Merge remote-tracking branch 'remotes/mst/tags/for_upstream' into staging

virtio, pc: fixes, features

New virtio iommu.
Unrealize memory leaks.
In-band kick/call support.
Bugfixes, documentation all over the place.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

# gpg: Signature made Thu 27 Feb 2020 08:46:33 GMT
# gpg:                using RSA key 5D09FD0871C8F85B94CA8A0D281F0DB8D28D5469
# gpg:                issuer "mst@redhat.com"
# gpg: Good signature from "Michael S. Tsirkin <mst@kernel.org>" [full]
# gpg:                 aka "Michael S. Tsirkin <mst@redhat.com>" [full]
# Primary key fingerprint: 0270 606B 6F3C DF3D 0B17  0970 C350 3912 AFBE 8E67
#      Subkey fingerprint: 5D09 FD08 71C8 F85B 94CA  8A0D 281F 0DB8 D28D 5469

* remotes/mst/tags/for_upstream: (30 commits)
  Fixed assert in vhost_user_set_mem_table_postcopy
  vhost-user: only set slave channel for first vq
  acpi: cpuhp: document CPHP_GET_CPU_ID_CMD command
  libvhost-user: implement in-band notifications
  docs: vhost-user: add in-band kick/call messages
  libvhost-user: handle NOFD flag in call/kick/err better
  libvhost-user-glib: use g_main_context_get_thread_default()
  libvhost-user-glib: fix VugDev main fd cleanup
  libvhost-user: implement VHOST_USER_PROTOCOL_F_REPLY_ACK
  MAINTAINERS: add virtio-iommu related files
  hw/arm/virt: Add the virtio-iommu device tree mappings
  virtio-iommu-pci: Add virtio iommu pci support
  virtio-iommu: Support migration
  virtio-iommu: Implement fault reporting
  virtio-iommu: Implement translate
  virtio-iommu: Implement map/unmap
  virtio-iommu: Implement attach/detach command
  virtio-iommu: Decode the command payload
  virtio-iommu: Add skeleton
  virtio: gracefully handle invalid region caches
  ...

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

10 files changed, 1135 insertions, 15 deletions

diff --git a/hw/virtio/Kconfig b/hw/virtio/Kconfig
index f87def27a6..d29525b36f 100644
--- a/hw/virtio/Kconfig
+++ b/hw/virtio/Kconfig
@@ -9,6 +9,11 @@ config VIRTIO_RNG
     default y
     depends on VIRTIO
 
+config VIRTIO_IOMMU
+    bool
+    default y
+    depends on VIRTIO
+
 config VIRTIO_PCI
     bool
     default y if PCI_DEVICES
diff --git a/hw/virtio/Makefile.objs b/hw/virtio/Makefile.objs
index de0f5fc39b..4e4d39a0a4 100644
--- a/hw/virtio/Makefile.objs
+++ b/hw/virtio/Makefile.objs
@@ -16,6 +16,7 @@ obj-$(call land,$(CONFIG_VIRTIO_CRYPTO),$(CONFIG_VIRTIO_PCI)) += virtio-crypto-p
 obj-$(CONFIG_VIRTIO_PMEM) += virtio-pmem.o
 common-obj-$(call land,$(CONFIG_VIRTIO_PMEM),$(CONFIG_VIRTIO_PCI)) += virtio-pmem-pci.o
 obj-$(call land,$(CONFIG_VHOST_USER_FS),$(CONFIG_VIRTIO_PCI)) += vhost-user-fs-pci.o
+obj-$(CONFIG_VIRTIO_IOMMU) += virtio-iommu.o
 obj-$(CONFIG_VHOST_VSOCK) += vhost-vsock.o
 
 ifeq ($(CONFIG_VIRTIO_PCI),y)
@@ -28,6 +29,7 @@ obj-$(CONFIG_VIRTIO_INPUT_HOST) += virtio-input-host-pci.o
 obj-$(CONFIG_VIRTIO_INPUT) += virtio-input-pci.o
 obj-$(CONFIG_VIRTIO_RNG) += virtio-rng-pci.o
 obj-$(CONFIG_VIRTIO_BALLOON) += virtio-balloon-pci.o
+obj-$(CONFIG_VIRTIO_IOMMU) += virtio-iommu-pci.o
 obj-$(CONFIG_VIRTIO_9P) += virtio-9p-pci.o
 obj-$(CONFIG_VIRTIO_SCSI) += virtio-scsi-pci.o
 obj-$(CONFIG_VIRTIO_BLK) += virtio-blk-pci.o
diff --git a/hw/virtio/trace-events b/hw/virtio/trace-events
index e28ba48da6..e83500bee9 100644
--- a/hw/virtio/trace-events
+++ b/hw/virtio/trace-events
@@ -53,3 +53,23 @@ virtio_mmio_write_offset(uint64_t offset, uint64_t value) "virtio_mmio_write off
 virtio_mmio_guest_page(uint64_t size, int shift) "guest page size 0x%" PRIx64 " shift %d"
 virtio_mmio_queue_write(uint64_t value, int max_size) "mmio_queue write 0x%" PRIx64 " max %d"
 virtio_mmio_setting_irq(int level) "virtio_mmio setting IRQ %d"
+
+# hw/virtio/virtio-iommu.c
+virtio_iommu_device_reset(void) "reset!"
+virtio_iommu_get_features(uint64_t features) "device supports features=0x%"PRIx64
+virtio_iommu_device_status(uint8_t status) "driver status = %d"
+virtio_iommu_get_config(uint64_t page_size_mask, uint64_t start, uint64_t end, uint32_t domain_range, uint32_t probe_size) "page_size_mask=0x%"PRIx64" start=0x%"PRIx64" end=0x%"PRIx64" domain_range=%d probe_size=0x%x"
+virtio_iommu_set_config(uint64_t page_size_mask, uint64_t start, uint64_t end, uint32_t domain_range, uint32_t probe_size) "page_size_mask=0x%"PRIx64" start=0x%"PRIx64" end=0x%"PRIx64" domain_bits=%d probe_size=0x%x"
+virtio_iommu_attach(uint32_t domain_id, uint32_t ep_id) "domain=%d endpoint=%d"
+virtio_iommu_detach(uint32_t domain_id, uint32_t ep_id) "domain=%d endpoint=%d"
+virtio_iommu_map(uint32_t domain_id, uint64_t virt_start, uint64_t virt_end, uint64_t phys_start, uint32_t flags) "domain=%d virt_start=0x%"PRIx64" virt_end=0x%"PRIx64 " phys_start=0x%"PRIx64" flags=%d"
+virtio_iommu_unmap(uint32_t domain_id, uint64_t virt_start, uint64_t virt_end) "domain=%d virt_start=0x%"PRIx64" virt_end=0x%"PRIx64
+virtio_iommu_unmap_done(uint32_t domain_id, uint64_t virt_start, uint64_t virt_end) "domain=%d virt_start=0x%"PRIx64" virt_end=0x%"PRIx64
+virtio_iommu_translate(const char *name, uint32_t rid, uint64_t iova, int flag) "mr=%s rid=%d addr=0x%"PRIx64" flag=%d"
+virtio_iommu_init_iommu_mr(char *iommu_mr) "init %s"
+virtio_iommu_get_endpoint(uint32_t ep_id) "Alloc endpoint=%d"
+virtio_iommu_put_endpoint(uint32_t ep_id) "Free endpoint=%d"
+virtio_iommu_get_domain(uint32_t domain_id) "Alloc domain=%d"
+virtio_iommu_put_domain(uint32_t domain_id) "Free domain=%d"
+virtio_iommu_translate_out(uint64_t virt_addr, uint64_t phys_addr, uint32_t sid) "0x%"PRIx64" -> 0x%"PRIx64 " for sid=%d"
+virtio_iommu_report_fault(uint8_t reason, uint32_t flags, uint32_t endpoint, uint64_t addr) "FAULT reason=%d flags=%d endpoint=%d address =0x%"PRIx64
diff --git a/hw/virtio/vhost-user-fs.c b/hw/virtio/vhost-user-fs.c
index 33b17848c2..6136768875 100644
--- a/hw/virtio/vhost-user-fs.c
+++ b/hw/virtio/vhost-user-fs.c
@@ -209,11 +209,12 @@ static void vuf_device_realize(DeviceState *dev, Error **errp)
                 sizeof(struct virtio_fs_config));
 
