Skip to contentSkip to conversation

vfio-user server in QEMU

PATCH Superseded Jagannathan Raman 4 participants 46 messages Apr 19, 2022 – May 5, 2022 Arm/Amlogic Meson SoC support

v8 of 12 Superseded Latest is v12 →

lore
Share & tools
lore
#0 lore raw

v8 cover vfio-user server in QEMU

Hi,

This is v8 of the server side changes to enable vfio-user in QEMU.

Thank you very much for reviewing the last revision of this series!

We've made the following changes in this revision:

[PATCH v8 06/17] vfio-user: build library
- updated libvfio-user to the latest

[PATCH v8 07/17] vfio-user: define vfio-user-server object
- changed auto_shutdown to a per-instance property than a per-class property

[PATCH v8 12/17] vfio-user: IOMMU support for remote device
- lock mutex while looking up hash table
- removed global hash table - added RemoteIommu object to house this variable
- added unplug handler to remove per-device IOMMU entry when a PCIDevice is unplugged

[PATCH v8 14/17] vfio-user: handle PCI BAR accesses
- refactored vfu_object_bar_rw()
- vfu_object_bar_rw() handles short sections returned by memory_region_find()

[PATCH v8 15/17] vfio-user: handle device interrupts
- removed callbacks for msi_notify() and msix_notify()
- added callbacks for msi_send_message() and msi(x)_get_message() operations

Thank you!

Jagannathan Raman (17):
tests/avocado: Specify target VM argument to helper routines
qdev: unplug blocker for devices
remote/machine: add HotplugHandler for remote machine remote/machine: add vfio-user property
configure: require cmake 3.19 or newer
vfio-user: build library
vfio-user: define vfio-user-server object
vfio-user: instantiate vfio-user context
vfio-user: find and init PCI device
vfio-user: run vfio-user context
vfio-user: handle PCI config space accesses
vfio-user: IOMMU support for remote device
vfio-user: handle DMA mappings
vfio-user: handle PCI BAR accesses
vfio-user: handle device interrupts
vfio-user: handle reset of remote device
vfio-user: avocado tests for vfio-user

 configure                                  |  36 +-
 meson.build                                |  44 +-
 qapi/misc.json                             |  23 +
 qapi/qom.json                              |  20 +-
 include/exec/memory.h                      |   3 +
 include/hw/pci/pci.h                       |  13 +
 include/hw/qdev-core.h                     |  29 +
 include/hw/remote/iommu.h                  |  40 +
 include/hw/remote/machine.h                |   4 +
 include/hw/remote/vfio-user-obj.h          |   6 +
 hw/core/qdev.c                             |  24 +
 hw/pci/msi.c                               |  16 +-
 hw/pci/msix.c                              |  10 +-
 hw/pci/pci.c                               |  13 +
 hw/remote/iommu.c                          | 114 +++
 hw/remote/machine.c                        |  88 +-
 hw/remote/vfio-user-obj.c                  | 891 +++++++++++++++++++++
 softmmu/physmem.c                          |   4 +-
 softmmu/qdev-monitor.c                     |   4 +
 stubs/vfio-user-obj.c                      |   6 +
 tests/qtest/fuzz/generic_fuzz.c            |   9 +-
 .gitlab-ci.d/buildtest.yml                 |   2 +
 .gitmodules                                |   3 +
 Kconfig.host                               |   4 +
 MAINTAINERS                                |   6 +
 hw/remote/Kconfig                          |   4 +
 hw/remote/meson.build                      |   4 +
 hw/remote/trace-events                     |  11 +
 meson_options.txt                          |   3 +
 stubs/meson.build                          |   1 +
 subprojects/libvfio-user                   |   1 +
 tests/avocado/avocado_qemu/__init__.py     |  14 +-
 tests/avocado/vfio-user.py                 | 164 ++++
 tests/docker/dockerfiles/centos8.docker    |   2 +
 tests/docker/dockerfiles/ubuntu2004.docker |   2 +
 35 files changed, 1594 insertions(+), 24 deletions(-)
 create mode 100644 include/hw/remote/iommu.h
 create mode 100644 include/hw/remote/vfio-user-obj.h
 create mode 100644 hw/remote/iommu.c
 create mode 100644 hw/remote/vfio-user-obj.c
 create mode 100644 stubs/vfio-user-obj.c
 create mode 160000 subprojects/libvfio-user
 create mode 100644 tests/avocado/vfio-user.py
Signature · 1 line
2.20.1
#1 lore raw

PATCH v8 1/17 tests/avocado: Specify target VM argument to helper routines

Specify target VM for exec_command and
exec_command_and_wait_for_pattern routines

---
 tests/avocado/avocado_qemu/__init__.py | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)
tests/avocado/avocado_qemu/__init__.py +10 -4
index ac85e36a4d..18a34a798c 100644
@@ -198,7 +198,7 @@ def wait_for_console_pattern(test, success_message, failure_message=None,
     """
     _console_interaction(test, success_message, failure_message, None, vm=vm)
 
-def exec_command(test, command):
+def exec_command(test, command, vm=None):
     """
     Send a command to a console (appending CRLF characters), while logging
     the content.
@@ -207,11 +207,14 @@ def exec_command(test, command):
     :type test: :class:`avocado_qemu.QemuSystemTest`
     :param command: the command to send
     :type command: str
+    :param vm: target vm
+    :type vm: :class:`qemu.machine.QEMUMachine`
     """
-    _console_interaction(test, None, None, command + '\r')
+    _console_interaction(test, None, None, command + '\r', vm=vm)
 
 def exec_command_and_wait_for_pattern(test, command,
-                                      success_message, failure_message=None):
+                                      success_message, failure_message=None,
+                                      vm=None):
     """
     Send a command to a console (appending CRLF characters), then wait
     for success_message to appear on the console, while logging the.
@@ -223,8 +226,11 @@ def exec_command_and_wait_for_pattern(test, command,
     :param command: the command to send
     :param success_message: if this message appears, test succeeds
     :param failure_message: if this message appears, test fails
+    :param vm: target vm
+    :type vm: :class:`qemu.machine.QEMUMachine`
     """
-    _console_interaction(test, success_message, failure_message, command + '\r')
+    _console_interaction(test, success_message, failure_message, command + '\r',
+                         vm=vm)
 
 class QemuBaseTest(avocado.Test):
     def _get_unique_tag_val(self, tag_name):
Signature · 1 line
2.20.1
#2 lore raw

PATCH v8 2/17 qdev: unplug blocker for devices

In reply to Jagannathan Raman

Add blocker to prevent hot-unplug of devices

---
 include/hw/qdev-core.h | 29 +++++++++++++++++++++++++++++
 hw/core/qdev.c         | 24 ++++++++++++++++++++++++
 softmmu/qdev-monitor.c |  4 ++++
 3 files changed, 57 insertions(+)
include/hw/qdev-core.h +29 -0
index 92c3d65208..1b9fa25e5c 100644
@@ -193,6 +193,7 @@ struct DeviceState {
     int instance_id_alias;
     int alias_required_for_version;
     ResettableState reset;
+    GSList *unplug_blockers;
 };
 
 struct DeviceListener {
@@ -419,6 +420,34 @@ void qdev_simple_device_unplug_cb(HotplugHandler *hotplug_dev,
 void qdev_machine_creation_done(void);
 bool qdev_machine_modified(void);
 
+/*
+ * qdev_add_unplug_blocker: Adds an unplug blocker to a device
+ *
+ * @dev: Device to be blocked from unplug
+ * @reason: Reason for blocking
+ */
+void qdev_add_unplug_blocker(DeviceState *dev, Error *reason);
+
+/*
+ * qdev_del_unplug_blocker: Removes an unplug blocker from a device
+ *
+ * @dev: Device to be unblocked
+ * @reason: Pointer to the Error used with qdev_add_unplug_blocker.
+ *          Used as a handle to lookup the blocker for deletion.
+ */
+void qdev_del_unplug_blocker(DeviceState *dev, Error *reason);
+
+/*
+ * qdev_unplug_blocked: Confirms if a device is blocked from unplug
+ *
+ * @dev: Device to be tested
+ * @reason: Returns one of the reasons why the device is blocked,
+ *          if any
+ *
+ * Returns: true if device is blocked from unplug, false otherwise
+ */
+bool qdev_unplug_blocked(DeviceState *dev, Error **errp);
+
 /**
  * GpioPolarity: Polarity of a GPIO line
  *
hw/core/qdev.c +24 -0
index 84f3019440..0806d8fcaa 100644
@@ -468,6 +468,28 @@ char *qdev_get_dev_path(DeviceState *dev)
     return NULL;
 }
 
+void qdev_add_unplug_blocker(DeviceState *dev, Error *reason)
+{
+    dev->unplug_blockers = g_slist_prepend(dev->unplug_blockers, reason);
+}
+
+void qdev_del_unplug_blocker(DeviceState *dev, Error *reason)
+{
+    dev->unplug_blockers = g_slist_remove(dev->unplug_blockers, reason);
+}
+
+bool qdev_unplug_blocked(DeviceState *dev, Error **errp)
+{
+    ERRP_GUARD();
+
+    if (dev->unplug_blockers) {
+        error_propagate(errp, error_copy(dev->unplug_blockers->data));
+        return true;
+    }
+
+    return false;
+}
+
 static bool device_get_realized(Object *obj, Error **errp)
 {
     DeviceState *dev = DEVICE(obj);
@@ -704,6 +726,8 @@ static void device_finalize(Object *obj)
 
     DeviceState *dev = DEVICE(obj);
 
+    g_assert(!dev->unplug_blockers);
+
     QLIST_FOREACH_SAFE(ngl, &dev->gpios, node, next) {
         QLIST_REMOVE(ngl, node);
         qemu_free_irqs(ngl->in, ngl->num_in);
softmmu/qdev-monitor.c +4 -0
index 12fe60c467..9cfd59d17c 100644
@@ -898,6 +898,10 @@ void qdev_unplug(DeviceState *dev, Error **errp)
     HotplugHandlerClass *hdc;
     Error *local_err = NULL;
 
+    if (qdev_unplug_blocked(dev, errp)) {
+        return;
+    }
+
     if (dev->parent_bus && !qbus_is_hotpluggable(dev->parent_bus)) {
         error_setg(errp, QERR_BUS_NO_HOTPLUG, dev->parent_bus->name);
         return;
Signature · 1 line
2.20.1
#3 lore raw

Re: [PATCH v8 02/17] qdev: unplug blocker for devices

Quoted 1 line · Jagannathan Raman

Add blocker to prevent hot-unplug of devices

Why do you need this? I'm not doubting you do, I just want to read your reasons here :)

Quoted 76 lines
---
 include/hw/qdev-core.h | 29 +++++++++++++++++++++++++++++
 hw/core/qdev.c         | 24 ++++++++++++++++++++++++
 softmmu/qdev-monitor.c |  4 ++++
 3 files changed, 57 insertions(+)
include/hw/qdev-core.h +29 -0
index 92c3d65208..1b9fa25e5c 100644
@@ -193,6 +193,7 @@ struct DeviceState {
     int instance_id_alias;
     int alias_required_for_version;
     ResettableState reset;
+    GSList *unplug_blockers;
 };
 
 struct DeviceListener {
@@ -419,6 +420,34 @@ void qdev_simple_device_unplug_cb(HotplugHandler *hotplug_dev,
 void qdev_machine_creation_done(void);
 bool qdev_machine_modified(void);
 
+/*
+ * qdev_add_unplug_blocker: Adds an unplug blocker to a device
+ *
+ * @dev: Device to be blocked from unplug
+ * @reason: Reason for blocking
+ */
+void qdev_add_unplug_blocker(DeviceState *dev, Error *reason);
+
+/*
+ * qdev_del_unplug_blocker: Removes an unplug blocker from a device
+ *
+ * @dev: Device to be unblocked
+ * @reason: Pointer to the Error used with qdev_add_unplug_blocker.
+ *          Used as a handle to lookup the blocker for deletion.
+ */
+void qdev_del_unplug_blocker(DeviceState *dev, Error *reason);
+
+/*
+ * qdev_unplug_blocked: Confirms if a device is blocked from unplug
+ *
+ * @dev: Device to be tested
+ * @reason: Returns one of the reasons why the device is blocked,
+ *          if any
+ *
+ * Returns: true if device is blocked from unplug, false otherwise
+ */
+bool qdev_unplug_blocked(DeviceState *dev, Error **errp);
+
 /**
  * GpioPolarity: Polarity of a GPIO line
  *
hw/core/qdev.c +11 -0
index 84f3019440..0806d8fcaa 100644
@@ -468,6 +468,28 @@ char *qdev_get_dev_path(DeviceState *dev)
     return NULL;
 }
 
+void qdev_add_unplug_blocker(DeviceState *dev, Error *reason)
+{
+    dev->unplug_blockers = g_slist_prepend(dev->unplug_blockers, reason);
+}
+
+void qdev_del_unplug_blocker(DeviceState *dev, Error *reason)
+{
+    dev->unplug_blockers = g_slist_remove(dev->unplug_blockers, reason);
+}
+
+bool qdev_unplug_blocked(DeviceState *dev, Error **errp)

hw/core/qdev.c char *qdev_get_dev_path(DeviceState *dev) · line 491

Quoted 10 lines
hw/core/qdev.c +10 -0
+{
+    ERRP_GUARD();
+
+    if (dev->unplug_blockers) {
+        error_propagate(errp, error_copy(dev->unplug_blockers->data));
+        return true;
+    }
+
+    return false;
+}

This cites the most recently added blocker as reason. Your function comment covers it: "Returns one of the reasons". Okay.

Quoted 28 lines
hw/core/qdev.c +3 -0
+
 static bool device_get_realized(Object *obj, Error **errp)
 {
     DeviceState *dev = DEVICE(obj);
@@ -704,6 +726,8 @@ static void device_finalize(Object *obj)
 
     DeviceState *dev = DEVICE(obj);
 
+    g_assert(!dev->unplug_blockers);
+
     QLIST_FOREACH_SAFE(ngl, &dev->gpios, node, next) {
         QLIST_REMOVE(ngl, node);
         qemu_free_irqs(ngl->in, ngl->num_in);
softmmu/qdev-monitor.c +4 -0
index 12fe60c467..9cfd59d17c 100644
@@ -898,6 +898,10 @@ void qdev_unplug(DeviceState *dev, Error **errp)
     HotplugHandlerClass *hdc;
     Error *local_err = NULL;
 
+    if (qdev_unplug_blocked(dev, errp)) {
+        return;
+    }
+
     if (dev->parent_bus && !qbus_is_hotpluggable(dev->parent_bus)) {
         error_setg(errp, QERR_BUS_NO_HOTPLUG, dev->parent_bus->name);
         return;
author
#4 lore raw

Re: [PATCH v8 02/17] qdev: unplug blocker for devices

Quoted 10 lines · Markus Armbruster
Quoted 1 line · Jagannathan Raman

Add blocker to prevent hot-unplug of devices

Why do you need this? I'm not doubting you do, I just want to read your reasons here :)

Hi Markus, :)

The x-vfio-user-server depends on an attached PCIDevice. As long as x-vfio-user-server is used, we don’t want the PCIDevice to be unplugged. This blocker prevents an user from removing PCIDevice while the vfio-user server is in use.

Quoted 5 lines
Quoted 5 lines

I recall receiving a “Reviewed-by” from Stefan previously.

I’m very sorry I didn’t add that here. I’ll go over all the patches once again to confirm that the “Reviewed-by” status reflects accurately.

Quoted 77 lines
Quoted 77 lines

---

include/hw/qdev-core.h | 29 +++++++++++++++++++++++++++++
hw/core/qdev.c         | 24 ++++++++++++++++++++++++
softmmu/qdev-monitor.c |  4 ++++
3 files changed, 57 insertions(+)
include/hw/qdev-core.h +29 -0
index 92c3d65208..1b9fa25e5c 100644
@@ -193,6 +193,7 @@ struct DeviceState {
    int instance_id_alias;
    int alias_required_for_version;
    ResettableState reset;
+    GSList *unplug_blockers;
 };
 
 struct DeviceListener {
@@ -419,6 +420,34 @@ void qdev_simple_device_unplug_cb(HotplugHandler *hotplug_dev,
 void qdev_machine_creation_done(void);
 bool qdev_machine_modified(void);
 
+/*
+ * qdev_add_unplug_blocker: Adds an unplug blocker to a device
+ *
+ * @dev: Device to be blocked from unplug
+ * @reason: Reason for blocking
+ */
+void qdev_add_unplug_blocker(DeviceState *dev, Error *reason);
+
+/*
+ * qdev_del_unplug_blocker: Removes an unplug blocker from a device
+ *
+ * @dev: Device to be unblocked
+ * @reason: Pointer to the Error used with qdev_add_unplug_blocker.
+ *          Used as a handle to lookup the blocker for deletion.
+ */
+void qdev_del_unplug_blocker(DeviceState *dev, Error *reason);
+
+/*
+ * qdev_unplug_blocked: Confirms if a device is blocked from unplug
+ *
+ * @dev: Device to be tested
+ * @reason: Returns one of the reasons why the device is blocked,
+ *          if any
+ *
+ * Returns: true if device is blocked from unplug, false otherwise
+ */
+bool qdev_unplug_blocked(DeviceState *dev, Error **errp);
+
 /**
 * GpioPolarity: Polarity of a GPIO line
 *
hw/core/qdev.c +16 -0
index 84f3019440..0806d8fcaa 100644
@@ -468,6 +468,28 @@ char *qdev_get_dev_path(DeviceState *dev)
    return NULL;
 }
 
+void qdev_add_unplug_blocker(DeviceState *dev, Error *reason)
+{
+    dev->unplug_blockers = g_slist_prepend(dev->unplug_blockers, reason);
+}
+
+void qdev_del_unplug_blocker(DeviceState *dev, Error *reason)
+{
+    dev->unplug_blockers = g_slist_remove(dev->unplug_blockers, reason);
+}
+
+bool qdev_unplug_blocked(DeviceState *dev, Error **errp)
+{
+    ERRP_GUARD();
+
+    if (dev->unplug_blockers) {
+        error_propagate(errp, error_copy(dev->unplug_blockers->data));
Quoted 8 lines

hw/core/qdev.c char *qdev_get_dev_path(DeviceState *dev) · line 491

Quoted 5 lines
hw/core/qdev.c +5 -0
+        return true;
+    }
+
+    return false;
+}

This cites the most recently added blocker as reason. Your function comment covers it: "Returns one of the reasons". Okay.

