Skip to contentSkip to conversation

Refactor PCI/SHPC/PCIE hotplug to use a more generic hotplug API

PATCH Superseded Igor Mammedov 6 participants 33 messages Dec 13, 2013 – Dec 18, 2013

v3 of 6 Superseded Latest is v6 →

lore
Share & tools
lore
#0 lore raw

v3 cover Refactor PCI/SHPC/PCIE hotplug to use a more generic hotplug API

changes since v2:
* s/hotplugable/hotpluggable/
* move hotplug check to an earlier patch:
"qdev: add "hotpluggable" property to Device"
--
Refactor PCI specific hotplug API to a more generic/reusable one. Model it after SCSI-BUS like hotplug API replacing single hotplug callback with hotplug/hot_unplug pair of callbacks as suggested by Paolo.
Difference between SCSI-BUS and this approach is that the former is BUS centric while the latter is device centred. Which is evolved from the fact that hotplug callbacks used by ACPI/SHPC/PCIE are implemented by devices rather than by bus and bus serves only as a proxy to forward event to hotplug device.
Memory hotplug also exposes tha same usage pattern hence an attempt to generalize hotplug API.

Refactoring also simplifies wiring of a hotplug device with a bus, all it needs is to set "hotplug-device" link on bus, which would potentially allow to do it from configuration file, there is not need to setup hotplug device callbacks on bus synce it can get them via HOTPLUG_DEVICE API of "hotplug-device" target.

In addition device centred hotplug API may be used by bus-less hotplug implementations as well if it's decided to use link<foo...> instead of bus.

Patches 8-11 are should be merged as one and are split only for simplifying review (they compile fine but PCI hotplug is broken until the last patch is applyed).

git tree for testing:
https://github.com/imammedo/qemu/commits/hotplug_dev_inf_v3

tested only ACPI and PCIE hotplug.

Hervé Poussineau (1):
qom: detect bad reentrance during object_class_foreach

Igor Mammedov (9):
define hotplug interface
qdev: add to BusState "hotplug-handler" link
qdev: add "hotpluggable" property to Device
hw/acpi: move typeinfo to the file end
qdev:pci: refactor PCIDevice to use generic "hotpluggable" property acpi/piix4pm: convert ACPI PCI hotplug to use hotplug-handler API pci/shpc: convert SHPC hotplug to use hotplug-handler API pci/pcie: convert PCIE hotplug to use hotplug-handler API hw/pci: switch to a generic hotplug handling for PCIDevice

Paolo Bonzini (1):
qom: do not register interface "types" in the type table

 hw/acpi/piix4.c                | 151 ++++++++++++++++++++++-------------------
 hw/core/Makefile.objs          |   1 +
 hw/core/hotplug.c              |  48 +++++++++++++
 hw/core/qdev.c                 |  50 ++++++++++++--
 hw/display/cirrus_vga.c        |   2 +-
 hw/display/qxl.c               |   2 +-
 hw/display/vga-pci.c           |   2 +-
 hw/display/vmware_vga.c        |   2 +-
 hw/i386/acpi-build.c           |   6 +-
 hw/ide/piix.c                  |   4 +-
 hw/isa/piix4.c                 |   2 +-
 hw/pci-bridge/pci_bridge_dev.c |   9 +++
 hw/pci-host/piix.c             |   6 +-
 hw/pci/pci.c                   |  40 +----------
 hw/pci/pcie.c                  |  73 +++++++++++++-------
 hw/pci/pcie_port.c             |   8 +++
 hw/pci/shpc.c                  | 133 +++++++++++++++++++++++-------------
 hw/usb/hcd-ehci-pci.c          |   2 +-
 hw/usb/hcd-ohci.c              |   2 +-
 hw/usb/hcd-uhci.c              |   2 +-
 hw/usb/hcd-xhci.c              |   2 +-
 include/hw/hotplug.h           |  75 ++++++++++++++++++++
 include/hw/pci/pci.h           |  13 ----
 include/hw/pci/pci_bus.h       |   2 -
 include/hw/pci/pcie.h          |   5 ++
 include/hw/pci/shpc.h          |   8 +++
 include/hw/qdev-core.h         |   8 +++
 qom/object.c                   |  17 ++++-
 28 files changed, 455 insertions(+), 220 deletions(-)
 create mode 100644 hw/core/hotplug.c
 create mode 100644 include/hw/hotplug.h
Signature · 1 line
1.8.3.1
#1 lore raw

PATCH v3 1/11 qom: do not register interface "types" in the type table

From: Paolo Bonzini <[email protected]>

There should be no need to look them up nor enumerate the interface "types", whose "classes" are really just vtables. Just create the types and add them to the interface list of the parent type.