     /* Hiprio queue */
-    virtio_add_queue(vdev, fs->conf.queue_size, vuf_handle_output);
+    fs->hiprio_vq = virtio_add_queue(vdev, fs->conf.queue_size, vuf_handle_output);
 
     /* Request queues */
+    fs->req_vqs = g_new(VirtQueue *, fs->conf.num_request_queues);
     for (i = 0; i < fs->conf.num_request_queues; i++) {
-        virtio_add_queue(vdev, fs->conf.queue_size, vuf_handle_output);
+        fs->req_vqs[i] = virtio_add_queue(vdev, fs->conf.queue_size, vuf_handle_output);
     }
 
     /* 1 high prio queue, plus the number configured */
@@ -230,6 +231,11 @@ static void vuf_device_realize(DeviceState *dev, Error **errp)
 
 err_virtio:
     vhost_user_cleanup(&fs->vhost_user);
+    virtio_delete_queue(fs->hiprio_vq);
+    for (i = 0; i < fs->conf.num_request_queues; i++) {
+        virtio_delete_queue(fs->req_vqs[i]);
+    }
+    g_free(fs->req_vqs);
     virtio_cleanup(vdev);
     g_free(fs->vhost_dev.vqs);
     return;
@@ -239,6 +245,7 @@ static void vuf_device_unrealize(DeviceState *dev, Error **errp)
 {
     VirtIODevice *vdev = VIRTIO_DEVICE(dev);
     VHostUserFS *fs = VHOST_USER_FS(dev);
+    int i;
 
     /* This will stop vhost backend if appropriate. */
     vuf_set_status(vdev, 0);
@@ -247,6 +254,11 @@ static void vuf_device_unrealize(DeviceState *dev, Error **errp)
 
     vhost_user_cleanup(&fs->vhost_user);
 
+    virtio_delete_queue(fs->hiprio_vq);
+    for (i = 0; i < fs->conf.num_request_queues; i++) {
+        virtio_delete_queue(fs->req_vqs[i]);
+    }
+    g_free(fs->req_vqs);
     virtio_cleanup(vdev);
     g_free(fs->vhost_dev.vqs);
     fs->vhost_dev.vqs = NULL;
diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
index 2e81f5514f..08e7e63790 100644
--- a/hw/virtio/vhost-user.c
+++ b/hw/virtio/vhost-user.c
@@ -443,6 +443,7 @@ static int vhost_user_set_mem_table_postcopy(struct vhost_dev *dev,
                                      &offset);
         fd = memory_region_get_fd(mr);
         if (fd > 0) {
+            assert(fd_num < VHOST_MEMORY_MAX_NREGIONS);
             trace_vhost_user_set_mem_table_withfd(fd_num, mr->name,
                                                   reg->memory_size,
                                                   reg->guest_phys_addr,
@@ -455,7 +456,6 @@ static int vhost_user_set_mem_table_postcopy(struct vhost_dev *dev,
             msg.payload.memory.regions[fd_num].guest_phys_addr =
                 reg->guest_phys_addr;
             msg.payload.memory.regions[fd_num].mmap_offset = offset;
-            assert(fd_num < VHOST_MEMORY_MAX_NREGIONS);
             fds[fd_num++] = fd;
         } else {
             u->region_rb_offset[i] = 0;
@@ -1458,9 +1458,11 @@ static int vhost_user_backend_init(struct vhost_dev *dev, void *opaque)
                    "VHOST_USER_PROTOCOL_F_LOG_SHMFD feature.");
     }
 
-    err = vhost_setup_slave_channel(dev);
-    if (err < 0) {
-        return err;
+    if (dev->vq_index == 0) {
+        err = vhost_setup_slave_channel(dev);
+        if (err < 0) {
+            return err;
+        }
     }
 
     u->postcopy_notifier.notify = vhost_user_postcopy_notifier;
diff --git a/hw/virtio/virtio-crypto.c b/hw/virtio/virtio-crypto.c
index 7351ab0a19..4c65114de5 100644
--- a/hw/virtio/virtio-crypto.c
+++ b/hw/virtio/virtio-crypto.c
@@ -831,12 +831,13 @@ static void virtio_crypto_device_unrealize(DeviceState *dev, Error **errp)
 
     max_queues = vcrypto->multiqueue ? vcrypto->max_queues : 1;
     for (i = 0; i < max_queues; i++) {
-        virtio_del_queue(vdev, i);
+        virtio_delete_queue(vcrypto->vqs[i].dataq);
         q = &vcrypto->vqs[i];
         qemu_bh_delete(q->dataq_bh);
     }
 
     g_free(vcrypto->vqs);
+    virtio_delete_queue(vcrypto->ctrl_vq);
 