I could change the comment to say that it returns the recently added reason.

Thank you!
--
Jag

Quoted 30 lines
Quoted 29 lines
hw/core/qdev.c +3 -0
+
 static bool device_get_realized(Object *obj, Error **errp)
 {
    DeviceState *dev = DEVICE(obj);
@@ -704,6 +726,8 @@ static void device_finalize(Object *obj)
 
    DeviceState *dev = DEVICE(obj);
 
+    g_assert(!dev->unplug_blockers);
+
    QLIST_FOREACH_SAFE(ngl, &dev->gpios, node, next) {
        QLIST_REMOVE(ngl, node);
        qemu_free_irqs(ngl->in, ngl->num_in);
softmmu/qdev-monitor.c +4 -0
index 12fe60c467..9cfd59d17c 100644
@@ -898,6 +898,10 @@ void qdev_unplug(DeviceState *dev, Error **errp)
    HotplugHandlerClass *hdc;
    Error *local_err = NULL;
 
+    if (qdev_unplug_blocked(dev, errp)) {
+        return;
+    }
+
    if (dev->parent_bus && !qbus_is_hotpluggable(dev->parent_bus)) {
        error_setg(errp, QERR_BUS_NO_HOTPLUG, dev->parent_bus->name);
        return;
#5 lore raw

Re: [PATCH v8 02/17] qdev: unplug blocker for devices

Quoted 6 lines · Jag Raman
Quoted 6 lines · Markus Armbruster
Quoted 1 line · Jagannathan Raman

Add blocker to prevent hot-unplug of devices

Quoted 8 lines
Quoted 2 lines

Why do you need this? I'm not doubting you do, I just want to read your reasons here :)

Hi Markus, :)

The x-vfio-user-server depends on an attached PCIDevice. As long as x-vfio-user-server is used, we don’t want the PCIDevice to be unplugged. This blocker prevents an user from removing PCIDevice while the vfio-user server is in use.

Please work that into your commit message. Perhaps along the lines of

One of the next commits will do <stuff>. <badness> will happen when the PCI device is unplugged. Create the means to prevent that.

Quoted 95 lines
Quoted 3 lines
Quoted 3 lines

I recall receiving a “Reviewed-by” from Stefan previously.

I’m very sorry I didn’t add that here. I’ll go over all the patches once again to confirm that the “Reviewed-by” status reflects accurately.

Quoted 77 lines
Quoted 77 lines

---

include/hw/qdev-core.h | 29 +++++++++++++++++++++++++++++
hw/core/qdev.c         | 24 ++++++++++++++++++++++++
softmmu/qdev-monitor.c |  4 ++++
3 files changed, 57 insertions(+)
include/hw/qdev-core.h +29 -0
index 92c3d65208..1b9fa25e5c 100644
@@ -193,6 +193,7 @@ struct DeviceState {
    int instance_id_alias;
    int alias_required_for_version;
    ResettableState reset;
+    GSList *unplug_blockers;
 };
 
 struct DeviceListener {
@@ -419,6 +420,34 @@ void qdev_simple_device_unplug_cb(HotplugHandler *hotplug_dev,
 void qdev_machine_creation_done(void);
 bool qdev_machine_modified(void);
 
+/*
+ * qdev_add_unplug_blocker: Adds an unplug blocker to a device
+ *
+ * @dev: Device to be blocked from unplug
+ * @reason: Reason for blocking
+ */
+void qdev_add_unplug_blocker(DeviceState *dev, Error *reason);
+
+/*
+ * qdev_del_unplug_blocker: Removes an unplug blocker from a device
+ *
+ * @dev: Device to be unblocked
+ * @reason: Pointer to the Error used with qdev_add_unplug_blocker.
+ *          Used as a handle to lookup the blocker for deletion.
+ */
+void qdev_del_unplug_blocker(DeviceState *dev, Error *reason);
+
+/*
+ * qdev_unplug_blocked: Confirms if a device is blocked from unplug
+ *
+ * @dev: Device to be tested
+ * @reason: Returns one of the reasons why the device is blocked,
+ *          if any
+ *
+ * Returns: true if device is blocked from unplug, false otherwise
+ */
+bool qdev_unplug_blocked(DeviceState *dev, Error **errp);
+
 /**
 * GpioPolarity: Polarity of a GPIO line
 *
hw/core/qdev.c +16 -0
index 84f3019440..0806d8fcaa 100644
@@ -468,6 +468,28 @@ char *qdev_get_dev_path(DeviceState *dev)
    return NULL;
 }
 
+void qdev_add_unplug_blocker(DeviceState *dev, Error *reason)
+{
+    dev->unplug_blockers = g_slist_prepend(dev->unplug_blockers, reason);
+}
+
+void qdev_del_unplug_blocker(DeviceState *dev, Error *reason)
+{
+    dev->unplug_blockers = g_slist_remove(dev->unplug_blockers, reason);
+}
+
+bool qdev_unplug_blocked(DeviceState *dev, Error **errp)
+{
+    ERRP_GUARD();
+
+    if (dev->unplug_blockers) {
+        error_propagate(errp, error_copy(dev->unplug_blockers->data));
Quoted 8 lines

hw/core/qdev.c char *qdev_get_dev_path(DeviceState *dev) · line 491

Quoted 5 lines
hw/core/qdev.c +5 -0
+        return true;
+    }
+
+    return false;
+}

This cites the most recently added blocker as reason. Your function comment covers it: "Returns one of the reasons". Okay.

Quoted 1 line

I could change the comment to say that it returns the recently added reason.

Up to you.

author
#6 lore raw

Re: [PATCH v8 02/17] qdev: unplug blocker for devices

Quoted 17 lines · Markus Armbruster
Quoted 11 lines · Jag Raman
Quoted 8 lines · Markus Armbruster
Quoted 1 line · Jagannathan Raman

Add blocker to prevent hot-unplug of devices

Why do you need this? I'm not doubting you do, I just want to read your reasons here :)

Hi Markus, :)

Quoted 8 lines
Quoted 3 lines

The x-vfio-user-server depends on an attached PCIDevice. As long as x-vfio-user-server is used, we don’t want the PCIDevice to be unplugged. This blocker prevents an user from removing PCIDevice while the vfio-user server is in use.

Please work that into your commit message. Perhaps along the lines of

One of the next commits will do <stuff>. <badness> will happen when the PCI device is unplugged. Create the means to prevent that.

OK, will do.

Thank you!
--
Jag

Quoted 100 lines
Quoted 96 lines
Quoted 4 lines
Quoted 4 lines

I recall receiving a “Reviewed-by” from Stefan previously.

I’m very sorry I didn’t add that here. I’ll go over all the patches once again to confirm that the “Reviewed-by” status reflects accurately.

Quoted 77 lines
Quoted 77 lines

---

include/hw/qdev-core.h | 29 +++++++++++++++++++++++++++++
hw/core/qdev.c         | 24 ++++++++++++++++++++++++
softmmu/qdev-monitor.c |  4 ++++
3 files changed, 57 insertions(+)
include/hw/qdev-core.h +29 -0
index 92c3d65208..1b9fa25e5c 100644
@@ -193,6 +193,7 @@ struct DeviceState {
   int instance_id_alias;
   int alias_required_for_version;
   ResettableState reset;
+    GSList *unplug_blockers;
 };
 
 struct DeviceListener {
@@ -419,6 +420,34 @@ void qdev_simple_device_unplug_cb(HotplugHandler *hotplug_dev,
 void qdev_machine_creation_done(void);
 bool qdev_machine_modified(void);
 
+/*
+ * qdev_add_unplug_blocker: Adds an unplug blocker to a device
+ *
+ * @dev: Device to be blocked from unplug
+ * @reason: Reason for blocking
+ */
+void qdev_add_unplug_blocker(DeviceState *dev, Error *reason);
+
+/*
+ * qdev_del_unplug_blocker: Removes an unplug blocker from a device
+ *
+ * @dev: Device to be unblocked
+ * @reason: Pointer to the Error used with qdev_add_unplug_blocker.
+ *          Used as a handle to lookup the blocker for deletion.
+ */
+void qdev_del_unplug_blocker(DeviceState *dev, Error *reason);
+
+/*
+ * qdev_unplug_blocked: Confirms if a device is blocked from unplug
+ *
+ * @dev: Device to be tested
+ * @reason: Returns one of the reasons why the device is blocked,
+ *          if any
+ *
+ * Returns: true if device is blocked from unplug, false otherwise
+ */
+bool qdev_unplug_blocked(DeviceState *dev, Error **errp);
+
 /**
 * GpioPolarity: Polarity of a GPIO line
 *
hw/core/qdev.c +16 -0
index 84f3019440..0806d8fcaa 100644
@@ -468,6 +468,28 @@ char *qdev_get_dev_path(DeviceState *dev)
   return NULL;
 }
 
+void qdev_add_unplug_blocker(DeviceState *dev, Error *reason)
+{
+    dev->unplug_blockers = g_slist_prepend(dev->unplug_blockers, reason);
+}
+
+void qdev_del_unplug_blocker(DeviceState *dev, Error *reason)
+{
+    dev->unplug_blockers = g_slist_remove(dev->unplug_blockers, reason);
+}
+
+bool qdev_unplug_blocked(DeviceState *dev, Error **errp)
+{
+    ERRP_GUARD();
+
+    if (dev->unplug_blockers) {
+        error_propagate(errp, error_copy(dev->unplug_blockers->data));
Quoted 8 lines

hw/core/qdev.c char *qdev_get_dev_path(DeviceState *dev) · line 491

Quoted 5 lines
hw/core/qdev.c +5 -0
+        return true;
+    }
+
+    return false;
+}

This cites the most recently added blocker as reason. Your function comment covers it: "Returns one of the reasons". Okay.

Quoted 1 line

I could change the comment to say that it returns the recently added reason.

Up to you.

#7 lore raw

PATCH v8 3/17 remote/machine: add HotplugHandler for remote machine

In reply to Jagannathan Raman

Allow hotplugging of PCI(e) devices to remote machine

---
 hw/remote/machine.c | 10 ++++++++++
 1 file changed, 10 insertions(+)
hw/remote/machine.c +10 -0
index 952105eab5..0c5bd4f923 100644
@@ -21,6 +21,7 @@
 #include "qapi/error.h"
 #include "hw/pci/pci_host.h"
 #include "hw/remote/iohub.h"
+#include "hw/qdev-core.h"
 
 static void remote_machine_init(MachineState *machine)
 {
@@ -54,14 +55,19 @@ static void remote_machine_init(MachineState *machine)
 
     pci_bus_irqs(pci_host->bus, remote_iohub_set_irq, remote_iohub_map_irq,
                  &s->iohub, REMOTE_IOHUB_NB_PIRQS);
+
+    qbus_set_hotplug_handler(BUS(pci_host->bus), OBJECT(s));
 }
 
 static void remote_machine_class_init(ObjectClass *oc, void *data)
 {
     MachineClass *mc = MACHINE_CLASS(oc);
+    HotplugHandlerClass *hc = HOTPLUG_HANDLER_CLASS(oc);
 
     mc->init = remote_machine_init;
     mc->desc = "Experimental remote machine";
+
+    hc->unplug = qdev_simple_device_unplug_cb;
 }
 
 static const TypeInfo remote_machine = {
@@ -69,6 +75,10 @@ static const TypeInfo remote_machine = {
     .parent = TYPE_MACHINE,
     .instance_size = sizeof(RemoteMachineState),
     .class_init = remote_machine_class_init,
+    .interfaces = (InterfaceInfo[]) {
+        { TYPE_HOTPLUG_HANDLER },
+        { }
+    }
 };
 
 static void remote_machine_register_types(void)
Signature · 1 line
2.20.1
#8 lore raw

PATCH v8 4/17 remote/machine: add vfio-user property

In reply to Jagannathan Raman

Add vfio-user to x-remote machine. It is a boolean, which indicates if the machine supports vfio-user protocol. The machine configures the bus differently vfio-user and multiprocess protocols, so this property informs it on how to configure the bus.

This property should be short lived. Once vfio-user fully replaces multiprocess, this property could be removed.

---
 include/hw/remote/machine.h |  2 ++
 hw/remote/machine.c         | 23 +++++++++++++++++++++++
 2 files changed, 25 insertions(+)
include/hw/remote/machine.h +2 -0
index 2a2a33c4b2..8d0fa98d33 100644
@@ -22,6 +22,8 @@ struct RemoteMachineState {
 
     RemotePCIHost *host;
     RemoteIOHubState iohub;
+
+    bool vfio_user;
 };
 
 /* Used to pass to co-routine device and ioc. */
hw/remote/machine.c +23 -0
index 0c5bd4f923..a9a75e170f 100644
@@ -59,6 +59,25 @@ static void remote_machine_init(MachineState *machine)
     qbus_set_hotplug_handler(BUS(pci_host->bus), OBJECT(s));
 }
 
+static bool remote_machine_get_vfio_user(Object *obj, Error **errp)
+{
+    RemoteMachineState *s = REMOTE_MACHINE(obj);
+
+    return s->vfio_user;
+}
+
+static void remote_machine_set_vfio_user(Object *obj, bool value, Error **errp)
+{
+    RemoteMachineState *s = REMOTE_MACHINE(obj);
+
+    if (phase_check(PHASE_MACHINE_CREATED)) {
+        error_setg(errp, "Error enabling vfio-user - machine already created");
+        return;
+    }
+
+    s->vfio_user = value;
+}
+
 static void remote_machine_class_init(ObjectClass *oc, void *data)
 {
     MachineClass *mc = MACHINE_CLASS(oc);
@@ -68,6 +87,10 @@ static void remote_machine_class_init(ObjectClass *oc, void *data)
     mc->desc = "Experimental remote machine";
 
     hc->unplug = qdev_simple_device_unplug_cb;
+
+    object_class_property_add_bool(oc, "vfio-user",
+                                   remote_machine_get_vfio_user,
+                                   remote_machine_set_vfio_user);
 }
 
 static const TypeInfo remote_machine = {
Signature · 1 line
2.20.1
#9 lore raw

PATCH v8 5/17 configure: require cmake 3.19 or newer

In reply to Jagannathan Raman

cmake needs to accept the compiler flags specified with CMAKE_<LANG>_COMPILER variable. It does so starting with version 3.19

---
 configure | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)
configure +16 -0
index 7c08c18358..7a1a98bddf 100755
@@ -250,6 +250,7 @@ stack_protector=""
 safe_stack=""
 use_containers="yes"
 gdb_bin=$(command -v "gdb-multiarch" || command -v "gdb")
+cmake_required="no"
 
 if test -e "$source_path/.git"
 then
@@ -2777,6 +2778,21 @@ if !(GIT="$git" "$source_path/scripts/git-submodule.sh" "$git_submodules_action"
     exit 1
 fi
 
+# Per cmake spec, CMAKE_<LANG>_COMPILER variable may include "mandatory" compiler
+# flags. QEMU needs to specify these flags to correctly configure the build
+# environment. cmake 3.19 allows specifying these mandatory compiler flags,
+# and as such 3.19 or newer is required to build QEMU.
+if test "$cmake_required" = "yes" ; then
+    cmake_bin=$(command -v "cmake")
+    if [ -z "$cmake_bin" ]; then
+        error_exit "cmake not found"
+    fi
+    cmake_version=$($cmake_bin --version | head -n 1)
+    if ! version_ge ${cmake_version##* } 3.19; then
+        error_exit "QEMU needs cmake 3.19 or newer"
+    fi
+fi
+
 config_host_mak="config-host.mak"
 
 echo "# Automatically generated by configure - do not modify" > $config_host_mak
Signature · 1 line
2.20.1
#10 lore raw

PATCH v8 6/17 vfio-user: build library

In reply to Jagannathan Raman

add the libvfio-user library as a submodule. build it as a cmake subproject.

