qemu: Implement support for associating iommufd to hostdev

Implement a new iommufd attribute under hostdevs' PCI
subsystem driver that can be used to specify associated
iommufd object when launching a qemu VM.

Signed-off-by: Ján Tomko <jtomko@redhat.com>
Signed-off-by: Nathan Chen <nathanc@nvidia.com>
Reviewed-by: Pavel Hrdina <phrdina@redhat.com>

diff --git a/docs/formatdomain.rst b/docs/formatdomain.rst
index ad74d919500bb568e798e4c704e6cd95e611f640..0fa9d2e08b3b24483922533a56e867d0c30ccc05 100644 (file)
--- a/docs/formatdomain.rst
+++ b/docs/formatdomain.rst
@@ -4924,6 +4924,12 @@ or:
    found is "problematic" in some way, the generic vfio-pci driver
    similarly be forced.
 
+   :since:`Since 12.1.0 (QEMU and KVM only)`, the ``iommufd`` element
+   can be used to enable IOMMUFD backend for VFIO device. This
+   provides an interface to propagate DMA mappings to kernel for
+   assigned devices. Libvirt will open the /dev/iommu and VFIO device
+   cdev and pass associated file descriptors to QEMU.
+
    (Note: :since:`Since 1.0.5`, the ``name`` attribute has been
    described to be used to select the type of PCI device assignment
    ("vfio", "kvm", or "xen"), but those values have been mostly

diff --git a/src/conf/device_conf.c b/src/conf/device_conf.c
index c278b8165212e62ab99ce4774ced4e9a7879b3f9..d68232a4f40876fc96cdc5458f6ae13570898db4 100644 (file)
--- a/src/conf/device_conf.c
+++ b/src/conf/device_conf.c
@@ -67,6 +67,11 @@ virDeviceHostdevPCIDriverInfoParseXML(xmlNodePtr node,
         return -1;
     }
 
+    if (virXMLPropTristateBool(node, "iommufd",
+                               VIR_XML_PROP_NONE,
+                               &driver->iommufd) < 0)
+        return -1;
+
     driver->model = virXMLPropString(node, "model");
     return 0;
 }
@@ -93,6 +98,12 @@ virDeviceHostdevPCIDriverInfoFormat(virBuffer *buf,
 
     virBufferEscapeString(&driverAttrBuf, " model='%s'", driver->model);
 
+    if (driver->iommufd == VIR_TRISTATE_BOOL_YES) {
+        virBufferAddLit(&driverAttrBuf, " iommufd='yes'");
+    } else if (driver->iommufd == VIR_TRISTATE_BOOL_NO) {
+        virBufferAddLit(&driverAttrBuf, " iommufd='no'");
+    }
+
     virXMLFormatElement(buf, "driver", &driverAttrBuf, NULL);
     return 0;
 }

diff --git a/src/conf/device_conf.h b/src/conf/device_conf.h
index e570f518242f313a7bafd24251cbe5420d647f6f..116b95914389c6fa1ceb8adb4d4050f910b90726 100644 (file)
--- a/src/conf/device_conf.h
+++ b/src/conf/device_conf.h
@@ -47,6 +47,7 @@ VIR_ENUM_DECL(virDeviceHostdevPCIDriverName);
 struct _virDeviceHostdevPCIDriverInfo {
     virDeviceHostdevPCIDriverName name;
     char *model;
+    virTristateBool iommufd;
 };
 
 typedef enum {

diff --git a/src/conf/schemas/basictypes.rng b/src/conf/schemas/basictypes.rng
index 5689170fad155327b65d6a8b3d8e578f2976b555..381e0ac24ffdf1c860046bacda857332fcaf24d0 100644 (file)
--- a/src/conf/schemas/basictypes.rng
+++ b/src/conf/schemas/basictypes.rng
@@ -673,6 +673,11 @@
           <ref name="genericName"/>
         </attribute>
       </optional>
+      <optional>
+        <attribute name="iommufd">
+          <ref name="virYesNo"/>
+        </attribute>
+      </optional>
       <empty/>
     </element>
   </define>

diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c
index fde2e30924ac86939bdedea47af239fcc78a423c..9066149df5793a2fe7f97f3502b0e93dd31e58d0 100644 (file)
--- a/src/qemu/qemu_command.c
+++ b/src/qemu/qemu_command.c
@@ -4761,6 +4761,7 @@ qemuBuildPCIHostdevDevProps(const virDomainDef *def,
     g_autofree char *host = virPCIDeviceAddressAsString(&pcisrc->addr);
     const char *failover_pair_id = NULL;
     const char *driver = NULL;
+    const char *iommufdId = NULL;
     /* 'ramfb' property must be omitted unless it's to be enabled */
     bool ramfb = pcisrc->ramfb == VIR_TRISTATE_SWITCH_ON;
 
@@ -4794,6 +4795,9 @@ qemuBuildPCIHostdevDevProps(const virDomainDef *def,
         teaming->persistent)
         failover_pair_id = teaming->persistent;
 
+    if (pcisrc->driver.iommufd == VIR_TRISTATE_BOOL_YES)
+        iommufdId = "iommufd0";
+
     if (virJSONValueObjectAdd(&props,
                               "s:driver", driver,
                               "s:host", host,
@@ -4802,6 +4806,7 @@ qemuBuildPCIHostdevDevProps(const virDomainDef *def,
                               "S:failover_pair_id", failover_pair_id,
                               "S:display", qemuOnOffAuto(pcisrc->display),
                               "B:ramfb", ramfb,
+                              "S:iommufd", iommufdId,
                               NULL) < 0)
         return NULL;
 
@@ -5321,6 +5326,44 @@ qemuBuildHostdevCommandLine(virCommand *cmd,
 }
 
 
+static int
+qemuBuildIOMMUFDCommandLine(virCommand *cmd,
+                            const virDomainDef *def)
+{
+    size_t i;
+
+    for (i = 0; i < def->nhostdevs; i++) {
+        virDomainHostdevDef *hostdev = def->hostdevs[i];
+        virDomainHostdevSubsys *subsys = &hostdev->source.subsys;
+        g_autoptr(virJSONValue) props = NULL;
+
+        if (hostdev->mode != VIR_DOMAIN_HOSTDEV_MODE_SUBSYS)
+            continue;
+
+        if (subsys->type != VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI)
+            continue;
+
+        if (hostdev->info->type == VIR_DOMAIN_DEVICE_ADDRESS_TYPE_UNASSIGNED)
+            continue;
+
+        if (subsys->u.pci.driver.iommufd != VIR_TRISTATE_BOOL_YES)
+            continue;
+
+        if (qemuMonitorCreateObjectProps(&props, "iommufd",
+                                         "iommufd0",
+                                         NULL) < 0)
+            return -1;
+
+        if (qemuBuildObjectCommandlineFromJSON(cmd, props) < 0)
+            return -1;
+
+        break;
+    }
+
+    return 0;
+}
+
+
 static int
 qemuBuildMonitorCommandLine(virCommand *cmd,
                             qemuDomainObjPrivate *priv)
@@ -10942,6 +10985,9 @@ qemuBuildCommandLine(virDomainObj *vm,
     if (qemuBuildRedirdevCommandLine(cmd, def, qemuCaps) < 0)
         return NULL;
 
+    if (qemuBuildIOMMUFDCommandLine(cmd, def) < 0)
+        return NULL;
+
     if (qemuBuildHostdevCommandLine(cmd, def, qemuCaps) < 0)
         return NULL;