     virtio_cleanup(vdev);
     cryptodev_backend_set_used(vcrypto->cryptodev, false);
diff --git a/hw/virtio/virtio-iommu-pci.c b/hw/virtio/virtio-iommu-pci.c
new file mode 100644
index 0000000000..3dfbf55b47
--- /dev/null
+++ b/hw/virtio/virtio-iommu-pci.c
@@ -0,0 +1,104 @@
+/*
+ * Virtio IOMMU PCI Bindings
+ *
+ * Copyright (c) 2019 Red Hat, Inc.
+ * Written by Eric Auger
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License version 2 or
+ *  (at your option) any later version.
+ */
+
+#include "qemu/osdep.h"
+
+#include "virtio-pci.h"
+#include "hw/virtio/virtio-iommu.h"
+#include "hw/qdev-properties.h"
+#include "qapi/error.h"
+#include "hw/boards.h"
+
+typedef struct VirtIOIOMMUPCI VirtIOIOMMUPCI;
+
+/*
+ * virtio-iommu-pci: This extends VirtioPCIProxy.
+ *
+ */
+#define VIRTIO_IOMMU_PCI(obj) \
+        OBJECT_CHECK(VirtIOIOMMUPCI, (obj), TYPE_VIRTIO_IOMMU_PCI)
+
+struct VirtIOIOMMUPCI {
+    VirtIOPCIProxy parent_obj;
+    VirtIOIOMMU vdev;
+};
+
+static Property virtio_iommu_pci_properties[] = {
+    DEFINE_PROP_UINT32("class", VirtIOPCIProxy, class_code, 0),
+    DEFINE_PROP_END_OF_LIST(),
+};
+
+static void virtio_iommu_pci_realize(VirtIOPCIProxy *vpci_dev, Error **errp)
+{
+    VirtIOIOMMUPCI *dev = VIRTIO_IOMMU_PCI(vpci_dev);
+    DeviceState *vdev = DEVICE(&dev->vdev);
+
+    if (!qdev_get_machine_hotplug_handler(DEVICE(vpci_dev))) {
+        MachineClass *mc = MACHINE_GET_CLASS(qdev_get_machine());
+
+        error_setg(errp,
+                   "%s machine fails to create iommu-map device tree bindings",
+                   mc->name);
+        error_append_hint(errp,
+                          "Check you machine implements a hotplug handler "
+                          "for the virtio-iommu-pci device\n");
+        error_append_hint(errp, "Check the guest is booted without FW or with "
+                          "-no-acpi\n");
+        return;
+    }
+    qdev_set_parent_bus(vdev, BUS(&vpci_dev->bus));
+    object_property_set_link(OBJECT(dev),
+                             OBJECT(pci_get_bus(&vpci_dev->pci_dev)),
+                             "primary-bus", errp);
+    object_property_set_bool(OBJECT(vdev), true, "realized", errp);
+}
+
+static void virtio_iommu_pci_class_init(ObjectClass *klass, void *data)
+{
+    DeviceClass *dc = DEVICE_CLASS(klass);
+    VirtioPCIClass *k = VIRTIO_PCI_CLASS(klass);
+    PCIDeviceClass *pcidev_k = PCI_DEVICE_CLASS(klass);
+    k->realize = virtio_iommu_pci_realize;
+    set_bit(DEVICE_CATEGORY_MISC, dc->categories);
+    device_class_set_props(dc, virtio_iommu_pci_properties);
+    pcidev_k->vendor_id = PCI_VENDOR_ID_REDHAT_QUMRANET;
+    pcidev_k->device_id = PCI_DEVICE_ID_VIRTIO_IOMMU;
+    pcidev_k->revision = VIRTIO_PCI_ABI_VERSION;
+    pcidev_k->class_id = PCI_CLASS_OTHERS;
+    dc->hotpluggable = false;
+}
+
+static void virtio_iommu_pci_instance_init(Object *obj)
+{
+    VirtIOIOMMUPCI *dev = VIRTIO_IOMMU_PCI(obj);
+
+    virtio_instance_init_common(obj, &dev->vdev, sizeof(dev->vdev),
+                                TYPE_VIRTIO_IOMMU);
+}
+
+static const VirtioPCIDeviceTypeInfo virtio_iommu_pci_info = {
+    .base_name             = TYPE_VIRTIO_IOMMU_PCI,
+    .generic_name          = "virtio-iommu-pci",
+    .transitional_name     = "virtio-iommu-pci-transitional",
+    .non_transitional_name = "virtio-iommu-pci-non-transitional",
+    .instance_size = sizeof(VirtIOIOMMUPCI),
+    .instance_init = virtio_iommu_pci_instance_init,
+    .class_init    = virtio_iommu_pci_class_init,
+};
+
+static void virtio_iommu_pci_register(void)
+{
+    virtio_pci_types_register(&virtio_iommu_pci_info);
+}
+
+type_init(virtio_iommu_pci_register)
+
+
diff --git a/hw/virtio/virtio-iommu.c b/hw/virtio/virtio-iommu.c
new file mode 100644
index 0000000000..4cee8083bc
--- /dev/null
+++ b/hw/virtio/virtio-iommu.c
@@ -0,0 +1,890 @@
+/*
+ * virtio-iommu device
+ *
+ * Copyright (c) 2020 Red Hat, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2 or later, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ *
+ */
+
+#include "qemu/osdep.h"
+#include "qemu/log.h"
+#include "qemu/iov.h"
+#include "qemu-common.h"
+#include "hw/qdev-properties.h"
+#include "hw/virtio/virtio.h"
+#include "sysemu/kvm.h"
+#include "qapi/error.h"
+#include "qemu/error-report.h"
+#include "trace.h"
+
+#include "standard-headers/linux/virtio_ids.h"
+
+#include "hw/virtio/virtio-bus.h"
+#include "hw/virtio/virtio-access.h"
+#include "hw/virtio/virtio-iommu.h"
+#include "hw/pci/pci_bus.h"
+#include "hw/pci/pci.h"
+
+/* Max size */
+#define VIOMMU_DEFAULT_QUEUE_SIZE 256
+
+typedef struct VirtIOIOMMUDomain {
+    uint32_t id;
+    GTree *mappings;
+    QLIST_HEAD(, VirtIOIOMMUEndpoint) endpoint_list;
+} VirtIOIOMMUDomain;
+
+typedef struct VirtIOIOMMUEndpoint {
+    uint32_t id;
+    VirtIOIOMMUDomain *domain;
+    QLIST_ENTRY(VirtIOIOMMUEndpoint) next;
+} VirtIOIOMMUEndpoint;
+
+typedef struct VirtIOIOMMUInterval {
+    uint64_t low;
+    uint64_t high;
+} VirtIOIOMMUInterval;
+
+typedef struct VirtIOIOMMUMapping {
+    uint64_t phys_addr;
+    uint32_t flags;
+} VirtIOIOMMUMapping;
+
+static inline uint16_t virtio_iommu_get_bdf(IOMMUDevice *dev)
+{
+    return PCI_BUILD_BDF(pci_bus_num(dev->bus), dev->devfn);
+}
+
+/**
+ * The bus number is used for lookup when SID based operations occur.
+ * In that case we lazily populate the IOMMUPciBus array from the bus hash
+ * table. At the time the IOMMUPciBus is created (iommu_find_add_as), the bus
+ * numbers may not be always initialized yet.
+ */
+static IOMMUPciBus *iommu_find_iommu_pcibus(VirtIOIOMMU *s, uint8_t bus_num)
+{
+    IOMMUPciBus *iommu_pci_bus = s->iommu_pcibus_by_bus_num[bus_num];
+
+    if (!iommu_pci_bus) {
+        GHashTableIter iter;
+
+        g_hash_table_iter_init(&iter, s->as_by_busptr);
+        while (g_hash_table_iter_next(&iter, NULL, (void **)&iommu_pci_bus)) {
+            if (pci_bus_num(iommu_pci_bus->bus) == bus_num) {
+                s->iommu_pcibus_by_bus_num[bus_num] = iommu_pci_bus;
+                return iommu_pci_bus;
+            }
+        }
+        return NULL;
+    }
+    return iommu_pci_bus;
+}
+
+static IOMMUMemoryRegion *virtio_iommu_mr(VirtIOIOMMU *s, uint32_t sid)
+{
+    uint8_t bus_n, devfn;
+    IOMMUPciBus *iommu_pci_bus;
+    IOMMUDevice *dev;
+
+    bus_n = PCI_BUS_NUM(sid);
+    iommu_pci_bus = iommu_find_iommu_pcibus(s, bus_n);
+    if (iommu_pci_bus) {
+        devfn = sid & PCI_DEVFN_MAX;
+        dev = iommu_pci_bus->pbdev[devfn];
+        if (dev) {
+            return &dev->iommu_mr;
+        }
+    }
+    return NULL;
+}
+
+static gint interval_cmp(gconstpointer a, gconstpointer b, gpointer user_data)
+{
+    VirtIOIOMMUInterval *inta = (VirtIOIOMMUInterval *)a;
+    VirtIOIOMMUInterval *intb = (VirtIOIOMMUInterval *)b;
+
+    if (inta->high < intb->low) {
+        return -1;
+    } else if (intb->high < inta->low) {
+        return 1;
+    } else {
+        return 0;
+    }
+}
+
+static void virtio_iommu_detach_endpoint_from_domain(VirtIOIOMMUEndpoint *ep)
+{
+    if (!ep->domain) {
+        return;
+    }
+    QLIST_REMOVE(ep, next);
+    ep->domain = NULL;
+}
+
+static VirtIOIOMMUEndpoint *virtio_iommu_get_endpoint(VirtIOIOMMU *s,
+                                                      uint32_t ep_id)
+{
+    VirtIOIOMMUEndpoint *ep;
+
+    ep = g_tree_lookup(s->endpoints, GUINT_TO_POINTER(ep_id));
+    if (ep) {
+        return ep;
+    }
+    if (!virtio_iommu_mr(s, ep_id)) {
+        return NULL;
+    }
+    ep = g_malloc0(sizeof(*ep));
+    ep->id = ep_id;
+    trace_virtio_iommu_get_endpoint(ep_id);
+    g_tree_insert(s->endpoints, GUINT_TO_POINTER(ep_id), ep);
+    return ep;
+}
+
+static void virtio_iommu_put_endpoint(gpointer data)
+{
+    VirtIOIOMMUEndpoint *ep = (VirtIOIOMMUEndpoint *)data;
+
+    if (ep->domain) {
+        virtio_iommu_detach_endpoint_from_domain(ep);
+    }
+
+    trace_virtio_iommu_put_endpoint(ep->id);
+    g_free(ep);
+}
+
+static VirtIOIOMMUDomain *virtio_iommu_get_domain(VirtIOIOMMU *s,
+                                                  uint32_t domain_id)
+{
+    VirtIOIOMMUDomain *domain;
+
+    domain = g_tree_lookup(s->domains, GUINT_TO_POINTER(domain_id));
+    if (domain) {
+        return domain;
+    }