---
 configure                                  | 20 +++++++++-
 meson.build                                | 44 +++++++++++++++++++++-
 .gitlab-ci.d/buildtest.yml                 |  2 +
 .gitmodules                                |  3 ++
 Kconfig.host                               |  4 ++
 MAINTAINERS                                |  1 +
 hw/remote/Kconfig                          |  4 ++
 hw/remote/meson.build                      |  2 +
 meson_options.txt                          |  3 ++
 subprojects/libvfio-user                   |  1 +
 tests/docker/dockerfiles/centos8.docker    |  2 +
 tests/docker/dockerfiles/ubuntu2004.docker |  2 +
 12 files changed, 86 insertions(+), 2 deletions(-)
 create mode 160000 subprojects/libvfio-user
configure +19 -1
index 7a1a98bddf..c4fd7a42d4 100755
@@ -333,6 +333,7 @@ meson_args=""
 ninja=""
 gio="$default_feature"
 skip_meson=no
+vfio_user_server="disabled"
 
 # The following Meson options are handled manually (still they
 # are included in the automatically generated help message)
@@ -1044,6 +1045,11 @@ for opt do
   ;;
   --disable-blobs) meson_option_parse --disable-install-blobs ""
   ;;
+  --enable-vfio-user-server) vfio_user_server="enabled"
+                             cmake_required="yes"
+  ;;
+  --disable-vfio-user-server) vfio_user_server="disabled"
+  ;;
   --enable-tcmalloc) meson_option_parse --enable-malloc=tcmalloc tcmalloc
   ;;
   --enable-jemalloc) meson_option_parse --enable-malloc=jemalloc jemalloc
@@ -1267,6 +1273,7 @@ cat << EOF
   vhost-vdpa      vhost-vdpa kernel backend support
   opengl          opengl support
   gio             libgio support
+  vfio-user-server    vfio-user server support
 
 NOTE: The object files are built at the place where configure is launched
 EOF
@@ -2622,6 +2629,17 @@ but not implemented on your system"
     fi
 fi
 
+##########################################
+# check for vfio_user_server
+
+case "$vfio_user_server" in
+  auto | enabled )
+    if test "$git_submodules_action" != "ignore"; then
+      git_submodules="${git_submodules} subprojects/libvfio-user"
+    fi
+    ;;
+esac
+
 ##########################################
 # End of CC checks
 # After here, no more $cc or $ld runs
@@ -3185,7 +3203,7 @@ if test "$skip_meson" = no; then
         -Db_pie=$(if test "$pie" = yes; then echo true; else echo false; fi) \
         -Db_coverage=$(if test "$gcov" = yes; then echo true; else echo false; fi) \
         -Db_lto=$lto -Dcfi=$cfi -Dtcg=$tcg -Dxen=$xen \
-        -Dcapstone=$capstone -Dfdt=$fdt -Dslirp=$slirp \
+        -Dcapstone=$capstone -Dfdt=$fdt -Dslirp=$slirp -Dvfio_user_server=$vfio_user_server \
         $(test -n "${LIB_FUZZING_ENGINE+xxx}" && echo "-Dfuzzing_engine=$LIB_FUZZING_ENGINE") \
         $(if test "$default_feature" = no; then echo "-Dauto_features=disabled"; fi) \
         "$@" $cross_arg "$PWD" "$source_path"
meson.build +43 -1

68 lines not on this page · raw message

.gitlab-ci.d/buildtest.yml +2 -0
index 0aea7ab84c..671a5d8fa4 100644
@@ -42,6 +42,7 @@ build-system-ubuntu:
   variables:
     IMAGE: ubuntu2004
     CONFIGURE_ARGS: --enable-docs --enable-fdt=system --enable-slirp=system
+                    --enable-vfio-user-server
     TARGETS: aarch64-softmmu alpha-softmmu cris-softmmu hppa-softmmu
       microblazeel-softmmu mips64el-softmmu
     MAKE_CHECK_ARGS: check-build
@@ -165,6 +166,7 @@ build-system-centos:
     IMAGE: centos8
     CONFIGURE_ARGS: --disable-nettle --enable-gcrypt --enable-fdt=system
       --enable-modules --enable-trace-backends=dtrace --enable-docs
+      --enable-vfio-user-server
     TARGETS: ppc64-softmmu or1k-softmmu s390x-softmmu
       x86_64-softmmu rx-softmmu sh4-softmmu nios2-softmmu
     MAKE_CHECK_ARGS: check-build
.gitmodules +3 -0
index f4b6a9b401..d66af96dc9 100644
@@ -67,3 +67,6 @@
 [submodule "tests/lcitool/libvirt-ci"]
 	path = tests/lcitool/libvirt-ci
 	url = https://gitlab.com/libvirt/libvirt-ci.git
+[submodule "subprojects/libvfio-user"]
+	path = subprojects/libvfio-user
+	url = https://github.com/nutanix/libvfio-user.git
Kconfig.host +4 -0
index 60b9c07b5e..f2da8bcf8a 100644
@@ -45,3 +45,7 @@ config MULTIPROCESS_ALLOWED
 config FUZZ
     bool
     select SPARSE_MEM
+
+config VFIO_USER_SERVER_ALLOWED
+    bool
+    imply VFIO_USER_SERVER
MAINTAINERS +1 -0
index 4ad2451e03..dca81a2d22 100644
@@ -3597,6 +3597,7 @@ F: hw/remote/proxy-memory-listener.c
 F: include/hw/remote/proxy-memory-listener.h
 F: hw/remote/iohub.c
 F: include/hw/remote/iohub.h
+F: subprojects/libvfio-user
 
 EBPF:
 M: Jason Wang <[email protected]>
hw/remote/Kconfig +4 -0
index 08c16e235f..2d6b4f4cf4 100644
@@ -2,3 +2,7 @@ config MULTIPROCESS
     bool
     depends on PCI && PCI_EXPRESS && KVM
     select REMOTE_PCIHOST
+
+config VFIO_USER_SERVER
+    bool
+    depends on MULTIPROCESS
hw/remote/meson.build +2 -0
index e6a5574242..dfea6b533b 100644
@@ -7,6 +7,8 @@ remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('remote-obj.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('proxy.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('iohub.c'))
 
+remote_ss.add(when: 'CONFIG_VFIO_USER_SERVER', if_true: vfiouser)
+
 specific_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('memory.c'))
 specific_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('proxy-memory-listener.c'))
 
meson_options.txt +3 -0
index 52b11cead4..bd531fd545 100644
@@ -91,6 +91,9 @@ option('avx2', type: 'feature', value: 'auto',
 option('avx512f', type: 'feature', value: 'disabled',
        description: 'AVX512F optimizations')
 
+option('vfio_user_server', type: 'feature', value: 'auto',
+       description: 'vfio-user server support')
+
 option('attr', type : 'feature', value : 'auto',
        description: 'attr/xattr support')
 option('auth_pam', type : 'feature', value : 'auto',
subprojects/libvfio-user new file +1 -0
new file mode 160000
index 0000000000..030d2f6e79
@@ -0,0 +1 @@
+Subproject commit 030d2f6e7978b8ca7577b81d4f48e2771bcd8f47
tests/docker/dockerfiles/centos8.docker +2 -0
index 3ede55d09b..b6b4aa9626 100644
@@ -23,6 +23,7 @@ RUN dnf update -y && \
         capstone-devel \
         ccache \
         clang \
+        cmake \
         ctags \
         cyrus-sasl-devel \
         daxctl-devel \
@@ -45,6 +46,7 @@ RUN dnf update -y && \
         gtk3-devel \
         hostname \
         jemalloc-devel \
+        json-c-devel \
         libaio-devel \
         libasan \
         libattr-devel \
tests/docker/dockerfiles/ubuntu2004.docker +2 -0
index b9d06cb040..2422cc5574 100644
@@ -18,6 +18,7 @@ RUN export DEBIAN_FRONTEND=noninteractive && \
             ca-certificates \
             ccache \
             clang \
+            cmake \
             dbus \
             debianutils \
             diffutils \
@@ -58,6 +59,7 @@ RUN export DEBIAN_FRONTEND=noninteractive && \
             libiscsi-dev \
             libjemalloc-dev \
             libjpeg-turbo8-dev \
+            libjson-c-dev \
             liblttng-ust-dev \
             liblzo2-dev \
             libncursesw5-dev \
Signature · 1 line
2.20.1
Reviewed-by on v8 6/17
#11 lore raw

Re: [PATCH v8 06/17] vfio-user: build library

Quoted 6 lines · Jagannathan Raman

add the libvfio-user library as a submodule. build it as a cmake subproject.

Quoted 15 lines
---
 configure                                  | 20 +++++++++-
 meson.build                                | 44 +++++++++++++++++++++-
 .gitlab-ci.d/buildtest.yml                 |  2 +
 .gitmodules                                |  3 ++
 Kconfig.host                               |  4 ++
 MAINTAINERS                                |  1 +
 hw/remote/Kconfig                          |  4 ++
 hw/remote/meson.build                      |  2 +
 meson_options.txt                          |  3 ++
 subprojects/libvfio-user                   |  1 +
 tests/docker/dockerfiles/centos8.docker    |  2 +
 tests/docker/dockerfiles/ubuntu2004.docker |  2 +
 12 files changed, 86 insertions(+), 2 deletions(-)
 create mode 160000 subprojects/libvfio-user

libvfio-user is BSD-3-Clause and libjson-c is MIT (Expat).

#12 lore raw

PATCH v8 7/17 vfio-user: define vfio-user-server object

In reply to Jagannathan Raman

Define vfio-user object which is remote process server for QEMU. Setup object initialization functions and properties necessary to instantiate the object

---
 qapi/qom.json               |  20 +++-
 include/hw/remote/machine.h |   2 +
 hw/remote/machine.c         |  27 +++++
 hw/remote/vfio-user-obj.c   | 211 ++++++++++++++++++++++++++++++++++++
 MAINTAINERS                 |   1 +
 hw/remote/meson.build       |   1 +
 hw/remote/trace-events      |   3 +
 7 files changed, 263 insertions(+), 2 deletions(-)
 create mode 100644 hw/remote/vfio-user-obj.c
qapi/qom.json +18 -2
index eeb5395ff3..e7b1758a11 100644
@@ -703,6 +703,20 @@
 { 'struct': 'RemoteObjectProperties',
   'data': { 'fd': 'str', 'devid': 'str' } }
 
+##
+# @VfioUserServerProperties:
+#
+# Properties for x-vfio-user-server objects.
+#
+# @socket: socket to be used by the libvfiouser library
+#
+# @device: the id of the device to be emulated at the server
+#
+# Since: 7.1
+##
+{ 'struct': 'VfioUserServerProperties',
+  'data': { 'socket': 'SocketAddress', 'device': 'str' } }
+
 ##
 # @RngProperties:
 #
@@ -842,7 +856,8 @@
     'tls-creds-psk',
     'tls-creds-x509',
     'tls-cipher-suites',
-    { 'name': 'x-remote-object', 'features': [ 'unstable' ] }
+    { 'name': 'x-remote-object', 'features': [ 'unstable' ] },
+    { 'name': 'x-vfio-user-server', 'features': [ 'unstable' ] }
   ] }
 
 ##
@@ -905,7 +920,8 @@
       'tls-creds-psk':              'TlsCredsPskProperties',
       'tls-creds-x509':             'TlsCredsX509Properties',
       'tls-cipher-suites':          'TlsCredsProperties',
-      'x-remote-object':            'RemoteObjectProperties'
+      'x-remote-object':            'RemoteObjectProperties',
+      'x-vfio-user-server':         'VfioUserServerProperties'
   } }
 
 ##
include/hw/remote/machine.h +2 -0
index 8d0fa98d33..ac32fda387 100644
@@ -24,6 +24,8 @@ struct RemoteMachineState {
     RemoteIOHubState iohub;
 
     bool vfio_user;
+
+    bool auto_shutdown;
 };
 
 /* Used to pass to co-routine device and ioc. */
hw/remote/machine.c +27 -0
index a9a75e170f..ed91659794 100644
@@ -78,6 +78,28 @@ static void remote_machine_set_vfio_user(Object *obj, bool value, Error **errp)
     s->vfio_user = value;
 }
 
+static bool remote_machine_get_auto_shutdown(Object *obj, Error **errp)
+{
+    RemoteMachineState *s = REMOTE_MACHINE(obj);
+
+    return s->auto_shutdown;
+}
+
+static void remote_machine_set_auto_shutdown(Object *obj, bool value,
+                                             Error **errp)
+{
+    RemoteMachineState *s = REMOTE_MACHINE(obj);
+
+    s->auto_shutdown = value;
+}
+
+static void remote_machine_instance_init(Object *obj)
+{
+    RemoteMachineState *s = REMOTE_MACHINE(obj);
+
+    s->auto_shutdown = true;
+}
+
 static void remote_machine_class_init(ObjectClass *oc, void *data)
 {
     MachineClass *mc = MACHINE_CLASS(oc);
@@ -91,12 +113,17 @@ static void remote_machine_class_init(ObjectClass *oc, void *data)
     object_class_property_add_bool(oc, "vfio-user",
                                    remote_machine_get_vfio_user,
                                    remote_machine_set_vfio_user);
+
+    object_class_property_add_bool(oc, "auto-shutdown",
+                                   remote_machine_get_auto_shutdown,
+                                   remote_machine_set_auto_shutdown);
 }
 
 static const TypeInfo remote_machine = {
     .name = TYPE_REMOTE_MACHINE,
     .parent = TYPE_MACHINE,
     .instance_size = sizeof(RemoteMachineState),
+    .instance_init = remote_machine_instance_init,
     .class_init = remote_machine_class_init,
     .interfaces = (InterfaceInfo[]) {
         { TYPE_HOTPLUG_HANDLER },
hw/remote/vfio-user-obj.c new file +211 -0

211 lines not on this page · raw message

MAINTAINERS +1 -0
index dca81a2d22..37afdecc61 100644
@@ -3598,6 +3598,7 @@ F: include/hw/remote/proxy-memory-listener.h
 F: hw/remote/iohub.c
 F: include/hw/remote/iohub.h
 F: subprojects/libvfio-user
+F: hw/remote/vfio-user-obj.c
 
 EBPF:
 M: Jason Wang <[email protected]>
hw/remote/meson.build +1 -0
index dfea6b533b..534ac5df79 100644
@@ -6,6 +6,7 @@ remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('message.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('remote-obj.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('proxy.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('iohub.c'))
+remote_ss.add(when: 'CONFIG_VFIO_USER_SERVER', if_true: files('vfio-user-obj.c'))
 
 remote_ss.add(when: 'CONFIG_VFIO_USER_SERVER', if_true: vfiouser)
 
hw/remote/trace-events +3 -0
index 0b23974f90..7da12f0d96 100644
@@ -2,3 +2,6 @@
 
 mpqemu_send_io_error(int cmd, int size, int nfds) "send command %d size %d, %d file descriptors to remote process"
 mpqemu_recv_io_error(int cmd, int size, int nfds) "failed to receive %d size %d, %d file descriptors to remote process"
+
+# vfio-user-obj.c
+vfu_prop(const char *prop, const char *val) "vfu: setting %s as %s"
Signature · 1 line
2.20.1
#13 lore raw

Re: [PATCH v8 07/17] vfio-user: define vfio-user-server object

Quoted 5 lines · Jagannathan Raman
qapi/qom.json +0 -0
index eeb5395ff3..e7b1758a11 100644

qapi/qom.json line 711

Quoted 9 lines
qapi/qom.json +6 -0
@@ -703,6 +703,20 @@
 { 'struct': 'RemoteObjectProperties',
   'data': { 'fd': 'str', 'devid': 'str' } }
 
+##
+# @VfioUserServerProperties:
+#
+# Properties for x-vfio-user-server objects.
+#
+# @socket: socket to be used by the libvfiouser library

For consistency: libvfio-user library

#14 lore raw

PATCH v8 8/17 vfio-user: instantiate vfio-user context

In reply to Jagannathan Raman

create a context with the vfio-user library to run a PCI device

---
 hw/remote/vfio-user-obj.c | 82 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 82 insertions(+)
hw/remote/vfio-user-obj.c +82 -0
index c4d59b4d9d..d46acd5b63 100644
@@ -41,6 +41,9 @@
 #include "hw/remote/machine.h"
 #include "qapi/error.h"
 #include "qapi/qapi-visit-sockets.h"
+#include "qemu/notify.h"
+#include "sysemu/sysemu.h"
+#include "libvfio-user.h"
 
 #define TYPE_VFU_OBJECT "x-vfio-user-server"
 OBJECT_DECLARE_TYPE(VfuObject, VfuObjectClass, VFU_OBJECT)
@@ -74,8 +77,14 @@ struct VfuObject {
     char *device;
 
     Error *err;
+
+    Notifier machine_done;
+
+    vfu_ctx_t *vfu_ctx;
 };
 
+static void vfu_object_init_ctx(VfuObject *o, Error **errp);
+
 static bool vfu_object_auto_shutdown(void)
 {
     bool auto_shutdown = true;
@@ -108,6 +117,11 @@ static void vfu_object_set_socket(Object *obj, Visitor *v, const char *name,
 {
     VfuObject *o = VFU_OBJECT(obj);
 
+    if (o->vfu_ctx) {
+        error_setg(errp, "vfu: Unable to set socket property - server busy");
+        return;
+    }
+
     qapi_free_SocketAddress(o->socket);
 
     o->socket = NULL;
@@ -123,17 +137,69 @@ static void vfu_object_set_socket(Object *obj, Visitor *v, const char *name,
     }
 
     trace_vfu_prop("socket", o->socket->u.q_unix.path);
+
+    vfu_object_init_ctx(o, errp);
 }
 
 static void vfu_object_set_device(Object *obj, const char *str, Error **errp)
 {
     VfuObject *o = VFU_OBJECT(obj);
 