---
 qom/object.c | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)
qom/object.c +10 -3
index fc19cf6..3a43186 100644
@@ -88,7 +88,7 @@ static TypeImpl *type_table_lookup(const char *name)
     return g_hash_table_lookup(type_table_get(), name);
 }
 
-static TypeImpl *type_register_internal(const TypeInfo *info)
+static TypeImpl *type_new(const TypeInfo *info)
 {
     TypeImpl *ti = g_malloc0(sizeof(*ti));
     int i;
@@ -122,8 +122,15 @@ static TypeImpl *type_register_internal(const TypeInfo *info)
     }
     ti->num_interfaces = i;
 
-    type_table_add(ti);
+    return ti;
+}
 
+static TypeImpl *type_register_internal(const TypeInfo *info)
+{
+    TypeImpl *ti;
+    ti = type_new(info);
+
+    type_table_add(ti);
     return ti;
 }
 
@@ -216,7 +223,7 @@ static void type_initialize_interface(TypeImpl *ti, const char *parent)
     info.name = g_strdup_printf("%s::%s", ti->name, info.parent);
     info.abstract = true;
 
-    iface_impl = type_register(&info);
+    iface_impl = type_new(&info);
     type_initialize(iface_impl);
     g_free((char *)info.name);
 
Signature · 1 line
1.8.3.1
#2 lore raw

PATCH v3 2/11 qom: detect bad reentrance during object_class_foreach

In reply to Igor Mammedov

From: Hervé Poussineau <[email protected]>

We should not modify the type hash table while it is being iterated on. Assert that it does not happen.

---
v2:
* make ver more descriptinve s/enumerating/enumerating_classes/ [asked-by: Peter Crosthwaite]

---
 qom/object.c | 4 ++++
 1 file changed, 4 insertions(+)
qom/object.c +4 -0
index 3a43186..4a0fb86 100644
@@ -78,8 +78,10 @@ static GHashTable *type_table_get(void)
     return type_table;
 }
 
+static bool enumerating_classes = false;
 static void type_table_add(TypeImpl *ti)
 {
+    assert(!enumerating_classes);
     g_hash_table_insert(type_table_get(), (void *)ti->name, ti);
 }
 
@@ -666,7 +668,9 @@ void object_class_foreach(void (*fn)(ObjectClass *klass, void *opaque),
 {
     OCFData data = { fn, implements_type, include_abstract, opaque };
 
+    enumerating_classes = true;
     g_hash_table_foreach(type_table_get(), object_class_foreach_tramp, &data);
+    enumerating_classes = false;
 }
 
 int object_child_foreach(Object *obj, int (*fn)(Object *child, void *opaque),
Signature · 1 line
1.8.3.1
Reviewed-by on v3 2/11
#3 lore raw

Re: [Qemu-devel] [PATCH 02/11] qom: detect bad reentrance during object_class_foreach

Quoted 21 lines · Igor Mammedov

From: Hervé Poussineau <[email protected]>

We should not modify the type hash table while it is being iterated on. Assert that it does not happen.

---
v2:
* make ver more descriptinve s/enumerating/enumerating_classes/ [asked-by: Peter Crosthwaite]

---
 qom/object.c | 4 ++++
 1 file changed, 4 insertions(+)
qom/object.c +0 -0
index 3a43186..4a0fb86 100644

qom/object.c static GHashTable *type_table_get(void) · line 81

Quoted 4 lines
qom/object.c +1 -0
@@ -78,8 +78,10 @@ static GHashTable *type_table_get(void)
     return type_table;
 }
 
+static bool enumerating_classes = false;

This seems to be a change in terminology i.e. s/type/class. Should it be enumerating_types or enumerating_type_table?