+    domain = g_malloc0(sizeof(*domain));
+    domain->id = domain_id;
+    domain->mappings = g_tree_new_full((GCompareDataFunc)interval_cmp,
+                                   NULL, (GDestroyNotify)g_free,
+                                   (GDestroyNotify)g_free);
+    g_tree_insert(s->domains, GUINT_TO_POINTER(domain_id), domain);
+    QLIST_INIT(&domain->endpoint_list);
+    trace_virtio_iommu_get_domain(domain_id);
+    return domain;
+}
+
+static void virtio_iommu_put_domain(gpointer data)
+{
+    VirtIOIOMMUDomain *domain = (VirtIOIOMMUDomain *)data;
+    VirtIOIOMMUEndpoint *iter, *tmp;
+
+    QLIST_FOREACH_SAFE(iter, &domain->endpoint_list, next, tmp) {
+        virtio_iommu_detach_endpoint_from_domain(iter);
+    }
+    g_tree_destroy(domain->mappings);
+    trace_virtio_iommu_put_domain(domain->id);
+    g_free(domain);
+}
+
+static AddressSpace *virtio_iommu_find_add_as(PCIBus *bus, void *opaque,
+                                              int devfn)
+{
+    VirtIOIOMMU *s = opaque;
+    IOMMUPciBus *sbus = g_hash_table_lookup(s->as_by_busptr, bus);
+    static uint32_t mr_index;
+    IOMMUDevice *sdev;
+
+    if (!sbus) {
+        sbus = g_malloc0(sizeof(IOMMUPciBus) +
+                         sizeof(IOMMUDevice *) * PCI_DEVFN_MAX);
+        sbus->bus = bus;
+        g_hash_table_insert(s->as_by_busptr, bus, sbus);
+    }
+
+    sdev = sbus->pbdev[devfn];
+    if (!sdev) {
+        char *name = g_strdup_printf("%s-%d-%d",
+                                     TYPE_VIRTIO_IOMMU_MEMORY_REGION,
+                                     mr_index++, devfn);
+        sdev = sbus->pbdev[devfn] = g_malloc0(sizeof(IOMMUDevice));
+
+        sdev->viommu = s;
+        sdev->bus = bus;
+        sdev->devfn = devfn;
+
+        trace_virtio_iommu_init_iommu_mr(name);
+
+        memory_region_init_iommu(&sdev->iommu_mr, sizeof(sdev->iommu_mr),
+                                 TYPE_VIRTIO_IOMMU_MEMORY_REGION,
+                                 OBJECT(s), name,
+                                 UINT64_MAX);
+        address_space_init(&sdev->as,
+                           MEMORY_REGION(&sdev->iommu_mr), TYPE_VIRTIO_IOMMU);
+        g_free(name);
+    }
+    return &sdev->as;
+}
+
+static int virtio_iommu_attach(VirtIOIOMMU *s,
+                               struct virtio_iommu_req_attach *req)
+{
+    uint32_t domain_id = le32_to_cpu(req->domain);
+    uint32_t ep_id = le32_to_cpu(req->endpoint);
+    VirtIOIOMMUDomain *domain;
+    VirtIOIOMMUEndpoint *ep;
+
+    trace_virtio_iommu_attach(domain_id, ep_id);
+
+    ep = virtio_iommu_get_endpoint(s, ep_id);
+    if (!ep) {
+        return VIRTIO_IOMMU_S_NOENT;
+    }
+
+    if (ep->domain) {
+        VirtIOIOMMUDomain *previous_domain = ep->domain;
+        /*
+         * the device is already attached to a domain,
+         * detach it first
+         */
+        virtio_iommu_detach_endpoint_from_domain(ep);
+        if (QLIST_EMPTY(&previous_domain->endpoint_list)) {
+            g_tree_remove(s->domains, GUINT_TO_POINTER(previous_domain->id));
+        }
+    }
+
+    domain = virtio_iommu_get_domain(s, domain_id);
+    QLIST_INSERT_HEAD(&domain->endpoint_list, ep, next);
+
+    ep->domain = domain;
+
+    return VIRTIO_IOMMU_S_OK;
+}
+
+static int virtio_iommu_detach(VirtIOIOMMU *s,
+                               struct virtio_iommu_req_detach *req)
+{
+    uint32_t domain_id = le32_to_cpu(req->domain);
+    uint32_t ep_id = le32_to_cpu(req->endpoint);
+    VirtIOIOMMUDomain *domain;
+    VirtIOIOMMUEndpoint *ep;
+
+    trace_virtio_iommu_detach(domain_id, ep_id);
+
+    ep = g_tree_lookup(s->endpoints, GUINT_TO_POINTER(ep_id));
+    if (!ep) {
+        return VIRTIO_IOMMU_S_NOENT;
+    }
+
+    domain = ep->domain;
+
+    if (!domain || domain->id != domain_id) {
+        return VIRTIO_IOMMU_S_INVAL;
+    }
+
+    virtio_iommu_detach_endpoint_from_domain(ep);
+
+    if (QLIST_EMPTY(&domain->endpoint_list)) {
+        g_tree_remove(s->domains, GUINT_TO_POINTER(domain->id));
+    }
+    return VIRTIO_IOMMU_S_OK;
+}
+
+static int virtio_iommu_map(VirtIOIOMMU *s,
+                            struct virtio_iommu_req_map *req)
+{
+    uint32_t domain_id = le32_to_cpu(req->domain);
+    uint64_t phys_start = le64_to_cpu(req->phys_start);
+    uint64_t virt_start = le64_to_cpu(req->virt_start);
+    uint64_t virt_end = le64_to_cpu(req->virt_end);
+    uint32_t flags = le32_to_cpu(req->flags);
+    VirtIOIOMMUDomain *domain;
+    VirtIOIOMMUInterval *interval;
+    VirtIOIOMMUMapping *mapping;
+
+    if (flags & ~VIRTIO_IOMMU_MAP_F_MASK) {
+        return VIRTIO_IOMMU_S_INVAL;
+    }
+
+    domain = g_tree_lookup(s->domains, GUINT_TO_POINTER(domain_id));
+    if (!domain) {
+        return VIRTIO_IOMMU_S_NOENT;
+    }
+
+    interval = g_malloc0(sizeof(*interval));
+
+    interval->low = virt_start;
+    interval->high = virt_end;
+
+    mapping = g_tree_lookup(domain->mappings, (gpointer)interval);
+    if (mapping) {
+        g_free(interval);
+        return VIRTIO_IOMMU_S_INVAL;
+    }
+
+    trace_virtio_iommu_map(domain_id, virt_start, virt_end, phys_start, flags);
+
+    mapping = g_malloc0(sizeof(*mapping));
+    mapping->phys_addr = phys_start;
+    mapping->flags = flags;
+
+    g_tree_insert(domain->mappings, interval, mapping);
+
+    return VIRTIO_IOMMU_S_OK;
+}
+
+static int virtio_iommu_unmap(VirtIOIOMMU *s,
+                              struct virtio_iommu_req_unmap *req)
+{
+    uint32_t domain_id = le32_to_cpu(req->domain);
+    uint64_t virt_start = le64_to_cpu(req->virt_start);
+    uint64_t virt_end = le64_to_cpu(req->virt_end);
+    VirtIOIOMMUMapping *iter_val;
+    VirtIOIOMMUInterval interval, *iter_key;
+    VirtIOIOMMUDomain *domain;
+    int ret = VIRTIO_IOMMU_S_OK;
+
+    trace_virtio_iommu_unmap(domain_id, virt_start, virt_end);
+
+    domain = g_tree_lookup(s->domains, GUINT_TO_POINTER(domain_id));
+    if (!domain) {
+        return VIRTIO_IOMMU_S_NOENT;
+    }
+    interval.low = virt_start;
+    interval.high = virt_end;
+
+    while (g_tree_lookup_extended(domain->mappings, &interval,
+                                  (void **)&iter_key, (void**)&iter_val)) {
+        uint64_t current_low = iter_key->low;
+        uint64_t current_high = iter_key->high;
+
+        if (interval.low <= current_low && interval.high >= current_high) {
+            g_tree_remove(domain->mappings, iter_key);
+            trace_virtio_iommu_unmap_done(domain_id, current_low, current_high);
+        } else {
+            ret = VIRTIO_IOMMU_S_RANGE;
+            break;
+        }
+    }
+    return ret;
+}
+
+static int virtio_iommu_iov_to_req(struct iovec *iov,
+                                   unsigned int iov_cnt,
+                                   void *req, size_t req_sz)
+{
+    size_t sz, payload_sz = req_sz - sizeof(struct virtio_iommu_req_tail);
+
+    sz = iov_to_buf(iov, iov_cnt, 0, req, payload_sz);
+    if (unlikely(sz != payload_sz)) {
+        return VIRTIO_IOMMU_S_INVAL;
+    }
+    return 0;
+}
+
+#define virtio_iommu_handle_req(__req)                                  \
+static int virtio_iommu_handle_ ## __req(VirtIOIOMMU *s,                \
+                                         struct iovec *iov,             \
+                                         unsigned int iov_cnt)          \
+{                                                                       \
+    struct virtio_iommu_req_ ## __req req;                              \
+    int ret = virtio_iommu_iov_to_req(iov, iov_cnt, &req, sizeof(req)); \
+                                                                        \
+    return ret ? ret : virtio_iommu_ ## __req(s, &req);                 \
+}
+
+virtio_iommu_handle_req(attach)
+virtio_iommu_handle_req(detach)
+virtio_iommu_handle_req(map)
+virtio_iommu_handle_req(unmap)
+
+static void virtio_iommu_handle_command(VirtIODevice *vdev, VirtQueue *vq)
+{
+    VirtIOIOMMU *s = VIRTIO_IOMMU(vdev);
+    struct virtio_iommu_req_head head;
+    struct virtio_iommu_req_tail tail = {};
+    VirtQueueElement *elem;
+    unsigned int iov_cnt;
+    struct iovec *iov;
+    size_t sz;
+
+    for (;;) {
+        elem = virtqueue_pop(vq, sizeof(VirtQueueElement));
+        if (!elem) {
+            return;
+        }