+    if (o->vfu_ctx) {
+        error_setg(errp, "vfu: Unable to set device property - server busy");
+        return;
+    }
+
     g_free(o->device);
 
     o->device = g_strdup(str);
 
     trace_vfu_prop("device", str);
+
+    vfu_object_init_ctx(o, errp);
+}
+
+/*
+ * TYPE_VFU_OBJECT depends on the availability of the 'socket' and 'device'
+ * properties. It also depends on devices instantiated in QEMU. These
+ * dependencies are not available during the instance_init phase of this
+ * object's life-cycle. As such, the server is initialized after the
+ * machine is setup. machine_init_done_notifier notifies TYPE_VFU_OBJECT
+ * when the machine is setup, and the dependencies are available.
+ */
+static void vfu_object_machine_done(Notifier *notifier, void *data)
+{
+    VfuObject *o = container_of(notifier, VfuObject, machine_done);
+    Error *err = NULL;
+
+    vfu_object_init_ctx(o, &err);
+
+    if (err) {
+        error_propagate(&error_abort, err);
+    }
+}
+
+static void vfu_object_init_ctx(VfuObject *o, Error **errp)
+{
+    ERRP_GUARD();
+
+    if (o->vfu_ctx || !o->socket || !o->device ||
+            !phase_check(PHASE_MACHINE_READY)) {
+        return;
+    }
+
+    if (o->err) {
+        error_propagate(errp, o->err);
+        o->err = NULL;
+        return;
+    }
+
+    o->vfu_ctx = vfu_create_ctx(VFU_TRANS_SOCK, o->socket->u.q_unix.path, 0,
+                                o, VFU_DEV_TYPE_PCI);
+    if (o->vfu_ctx == NULL) {
+        error_setg(errp, "vfu: Failed to create context - %s", strerror(errno));
+        return;
+    }
 }
 
 static void vfu_object_init(Object *obj)
@@ -148,6 +214,12 @@ static void vfu_object_init(Object *obj)
                    TYPE_VFU_OBJECT, TYPE_REMOTE_MACHINE);
         return;
     }
+
+    if (!phase_check(PHASE_MACHINE_READY)) {
+        o->machine_done.notify = vfu_object_machine_done;
+        qemu_add_machine_init_done_notifier(&o->machine_done);
+    }
+
 }
 
 static void vfu_object_finalize(Object *obj)
@@ -161,6 +233,11 @@ static void vfu_object_finalize(Object *obj)
 
     o->socket = NULL;
 
+    if (o->vfu_ctx) {
+        vfu_destroy_ctx(o->vfu_ctx);
+        o->vfu_ctx = NULL;
+    }
+
     g_free(o->device);
 
     o->device = NULL;
@@ -168,6 +245,11 @@ static void vfu_object_finalize(Object *obj)
     if (!k->nr_devs && vfu_object_auto_shutdown()) {
         qemu_system_shutdown_request(SHUTDOWN_CAUSE_GUEST_SHUTDOWN);
     }
+
+    if (o->machine_done.notify) {
+        qemu_remove_machine_init_done_notifier(&o->machine_done);
+        o->machine_done.notify = NULL;
+    }
 }
 
 static void vfu_object_class_init(ObjectClass *klass, void *data)
Signature · 1 line
2.20.1
#15 lore raw

PATCH v8 9/17 vfio-user: find and init PCI device

In reply to Jagannathan Raman

Find the PCI device with specified id. Initialize the device context with the QEMU PCI device

---
 hw/remote/vfio-user-obj.c | 67 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 67 insertions(+)
hw/remote/vfio-user-obj.c +67 -0
index d46acd5b63..15f6fe3a1a 100644
@@ -44,6 +44,8 @@
 #include "qemu/notify.h"
 #include "sysemu/sysemu.h"
 #include "libvfio-user.h"
+#include "hw/qdev-core.h"
+#include "hw/pci/pci.h"
 
 #define TYPE_VFU_OBJECT "x-vfio-user-server"
 OBJECT_DECLARE_TYPE(VfuObject, VfuObjectClass, VFU_OBJECT)
@@ -81,6 +83,10 @@ struct VfuObject {
     Notifier machine_done;
 
     vfu_ctx_t *vfu_ctx;
+
+    PCIDevice *pci_dev;
+
+    Error *unplug_blocker;
 };
 
 static void vfu_object_init_ctx(VfuObject *o, Error **errp);
@@ -182,6 +188,9 @@ static void vfu_object_machine_done(Notifier *notifier, void *data)
 static void vfu_object_init_ctx(VfuObject *o, Error **errp)
 {
     ERRP_GUARD();
+    DeviceState *dev = NULL;
+    vfu_pci_type_t pci_type = VFU_PCI_TYPE_CONVENTIONAL;
+    int ret;
 
     if (o->vfu_ctx || !o->socket || !o->device ||
             !phase_check(PHASE_MACHINE_READY)) {
@@ -200,6 +209,53 @@ static void vfu_object_init_ctx(VfuObject *o, Error **errp)
         error_setg(errp, "vfu: Failed to create context - %s", strerror(errno));
         return;
     }
+
+    dev = qdev_find_recursive(sysbus_get_default(), o->device);
+    if (dev == NULL) {
+        error_setg(errp, "vfu: Device %s not found", o->device);
+        goto fail;
+    }
+
+    if (!object_dynamic_cast(OBJECT(dev), TYPE_PCI_DEVICE)) {
+        error_setg(errp, "vfu: %s not a PCI device", o->device);
+        goto fail;
+    }
+
+    o->pci_dev = PCI_DEVICE(dev);
+
+    object_ref(OBJECT(o->pci_dev));
+
+    if (pci_is_express(o->pci_dev)) {
+        pci_type = VFU_PCI_TYPE_EXPRESS;
+    }
+
+    ret = vfu_pci_init(o->vfu_ctx, pci_type, PCI_HEADER_TYPE_NORMAL, 0);
+    if (ret < 0) {
+        error_setg(errp,
+                   "vfu: Failed to attach PCI device %s to context - %s",
+                   o->device, strerror(errno));
+        goto fail;
+    }
+
+    error_setg(&o->unplug_blocker,
+               "vfu: %s for %s must be deleted before unplugging",
+               TYPE_VFU_OBJECT, o->device);
+    qdev_add_unplug_blocker(DEVICE(o->pci_dev), o->unplug_blocker);
+
+    return;
+
+fail:
+    vfu_destroy_ctx(o->vfu_ctx);
+    if (o->unplug_blocker && o->pci_dev) {
+        qdev_del_unplug_blocker(DEVICE(o->pci_dev), o->unplug_blocker);
+        error_free(o->unplug_blocker);
+        o->unplug_blocker = NULL;
+    }
+    if (o->pci_dev) {
+        object_unref(OBJECT(o->pci_dev));
+        o->pci_dev = NULL;
+    }
+    o->vfu_ctx = NULL;
 }
 
 static void vfu_object_init(Object *obj)
@@ -242,6 +298,17 @@ static void vfu_object_finalize(Object *obj)
 
     o->device = NULL;
 
+    if (o->unplug_blocker && o->pci_dev) {
+        qdev_del_unplug_blocker(DEVICE(o->pci_dev), o->unplug_blocker);
+        error_free(o->unplug_blocker);
+        o->unplug_blocker = NULL;
+    }
+
+    if (o->pci_dev) {
+        object_unref(OBJECT(o->pci_dev));
+        o->pci_dev = NULL;
+    }
+
     if (!k->nr_devs && vfu_object_auto_shutdown()) {
         qemu_system_shutdown_request(SHUTDOWN_CAUSE_GUEST_SHUTDOWN);
     }
Signature · 1 line
2.20.1
#16 lore raw

PATCH v8 10/17 vfio-user: run vfio-user context

In reply to Jagannathan Raman

Setup a handler to run vfio-user context. The context is driven by messages to the file descriptor associated with it - get the fd for the context and hook up the handler with it

---
 qapi/misc.json            | 23 ++++++++++
 hw/remote/vfio-user-obj.c | 95 ++++++++++++++++++++++++++++++++++++++-
 2 files changed, 117 insertions(+), 1 deletion(-)
qapi/misc.json +23 -0
index b83cc39029..f3cc4a4854 100644
@@ -553,3 +553,26 @@
 ##
 { 'event': 'RTC_CHANGE',
   'data': { 'offset': 'int', 'qom-path': 'str' } }
+
+##
+# @VFU_CLIENT_HANGUP:
+#
+# Emitted when the client of a TYPE_VFIO_USER_SERVER closes the
+# communication channel
+#
+# @id: ID of the TYPE_VFIO_USER_SERVER object
+#
+# @device: ID of attached PCI device
+#
+# Since: 7.1
+#
+# Example:
+#
+# <- { "event": "VFU_CLIENT_HANGUP",
+#      "data": { "id": "vfu1",
+#                "device": "lsi1" },
+#      "timestamp": { "seconds": 1265044230, "microseconds": 450486 } }
+#
+##
+{ 'event': 'VFU_CLIENT_HANGUP',
+  'data': { 'id': 'str', 'device': 'str' } }
hw/remote/vfio-user-obj.c +94 -1

148 lines not on this page · raw message

Signature · 1 line
2.20.1
#17 lore raw

Re: [PATCH v8 10/17] vfio-user: run vfio-user context

Quoted 21 lines · Jagannathan Raman

Setup a handler to run vfio-user context. The context is driven by messages to the file descriptor associated with it - get the fd for the context and hook up the handler with it

---
 qapi/misc.json            | 23 ++++++++++
 hw/remote/vfio-user-obj.c | 95 ++++++++++++++++++++++++++++++++++++++-
 2 files changed, 117 insertions(+), 1 deletion(-)
qapi/misc.json +0 -0
index b83cc39029..f3cc4a4854 100644
@@ -553,3 +553,26 @@
 ##
 { 'event': 'RTC_CHANGE',
   'data': { 'offset': 'int', 'qom-path': 'str' } }

qapi/misc.json line 565

Quoted 10 lines
qapi/misc.json +10 -0
+
+##
+# @VFU_CLIENT_HANGUP:
+#
+# Emitted when the client of a TYPE_VFIO_USER_SERVER closes the
+# communication channel
+#
+# @id: ID of the TYPE_VFIO_USER_SERVER object
+#
+# @device: ID of attached PCI device

Is this the ID set with -device id=... and such?

Quoted 13 lines
qapi/misc.json +13 -0
+#
+# Since: 7.1
+#
+# Example:
+#
+# <- { "event": "VFU_CLIENT_HANGUP",
+#      "data": { "id": "vfu1",
+#                "device": "lsi1" },
+#      "timestamp": { "seconds": 1265044230, "microseconds": 450486 } }
+#
+##
+{ 'event': 'VFU_CLIENT_HANGUP',
+  'data': { 'id': 'str', 'device': 'str' } }
author
#18 lore raw

Re: [PATCH v8 10/17] vfio-user: run vfio-user context

Quoted 31 lines · Markus Armbruster
Quoted 25 lines · Jagannathan Raman

Setup a handler to run vfio-user context. The context is driven by messages to the file descriptor associated with it - get the fd for the context and hook up the handler with it

---
qapi/misc.json | 23 ++++++++++
hw/remote/vfio-user-obj.c | 95 ++++++++++++++++++++++++++++++++++++++-

2 files changed, 117 insertions(+), 1 deletion(-)
qapi/misc.json +4 -0
index b83cc39029..f3cc4a4854 100644
@@ -553,3 +553,26 @@
 ##
 { 'event': 'RTC_CHANGE',
  'data': { 'offset': 'int', 'qom-path': 'str' } }
+
+##
+# @VFU_CLIENT_HANGUP:
+#
Quoted 8 lines

qapi/misc.json line 565

Quoted 6 lines
qapi/misc.json +6 -0
+# Emitted when the client of a TYPE_VFIO_USER_SERVER closes the
+# communication channel
+#
+# @id: ID of the TYPE_VFIO_USER_SERVER object
+#
+# @device: ID of attached PCI device

Is this the ID set with -device id=... and such?

Yes, that is correct. It’s the ID set with the “-device id=…” option/

Thank you!
--
Jag

Quoted 15 lines
Quoted 14 lines
qapi/misc.json +13 -0
+#
+# Since: 7.1
+#
+# Example:
+#
+# <- { "event": "VFU_CLIENT_HANGUP",
+#      "data": { "id": "vfu1",
+#                "device": "lsi1" },
+#      "timestamp": { "seconds": 1265044230, "microseconds": 450486 } }
+#
+##
+{ 'event': 'VFU_CLIENT_HANGUP',
+  'data': { 'id': 'str', 'device': 'str' } }
#19 lore raw

Re: [PATCH v8 10/17] vfio-user: run vfio-user context

Quoted 38 lines · Jag Raman
Quoted 29 lines · Markus Armbruster
Quoted 25 lines · Jagannathan Raman

Setup a handler to run vfio-user context. The context is driven by messages to the file descriptor associated with it - get the fd for the context and hook up the handler with it

---
qapi/misc.json | 23 ++++++++++
hw/remote/vfio-user-obj.c | 95 ++++++++++++++++++++++++++++++++++++++-

2 files changed, 117 insertions(+), 1 deletion(-)
qapi/misc.json +4 -0
index b83cc39029..f3cc4a4854 100644
@@ -553,3 +553,26 @@
 ##
 { 'event': 'RTC_CHANGE',
  'data': { 'offset': 'int', 'qom-path': 'str' } }
+
+##
+# @VFU_CLIENT_HANGUP:
+#
Quoted 8 lines

qapi/misc.json line 565

Quoted 6 lines
qapi/misc.json +6 -0
+# Emitted when the client of a TYPE_VFIO_USER_SERVER closes the
+# communication channel
+#
+# @id: ID of the TYPE_VFIO_USER_SERVER object
+#
+# @device: ID of attached PCI device

Is this the ID set with -device id=... and such?

Quoted 1 line

Yes, that is correct. It’s the ID set with the “-device id=…” option/

What happens when the device was added *without* id=...? DeviceState member @id is null then.

I figure we need to make @device optional here, present if the device has an ID. I recommend to also add a member @qom-path, like we did for MEMORY_DEVICE_SIZE_CHANGE in commit d89dd28f0e2.

author
#20 lore raw

Re: [PATCH v8 10/17] vfio-user: run vfio-user context

Quoted 44 lines · Markus Armbruster
Quoted 38 lines · Jag Raman
Quoted 37 lines · Markus Armbruster
Quoted 21 lines · Jagannathan Raman

Setup a handler to run vfio-user context. The context is driven by messages to the file descriptor associated with it - get the fd for the context and hook up the handler with it

---
qapi/misc.json | 23 ++++++++++
hw/remote/vfio-user-obj.c | 95 ++++++++++++++++++++++++++++++++++++++-

2 files changed, 117 insertions(+), 1 deletion(-)
qapi/misc.json +0 -0
index b83cc39029..f3cc4a4854 100644
@@ -553,3 +553,26 @@
 ##
 { 'event': 'RTC_CHANGE',
 'data': { 'offset': 'int', 'qom-path': 'str' } }

qapi/misc.json line 565

Quoted 10 lines
qapi/misc.json +10 -0
+
+##
+# @VFU_CLIENT_HANGUP:
+#
+# Emitted when the client of a TYPE_VFIO_USER_SERVER closes the
+# communication channel
+#
+# @id: ID of the TYPE_VFIO_USER_SERVER object
+#
+# @device: ID of attached PCI device

Is this the ID set with -device id=... and such?

Quoted 4 lines
Quoted 1 line

Yes, that is correct. It’s the ID set with the “-device id=…” option/

What happens when the device was added *without* id=...? DeviceState member @id is null then.

I’m sorry, the @id member in the event is the same as DeviceState->id, but it’s not directly derived from it. It derived from VfuObject->device, which is a property of TYPE_VFU_OBJECT that the user sets.

This will not be NULL, as TYPE_VFU_OBJECT will not initialize without it.

Quoted 4 lines

I figure we need to make @device optional here, present if the device has an ID. I recommend to also add a member @qom-path, like we did for MEMORY_DEVICE_SIZE_CHANGE in commit d89dd28f0e2.

OK, will add it.

Thank you!
--
Jag

Quoted 1 line
#21 lore raw

PATCH v8 11/17 vfio-user: handle PCI config space accesses

In reply to Jagannathan Raman

Define and register handlers for PCI config space accesses

---
 hw/remote/vfio-user-obj.c | 51 +++++++++++++++++++++++++++++++++++++++
 hw/remote/trace-events    |  2 ++
 2 files changed, 53 insertions(+)
hw/remote/vfio-user-obj.c +51 -0
index 06d99a8698..7b863dec4f 100644
@@ -47,6 +47,7 @@
 #include "qapi/qapi-events-misc.h"
 #include "qemu/notify.h"
 #include "qemu/thread.h"
+#include "qemu/main-loop.h"
 #include "sysemu/sysemu.h"
 #include "libvfio-user.h"
 #include "hw/qdev-core.h"
@@ -236,6 +237,45 @@ retry_attach:
     qemu_set_fd_handler(o->vfu_poll_fd, vfu_object_ctx_run, NULL, o);
 }
 
+static ssize_t vfu_object_cfg_access(vfu_ctx_t *vfu_ctx, char * const buf,
+                                     size_t count, loff_t offset,
+                                     const bool is_write)
+{
+    VfuObject *o = vfu_get_private(vfu_ctx);
+    uint32_t pci_access_width = sizeof(uint32_t);
+    size_t bytes = count;
+    uint32_t val = 0;
+    char *ptr = buf;
+    int len;
+
+    /*
+     * Writes to the BAR registers would trigger an update to the
+     * global Memory and IO AddressSpaces. But the remote device
+     * never uses the global AddressSpaces, therefore overlapping
+     * memory regions are not a problem
+     */
+    while (bytes > 0) {
+        len = (bytes > pci_access_width) ? pci_access_width : bytes;
+        if (is_write) {
+            memcpy(&val, ptr, len);
+            pci_host_config_write_common(o->pci_dev, offset,
+                                         pci_config_size(o->pci_dev),
+                                         val, len);
+            trace_vfu_cfg_write(offset, val);
+        } else {
+            val = pci_host_config_read_common(o->pci_dev, offset,
+                                              pci_config_size(o->pci_dev), len);
+            memcpy(ptr, &val, len);
+            trace_vfu_cfg_read(offset, val);
+        }
+        offset += len;
+        ptr += len;
+        bytes -= len;
+    }
+
+    return count;
+}
+
 /*
  * TYPE_VFU_OBJECT depends on the availability of the 'socket' and 'device'
  * properties. It also depends on devices instantiated in QEMU. These
@@ -314,6 +354,17 @@ static void vfu_object_init_ctx(VfuObject *o, Error **errp)
                TYPE_VFU_OBJECT, o->device);
     qdev_add_unplug_blocker(DEVICE(o->pci_dev), o->unplug_blocker);
 
+    ret = vfu_setup_region(o->vfu_ctx, VFU_PCI_DEV_CFG_REGION_IDX,
+                           pci_config_size(o->pci_dev), &vfu_object_cfg_access,
+                           VFU_REGION_FLAG_RW | VFU_REGION_FLAG_ALWAYS_CB,
+                           NULL, 0, -1, 0);
+    if (ret < 0) {
+        error_setg(errp,
+                   "vfu: Failed to setup config space handlers for %s- %s",
+                   o->device, strerror(errno));
+        goto fail;
+    }
+
     ret = vfu_realize_ctx(o->vfu_ctx);
     if (ret < 0) {
         error_setg(errp, "vfu: Failed to realize device %s- %s",
hw/remote/trace-events +2 -0
index 7da12f0d96..2ef7884346 100644
@@ -5,3 +5,5 @@ mpqemu_recv_io_error(int cmd, int size, int nfds) "failed to receive %d size %d,
 
 # vfio-user-obj.c
 vfu_prop(const char *prop, const char *val) "vfu: setting %s as %s"
+vfu_cfg_read(uint32_t offset, uint32_t val) "vfu: cfg: 0x%u -> 0x%x"
+vfu_cfg_write(uint32_t offset, uint32_t val) "vfu: cfg: 0x%u <- 0x%x"
Signature · 1 line
2.20.1
#22 lore raw

PATCH v8 12/17 vfio-user: IOMMU support for remote device

In reply to Jagannathan Raman

Assign separate address space for each device in the remote processes.