Also blank line here between global var decl and fn.

Otherwise,

Regards,
Peter

Quoted 20 lines
qom/object.c +3 -1
 static void type_table_add(TypeImpl *ti)
 {
+    assert(!enumerating_classes);
     g_hash_table_insert(type_table_get(), (void *)ti->name, ti);
 }
 
@@ -666,7 +668,9 @@ void object_class_foreach(void (*fn)(ObjectClass *klass, void *opaque),
 {
     OCFData data = { fn, implements_type, include_abstract, opaque };
 
+    enumerating_classes = true;
     g_hash_table_foreach(type_table_get(), object_class_foreach_tramp, &data);
+    enumerating_classes = false;
 }
 
 int object_child_foreach(Object *obj, int (*fn)(Object *child, void *opaque),
--

1.8.3.1

#4 lore raw

Re: [Qemu-devel] [PATCH 02/11] qom: detect bad reentrance during object_class_foreach

Am 14.12.2013 07:53, schrieb Peter Crosthwaite:

Quoted 25 lines · Igor Mammedov
Quoted 24 lines

From: Hervé Poussineau <[email protected]>

We should not modify the type hash table while it is being iterated on. Assert that it does not happen.

---
v2:
* make ver more descriptinve s/enumerating/enumerating_classes/ [asked-by: Peter Crosthwaite]

---
 qom/object.c | 4 ++++
 1 file changed, 4 insertions(+)
qom/object.c +0 -0
index 3a43186..4a0fb86 100644
@@ -78,8 +78,10 @@ static GHashTable *type_table_get(void)
     return type_table;
 }
 
Quoted 4 lines

qom/object.c static GHashTable *type_table_get(void) · line 81

Quoted 1 line
qom/object.c +1 -0
+static bool enumerating_classes = false;

This seems to be a change in terminology i.e. s/type/class. Should it be enumerating_types or enumerating_type_table?

Too many cooks... While offline, I picked up the original version, changing to enumerating_types, which Alexey also happened to choose, and then adopted his drop of "false" assignment.

Quoted 1 line

Also blank line here between global var decl and fn.

Same thought here, already did in my version. :)

Quoted 3 lines

Otherwise,

Thanks, will update while rebasing.

Andreas

Signature · 2 lines
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg
#5 lore raw

PATCH v3 3/11 define hotplug interface

In reply to Igor Mammedov

Provide generic hotplug interface for devices. Intended for replacing hotplug mechanism used by PCI/PCIE/SHPC code and will be used for memory hotplug.

---
v2:
* s/device/handler/
* add hotplug_handler_plug/hotplug_handler_unplug API v1:
it's scsi-bus like interface, but abstracted from bus altogether since all current users care about in hotplug handlers, it's hotplug device and hotplugged device and bus only serves as a means to get access to hotplug device and it's callbacks.

---
 hw/core/Makefile.objs |  1 +
 hw/core/hotplug.c     | 48 +++++++++++++++++++++++++++++++++
 include/hw/hotplug.h  | 75 +++++++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 124 insertions(+)
 create mode 100644 hw/core/hotplug.c
 create mode 100644 include/hw/hotplug.h
hw/core/Makefile.objs +1 -0
index 950146c..9e324be 100644
@@ -2,6 +2,7 @@
 common-obj-y += qdev.o qdev-properties.o
 # irq.o needed for qdev GPIO handling:
 common-obj-y += irq.o
+common-obj-y += hotplug.o
 
 common-obj-$(CONFIG_EMPTY_SLOT) += empty_slot.o
 common-obj-$(CONFIG_XILINX_AXI) += stream.o
hw/core/hotplug.c new file +48 -0

48 lines not on this page · raw message

include/hw/hotplug.h new file +75 -0

75 lines not on this page · raw message

Signature · 1 line
1.8.3.1
#6 lore raw

Re: [Qemu-devel] [PATCH 03/11] define hotplug interface

Quoted 1 line · Igor Mammedov

Provide generic hotplug interface for devices.

"Provide a". s/devices/hotplug handlers would be cleaner too, to match your v2 changes.

Quoted 52 lines

Intended for replacing hotplug mechanism used by PCI/PCIE/SHPC code and will be used for memory hotplug.