+
+        if (iov_size(elem->in_sg, elem->in_num) < sizeof(tail) ||
+            iov_size(elem->out_sg, elem->out_num) < sizeof(head)) {
+            virtio_error(vdev, "virtio-iommu bad head/tail size");
+            virtqueue_detach_element(vq, elem, 0);
+            g_free(elem);
+            break;
+        }
+
+        iov_cnt = elem->out_num;
+        iov = elem->out_sg;
+        sz = iov_to_buf(iov, iov_cnt, 0, &head, sizeof(head));
+        if (unlikely(sz != sizeof(head))) {
+            tail.status = VIRTIO_IOMMU_S_DEVERR;
+            goto out;
+        }
+        qemu_mutex_lock(&s->mutex);
+        switch (head.type) {
+        case VIRTIO_IOMMU_T_ATTACH:
+            tail.status = virtio_iommu_handle_attach(s, iov, iov_cnt);
+            break;
+        case VIRTIO_IOMMU_T_DETACH:
+            tail.status = virtio_iommu_handle_detach(s, iov, iov_cnt);
+            break;
+        case VIRTIO_IOMMU_T_MAP:
+            tail.status = virtio_iommu_handle_map(s, iov, iov_cnt);
+            break;
+        case VIRTIO_IOMMU_T_UNMAP:
+            tail.status = virtio_iommu_handle_unmap(s, iov, iov_cnt);
+            break;
+        default:
+            tail.status = VIRTIO_IOMMU_S_UNSUPP;
+        }
+        qemu_mutex_unlock(&s->mutex);
+
+out:
+        sz = iov_from_buf(elem->in_sg, elem->in_num, 0,
+                          &tail, sizeof(tail));
+        assert(sz == sizeof(tail));
+
+        virtqueue_push(vq, elem, sizeof(tail));
+        virtio_notify(vdev, vq);
+        g_free(elem);
+    }
+}
+
+static void virtio_iommu_report_fault(VirtIOIOMMU *viommu, uint8_t reason,
+                                      int flags, uint32_t endpoint,
+                                      uint64_t address)
+{
+    VirtIODevice *vdev = &viommu->parent_obj;
+    VirtQueue *vq = viommu->event_vq;
+    struct virtio_iommu_fault fault;
+    VirtQueueElement *elem;
+    size_t sz;
+
+    memset(&fault, 0, sizeof(fault));
+    fault.reason = reason;
+    fault.flags = cpu_to_le32(flags);
+    fault.endpoint = cpu_to_le32(endpoint);
+    fault.address = cpu_to_le64(address);
+
+    elem = virtqueue_pop(vq, sizeof(VirtQueueElement));
+
+    if (!elem) {
+        error_report_once(
+            "no buffer available in event queue to report event");
+        return;
+    }
+
+    if (iov_size(elem->in_sg, elem->in_num) < sizeof(fault)) {
+        virtio_error(vdev, "error buffer of wrong size");
+        virtqueue_detach_element(vq, elem, 0);
+        g_free(elem);
+        return;
+    }
+
+    sz = iov_from_buf(elem->in_sg, elem->in_num, 0,
+                      &fault, sizeof(fault));
+    assert(sz == sizeof(fault));
+
+    trace_virtio_iommu_report_fault(reason, flags, endpoint, address);
+    virtqueue_push(vq, elem, sz);
+    virtio_notify(vdev, vq);
+    g_free(elem);
+
+}
+
+static IOMMUTLBEntry virtio_iommu_translate(IOMMUMemoryRegion *mr, hwaddr addr,
+                                            IOMMUAccessFlags flag,
+                                            int iommu_idx)
+{
+    IOMMUDevice *sdev = container_of(mr, IOMMUDevice, iommu_mr);
+    VirtIOIOMMUInterval interval, *mapping_key;
+    VirtIOIOMMUMapping *mapping_value;
+    VirtIOIOMMU *s = sdev->viommu;
+    bool read_fault, write_fault;
+    VirtIOIOMMUEndpoint *ep;
+    uint32_t sid, flags;
+    bool bypass_allowed;
+    bool found;
+
+    interval.low = addr;
+    interval.high = addr + 1;
+
+    IOMMUTLBEntry entry = {
+        .target_as = &address_space_memory,
+        .iova = addr,
+        .translated_addr = addr,
+        .addr_mask = (1 << ctz32(s->config.page_size_mask)) - 1,
+        .perm = IOMMU_NONE,
+    };
+
+    bypass_allowed = virtio_vdev_has_feature(&s->parent_obj,
+                                             VIRTIO_IOMMU_F_BYPASS);
+
+    sid = virtio_iommu_get_bdf(sdev);
+
+    trace_virtio_iommu_translate(mr->parent_obj.name, sid, addr, flag);
+    qemu_mutex_lock(&s->mutex);
+
+    ep = g_tree_lookup(s->endpoints, GUINT_TO_POINTER(sid));
+    if (!ep) {
+        if (!bypass_allowed) {
+            error_report_once("%s sid=%d is not known!!", __func__, sid);
+            virtio_iommu_report_fault(s, VIRTIO_IOMMU_FAULT_R_UNKNOWN,
+                                      VIRTIO_IOMMU_FAULT_F_ADDRESS,
+                                      sid, addr);
+        } else {
+            entry.perm = flag;
+        }
+        goto unlock;
+    }
+
+    if (!ep->domain) {
+        if (!bypass_allowed) {
+            error_report_once("%s %02x:%02x.%01x not attached to any domain",
+                              __func__, PCI_BUS_NUM(sid),
+                              PCI_SLOT(sid), PCI_FUNC(sid));
+            virtio_iommu_report_fault(s, VIRTIO_IOMMU_FAULT_R_DOMAIN,
+                                      VIRTIO_IOMMU_FAULT_F_ADDRESS,
+                                      sid, addr);
+        } else {
+            entry.perm = flag;
+        }
+        goto unlock;
+    }
+
+    found = g_tree_lookup_extended(ep->domain->mappings, (gpointer)(&interval),
+                                   (void **)&mapping_key,
+                                   (void **)&mapping_value);
+    if (!found) {
+        error_report_once("%s no mapping for 0x%"PRIx64" for sid=%d",
+                          __func__, addr, sid);
+        virtio_iommu_report_fault(s, VIRTIO_IOMMU_FAULT_R_MAPPING,
+                                  VIRTIO_IOMMU_FAULT_F_ADDRESS,
+                                  sid, addr);
+        goto unlock;
+    }
+
+    read_fault = (flag & IOMMU_RO) &&
+                    !(mapping_value->flags & VIRTIO_IOMMU_MAP_F_READ);
+    write_fault = (flag & IOMMU_WO) &&
+                    !(mapping_value->flags & VIRTIO_IOMMU_MAP_F_WRITE);
+
+    flags = read_fault ? VIRTIO_IOMMU_FAULT_F_READ : 0;
+    flags |= write_fault ? VIRTIO_IOMMU_FAULT_F_WRITE : 0;
+    if (flags) {
+        error_report_once("%s permission error on 0x%"PRIx64"(%d): allowed=%d",
+                          __func__, addr, flag, mapping_value->flags);
+        flags |= VIRTIO_IOMMU_FAULT_F_ADDRESS;
+        virtio_iommu_report_fault(s, VIRTIO_IOMMU_FAULT_R_MAPPING,
+                                  flags | VIRTIO_IOMMU_FAULT_F_ADDRESS,
+                                  sid, addr);
+        goto unlock;
+    }
+    entry.translated_addr = addr - mapping_key->low + mapping_value->phys_addr;
+    entry.perm = flag;
+    trace_virtio_iommu_translate_out(addr, entry.translated_addr, sid);
+
+unlock:
+    qemu_mutex_unlock(&s->mutex);
+    return entry;
+}
+
+static void virtio_iommu_get_config(VirtIODevice *vdev, uint8_t *config_data)
+{
+    VirtIOIOMMU *dev = VIRTIO_IOMMU(vdev);
+    struct virtio_iommu_config *config = &dev->config;
+
+    trace_virtio_iommu_get_config(config->page_size_mask,
+                                  config->input_range.start,
+                                  config->input_range.end,
+                                  config->domain_range.end,
+                                  config->probe_size);
+    memcpy(config_data, &dev->config, sizeof(struct virtio_iommu_config));
+}
+
+static void virtio_iommu_set_config(VirtIODevice *vdev,
+                                      const uint8_t *config_data)
+{
+    struct virtio_iommu_config config;
+
+    memcpy(&config, config_data, sizeof(struct virtio_iommu_config));
+    trace_virtio_iommu_set_config(config.page_size_mask,
+                                  config.input_range.start,
+                                  config.input_range.end,
+                                  config.domain_range.end,
+                                  config.probe_size);
+}
+
+static uint64_t virtio_iommu_get_features(VirtIODevice *vdev, uint64_t f,
+                                          Error **errp)
+{
+    VirtIOIOMMU *dev = VIRTIO_IOMMU(vdev);
+
+    f |= dev->features;
+    trace_virtio_iommu_get_features(f);
+    return f;
+}
+
+static gint int_cmp(gconstpointer a, gconstpointer b, gpointer user_data)
+{
+    guint ua = GPOINTER_TO_UINT(a);
+    guint ub = GPOINTER_TO_UINT(b);
+    return (ua > ub) - (ua < ub);
+}
+
+static void virtio_iommu_device_realize(DeviceState *dev, Error **errp)
+{
+    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+    VirtIOIOMMU *s = VIRTIO_IOMMU(dev);