---
 include/hw/remote/iommu.h |  40 +++++++++++++
 hw/remote/iommu.c         | 114 ++++++++++++++++++++++++++++++++++++++
 hw/remote/machine.c       |  13 ++++-
 MAINTAINERS               |   2 +
 hw/remote/meson.build     |   1 +
 5 files changed, 169 insertions(+), 1 deletion(-)
 create mode 100644 include/hw/remote/iommu.h
 create mode 100644 hw/remote/iommu.c
include/hw/remote/iommu.h new file +40 -0
new file mode 100644
index 0000000000..33b68a8f4b
@@ -0,0 +1,40 @@
+/**
+ * Copyright © 2022 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef REMOTE_IOMMU_H
+#define REMOTE_IOMMU_H
+
+#include "hw/pci/pci_bus.h"
+#include "hw/pci/pci.h"
+
+#ifndef INT2VOIDP
+#define INT2VOIDP(i) (void *)(uintptr_t)(i)
+#endif
+
+typedef struct RemoteIommuElem {
+    MemoryRegion *mr;
+
+    AddressSpace as;
+} RemoteIommuElem;
+
+#define TYPE_REMOTE_IOMMU "x-remote-iommu"
+OBJECT_DECLARE_SIMPLE_TYPE(RemoteIommu, REMOTE_IOMMU)
+
+struct RemoteIommu {
+    Object parent;
+
+    GHashTable *elem_by_devfn;
+
+    QemuMutex lock;
+};
+
+void remote_iommu_setup(PCIBus *pci_bus);
+
+void remote_iommu_unplug_dev(PCIDevice *pci_dev);
+
+#endif
hw/remote/iommu.c new file +114 -0

114 lines not on this page · raw message

hw/remote/machine.c +12 -1
index ed91659794..cca5d25f50 100644
@@ -21,6 +21,7 @@
 #include "qapi/error.h"
 #include "hw/pci/pci_host.h"
 #include "hw/remote/iohub.h"
+#include "hw/remote/iommu.h"
 #include "hw/qdev-core.h"
 
 static void remote_machine_init(MachineState *machine)
@@ -100,6 +101,16 @@ static void remote_machine_instance_init(Object *obj)
     s->auto_shutdown = true;
 }
 
+static void remote_machine_dev_unplug_cb(HotplugHandler *hotplug_dev,
+                                         DeviceState *dev, Error **errp)
+{
+    qdev_unrealize(dev);
+
+    if (object_dynamic_cast(OBJECT(dev), TYPE_PCI_DEVICE)) {
+        remote_iommu_unplug_dev(PCI_DEVICE(dev));
+    }
+}
+
 static void remote_machine_class_init(ObjectClass *oc, void *data)
 {
     MachineClass *mc = MACHINE_CLASS(oc);
@@ -108,7 +119,7 @@ static void remote_machine_class_init(ObjectClass *oc, void *data)
     mc->init = remote_machine_init;
     mc->desc = "Experimental remote machine";
 
-    hc->unplug = qdev_simple_device_unplug_cb;
+    hc->unplug = remote_machine_dev_unplug_cb;
 
     object_class_property_add_bool(oc, "vfio-user",
                                    remote_machine_get_vfio_user,
MAINTAINERS +2 -0
index 37afdecc61..3e62ccab0a 100644
@@ -3599,6 +3599,8 @@ F: hw/remote/iohub.c
 F: include/hw/remote/iohub.h
 F: subprojects/libvfio-user
 F: hw/remote/vfio-user-obj.c
+F: hw/remote/iommu.c
+F: include/hw/remote/iommu.h
 
 EBPF:
 M: Jason Wang <[email protected]>
hw/remote/meson.build +1 -0
index 534ac5df79..bcef83c8cc 100644
@@ -6,6 +6,7 @@ remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('message.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('remote-obj.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('proxy.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('iohub.c'))
+remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('iommu.c'))
 remote_ss.add(when: 'CONFIG_VFIO_USER_SERVER', if_true: files('vfio-user-obj.c'))
 
 remote_ss.add(when: 'CONFIG_VFIO_USER_SERVER', if_true: vfiouser)
Signature · 1 line
2.20.1
author
#23 lore raw

Re: [PATCH v8 12/17] vfio-user: IOMMU support for remote device

Quoted 100 lines · Jag Raman

Assign separate address space for each device in the remote processes.

---

include/hw/remote/iommu.h |  40 +++++++++++++
hw/remote/iommu.c         | 114 ++++++++++++++++++++++++++++++++++++++
hw/remote/machine.c       |  13 ++++-
MAINTAINERS               |   2 +
hw/remote/meson.build     |   1 +
5 files changed, 169 insertions(+), 1 deletion(-)

create mode 100644 include/hw/remote/iommu.h
create mode 100644 hw/remote/iommu.c

include/hw/remote/iommu.h new file +40 -0
new file mode 100644
index 0000000000..33b68a8f4b
@@ -0,0 +1,40 @@
+/**
+ * Copyright © 2022 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef REMOTE_IOMMU_H
+#define REMOTE_IOMMU_H
+
+#include "hw/pci/pci_bus.h"
+#include "hw/pci/pci.h"
+
+#ifndef INT2VOIDP
+#define INT2VOIDP(i) (void *)(uintptr_t)(i)
+#endif
+
+typedef struct RemoteIommuElem {
+    MemoryRegion *mr;
+
+    AddressSpace as;
+} RemoteIommuElem;
+
+#define TYPE_REMOTE_IOMMU "x-remote-iommu"
+OBJECT_DECLARE_SIMPLE_TYPE(RemoteIommu, REMOTE_IOMMU)
+
+struct RemoteIommu {
+    Object parent;
+
+    GHashTable *elem_by_devfn;
+
+    QemuMutex lock;
+};
+
+void remote_iommu_setup(PCIBus *pci_bus);
+
+void remote_iommu_unplug_dev(PCIDevice *pci_dev);
+
+#endif
hw/remote/iommu.c new file +29 -0
new file mode 100644
index 0000000000..16c6b0834e
@@ -0,0 +1,114 @@
+/**
+ * IOMMU for remote device
+ *
+ * Copyright © 2022 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#include "qemu/osdep.h"
+#include "qemu-common.h"
+
+#include "hw/remote/iommu.h"
+#include "hw/pci/pci_bus.h"
+#include "hw/pci/pci.h"
+#include "exec/memory.h"
+#include "exec/address-spaces.h"
+#include "trace.h"
+
+static AddressSpace *remote_iommu_find_add_as(PCIBus *pci_bus,
+                                              void *opaque, int devfn)
+{
+    RemoteIommu *iommu = opaque;
+    RemoteIommuElem *elem = NULL;
+
+    qemu_mutex_lock(&iommu->lock);
+
+    elem = g_hash_table_lookup(iommu->elem_by_devfn, INT2VOIDP(devfn));

hw/remote/iommu.c line 39

Quoted 10 lines
hw/remote/iommu.c new file +10 -0
+
+    if (!elem) {
+        elem = g_malloc0(sizeof(RemoteIommuElem));
+        g_hash_table_insert(iommu->elem_by_devfn, INT2VOIDP(devfn), elem);
+    }
+
+    if (!elem->mr) {
+        elem->mr = MEMORY_REGION(object_new(TYPE_MEMORY_REGION));
+        memory_region_set_size(elem->mr, UINT64_MAX);
+        address_space_init(&elem->as, elem->mr, NULL);

Hi,

I’d like to add a note here.

We tried to add "elem->mr” as a child of PCIDevice. That way, when PCIDevice is unplugged, the child is also finalized.

However, there was some issue with hotplug. During the hotplug, there’s a window during initialization where we couldn’t lookup the PCIDevice by “devfn”.

do_pci_register_device() -> pci_init_bus_master() -> pci_device_iommu_address_space() happens before do_pci_register_device() -> “bus->devices[devfn] = pci_dev”. As such, pci_find_device() doesn’t work at this time.

Thank you!
--
Jag

Quoted 141 lines
hw/remote/iommu.c new file +75 -0
+    }
+
+    qemu_mutex_unlock(&iommu->lock);
+
+    return &elem->as;
+}
+
+void remote_iommu_unplug_dev(PCIDevice *pci_dev)
+{
+    AddressSpace *as = pci_device_iommu_address_space(pci_dev);
+    RemoteIommuElem *elem = NULL;
+
+    if (as == &address_space_memory) {
+        return;
+    }
+
+    elem = container_of(as, RemoteIommuElem, as);
+
+    address_space_destroy(&elem->as);
+
+    object_unref(elem->mr);
+
+    elem->mr = NULL;
+}
+
+static void remote_iommu_init(Object *obj)
+{
+    RemoteIommu *iommu = REMOTE_IOMMU(obj);
+
+    iommu->elem_by_devfn = g_hash_table_new_full(NULL, NULL, NULL, g_free);
+
+    qemu_mutex_init(&iommu->lock);
+}
+
+static void remote_iommu_finalize(Object *obj)
+{
+    RemoteIommu *iommu = REMOTE_IOMMU(obj);
+
+    qemu_mutex_destroy(&iommu->lock);
+
+    if (iommu->elem_by_devfn) {
+        g_hash_table_destroy(iommu->elem_by_devfn);
+        iommu->elem_by_devfn = NULL;
+    }
+}
+
+void remote_iommu_setup(PCIBus *pci_bus)
+{
+    RemoteIommu *iommu = NULL;
+
+    g_assert(pci_bus);
+
+    iommu = REMOTE_IOMMU(object_new(TYPE_REMOTE_IOMMU));
+
+    pci_setup_iommu(pci_bus, remote_iommu_find_add_as, iommu);
+
+    object_property_add_child(OBJECT(pci_bus), "remote-iommu", OBJECT(iommu));
+
+    object_unref(OBJECT(iommu));
+}
+
+static const TypeInfo remote_iommu_info = {
+    .name = TYPE_REMOTE_IOMMU,
+    .parent = TYPE_OBJECT,
+    .instance_size = sizeof(RemoteIommu),
+    .instance_init = remote_iommu_init,
+    .instance_finalize = remote_iommu_finalize,
+};
+
+static void remote_iommu_register_types(void)
+{
+    type_register_static(&remote_iommu_info);
+}
+
+type_init(remote_iommu_register_types)
hw/remote/machine.c +12 -1
index ed91659794..cca5d25f50 100644
@@ -21,6 +21,7 @@
 #include "qapi/error.h"
 #include "hw/pci/pci_host.h"
 #include "hw/remote/iohub.h"
+#include "hw/remote/iommu.h"
 #include "hw/qdev-core.h"
 
 static void remote_machine_init(MachineState *machine)
@@ -100,6 +101,16 @@ static void remote_machine_instance_init(Object *obj)
    s->auto_shutdown = true;
 }
 
+static void remote_machine_dev_unplug_cb(HotplugHandler *hotplug_dev,
+                                         DeviceState *dev, Error **errp)
+{
+    qdev_unrealize(dev);
+
+    if (object_dynamic_cast(OBJECT(dev), TYPE_PCI_DEVICE)) {
+        remote_iommu_unplug_dev(PCI_DEVICE(dev));
+    }
+}
+
 static void remote_machine_class_init(ObjectClass *oc, void *data)
 {
    MachineClass *mc = MACHINE_CLASS(oc);
@@ -108,7 +119,7 @@ static void remote_machine_class_init(ObjectClass *oc, void *data)
    mc->init = remote_machine_init;
    mc->desc = "Experimental remote machine";
 
-    hc->unplug = qdev_simple_device_unplug_cb;
+    hc->unplug = remote_machine_dev_unplug_cb;
 
    object_class_property_add_bool(oc, "vfio-user",
                                   remote_machine_get_vfio_user,
MAINTAINERS +0 -0
index 37afdecc61..3e62ccab0a 100644
@@ -3599,6 +3599,8 @@ F: hw/remote/iohub.c

F: include/hw/remote/iohub.h
F: subprojects/libvfio-user
F: hw/remote/vfio-user-obj.c

MAINTAINERS +2 -0
+F: hw/remote/iommu.c
+F: include/hw/remote/iommu.h

EBPF:
M: Jason Wang <[email protected]>

hw/remote/meson.build +1 -0
index 534ac5df79..bcef83c8cc 100644
@@ -6,6 +6,7 @@ remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('message.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('remote-obj.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('proxy.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('iohub.c'))
+remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('iommu.c'))
 remote_ss.add(when: 'CONFIG_VFIO_USER_SERVER', if_true: files('vfio-user-obj.c'))
 
 remote_ss.add(when: 'CONFIG_VFIO_USER_SERVER', if_true: vfiouser)
Signature · 1 line
2.20.1
#24 lore raw

Re: [PATCH v8 12/17] vfio-user: IOMMU support for remote device

Quoted 113 lines · Jag Raman
Quoted 100 lines · Jag Raman

Assign separate address space for each device in the remote processes.

---

include/hw/remote/iommu.h |  40 +++++++++++++
hw/remote/iommu.c         | 114 ++++++++++++++++++++++++++++++++++++++
hw/remote/machine.c       |  13 ++++-
MAINTAINERS               |   2 +
hw/remote/meson.build     |   1 +
5 files changed, 169 insertions(+), 1 deletion(-)

create mode 100644 include/hw/remote/iommu.h
create mode 100644 hw/remote/iommu.c

include/hw/remote/iommu.h new file +40 -0
new file mode 100644
index 0000000000..33b68a8f4b
@@ -0,0 +1,40 @@
+/**
+ * Copyright © 2022 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef REMOTE_IOMMU_H
+#define REMOTE_IOMMU_H
+
+#include "hw/pci/pci_bus.h"
+#include "hw/pci/pci.h"
+
+#ifndef INT2VOIDP
+#define INT2VOIDP(i) (void *)(uintptr_t)(i)
+#endif
+
+typedef struct RemoteIommuElem {
+    MemoryRegion *mr;
+
+    AddressSpace as;
+} RemoteIommuElem;
+
+#define TYPE_REMOTE_IOMMU "x-remote-iommu"
+OBJECT_DECLARE_SIMPLE_TYPE(RemoteIommu, REMOTE_IOMMU)
+
+struct RemoteIommu {
+    Object parent;
+
+    GHashTable *elem_by_devfn;
+
+    QemuMutex lock;
+};
+
+void remote_iommu_setup(PCIBus *pci_bus);
+
+void remote_iommu_unplug_dev(PCIDevice *pci_dev);
+
+#endif
hw/remote/iommu.c new file +29 -0
new file mode 100644
index 0000000000..16c6b0834e
@@ -0,0 +1,114 @@
+/**
+ * IOMMU for remote device
+ *
+ * Copyright © 2022 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#include "qemu/osdep.h"
+#include "qemu-common.h"
+
+#include "hw/remote/iommu.h"
+#include "hw/pci/pci_bus.h"
+#include "hw/pci/pci.h"
+#include "exec/memory.h"
+#include "exec/address-spaces.h"
+#include "trace.h"
+
+static AddressSpace *remote_iommu_find_add_as(PCIBus *pci_bus,
+                                              void *opaque, int devfn)
+{
+    RemoteIommu *iommu = opaque;
+    RemoteIommuElem *elem = NULL;
+
+    qemu_mutex_lock(&iommu->lock);
+
+    elem = g_hash_table_lookup(iommu->elem_by_devfn, INT2VOIDP(devfn));

hw/remote/iommu.c line 39

Quoted 10 lines
hw/remote/iommu.c new file +10 -0
+
+    if (!elem) {
+        elem = g_malloc0(sizeof(RemoteIommuElem));
+        g_hash_table_insert(iommu->elem_by_devfn, INT2VOIDP(devfn), elem);
+    }
+
+    if (!elem->mr) {
+        elem->mr = MEMORY_REGION(object_new(TYPE_MEMORY_REGION));
+        memory_region_set_size(elem->mr, UINT64_MAX);
+        address_space_init(&elem->as, elem->mr, NULL);

Hi,

Quoted 4 lines

I’d like to add a note here.

We tried to add "elem->mr” as a child of PCIDevice. That way, when PCIDevice is unplugged, the child is also finalized.