---
v2:
* s/device/handler/
* add hotplug_handler_plug/hotplug_handler_unplug API v1:
it's scsi-bus like interface, but abstracted from bus altogether since all current users care about in hotplug handlers, it's hotplug device and hotplugged device and bus only serves as a means to get access to hotplug device and it's callbacks.

---
 hw/core/Makefile.objs |  1 +
 hw/core/hotplug.c     | 48 +++++++++++++++++++++++++++++++++
 include/hw/hotplug.h  | 75 +++++++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 124 insertions(+)
 create mode 100644 hw/core/hotplug.c
 create mode 100644 include/hw/hotplug.h
hw/core/Makefile.objs +1 -0
index 950146c..9e324be 100644
@@ -2,6 +2,7 @@
 common-obj-y += qdev.o qdev-properties.o
 # irq.o needed for qdev GPIO handling:
 common-obj-y += irq.o
+common-obj-y += hotplug.o
 
 common-obj-$(CONFIG_EMPTY_SLOT) += empty_slot.o
 common-obj-$(CONFIG_XILINX_AXI) += stream.o
hw/core/hotplug.c new file +13 -0
new file mode 100644
index 0000000..5c3b5c9
@@ -0,0 +1,48 @@
+/*
+ * Hotplug handler interface.
+ *
+ * Copyright (c) 2013 Red Hat Inc.
+ *
+ * Authors:
+ *  Igor Mammedov <[email protected]>,
+ *
+ * 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 "hw/hotplug.h"
+#include "qemu/module.h"

hw/core/hotplug.c line 23

Quoted 10 lines
hw/core/hotplug.c new file +10 -0
+
+void hotplug_handler_plug(HotplugHandler *plug_handler,
+                          DeviceState *plugged_dev,
+                          Error **errp)
+{
+    HotplugHandlerClass *hdc = HOTPLUG_HANDLER_GET_CLASS(plug_handler);
+
+    if (hdc->plug) {
+        hdc->plug(plug_handler, plugged_dev, errp);
+    }

So does it mean anything to have a hotplug handler device that can't plug? Im thinking this API definition should be compusorly and throw-errors/assert rather than silent fail.

Quoted 49 lines
hw/core/hotplug.c new file +25 -0
+}
+
+void hotplug_handler_unplug(HotplugHandler *plug_handler,
+                            DeviceState *plugged_dev,
+                            Error **errp)
+{
+    HotplugHandlerClass *hdc = HOTPLUG_HANDLER_GET_CLASS(plug_handler);
+
+    if (hdc->unplug) {
+        hdc->unplug(plug_handler, plugged_dev, errp);
+    }
+}
+
+static const TypeInfo hotplug_handler_info = {
+    .name          = TYPE_HOTPLUG_HANDLER,
+    .parent        = TYPE_INTERFACE,
+    .class_size = sizeof(HotplugHandlerClass),
+};
+
+static void hotplug_handler_register_types(void)
+{
+    type_register_static(&hotplug_handler_info);
+}
+
+type_init(hotplug_handler_register_types)
include/hw/hotplug.h new file +18 -0
new file mode 100644
index 0000000..64ae6f7
@@ -0,0 +1,75 @@
+/*
+ * Hotplug handler interface.
+ *
+ * Copyright (c) 2013 Red Hat Inc.
+ *
+ * Authors:
+ *  Igor Mammedov <[email protected]>,
+ *
+ * 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 HOTPLUG_H
+#define HOTPLUG_H
+
+#include "qom/object.h"
+#include "qemu/typedefs.h"
+
+#define TYPE_HOTPLUG_HANDLER "hotplug-handler"

include/hw/hotplug.h line 28

Quoted 10 lines
include/hw/hotplug.h new file +10 -0
+
+#define HOTPLUG_HANDLER_CLASS(klass) \
+     OBJECT_CLASS_CHECK(HotplugHandlerClass, (klass), TYPE_HOTPLUG_HANDLER)
+#define HOTPLUG_HANDLER_GET_CLASS(obj) \
+     OBJECT_GET_CLASS(HotplugHandlerClass, (obj), TYPE_HOTPLUG_HANDLER)
+#define HOTPLUG_HANDLER(obj) \
+     INTERFACE_CHECK(HotplugHandler, (obj), TYPE_HOTPLUG_HANDLER)
+
+
+typedef struct HotplugHandler {