+
+    virtio_init(vdev, "virtio-iommu", VIRTIO_ID_IOMMU,
+                sizeof(struct virtio_iommu_config));
+
+    memset(s->iommu_pcibus_by_bus_num, 0, sizeof(s->iommu_pcibus_by_bus_num));
+
+    s->req_vq = virtio_add_queue(vdev, VIOMMU_DEFAULT_QUEUE_SIZE,
+                             virtio_iommu_handle_command);
+    s->event_vq = virtio_add_queue(vdev, VIOMMU_DEFAULT_QUEUE_SIZE, NULL);
+
+    s->config.page_size_mask = TARGET_PAGE_MASK;
+    s->config.input_range.end = -1UL;
+    s->config.domain_range.end = 32;
+
+    virtio_add_feature(&s->features, VIRTIO_RING_F_EVENT_IDX);
+    virtio_add_feature(&s->features, VIRTIO_RING_F_INDIRECT_DESC);
+    virtio_add_feature(&s->features, VIRTIO_F_VERSION_1);
+    virtio_add_feature(&s->features, VIRTIO_IOMMU_F_INPUT_RANGE);
+    virtio_add_feature(&s->features, VIRTIO_IOMMU_F_DOMAIN_RANGE);
+    virtio_add_feature(&s->features, VIRTIO_IOMMU_F_MAP_UNMAP);
+    virtio_add_feature(&s->features, VIRTIO_IOMMU_F_BYPASS);
+    virtio_add_feature(&s->features, VIRTIO_IOMMU_F_MMIO);
+
+    qemu_mutex_init(&s->mutex);
+
+    s->as_by_busptr = g_hash_table_new_full(NULL, NULL, NULL, g_free);
+
+    if (s->primary_bus) {
+        pci_setup_iommu(s->primary_bus, virtio_iommu_find_add_as, s);
+    } else {
+        error_setg(errp, "VIRTIO-IOMMU is not attached to any PCI bus!");
+    }
+}
+
+static void virtio_iommu_device_unrealize(DeviceState *dev, Error **errp)
+{
+    VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+    VirtIOIOMMU *s = VIRTIO_IOMMU(dev);
+
+    g_tree_destroy(s->domains);
+    g_tree_destroy(s->endpoints);
+
+    virtio_cleanup(vdev);
+}
+
+static void virtio_iommu_device_reset(VirtIODevice *vdev)
+{
+    VirtIOIOMMU *s = VIRTIO_IOMMU(vdev);
+
+    trace_virtio_iommu_device_reset();
+
+    if (s->domains) {
+        g_tree_destroy(s->domains);
+    }
+    if (s->endpoints) {
+        g_tree_destroy(s->endpoints);
+    }
+    s->domains = g_tree_new_full((GCompareDataFunc)int_cmp,
+                                 NULL, NULL, virtio_iommu_put_domain);
+    s->endpoints = g_tree_new_full((GCompareDataFunc)int_cmp,
+                                   NULL, NULL, virtio_iommu_put_endpoint);
+}
+
+static void virtio_iommu_set_status(VirtIODevice *vdev, uint8_t status)
+{
+    trace_virtio_iommu_device_status(status);
+}
+
+static void virtio_iommu_instance_init(Object *obj)
+{
+}
+
+#define VMSTATE_INTERVAL                               \
+{                                                      \
+    .name = "interval",                                \
+    .version_id = 1,                                   \
+    .minimum_version_id = 1,                           \
+    .fields = (VMStateField[]) {                       \
+        VMSTATE_UINT64(low, VirtIOIOMMUInterval),      \
+        VMSTATE_UINT64(high, VirtIOIOMMUInterval),     \
+        VMSTATE_END_OF_LIST()                          \
+    }                                                  \
+}
+
+#define VMSTATE_MAPPING                               \
+{                                                     \
+    .name = "mapping",                                \
+    .version_id = 1,                                  \
+    .minimum_version_id = 1,                          \
+    .fields = (VMStateField[]) {                      \
+        VMSTATE_UINT64(phys_addr, VirtIOIOMMUMapping),\
+        VMSTATE_UINT32(flags, VirtIOIOMMUMapping),    \
+        VMSTATE_END_OF_LIST()                         \
+    },                                                \
+}
+
+static const VMStateDescription vmstate_interval_mapping[2] = {
+    VMSTATE_MAPPING,   /* value */
+    VMSTATE_INTERVAL   /* key   */
+};
+
+static int domain_preload(void *opaque)
+{
+    VirtIOIOMMUDomain *domain = opaque;
+
+    domain->mappings = g_tree_new_full((GCompareDataFunc)interval_cmp,
+                                       NULL, g_free, g_free);
+    return 0;
+}
+
+static const VMStateDescription vmstate_endpoint = {
+    .name = "endpoint",
+    .version_id = 1,
+    .minimum_version_id = 1,
+    .fields = (VMStateField[]) {
+        VMSTATE_UINT32(id, VirtIOIOMMUEndpoint),
+        VMSTATE_END_OF_LIST()
+    }
+};
+
+static const VMStateDescription vmstate_domain = {
+    .name = "domain",
+    .version_id = 1,
+    .minimum_version_id = 1,
+    .pre_load = domain_preload,
+    .fields = (VMStateField[]) {
+        VMSTATE_UINT32(id, VirtIOIOMMUDomain),
+        VMSTATE_GTREE_V(mappings, VirtIOIOMMUDomain, 1,
+                        vmstate_interval_mapping,
+                        VirtIOIOMMUInterval, VirtIOIOMMUMapping),
+        VMSTATE_QLIST_V(endpoint_list, VirtIOIOMMUDomain, 1,
+                        vmstate_endpoint, VirtIOIOMMUEndpoint, next),
+        VMSTATE_END_OF_LIST()
+    }
+};
+
+static gboolean reconstruct_endpoints(gpointer key, gpointer value,
+                                      gpointer data)
+{
+    VirtIOIOMMU *s = (VirtIOIOMMU *)data;
+    VirtIOIOMMUDomain *d = (VirtIOIOMMUDomain *)value;
+    VirtIOIOMMUEndpoint *iter;
+
+    QLIST_FOREACH(iter, &d->endpoint_list, next) {
+        iter->domain = d;
+        g_tree_insert(s->endpoints, GUINT_TO_POINTER(iter->id), iter);
+    }
+    return false; /* continue the domain traversal */
+}
+
+static int iommu_post_load(void *opaque, int version_id)
+{
+    VirtIOIOMMU *s = opaque;
+
+    g_tree_foreach(s->domains, reconstruct_endpoints, s);
+    return 0;
+}
+
+static const VMStateDescription vmstate_virtio_iommu_device = {
+    .name = "virtio-iommu-device",
+    .minimum_version_id = 1,
+    .version_id = 1,
+    .post_load = iommu_post_load,
+    .fields = (VMStateField[]) {
+        VMSTATE_GTREE_DIRECT_KEY_V(domains, VirtIOIOMMU, 1,
+                                   &vmstate_domain, VirtIOIOMMUDomain),
+        VMSTATE_END_OF_LIST()
+    },
+};
+
+static const VMStateDescription vmstate_virtio_iommu = {
+    .name = "virtio-iommu",
+    .minimum_version_id = 1,
+    .priority = MIG_PRI_IOMMU,
+    .version_id = 1,
+    .fields = (VMStateField[]) {
+        VMSTATE_VIRTIO_DEVICE,
+        VMSTATE_END_OF_LIST()
+    },
+};
+
+static Property virtio_iommu_properties[] = {
+    DEFINE_PROP_LINK("primary-bus", VirtIOIOMMU, primary_bus, "PCI", PCIBus *),
+    DEFINE_PROP_END_OF_LIST(),
+};
+
+static void virtio_iommu_class_init(ObjectClass *klass, void *data)
+{
+    DeviceClass *dc = DEVICE_CLASS(klass);
+    VirtioDeviceClass *vdc = VIRTIO_DEVICE_CLASS(klass);
+
+    device_class_set_props(dc, virtio_iommu_properties);
+    dc->vmsd = &vmstate_virtio_iommu;
+
+    set_bit(DEVICE_CATEGORY_MISC, dc->categories);
+    vdc->realize = virtio_iommu_device_realize;
+    vdc->unrealize = virtio_iommu_device_unrealize;
+    vdc->reset = virtio_iommu_device_reset;
+    vdc->get_config = virtio_iommu_get_config;
+    vdc->set_config = virtio_iommu_set_config;
+    vdc->get_features = virtio_iommu_get_features;
+    vdc->set_status = virtio_iommu_set_status;
+    vdc->vmsd = &vmstate_virtio_iommu_device;
+}
+
+static void virtio_iommu_memory_region_class_init(ObjectClass *klass,
+                                                  void *data)
+{
+    IOMMUMemoryRegionClass *imrc = IOMMU_MEMORY_REGION_CLASS(klass);
+
+    imrc->translate = virtio_iommu_translate;
+}
+
+static const TypeInfo virtio_iommu_info = {
+    .name = TYPE_VIRTIO_IOMMU,
+    .parent = TYPE_VIRTIO_DEVICE,
+    .instance_size = sizeof(VirtIOIOMMU),
+    .instance_init = virtio_iommu_instance_init,
+    .class_init = virtio_iommu_class_init,
+};
+
+static const TypeInfo virtio_iommu_memory_region_info = {
+    .parent = TYPE_IOMMU_MEMORY_REGION,
+    .name = TYPE_VIRTIO_IOMMU_MEMORY_REGION,
+    .class_init = virtio_iommu_memory_region_class_init,
+};
+
+static void virtio_register_types(void)
+{
+    type_register_static(&virtio_iommu_info);
+    type_register_static(&virtio_iommu_memory_region_info);
+}
+
+type_init(virtio_register_types)
diff --git a/hw/virtio/virtio-pmem.c b/hw/virtio/virtio-pmem.c
index 97287e923b..43399522f5 100644
--- a/hw/virtio/virtio-pmem.c
+++ b/hw/virtio/virtio-pmem.c
@@ -130,6 +130,7 @@ static void virtio_pmem_unrealize(DeviceState *dev, Error **errp)
     VirtIOPMEM *pmem = VIRTIO_PMEM(dev);
 