Do you mean via a memory_region_init()-family function where a parent object is given? Or do you mean by adding a QOM child property?

Quoted 6 lines

However, there was some issue with hotplug. During the hotplug, there’s a window during initialization where we couldn’t lookup the PCIDevice by “devfn”.

do_pci_register_device() -> pci_init_bus_master() -> pci_device_iommu_address_space() happens before do_pci_register_device() -> “bus->devices[devfn] = pci_dev”. As such, pci_find_device() doesn’t work at this time.

I don't follow. What calls pci_find_device()?

Stefan

author
#25 lore raw

Re: [PATCH v8 12/17] vfio-user: IOMMU support for remote device

Quoted 118 lines · Stefan Hajnoczi
Quoted 113 lines · Jag Raman
Quoted 100 lines · Jag Raman

Assign separate address space for each device in the remote processes.

---

include/hw/remote/iommu.h |  40 +++++++++++++
hw/remote/iommu.c         | 114 ++++++++++++++++++++++++++++++++++++++
hw/remote/machine.c       |  13 ++++-
MAINTAINERS               |   2 +
hw/remote/meson.build     |   1 +
5 files changed, 169 insertions(+), 1 deletion(-)

create mode 100644 include/hw/remote/iommu.h
create mode 100644 hw/remote/iommu.c

include/hw/remote/iommu.h new file +40 -0
new file mode 100644
index 0000000000..33b68a8f4b
@@ -0,0 +1,40 @@
+/**
+ * Copyright © 2022 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef REMOTE_IOMMU_H
+#define REMOTE_IOMMU_H
+
+#include "hw/pci/pci_bus.h"
+#include "hw/pci/pci.h"
+
+#ifndef INT2VOIDP
+#define INT2VOIDP(i) (void *)(uintptr_t)(i)
+#endif
+
+typedef struct RemoteIommuElem {
+    MemoryRegion *mr;
+
+    AddressSpace as;
+} RemoteIommuElem;
+
+#define TYPE_REMOTE_IOMMU "x-remote-iommu"
+OBJECT_DECLARE_SIMPLE_TYPE(RemoteIommu, REMOTE_IOMMU)
+
+struct RemoteIommu {
+    Object parent;
+
+    GHashTable *elem_by_devfn;
+
+    QemuMutex lock;
+};
+
+void remote_iommu_setup(PCIBus *pci_bus);
+
+void remote_iommu_unplug_dev(PCIDevice *pci_dev);
+
+#endif
hw/remote/iommu.c new file +29 -0
new file mode 100644
index 0000000000..16c6b0834e
@@ -0,0 +1,114 @@
+/**
+ * IOMMU for remote device
+ *
+ * Copyright © 2022 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#include "qemu/osdep.h"
+#include "qemu-common.h"
+
+#include "hw/remote/iommu.h"
+#include "hw/pci/pci_bus.h"
+#include "hw/pci/pci.h"
+#include "exec/memory.h"
+#include "exec/address-spaces.h"
+#include "trace.h"
+
+static AddressSpace *remote_iommu_find_add_as(PCIBus *pci_bus,
+                                              void *opaque, int devfn)
+{
+    RemoteIommu *iommu = opaque;
+    RemoteIommuElem *elem = NULL;
+
+    qemu_mutex_lock(&iommu->lock);
+
+    elem = g_hash_table_lookup(iommu->elem_by_devfn, INT2VOIDP(devfn));

hw/remote/iommu.c line 39

Quoted 10 lines
hw/remote/iommu.c new file +10 -0
+
+    if (!elem) {
+        elem = g_malloc0(sizeof(RemoteIommuElem));
+        g_hash_table_insert(iommu->elem_by_devfn, INT2VOIDP(devfn), elem);
+    }
+
+    if (!elem->mr) {
+        elem->mr = MEMORY_REGION(object_new(TYPE_MEMORY_REGION));
+        memory_region_set_size(elem->mr, UINT64_MAX);
+        address_space_init(&elem->as, elem->mr, NULL);

Hi,

Quoted 7 lines
Quoted 4 lines

I’d like to add a note here.

We tried to add "elem->mr” as a child of PCIDevice. That way, when PCIDevice is unplugged, the child is also finalized.

Do you mean via a memory_region_init()-family function where a parent object is given? Or do you mean by adding a QOM child property?

I mean by adding “elem->mr” as a QOM child property of PCIDevice.

Quoted 9 lines
Quoted 7 lines

However, there was some issue with hotplug. During the hotplug, there’s a window during initialization where we couldn’t lookup the PCIDevice by “devfn”.

do_pci_register_device() -> pci_init_bus_master() -> pci_device_iommu_address_space() happens before do_pci_register_device() -> “bus->devices[devfn] = pci_dev”. As such, pci_find_device() doesn’t work at this time.

I don't follow. What calls pci_find_device()?

To add the MemoryRegion as a child of PCIDevice, remote_iommu_find_add_as() would need to lookup the PCIDevice using devfn. The function that looks up PCIDevice by devfn is pci_find_device().

The above note explains why we didn’t lookup the PCIDevice using pci_find_device() and then adding the MemoryRegion as its child.

Thank you!
--
Jag

Quoted 2 lines

Stefan

#26 lore raw

Re: [PATCH v8 12/17] vfio-user: IOMMU support for remote device

Quoted 136 lines · Jag Raman
Quoted 118 lines · Stefan Hajnoczi
Quoted 113 lines · Jag Raman
Quoted 100 lines · Jag Raman

Assign separate address space for each device in the remote processes.

---

include/hw/remote/iommu.h |  40 +++++++++++++
hw/remote/iommu.c         | 114 ++++++++++++++++++++++++++++++++++++++
hw/remote/machine.c       |  13 ++++-
MAINTAINERS               |   2 +
hw/remote/meson.build     |   1 +
5 files changed, 169 insertions(+), 1 deletion(-)

create mode 100644 include/hw/remote/iommu.h
create mode 100644 hw/remote/iommu.c

include/hw/remote/iommu.h new file +40 -0
new file mode 100644
index 0000000000..33b68a8f4b
@@ -0,0 +1,40 @@
+/**
+ * Copyright © 2022 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef REMOTE_IOMMU_H
+#define REMOTE_IOMMU_H
+
+#include "hw/pci/pci_bus.h"
+#include "hw/pci/pci.h"
+
+#ifndef INT2VOIDP
+#define INT2VOIDP(i) (void *)(uintptr_t)(i)
+#endif
+
+typedef struct RemoteIommuElem {
+    MemoryRegion *mr;
+
+    AddressSpace as;
+} RemoteIommuElem;
+
+#define TYPE_REMOTE_IOMMU "x-remote-iommu"
+OBJECT_DECLARE_SIMPLE_TYPE(RemoteIommu, REMOTE_IOMMU)
+
+struct RemoteIommu {
+    Object parent;
+
+    GHashTable *elem_by_devfn;
+
+    QemuMutex lock;
+};
+
+void remote_iommu_setup(PCIBus *pci_bus);
+
+void remote_iommu_unplug_dev(PCIDevice *pci_dev);
+
+#endif
hw/remote/iommu.c new file +29 -0
new file mode 100644
index 0000000000..16c6b0834e
@@ -0,0 +1,114 @@
+/**
+ * IOMMU for remote device
+ *
+ * Copyright © 2022 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#include "qemu/osdep.h"
+#include "qemu-common.h"
+
+#include "hw/remote/iommu.h"
+#include "hw/pci/pci_bus.h"
+#include "hw/pci/pci.h"
+#include "exec/memory.h"
+#include "exec/address-spaces.h"
+#include "trace.h"
+
+static AddressSpace *remote_iommu_find_add_as(PCIBus *pci_bus,
+                                              void *opaque, int devfn)
+{
+    RemoteIommu *iommu = opaque;
+    RemoteIommuElem *elem = NULL;
+
+    qemu_mutex_lock(&iommu->lock);
+
+    elem = g_hash_table_lookup(iommu->elem_by_devfn, INT2VOIDP(devfn));

hw/remote/iommu.c line 39

Quoted 10 lines
hw/remote/iommu.c new file +10 -0
+
+    if (!elem) {
+        elem = g_malloc0(sizeof(RemoteIommuElem));
+        g_hash_table_insert(iommu->elem_by_devfn, INT2VOIDP(devfn), elem);
+    }
+
+    if (!elem->mr) {
+        elem->mr = MEMORY_REGION(object_new(TYPE_MEMORY_REGION));
+        memory_region_set_size(elem->mr, UINT64_MAX);
+        address_space_init(&elem->as, elem->mr, NULL);

Hi,

Quoted 7 lines
Quoted 4 lines

I’d like to add a note here.

We tried to add "elem->mr” as a child of PCIDevice. That way, when PCIDevice is unplugged, the child is also finalized.

Do you mean via a memory_region_init()-family function where a parent object is given? Or do you mean by adding a QOM child property?

I mean by adding “elem->mr” as a QOM child property of PCIDevice.

Quoted 8 lines
Quoted 7 lines

However, there was some issue with hotplug. During the hotplug, there’s a window during initialization where we couldn’t lookup the PCIDevice by “devfn”.

do_pci_register_device() -> pci_init_bus_master() -> pci_device_iommu_address_space() happens before do_pci_register_device() -> “bus->devices[devfn] = pci_dev”. As such, pci_find_device() doesn’t work at this time.

Quoted 8 lines
Quoted 1 line

I don't follow. What calls pci_find_device()?

To add the MemoryRegion as a child of PCIDevice, remote_iommu_find_add_as() would need to lookup the PCIDevice using devfn. The function that looks up PCIDevice by devfn is pci_find_device().

The above note explains why we didn’t lookup the PCIDevice using pci_find_device() and then adding the MemoryRegion as its child.

If I understand correctly you're saying it's not possible to use memory_region_init(&elem->mr, OBJECT(pci_dev), ...) inside remote_iommu_find_add_as() because there is no way to find the PCIDevice?

It would be nice to automatically clean up the memory region but the AddressSpace needs to be destroyed too and there isn't an automatic way of doing that, so the approach in this patch is fine.

Stefan

#27 lore raw

Re: [PATCH v8 12/17] vfio-user: IOMMU support for remote device

In reply to Jagannathan Raman
Quoted 15 lines · Jagannathan Raman
hw/remote/iommu.c new file +15 -0
+static AddressSpace *remote_iommu_find_add_as(PCIBus *pci_bus,
+                                              void *opaque, int devfn)
+{
+    RemoteIommu *iommu = opaque;
+    RemoteIommuElem *elem = NULL;
+
+    qemu_mutex_lock(&iommu->lock);
+
+    elem = g_hash_table_lookup(iommu->elem_by_devfn, INT2VOIDP(devfn));
+
+    if (!elem) {
+        elem = g_malloc0(sizeof(RemoteIommuElem));
+        g_hash_table_insert(iommu->elem_by_devfn, INT2VOIDP(devfn), elem);
+    }
+

hw/remote/iommu.c line 45

Quoted 10 lines
hw/remote/iommu.c new file +10 -0
+    if (!elem->mr) {
+        elem->mr = MEMORY_REGION(object_new(TYPE_MEMORY_REGION));
+        memory_region_set_size(elem->mr, UINT64_MAX);
+        address_space_init(&elem->as, elem->mr, NULL);
+    }
+
+    qemu_mutex_unlock(&iommu->lock);
+
+    return &elem->as;
+}

A few comments that can be added to this file to explain the design:

- Each vfio-user server handles one PCIDevice on a PCIBus. There is one RemoteIommu per PCIBus, so the RemoteIommu tracks multiple PCIDevices by maintaining a ->elem_by_devfn mapping.

- memory_region_init_iommu() is not used because vfio-user MemoryRegions will be added to the elem->mr container instead. This is more natural than implementing the IOMMUMemoryRegionClass APIs since vfio-user provides something that is close to a full-fledged MemoryRegion and not like an IOMMU mapping.

- When a device is hot unplugged, the elem->mr reference is dropped so all vfio-user MemoryRegions associated with this vfio-user server are destroyed.

hw/remote/iommu.c line 80

Quoted 7 lines
hw/remote/iommu.c new file +7 -0
+static void remote_iommu_finalize(Object *obj)
+{
+    RemoteIommu *iommu = REMOTE_IOMMU(obj);
+
+    qemu_mutex_destroy(&iommu->lock);
+
+    if (iommu->elem_by_devfn) {

->init() and ->finalize() are a pair, so I don't think ->finalize() will ever be called with a NULL ->elem_by_devfn.

If ->elem_by_devfn can be NULL then there would probably need to be a check around qemu_mutex_destroy(&iommu->lock) too.

Quoted 4 lines
hw/remote/iommu.c new file +4 -0
+        g_hash_table_destroy(iommu->elem_by_devfn);
+        iommu->elem_by_devfn = NULL;
+    }
+}
author
#28 lore raw

Re: [PATCH v8 12/17] vfio-user: IOMMU support for remote device

Quoted 43 lines · Stefan Hajnoczi
Quoted 15 lines · Jagannathan Raman
hw/remote/iommu.c new file +15 -0
+static AddressSpace *remote_iommu_find_add_as(PCIBus *pci_bus,
+                                              void *opaque, int devfn)
+{
+    RemoteIommu *iommu = opaque;
+    RemoteIommuElem *elem = NULL;
+
+    qemu_mutex_lock(&iommu->lock);
+
+    elem = g_hash_table_lookup(iommu->elem_by_devfn, INT2VOIDP(devfn));
+
+    if (!elem) {
+        elem = g_malloc0(sizeof(RemoteIommuElem));
+        g_hash_table_insert(iommu->elem_by_devfn, INT2VOIDP(devfn), elem);
+    }
+

hw/remote/iommu.c line 45

Quoted 10 lines
hw/remote/iommu.c new file +10 -0
+    if (!elem->mr) {
+        elem->mr = MEMORY_REGION(object_new(TYPE_MEMORY_REGION));
+        memory_region_set_size(elem->mr, UINT64_MAX);
+        address_space_init(&elem->as, elem->mr, NULL);
+    }
+
+    qemu_mutex_unlock(&iommu->lock);
+
+    return &elem->as;
+}

A few comments that can be added to this file to explain the design:

- Each vfio-user server handles one PCIDevice on a PCIBus. There is one RemoteIommu per PCIBus, so the RemoteIommu tracks multiple PCIDevices by maintaining a ->elem_by_devfn mapping.

- memory_region_init_iommu() is not used because vfio-user MemoryRegions will be added to the elem->mr container instead. This is more natural than implementing the IOMMUMemoryRegionClass APIs since vfio-user provides something that is close to a full-fledged MemoryRegion and not like an IOMMU mapping.

Quoted 3 lines

- When a device is hot unplugged, the elem->mr reference is dropped so all vfio-user MemoryRegions associated with this vfio-user server are destroyed.

OK, will add this comment.

Quoted 11 lines

hw/remote/iommu.c line 80

Quoted 8 lines
hw/remote/iommu.c new file +7 -0
+static void remote_iommu_finalize(Object *obj)
+{
+    RemoteIommu *iommu = REMOTE_IOMMU(obj);
+
+    qemu_mutex_destroy(&iommu->lock);
+
+    if (iommu->elem_by_devfn) {

->init() and ->finalize() are a pair, so I don't think ->finalize() will ever be called with a NULL ->elem_by_devfn.

OK, got it.

Thanks!
--
Jag

Quoted 8 lines

If ->elem_by_devfn can be NULL then there would probably need to be a check around qemu_mutex_destroy(&iommu->lock) too.

Quoted 4 lines
hw/remote/iommu.c new file +4 -0
+        g_hash_table_destroy(iommu->elem_by_devfn);
+        iommu->elem_by_devfn = NULL;
+    }
+}
#29 lore raw

PATCH v8 13/17 vfio-user: handle DMA mappings

In reply to Jagannathan Raman

Define and register callbacks to manage the RAM regions used for device DMA

---
 hw/remote/machine.c       |  5 ++++
 hw/remote/vfio-user-obj.c | 55 +++++++++++++++++++++++++++++++++++++++
 hw/remote/trace-events    |  2 ++
 3 files changed, 62 insertions(+)
hw/remote/machine.c +5 -0
index cca5d25f50..7002d46980 100644
@@ -23,6 +23,7 @@
 #include "hw/remote/iohub.h"
 #include "hw/remote/iommu.h"
 #include "hw/qdev-core.h"
+#include "hw/remote/iommu.h"
 
 static void remote_machine_init(MachineState *machine)
 {
@@ -52,6 +53,10 @@ static void remote_machine_init(MachineState *machine)
 
     pci_host = PCI_HOST_BRIDGE(rem_host);
 
+    if (s->vfio_user) {
+        remote_iommu_setup(pci_host->bus);
+    }
+
     remote_iohub_init(&s->iohub);
 
     pci_bus_irqs(pci_host->bus, remote_iohub_set_irq, remote_iohub_map_irq,
hw/remote/vfio-user-obj.c +55 -0

67 lines not on this page · raw message

hw/remote/trace-events +2 -0
index 2ef7884346..f945c7e33b 100644
@@ -7,3 +7,5 @@ mpqemu_recv_io_error(int cmd, int size, int nfds) "failed to receive %d size %d,
 vfu_prop(const char *prop, const char *val) "vfu: setting %s as %s"
 vfu_cfg_read(uint32_t offset, uint32_t val) "vfu: cfg: 0x%u -> 0x%x"
 vfu_cfg_write(uint32_t offset, uint32_t val) "vfu: cfg: 0x%u <- 0x%x"
+vfu_dma_register(uint64_t gpa, size_t len) "vfu: registering GPA 0x%"PRIx64", %zu bytes"
+vfu_dma_unregister(uint64_t gpa) "vfu: unregistering GPA 0x%"PRIx64""
Signature · 1 line
2.20.1
#30 lore raw

Re: [PATCH v8 13/17] vfio-user: handle DMA mappings

Quoted 7 lines · Jagannathan Raman
hw/remote/vfio-user-obj.c +7 -0
+static void dma_unregister(vfu_ctx_t *vfu_ctx, vfu_dma_info_t *info)
+{
+    VfuObject *o = vfu_get_private(vfu_ctx);
+    AddressSpace *dma_as = NULL;
+    MemoryRegion *mr = NULL;
+    ram_addr_t offset;
+

hw/remote/vfio-user-obj.c static ssize_t vfu_object_cfg_access(vfu_ctx_t *vfu_ctx, char * const buf, · line 322

Quoted 10 lines
hw/remote/vfio-user-obj.c +10 -0
+    mr = memory_region_from_host(info->vaddr, &offset);
+    if (!mr) {
+        return;
+    }
+
+    dma_as = pci_device_iommu_address_space(o->pci_dev);
+
+    memory_region_del_subregion(dma_as->root, mr);
+
+    object_unparent((OBJECT(mr)));

Where is obj->parent set?

