID:10;      /* rx queue/ring ID */
-    u32        sop:1;        /* Start of Packet */
-    u32        eop:1;        /* End of Packet */
-    u32        ext1:2;
-    u32        rxdIdx:12;    /* Index of the RxDesc */
+            u32 ext2:1;
+            u32 cnc:1;        /* Checksum Not Calculated */
+            u32 rssType:4;    /* RSS hash type used */
+            u32 rqID:10;      /* rx queue/ring ID */
+            u32 sop:1;        /* Start of Packet */
+            u32 eop:1;        /* End of Packet */
+            u32 ext1:2;
+            u32 rxdIdx:12;    /* Index of the RxDesc */
 #else
-    u32        rxdIdx:12;    /* Index of the RxDesc */
-    u32        ext1:2;
-    u32        eop:1;        /* End of Packet */
-    u32        sop:1;        /* Start of Packet */
-    u32        rqID:10;      /* rx queue/ring ID */
-    u32        rssType:4;    /* RSS hash type used */
-    u32        cnc:1;        /* Checksum Not Calculated */
-    u32        ext2:1;
+            u32 rxdIdx:12;    /* Index of the RxDesc */
+            u32 ext1:2;
+            u32 eop:1;        /* End of Packet */
+            u32 sop:1;        /* Start of Packet */
+            u32 rqID:10;      /* rx queue/ring ID */
+            u32 rssType:4;    /* RSS hash type used */
+            u32 cnc:1;        /* Checksum Not Calculated */
+            u32 ext2:1;
 #endif  /* __BIG_ENDIAN_BITFIELD */
+        };
+        u32 val1;
+    };
 
     __le32        rssHash;      /* RSS hash value */
 
+    union {
+        struct {
 #ifdef __BIG_ENDIAN_BITFIELD
-    u32        tci:16;       /* Tag stripped */
-    u32        ts:1;         /* Tag is stripped */
-    u32        err:1;        /* Error */
-    u32        len:14;       /* data length */
+            u32 tci:16;       /* Tag stripped */
+            u32 ts:1;         /* Tag is stripped */
+            u32 err:1;        /* Error */
+            u32 len:14;       /* data length */
 #else
-    u32        len:14;       /* data length */
-    u32        err:1;        /* Error */
-    u32        ts:1;         /* Tag is stripped */
-    u32        tci:16;       /* Tag stripped */
+            u32 len:14;       /* data length */
+            u32 err:1;        /* Error */
+            u32 ts:1;         /* Tag is stripped */
+            u32 tci:16;       /* Tag stripped */
 #endif  /* __BIG_ENDIAN_BITFIELD */
+        };
+        u32 val2;
+    };
 
-
+    union {
+        struct {
 #ifdef __BIG_ENDIAN_BITFIELD
-    u32        gen:1;        /* generation bit */
-    u32        type:7;       /* completion type */
-    u32        fcs:1;        /* Frame CRC correct */
-    u32        frg:1;        /* IP Fragment */
-    u32        v4:1;         /* IPv4 */
-    u32        v6:1;         /* IPv6 */
-    u32        ipc:1;        /* IP Checksum Correct */
-    u32        tcp:1;        /* TCP packet */
-    u32        udp:1;        /* UDP packet */
-    u32        tuc:1;        /* TCP/UDP Checksum Correct */
-    u32        csum:16;
+            u32 gen:1;        /* generation bit */
+            u32 type:7;       /* completion type */
+            u32 fcs:1;        /* Frame CRC correct */
+            u32 frg:1;        /* IP Fragment */
+            u32 v4:1;         /* IPv4 */
+            u32 v6:1;         /* IPv6 */
+            u32 ipc:1;        /* IP Checksum Correct */
+            u32 tcp:1;        /* TCP packet */
+            u32 udp:1;        /* UDP packet */
+            u32 tuc:1;        /* TCP/UDP Checksum Correct */
+            u32 csum:16;
 #else
-    u32        csum:16;
-    u32        tuc:1;        /* TCP/UDP Checksum Correct */
-    u32        udp:1;        /* UDP packet */
-    u32        tcp:1;        /* TCP packet */
-    u32        ipc:1;        /* IP Checksum Correct */
-    u32        v6:1;         /* IPv6 */
-    u32        v4:1;         /* IPv4 */
-    u32        frg:1;        /* IP Fragment */
-    u32        fcs:1;        /* Frame CRC correct */
-    u32        type:7;       /* completion type */
-    u32        gen:1;        /* generation bit */
+            u32 csum:16;
+            u32 tuc:1;        /* TCP/UDP Checksum Correct */
+            u32 udp:1;        /* UDP packet */
+            u32 tcp:1;        /* TCP packet */
+            u32 ipc:1;        /* IP Checksum Correct */
+            u32 v6:1;         /* IPv6 */
+            u32 v4:1;         /* IPv4 */
+            u32 frg:1;        /* IP Fragment */
+            u32 fcs:1;        /* Frame CRC correct */
+            u32 type:7;       /* completion type */
+            u32 gen:1;        /* generation bit */
 #endif  /* __BIG_ENDIAN_BITFIELD */
+        };
+        u32 val3;
+    };
 };
 
 /* fields in RxCompDesc we access via Vmxnet3_GenericDesc.dword[3] */
-- 
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