     host_memory_backend_set_mapped(pmem->memdev, false);
+    virtio_delete_queue(pmem->rq_vq);
     virtio_cleanup(vdev);
 }
 
diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
index 9d06dbe3ef..b2d415e5dd 100644
--- a/hw/virtio/virtio.c
+++ b/hw/virtio/virtio.c
@@ -282,15 +282,19 @@ static void vring_packed_flags_write(VirtIODevice *vdev,
 /* Called within rcu_read_lock().  */
 static VRingMemoryRegionCaches *vring_get_region_caches(struct VirtQueue *vq)
 {
-    VRingMemoryRegionCaches *caches = atomic_rcu_read(&vq->vring.caches);
-    assert(caches != NULL);
-    return caches;
+    return atomic_rcu_read(&vq->vring.caches);
 }
+
 /* Called within rcu_read_lock().  */
 static inline uint16_t vring_avail_flags(VirtQueue *vq)
 {
     VRingMemoryRegionCaches *caches = vring_get_region_caches(vq);
     hwaddr pa = offsetof(VRingAvail, flags);
+
+    if (!caches) {
+        return 0;
+    }
+
     return virtio_lduw_phys_cached(vq->vdev, &caches->avail, pa);
 }
 
@@ -299,6 +303,11 @@ static inline uint16_t vring_avail_idx(VirtQueue *vq)
 {
     VRingMemoryRegionCaches *caches = vring_get_region_caches(vq);
     hwaddr pa = offsetof(VRingAvail, idx);
+
+    if (!caches) {
+        return 0;
+    }
+
     vq->shadow_avail_idx = virtio_lduw_phys_cached(vq->vdev, &caches->avail, pa);
     return vq->shadow_avail_idx;
 }
@@ -308,6 +317,11 @@ static inline uint16_t vring_avail_ring(VirtQueue *vq, int i)
 {
     VRingMemoryRegionCaches *caches = vring_get_region_caches(vq);
     hwaddr pa = offsetof(VRingAvail, ring[i]);
+
+    if (!caches) {
+        return 0;
+    }
+
     return virtio_lduw_phys_cached(vq->vdev, &caches->avail, pa);
 }
 