If it is not set then this call is a nop and mr is not freed:

  void object_unparent(Object *obj)
  {
      if (obj->parent) {
          object_property_del_child(obj->parent, obj);
      }
  }
author
#31 lore raw

Re: [PATCH v8 13/17] vfio-user: handle DMA mappings

Quoted 16 lines · Stefan Hajnoczi
Quoted 11 lines · Jagannathan Raman
hw/remote/vfio-user-obj.c +11 -0
+static void dma_unregister(vfu_ctx_t *vfu_ctx, vfu_dma_info_t *info)
+{
+    VfuObject *o = vfu_get_private(vfu_ctx);
+    AddressSpace *dma_as = NULL;
+    MemoryRegion *mr = NULL;
+    ram_addr_t offset;
+
+    mr = memory_region_from_host(info->vaddr, &offset);
+    if (!mr) {
+        return;
+    }
Quoted 8 lines

hw/remote/vfio-user-obj.c static ssize_t vfu_object_cfg_access(vfu_ctx_t *vfu_ctx, char * const buf, · line 322

Quoted 6 lines
hw/remote/vfio-user-obj.c +6 -0
+
+    dma_as = pci_device_iommu_address_space(o->pci_dev);
+
+    memory_region_del_subregion(dma_as->root, mr);
+
+    object_unparent((OBJECT(mr)));

Where is obj->parent set?

Yeah, it should be object_unref().

Thank you!
--
Jag

Quoted 9 lines

If it is not set then this call is a nop and mr is not freed:

void object_unparent(Object *obj)

 {
     if (obj->parent) {
         object_property_del_child(obj->parent, obj);
     }

}

author
#32 lore raw

Re: [PATCH v8 13/17] vfio-user: handle DMA mappings

Quoted 29 lines · Jag Raman
Quoted 17 lines · Stefan Hajnoczi
Quoted 11 lines · Jagannathan Raman
hw/remote/vfio-user-obj.c +11 -0
+static void dma_unregister(vfu_ctx_t *vfu_ctx, vfu_dma_info_t *info)
+{
+ VfuObject *o = vfu_get_private(vfu_ctx);
+ AddressSpace *dma_as = NULL;
+ MemoryRegion *mr = NULL;
+ ram_addr_t offset;
+
+ mr = memory_region_from_host(info->vaddr, &offset);
+ if (!mr) {
+ return;
+ }
Quoted 8 lines

hw/remote/vfio-user-obj.c static ssize_t vfu_object_cfg_access(vfu_ctx_t *vfu_ctx, char * const buf, · line 322

Quoted 6 lines
hw/remote/vfio-user-obj.c +6 -0
+
+ dma_as = pci_device_iommu_address_space(o->pci_dev);
+
+ memory_region_del_subregion(dma_as->root, mr);
+
+ object_unparent((OBJECT(mr)));

Where is obj->parent set?

Quoted 1 line

Yeah, it should be object_unref().

I think object_unparent() is the appropriate way to finalize the MemoryRegion in this case.

‘mr’ is an owner-less MemoryRegion created by dma_register(). owner-less MemoryRegions initialized using memory_region_init* functions are children of the “/unattached” object.

The parent is set by dma_register() -> memory_region_init_ram_ptr() -> memory_region_init() -> memory_region_do_init() -> object_property_add_child() -> object_property_try_add_child().

Thank you!
--
Jag

Quoted 14 lines

Thank you!
--
Jag

Quoted 9 lines

If it is not set then this call is a nop and mr is not freed:

void object_unparent(Object *obj)
{
if (obj->parent) {
object_property_del_child(obj->parent, obj);
}
}

#33 lore raw

PATCH v8 14/17 vfio-user: handle PCI BAR accesses

In reply to Jagannathan Raman

Determine the BARs used by the PCI device and register handlers to manage the access to the same.

---
 include/exec/memory.h           |   3 +
 hw/remote/vfio-user-obj.c       | 189 ++++++++++++++++++++++++++++++++
 softmmu/physmem.c               |   4 +-
 tests/qtest/fuzz/generic_fuzz.c |   9 +-
 hw/remote/trace-events          |   3 +
 5 files changed, 202 insertions(+), 6 deletions(-)
include/exec/memory.h +3 -0
index 4d5997e6bb..4b061e62d5 100644
@@ -2810,6 +2810,9 @@ MemTxResult address_space_write_cached_slow(MemoryRegionCache *cache,
                                             hwaddr addr, const void *buf,
                                             hwaddr len);
 
+int memory_access_size(MemoryRegion *mr, unsigned l, hwaddr addr);
+bool prepare_mmio_access(MemoryRegion *mr);
+
 static inline bool memory_access_is_direct(MemoryRegion *mr, bool is_write)
 {
     if (is_write) {
hw/remote/vfio-user-obj.c +189 -0

207 lines not on this page · raw message

softmmu/physmem.c +2 -2
index 4e1b27a20e..f9a68d1fe5 100644
@@ -2719,7 +2719,7 @@ void memory_region_flush_rom_device(MemoryRegion *mr, hwaddr addr, hwaddr size)
     invalidate_and_set_dirty(mr, addr, size);
 }
 
-static int memory_access_size(MemoryRegion *mr, unsigned l, hwaddr addr)
+int memory_access_size(MemoryRegion *mr, unsigned l, hwaddr addr)
 {
     unsigned access_size_max = mr->ops->valid.max_access_size;
 
@@ -2746,7 +2746,7 @@ static int memory_access_size(MemoryRegion *mr, unsigned l, hwaddr addr)
     return l;
 }
 
-static bool prepare_mmio_access(MemoryRegion *mr)
+bool prepare_mmio_access(MemoryRegion *mr)
 {
     bool release_lock = false;
 
tests/qtest/fuzz/generic_fuzz.c +5 -4
index dd7e25851c..77547fc1d8 100644
@@ -144,7 +144,7 @@ static void *pattern_alloc(pattern p, size_t len)
     return buf;
 }
 
-static int memory_access_size(MemoryRegion *mr, unsigned l, hwaddr addr)
+static int fuzz_memory_access_size(MemoryRegion *mr, unsigned l, hwaddr addr)
 {
     unsigned access_size_max = mr->ops->valid.max_access_size;
 
@@ -242,11 +242,12 @@ void fuzz_dma_read_cb(size_t addr, size_t len, MemoryRegion *mr)
 
         /*
          *  If mr1 isn't RAM, address_space_translate doesn't update l. Use
-         *  memory_access_size to identify the number of bytes that it is safe
-         *  to write without accidentally writing to another MemoryRegion.
+         *  fuzz_memory_access_size to identify the number of bytes that it
+         *  is safe to write without accidentally writing to another
+         *  MemoryRegion.
          */
         if (!memory_region_is_ram(mr1)) {
-            l = memory_access_size(mr1, l, addr1);
+            l = fuzz_memory_access_size(mr1, l, addr1);
         }
         if (memory_region_is_ram(mr1) ||
             memory_region_is_romd(mr1) ||
hw/remote/trace-events +3 -0
index f945c7e33b..847d50d88f 100644
@@ -9,3 +9,6 @@ vfu_cfg_read(uint32_t offset, uint32_t val) "vfu: cfg: 0x%u -> 0x%x"
 vfu_cfg_write(uint32_t offset, uint32_t val) "vfu: cfg: 0x%u <- 0x%x"
 vfu_dma_register(uint64_t gpa, size_t len) "vfu: registering GPA 0x%"PRIx64", %zu bytes"
 vfu_dma_unregister(uint64_t gpa) "vfu: unregistering GPA 0x%"PRIx64""
+vfu_bar_register(int i, uint64_t addr, uint64_t size) "vfu: BAR %d: addr 0x%"PRIx64" size 0x%"PRIx64""
+vfu_bar_rw_enter(const char *op, uint64_t addr) "vfu: %s request for BAR address 0x%"PRIx64""
+vfu_bar_rw_exit(const char *op, uint64_t addr) "vfu: Finished %s of BAR address 0x%"PRIx64""
Signature · 1 line
2.20.1
#34 lore raw

Re: [PATCH v8 14/17] vfio-user: handle PCI BAR accesses

Quoted 30 lines · Jagannathan Raman
hw/remote/vfio-user-obj.c +30 -0
+static size_t vfu_object_bar_rw(PCIDevice *pci_dev, int pci_bar,
+                                hwaddr bar_offset, char * const buf,
+                                hwaddr len, const bool is_write)
+{
+    MemoryRegionSection section = { 0 };
+    uint8_t *ptr = (uint8_t *)buf;
+    MemoryRegion *section_mr = NULL;
+    uint64_t section_size;
+    hwaddr section_offset;
+    hwaddr size = 0;
+
+    while (len) {
+        section = memory_region_find(pci_dev->io_regions[pci_bar].memory,
+                                     bar_offset, len);
+
+        if (!section.mr) {
+            warn_report("vfu: invalid address 0x%"PRIx64"", bar_offset);
+            return size;
+        }
+
+        section_mr = section.mr;
+        section_offset = section.offset_within_region;
+        section_size = int128_get64(section.size);
+
+        if (is_write && section_mr->readonly) {
+            warn_report("vfu: attempting to write to readonly region in "
+                        "bar %d - [0x%"PRIx64" - 0x%"PRIx64"]",
+                        pci_bar, bar_offset,
+                        (bar_offset + section_size));
+            memory_region_unref(section_mr);

hw/remote/vfio-user-obj.c static void dma_unregister(vfu_ctx_t *vfu_ctx, vfu_dma_info_t *info) · line 435

Quoted 10 lines
hw/remote/vfio-user-obj.c +10 -0
+            return size;
+        }
+
+        if (vfu_object_mr_rw(section_mr, ptr, section_offset,
+                             section_size, is_write)) {
+            warn_report("vfu: failed to %s "
+                        "[0x%"PRIx64" - 0x%"PRIx64"] in bar %d",
+                        is_write ? "write to" : "read from", bar_offset,
+                        (bar_offset + section_size), pci_bar);
+            return size;

Missing memory_region_unref(section_mr).

author
#35 lore raw

Re: [PATCH v8 14/17] vfio-user: handle PCI BAR accesses

Quoted 39 lines · Stefan Hajnoczi
Quoted 34 lines · Jagannathan Raman
hw/remote/vfio-user-obj.c +34 -0
+static size_t vfu_object_bar_rw(PCIDevice *pci_dev, int pci_bar,
+                                hwaddr bar_offset, char * const buf,
+                                hwaddr len, const bool is_write)
+{
+    MemoryRegionSection section = { 0 };
+    uint8_t *ptr = (uint8_t *)buf;
+    MemoryRegion *section_mr = NULL;
+    uint64_t section_size;
+    hwaddr section_offset;
+    hwaddr size = 0;
+
+    while (len) {
+        section = memory_region_find(pci_dev->io_regions[pci_bar].memory,
+                                     bar_offset, len);
+
+        if (!section.mr) {
+            warn_report("vfu: invalid address 0x%"PRIx64"", bar_offset);
+            return size;
+        }
+
+        section_mr = section.mr;
+        section_offset = section.offset_within_region;
+        section_size = int128_get64(section.size);
+
+        if (is_write && section_mr->readonly) {
+            warn_report("vfu: attempting to write to readonly region in "
+                        "bar %d - [0x%"PRIx64" - 0x%"PRIx64"]",
+                        pci_bar, bar_offset,
+                        (bar_offset + section_size));
+            memory_region_unref(section_mr);
+            return size;
+        }
+
+        if (vfu_object_mr_rw(section_mr, ptr, section_offset,
Quoted 8 lines

hw/remote/vfio-user-obj.c static void dma_unregister(vfu_ctx_t *vfu_ctx, vfu_dma_info_t *info) · line 435

Quoted 6 lines
hw/remote/vfio-user-obj.c +6 -0
+                             section_size, is_write)) {
+            warn_report("vfu: failed to %s "
+                        "[0x%"PRIx64" - 0x%"PRIx64"] in bar %d",
+                        is_write ? "write to" : "read from", bar_offset,
+                        (bar_offset + section_size), pci_bar);
+            return size;

Missing memory_region_unref(section_mr).

Ah got it.

Thanks!
--
Jag

#36 lore raw

PATCH v8 15/17 vfio-user: handle device interrupts

In reply to Jagannathan Raman

Forward remote device's interrupts to the guest

---
 include/hw/pci/pci.h              |  13 ++++
 include/hw/remote/vfio-user-obj.h |   6 ++
 hw/pci/msi.c                      |  16 ++--
 hw/pci/msix.c                     |  10 ++-
 hw/pci/pci.c                      |  13 ++++
 hw/remote/machine.c               |  14 +++-
 hw/remote/vfio-user-obj.c         | 123 ++++++++++++++++++++++++++++++
 stubs/vfio-user-obj.c             |   6 ++
 MAINTAINERS                       |   1 +
 hw/remote/trace-events            |   1 +
 stubs/meson.build                 |   1 +
 11 files changed, 193 insertions(+), 11 deletions(-)
 create mode 100644 include/hw/remote/vfio-user-obj.h
 create mode 100644 stubs/vfio-user-obj.c
include/hw/pci/pci.h +13 -0
index 3a32b8dd40..7595c05c98 100644
@@ -16,6 +16,7 @@ extern bool pci_available;
 #define PCI_SLOT(devfn)         (((devfn) >> 3) & 0x1f)
 #define PCI_FUNC(devfn)         ((devfn) & 0x07)
 #define PCI_BUILD_BDF(bus, devfn)     ((bus << 8) | (devfn))
+#define PCI_BDF_TO_DEVFN(x)     ((x) & 0xff)
 #define PCI_BUS_MAX             256
 #define PCI_DEVFN_MAX           256
 #define PCI_SLOT_MAX            32
@@ -127,6 +128,10 @@ typedef void PCIMapIORegionFunc(PCIDevice *pci_dev, int region_num,
                                 pcibus_t addr, pcibus_t size, int type);
 typedef void PCIUnregisterFunc(PCIDevice *pci_dev);
 
+typedef void MSITriggerFunc(PCIDevice *dev, MSIMessage msg);
+typedef MSIMessage MSIPrepareMessageFunc(PCIDevice *dev, unsigned vector);
+typedef MSIMessage MSIxPrepareMessageFunc(PCIDevice *dev, unsigned vector);
+
 typedef struct PCIIORegion {
     pcibus_t addr; /* current PCI mapping address. -1 means not mapped */
 #define PCI_BAR_UNMAPPED (~(pcibus_t)0)
@@ -321,6 +326,14 @@ struct PCIDevice {
     /* Space to store MSIX table & pending bit array */
     uint8_t *msix_table;
     uint8_t *msix_pba;
+
+    /* May be used by INTx or MSI during interrupt notification */
+    void *irq_opaque;
+
+    MSITriggerFunc *msi_trigger;
+    MSIPrepareMessageFunc *msi_prepare_message;
+    MSIxPrepareMessageFunc *msix_prepare_message;
+
     /* MemoryRegion container for msix exclusive BAR setup */
     MemoryRegion msix_exclusive_bar;
     /* Memory Regions for MSIX table and pending bit entries. */
include/hw/remote/vfio-user-obj.h new file +6 -0
new file mode 100644
index 0000000000..87ab78b875
@@ -0,0 +1,6 @@
+#ifndef VFIO_USER_OBJ_H
+#define VFIO_USER_OBJ_H
+
+void vfu_object_set_bus_irq(PCIBus *pci_bus);
+
+#endif
hw/pci/msi.c +10 -6
index 47d2b0f33c..d556e17a09 100644
@@ -134,7 +134,7 @@ void msi_set_message(PCIDevice *dev, MSIMessage msg)
     pci_set_word(dev->config + msi_data_off(dev, msi64bit), msg.data);
 }
 
-MSIMessage msi_get_message(PCIDevice *dev, unsigned int vector)
+static MSIMessage msi_prepare_message(PCIDevice *dev, unsigned int vector)
 {
     uint16_t flags = pci_get_word(dev->config + msi_flags_off(dev));
     bool msi64bit = flags & PCI_MSI_FLAGS_64BIT;
@@ -159,6 +159,11 @@ MSIMessage msi_get_message(PCIDevice *dev, unsigned int vector)
     return msg;
 }
 
+MSIMessage msi_get_message(PCIDevice *dev, unsigned int vector)
+{
+    return dev->msi_prepare_message(dev, vector);
+}
+
 bool msi_enabled(const PCIDevice *dev)
 {
     return msi_present(dev) &&
@@ -241,6 +246,8 @@ int msi_init(struct PCIDevice *dev, uint8_t offset,
                      0xffffffff >> (PCI_MSI_VECTORS_MAX - nr_vectors));
     }
 