@@ -323,6 +337,11 @@ static inline void vring_used_write(VirtQueue *vq, VRingUsedElem *uelem,
 {
     VRingMemoryRegionCaches *caches = vring_get_region_caches(vq);
     hwaddr pa = offsetof(VRingUsed, ring[i]);
+
+    if (!caches) {
+        return;
+    }
+
     virtio_tswap32s(vq->vdev, &uelem->id);
     virtio_tswap32s(vq->vdev, &uelem->len);
     address_space_write_cached(&caches->used, pa, uelem, sizeof(VRingUsedElem));
@@ -334,6 +353,11 @@ static uint16_t vring_used_idx(VirtQueue *vq)
 {
     VRingMemoryRegionCaches *caches = vring_get_region_caches(vq);
     hwaddr pa = offsetof(VRingUsed, idx);
+
+    if (!caches) {
+        return 0;
+    }
+
     return virtio_lduw_phys_cached(vq->vdev, &caches->used, pa);
 }
 
@@ -342,8 +366,12 @@ static inline void vring_used_idx_set(VirtQueue *vq, uint16_t val)
 {
     VRingMemoryRegionCaches *caches = vring_get_region_caches(vq);
     hwaddr pa = offsetof(VRingUsed, idx);
-    virtio_stw_phys_cached(vq->vdev, &caches->used, pa, val);
-    address_space_cache_invalidate(&caches->used, pa, sizeof(val));
+
+    if (caches) {
+        virtio_stw_phys_cached(vq->vdev, &caches->used, pa, val);
+        address_space_cache_invalidate(&caches->used, pa, sizeof(val));
+    }
+
     vq->used_idx = val;
 }
 
@@ -353,8 +381,13 @@ static inline void vring_used_flags_set_bit(VirtQueue *vq, int mask)
     VRingMemoryRegionCaches *caches = vring_get_region_caches(vq);
     VirtIODevice *vdev = vq->vdev;
     hwaddr pa = offsetof(VRingUsed, flags);
-    uint16_t flags = virtio_lduw_phys_cached(vq->vdev, &caches->used, pa);
+    uint16_t flags;
 
+    if (!caches) {
+        return;
+    }
+
+    flags = virtio_lduw_phys_cached(vq->vdev, &caches->used, pa);
     virtio_stw_phys_cached(vdev, &caches->used, pa, flags | mask);
     address_space_cache_invalidate(&caches->used, pa, sizeof(flags));
 }
@@ -365,8 +398,13 @@ static inline void vring_used_flags_unset_bit(VirtQueue *vq, int mask)
     VRingMemoryRegionCaches *caches = vring_get_region_caches(vq);
     VirtIODevice *vdev = vq->vdev;
     hwaddr pa = offsetof(VRingUsed, flags);
-    uint16_t flags = virtio_lduw_phys_cached(vq->vdev, &caches->used, pa);
+    uint16_t flags;
 
+    if (!caches) {
+        return;
+    }
+
+    flags = virtio_lduw_phys_cached(vq->vdev, &caches->used, pa);
     virtio_stw_phys_cached(vdev, &caches->used, pa, flags & ~mask);
     address_space_cache_invalidate(&caches->used, pa, sizeof(flags));
 }
@@ -381,6 +419,10 @@ static inline void vring_set_avail_event(VirtQueue *vq, uint16_t val)
     }
 
     caches = vring_get_region_caches(vq);
+    if (!caches) {
+        return;
+    }
+
     pa = offsetof(VRingUsed, ring[vq->vring.num]);
     virtio_stw_phys_cached(vq->vdev, &caches->used, pa, val);
     address_space_cache_invalidate(&caches->used, pa, sizeof(val));
@@ -410,7 +452,11 @@ static void virtio_queue_packed_set_notification(VirtQueue *vq, int enable)
     VRingMemoryRegionCaches *caches;
 
     RCU_READ_LOCK_GUARD();
-    caches  = vring_get_region_caches(vq);
+    caches = vring_get_region_caches(vq);
+    if (!caches) {
+        return;
+    }
+
     vring_packed_event_read(vq->vdev, &caches->used, &e);
 
     if (!enable) {
@@ -597,6 +643,10 @@ static int virtio_queue_packed_empty_rcu(VirtQueue *vq)
     }
 
     cache = vring_get_region_caches(vq);
+    if (!cache) {
+        return 1;
+    }
+
     vring_packed_desc_read_flags(vq->vdev, &desc.flags, &cache->desc,
                                  vq->last_avail_idx);
 
@@ -777,6 +827,10 @@ static void virtqueue_packed_fill_desc(VirtQueue *vq,
     }
 
     caches = vring_get_region_caches(vq);
+    if (!caches) {
+        return;
+    }
+
     vring_packed_desc_write(vq->vdev, &desc, &caches->desc, head, strict_order);
 }
 
@@ -949,6 +1003,10 @@ static void virtqueue_split_get_avail_bytes(VirtQueue *vq,
 
     max = vq->vring.num;
     caches = vring_get_region_caches(vq);
+    if (!caches) {
+        goto err;
+    }
+
     while ((rc = virtqueue_num_heads(vq, idx)) > 0) {
         MemoryRegionCache *desc_cache = &caches->desc;
         unsigned int num_bufs;
@@ -1089,6 +1147,9 @@ static void virtqueue_packed_get_avail_bytes(VirtQueue *vq,
 
     max = vq->vring.num;
     caches = vring_get_region_caches(vq);
+    if (!caches) {
+        goto err;
+    }
 
     for (;;) {
         unsigned int num_bufs = total_bufs;
@@ -1194,6 +1255,10 @@ void virtqueue_get_avail_bytes(VirtQueue *vq, unsigned int *in_bytes,
     }
 
     caches = vring_get_region_caches(vq);
+    if (!caches) {
+        goto err;
+    }
+
     desc_size = virtio_vdev_has_feature(vq->vdev, VIRTIO_F_RING_PACKED) ?
                                 sizeof(VRingPackedDesc) : sizeof(VRingDesc);
     if (caches->desc.len < vq->vring.num * desc_size) {
@@ -1388,6 +1453,11 @@ static void *virtqueue_split_pop(VirtQueue *vq, size_t sz)
     i = head;
 
     caches = vring_get_region_caches(vq);
+    if (!caches) {
+        virtio_error(vdev, "Region caches not initialized");
+        goto done;
+    }
+
     if (caches->desc.len < max * sizeof(VRingDesc)) {
         virtio_error(vdev, "Cannot map descriptor ring");
         goto done;
@@ -1510,6 +1580,11 @@ static void *virtqueue_packed_pop(VirtQueue *vq, size_t sz)
     i = vq->last_avail_idx;
 
     caches = vring_get_region_caches(vq);
+    if (!caches) {
+        virtio_error(vdev, "Region caches not initialized");
+        goto done;
+    }
+
     if (caches->desc.len < max * sizeof(VRingDesc)) {
         virtio_error(vdev, "Cannot map descriptor ring");
         goto done;
@@ -1629,6 +1704,10 @@ static unsigned int virtqueue_packed_drop_all(VirtQueue *vq)
     VRingPackedDesc desc;
 
     caches = vring_get_region_caches(vq);
+    if (!caches) {
+        return 0;
+    }
+
     desc_cache = &caches->desc;
 
     virtio_queue_set_notification(vq, 0);
@@ -2413,6 +2492,10 @@ static bool virtio_packed_should_notify(VirtIODevice *vdev, VirtQueue *vq)
     VRingMemoryRegionCaches *caches;
 
     caches = vring_get_region_caches(vq);
+    if (!caches) {
+        return false;
+    }
+
     vring_packed_event_read(vdev, &caches->avail, &e);
 
     old = vq->signalled_used;