+    dev->msi_prepare_message = msi_prepare_message;
+
     return 0;
 }
 
@@ -256,6 +263,7 @@ void msi_uninit(struct PCIDevice *dev)
     cap_size = msi_cap_sizeof(flags);
     pci_del_capability(dev, PCI_CAP_ID_MSI, cap_size);
     dev->cap_present &= ~QEMU_PCI_CAP_MSI;
+    dev->msi_prepare_message = NULL;
 
     MSI_DEV_PRINTF(dev, "uninit\n");
 }
@@ -334,11 +342,7 @@ void msi_notify(PCIDevice *dev, unsigned int vector)
 
 void msi_send_message(PCIDevice *dev, MSIMessage msg)
 {
-    MemTxAttrs attrs = {};
-
-    attrs.requester_id = pci_requester_id(dev);
-    address_space_stl_le(&dev->bus_master_as, msg.address, msg.data,
-                         attrs, NULL);
+    dev->msi_trigger(dev, msg);
 }
 
 /* Normally called by pci_default_write_config(). */
hw/pci/msix.c +9 -1
index ae9331cd0b..6f85192d6f 100644
@@ -31,7 +31,7 @@
 #define MSIX_ENABLE_MASK (PCI_MSIX_FLAGS_ENABLE >> 8)
 #define MSIX_MASKALL_MASK (PCI_MSIX_FLAGS_MASKALL >> 8)
 
-MSIMessage msix_get_message(PCIDevice *dev, unsigned vector)
+static MSIMessage msix_prepare_message(PCIDevice *dev, unsigned vector)
 {
     uint8_t *table_entry = dev->msix_table + vector * PCI_MSIX_ENTRY_SIZE;
     MSIMessage msg;
@@ -41,6 +41,11 @@ MSIMessage msix_get_message(PCIDevice *dev, unsigned vector)
     return msg;
 }
 
+MSIMessage msix_get_message(PCIDevice *dev, unsigned vector)
+{
+    return dev->msix_prepare_message(dev, vector);
+}
+
 /*
  * Special API for POWER to configure the vectors through
  * a side channel. Should never be used by devices.
@@ -344,6 +349,8 @@ int msix_init(struct PCIDevice *dev, unsigned short nentries,
                           "msix-pba", pba_size);
     memory_region_add_subregion(pba_bar, pba_offset, &dev->msix_pba_mmio);
 
+    dev->msix_prepare_message = msix_prepare_message;
+
     return 0;
 }
 
@@ -429,6 +436,7 @@ void msix_uninit(PCIDevice *dev, MemoryRegion *table_bar, MemoryRegion *pba_bar)
     g_free(dev->msix_entry_used);
     dev->msix_entry_used = NULL;
     dev->cap_present &= ~QEMU_PCI_CAP_MSIX;
+    dev->msix_prepare_message = NULL;
 }
 
 void msix_uninit_exclusive_bar(PCIDevice *dev)
hw/pci/pci.c +13 -0
index dae9119bfe..48d4bc5f69 100644
@@ -307,6 +307,15 @@ void pci_device_deassert_intx(PCIDevice *dev)
     }
 }
 
+static void pci_msi_trigger(PCIDevice *dev, MSIMessage msg)
+{
+    MemTxAttrs attrs = {};
+
+    attrs.requester_id = pci_requester_id(dev);
+    address_space_stl_le(&dev->bus_master_as, msg.address, msg.data,
+                         attrs, NULL);
+}
+
 static void pci_reset_regions(PCIDevice *dev)
 {
     int r;
@@ -1202,6 +1211,8 @@ static void pci_qdev_unrealize(DeviceState *dev)
 
     pci_device_deassert_intx(pci_dev);
     do_pci_unregister_device(pci_dev);
+
+    pci_dev->msi_trigger = NULL;
 }
 
 void pci_register_bar(PCIDevice *pci_dev, int region_num,
@@ -2236,6 +2247,8 @@ static void pci_qdev_realize(DeviceState *qdev, Error **errp)
     }
 
     pci_set_power(pci_dev, true);
+
+    pci_dev->msi_trigger = pci_msi_trigger;
 }
 
 PCIDevice *pci_new_multifunction(int devfn, bool multifunction,
hw/remote/machine.c +10 -4
index 7002d46980..37538edb2f 100644
@@ -24,6 +24,8 @@
 #include "hw/remote/iommu.h"
 #include "hw/qdev-core.h"
 #include "hw/remote/iommu.h"
+#include "hw/remote/vfio-user-obj.h"
+#include "hw/pci/msi.h"
 
 static void remote_machine_init(MachineState *machine)
 {
@@ -55,12 +57,16 @@ static void remote_machine_init(MachineState *machine)
 
     if (s->vfio_user) {
         remote_iommu_setup(pci_host->bus);
-    }
 
-    remote_iohub_init(&s->iohub);
+        msi_nonbroken = true;
+
+        vfu_object_set_bus_irq(pci_host->bus);
+    } else {
+        remote_iohub_init(&s->iohub);
 
-    pci_bus_irqs(pci_host->bus, remote_iohub_set_irq, remote_iohub_map_irq,
-                 &s->iohub, REMOTE_IOHUB_NB_PIRQS);
+        pci_bus_irqs(pci_host->bus, remote_iohub_set_irq, remote_iohub_map_irq,
+                     &s->iohub, REMOTE_IOHUB_NB_PIRQS);
+    }
 
     qbus_set_hotplug_handler(BUS(pci_host->bus), OBJECT(s));
 }
hw/remote/vfio-user-obj.c +123 -0

159 lines not on this page · raw message

stubs/vfio-user-obj.c new file +6 -0
new file mode 100644
index 0000000000..79100d768e
@@ -0,0 +1,6 @@
+#include "qemu/osdep.h"
+#include "hw/remote/vfio-user-obj.h"
+
+void vfu_object_set_bus_irq(PCIBus *pci_bus)
+{
+}
MAINTAINERS +1 -0
index 3e62ccab0a..ad51ec0dc8 100644
@@ -3599,6 +3599,7 @@ F: hw/remote/iohub.c
 F: include/hw/remote/iohub.h
 F: subprojects/libvfio-user
 F: hw/remote/vfio-user-obj.c
+F: include/hw/remote/vfio-user-obj.h
 F: hw/remote/iommu.c
 F: include/hw/remote/iommu.h
 
hw/remote/trace-events +1 -0
index 847d50d88f..c167b3c7a5 100644
@@ -12,3 +12,4 @@ vfu_dma_unregister(uint64_t gpa) "vfu: unregistering GPA 0x%"PRIx64""
 vfu_bar_register(int i, uint64_t addr, uint64_t size) "vfu: BAR %d: addr 0x%"PRIx64" size 0x%"PRIx64""
 vfu_bar_rw_enter(const char *op, uint64_t addr) "vfu: %s request for BAR address 0x%"PRIx64""
 vfu_bar_rw_exit(const char *op, uint64_t addr) "vfu: Finished %s of BAR address 0x%"PRIx64""
+vfu_interrupt(int pirq) "vfu: sending interrupt to device - PIRQ %d"
stubs/meson.build +1 -0
index 6f80fec761..d8f3fd5c44 100644
@@ -60,3 +60,4 @@ if have_system
 else
   stub_ss.add(files('qdev.c'))
 endif
+stub_ss.add(when: 'CONFIG_VFIO_USER_SERVER', if_false: files('vfio-user-obj.c'))
Signature · 1 line
2.20.1
#37 lore raw

Re: [PATCH v8 15/17] vfio-user: handle device interrupts

Quoted 7 lines · Jagannathan Raman
hw/remote/vfio-user-obj.c +7 -0
+static MSIMessage vfu_object_msi_prepare_msg(PCIDevice *pci_dev,
+                                             unsigned int vector)
+{
+    MSIMessage msg;
+
+    msg.address = 0;
+    msg.data = vector;

hw/remote/vfio-user-obj.c static void vfu_object_register_bars(vfu_ctx_t *vfu_ctx, PCIDevice *pdev) · line 572

Quoted 10 lines
hw/remote/vfio-user-obj.c +10 -0
+
+    return msg;
+}
+
+static void vfu_object_msi_trigger(PCIDevice *pci_dev, MSIMessage msg)
+{
+    vfu_ctx_t *vfu_ctx = pci_dev->irq_opaque;
+
+    vfu_irq_trigger(vfu_ctx, msg.data);
+}

Why did you switch to vfu_object_msi_prepare_msg() + vfu_object_msi_trigger() in this revision?

Stefan

author
#38 lore raw

Re: [PATCH v8 15/17] vfio-user: handle device interrupts

Quoted 17 lines · Stefan Hajnoczi
Quoted 12 lines · Jagannathan Raman
hw/remote/vfio-user-obj.c +12 -0
+static MSIMessage vfu_object_msi_prepare_msg(PCIDevice *pci_dev,
+                                             unsigned int vector)
+{
+    MSIMessage msg;
+
+    msg.address = 0;
+    msg.data = vector;
+
+    return msg;
+}
+
+static void vfu_object_msi_trigger(PCIDevice *pci_dev, MSIMessage msg)
Quoted 8 lines

hw/remote/vfio-user-obj.c static void vfu_object_register_bars(vfu_ctx_t *vfu_ctx, PCIDevice *pdev) · line 572

Quoted 5 lines
hw/remote/vfio-user-obj.c +5 -0
+{
+    vfu_ctx_t *vfu_ctx = pci_dev->irq_opaque;
+
+    vfu_irq_trigger(vfu_ctx, msg.data);
+}

Why did you switch to vfu_object_msi_prepare_msg() + vfu_object_msi_trigger() in this revision?

We previously did not do this switch because the server didn’t get updates to the MSIx table & PBA.

The latest client version (which is not part of this series) forwards accesses to the MSIx table & PBA over to the server. It also reads the PBA set by the server. These change make it possible for the server to make this switch.

Thank you!
--
Jag

Quoted 2 lines

Stefan

#39 lore raw

Re: [PATCH v8 15/17] vfio-user: handle device interrupts

Quoted 24 lines · Jag Raman
Quoted 15 lines · Stefan Hajnoczi
Quoted 12 lines · Jagannathan Raman
hw/remote/vfio-user-obj.c +12 -0
+static MSIMessage vfu_object_msi_prepare_msg(PCIDevice *pci_dev,
+                                             unsigned int vector)
+{
+    MSIMessage msg;
+
+    msg.address = 0;
+    msg.data = vector;
+
+    return msg;
+}
+
+static void vfu_object_msi_trigger(PCIDevice *pci_dev, MSIMessage msg)
Quoted 8 lines

hw/remote/vfio-user-obj.c static void vfu_object_register_bars(vfu_ctx_t *vfu_ctx, PCIDevice *pdev) · line 572

Quoted 5 lines
hw/remote/vfio-user-obj.c +5 -0
+{
+    vfu_ctx_t *vfu_ctx = pci_dev->irq_opaque;
+
+    vfu_irq_trigger(vfu_ctx, msg.data);
+}

Why did you switch to vfu_object_msi_prepare_msg() + vfu_object_msi_trigger() in this revision?

Quoted 6 lines

We previously did not do this switch because the server didn’t get updates to the MSIx table & PBA.

The latest client version (which is not part of this series) forwards accesses to the MSIx table & PBA over to the server. It also reads the PBA set by the server. These change make it possible for the server to make this switch.

Interesting. That's different from kernel VFIO. Before vfio-user commits to a new approach it would be worth checking with Alex that he agrees with the design.

I remember sending an email asking about why VFIO MSI-X PBA does not offer the full semantics described in the PCIe spec but didn't get a response from Alex (Message-Id:
[email protected]).

Stefan

#40 lore raw

Re: [PATCH v8 15/17] vfio-user: handle device interrupts

Quoted 32 lines · On Thu, 28 Apr 2022 10:54:04 +0100 Stefan Hajnoczi <[email protected]> wrote:
Quoted 24 lines · Jag Raman
Quoted 15 lines · Stefan Hajnoczi
Quoted 12 lines · Jagannathan Raman
hw/remote/vfio-user-obj.c +12 -0
+static MSIMessage vfu_object_msi_prepare_msg(PCIDevice *pci_dev,
+                                             unsigned int vector)
+{
+    MSIMessage msg;
+
+    msg.address = 0;
+    msg.data = vector;
+
+    return msg;
+}
+
+static void vfu_object_msi_trigger(PCIDevice *pci_dev, MSIMessage msg)
Quoted 8 lines

hw/remote/vfio-user-obj.c static void vfu_object_register_bars(vfu_ctx_t *vfu_ctx, PCIDevice *pdev) · line 572

Quoted 5 lines
hw/remote/vfio-user-obj.c +5 -0
+{
+    vfu_ctx_t *vfu_ctx = pci_dev->irq_opaque;
+
+    vfu_irq_trigger(vfu_ctx, msg.data);
+}  

Why did you switch to vfu_object_msi_prepare_msg() + vfu_object_msi_trigger() in this revision?

Quoted 6 lines

We previously did not do this switch because the server didn’t get updates to the MSIx table & PBA.

The latest client version (which is not part of this series) forwards accesses to the MSIx table & PBA over to the server. It also reads the PBA set by the server. These change make it possible for the server to make this switch.

Quoted 8 lines

Interesting. That's different from kernel VFIO. Before vfio-user commits to a new approach it would be worth checking with Alex that he agrees with the design.

I remember sending an email asking about why VFIO MSI-X PBA does not offer the full semantics described in the PCIe spec but didn't get a response from Alex (Message-Id:
[email protected]).

IIUC, the question is why we redirect the MSI-X interrupt from the KVM irqfd to be handled in QEMU when the vector is masked. This is largely to work around the fact that we haven't had a means to implement mask and unmask in the kernel, therefore we leave the vector enabled and only enable the emulated PBA if a masked vector fires. This works because nobody really cares about the PBA, nor operates in a mode where vectors are masked and the PBA is polled. Drivers that understand the device likely have better places to poll for service requests than the PBA.

Ideally, masking a vector would make use of the existing mask and unmask uAPI via the SET_IRQS ioctl, but we haven't been able to implement this due to lack of internal kernel APIs to support it. We may have those interfaces now, but lacking bandwidth, I haven't checked recently and we seem to be getting by ok as is. Thanks,

Alex

#41 lore raw

PATCH v8 16/17 vfio-user: handle reset of remote device

In reply to Jagannathan Raman

Adds handler to reset a remote device

---
 hw/remote/vfio-user-obj.c | 20 ++++++++++++++++++++
 1 file changed, 20 insertions(+)
hw/remote/vfio-user-obj.c +20 -0
index 70b4d8b9ce..8ca823aa01 100644
@@ -623,6 +623,20 @@ void vfu_object_set_bus_irq(PCIBus *pci_bus)
     pci_bus_irqs(pci_bus, vfu_object_set_irq, vfu_object_map_irq, pci_bus, 1);
 }
 
+static int vfu_object_device_reset(vfu_ctx_t *vfu_ctx, vfu_reset_type_t type)
+{
+    VfuObject *o = vfu_get_private(vfu_ctx);
+
+    /* vfu_object_ctx_run() handles lost connection */
+    if (type == VFU_RESET_LOST_CONN) {
+        return 0;
+    }
+
+    qdev_reset_all(DEVICE(o->pci_dev));
+
+    return 0;
+}
+
 /*
  * TYPE_VFU_OBJECT depends on the availability of the 'socket' and 'device'
  * properties. It also depends on devices instantiated in QEMU. These
@@ -728,6 +742,12 @@ static void vfu_object_init_ctx(VfuObject *o, Error **errp)
         goto fail;
     }
 
+    ret = vfu_setup_device_reset_cb(o->vfu_ctx, &vfu_object_device_reset);
+    if (ret < 0) {
+        error_setg(errp, "vfu: Failed to setup reset callback");
+        goto fail;
+    }
+
     ret = vfu_realize_ctx(o->vfu_ctx);
     if (ret < 0) {
         error_setg(errp, "vfu: Failed to realize device %s- %s",
Signature · 1 line
2.20.1
#43 lore raw

PATCH v8 17/17 vfio-user: avocado tests for vfio-user

In reply to Jagannathan Raman

Avocado tests for libvfio-user in QEMU - tests startup, hotplug and migration of the server object

---
 MAINTAINERS                |   1 +
 tests/avocado/vfio-user.py | 164 +++++++++++++++++++++++++++++++++++++
 2 files changed, 165 insertions(+)
 create mode 100644 tests/avocado/vfio-user.py
MAINTAINERS +1 -0
index ad51ec0dc8..8676f546e9 100644
@@ -3602,6 +3602,7 @@ F: hw/remote/vfio-user-obj.c
 F: include/hw/remote/vfio-user-obj.h
 F: hw/remote/iommu.c
 F: include/hw/remote/iommu.h
+F: tests/avocado/vfio-user.py
 
 EBPF:
 M: Jason Wang <[email protected]>
tests/avocado/vfio-user.py new file +164 -0

164 lines not on this page · raw message

Signature · 1 line
2.20.1
#44 lore raw
In reply to Jagannathan Raman
Quoted 3 lines · Jagannathan Raman

This is v8 of the server side changes to enable vfio-user in QEMU.

Thank you very much for reviewing the last revision of this series!

I posted some minor comments. I hope the next revision or the one after it will be merged because the code will benefit from being in-tree where more people can easily try it out.

Stefan

author
#45 lore raw
Quoted 12 lines · Stefan Hajnoczi
Quoted 3 lines · Jagannathan Raman

This is v8 of the server side changes to enable vfio-user in QEMU.

Thank you very much for reviewing the last revision of this series!

I posted some minor comments. I hope the next revision or the one after it will be merged because the code will benefit from being in-tree where more people can easily try it out.

Thank you very much!

Will send the next revision out shortly!

--
Jag

Quoted 2 lines

Stefan