/* <private> */

Quoted 9 lines
include/hw/hotplug.h new file +9 -0
+    Object Parent;
+} HotplugHandler;
+
+/**
+ * hotplug_fn:
+ * @plug_handler: a device performing plug/uplug action
+ * @plugged_dev: a device that has been (un)plugged
+ * @errp: returns an error if this function fails
+ */

include/hw/hotplug.h line 47

Quoted 10 lines
include/hw/hotplug.h new file +10 -0
+typedef void (*hotplug_fn)(HotplugHandler *plug_handler,
+                           DeviceState *plugged_dev, Error **errp);
+
+/**
+ * HotplugDeviceClass:
+ *
+ * Interface to be implemented by a device performing
+ * hardware (un)plug functions.
+ *
+ * @parent: Opaque parent interface.

Not sure if it makes sense to publicly document the QOM parent first-field. It should not be used in code from this context.

include/hw/hotplug.h line 51

Quoted 4 lines
include/hw/hotplug.h new file +4 -0
+ * @plug: plug callback.
+ * @unplug: unplug callback.
+ */
+typedef struct HotplugHandlerClass {

/* <private> */

include/hw/hotplug.h line 53

Quoted 2 lines
include/hw/hotplug.h new file +2 -0
+    InterfaceClass parent;
+

/* <public> */

Quoted 26 lines
include/hw/hotplug.h new file +22 -1
+    hotplug_fn plug;
+    hotplug_fn unplug;
+} HotplugHandlerClass;
+
+/**
+ * hotplug_handler_plug:
+ *
+ * Call #HotplugHandlerClass.plug callback of @plug_handler.
+ */
+void hotplug_handler_plug(HotplugHandler *plug_handler,
+                          DeviceState *plugged_dev,
+                          Error **errp);
+
+/**
+ * hotplug_handler_unplug:
+ *
+ * Call #HotplugHandlerClass.unplug callback of @plug_handler.
+ */
+void hotplug_handler_unplug(HotplugHandler *plug_handler,
+                            DeviceState *plugged_dev,
+                            Error **errp);
+#endif
--

1.8.3.1

#7 lore raw

Re: [Qemu-devel] [PATCH 03/11] define hotplug interface

Quoted 5 lines · On Sat, 14 Dec 2013 17:03:34 +1000 Peter Crosthwaite…
Quoted 1 line · Igor Mammedov

Provide generic hotplug interface for devices.

"Provide a". s/devices/hotplug handlers would be cleaner too, to match your v2 changes.

sure

Quoted 59 lines
Quoted 59 lines

Intended for replacing hotplug mechanism used by PCI/PCIE/SHPC code and will be used for memory hotplug.

---
v2:
* s/device/handler/
* add hotplug_handler_plug/hotplug_handler_unplug API v1:
it's scsi-bus like interface, but abstracted from bus altogether since all current users care about in hotplug handlers, it's hotplug device and hotplugged device and bus only serves as a means to get access to hotplug device and it's callbacks.

---
 hw/core/Makefile.objs |  1 +
 hw/core/hotplug.c     | 48 +++++++++++++++++++++++++++++++++
 include/hw/hotplug.h  | 75 +++++++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 124 insertions(+)
 create mode 100644 hw/core/hotplug.c
 create mode 100644 include/hw/hotplug.h
hw/core/Makefile.objs +1 -0
index 950146c..9e324be 100644
@@ -2,6 +2,7 @@
 common-obj-y += qdev.o qdev-properties.o
 # irq.o needed for qdev GPIO handling:
 common-obj-y += irq.o
+common-obj-y += hotplug.o
 
 common-obj-$(CONFIG_EMPTY_SLOT) += empty_slot.o
 common-obj-$(CONFIG_XILINX_AXI) += stream.o
hw/core/hotplug.c new file +19 -0
new file mode 100644
index 0000000..5c3b5c9
@@ -0,0 +1,48 @@
+/*
+ * Hotplug handler interface.
+ *
+ * Copyright (c) 2013 Red Hat Inc.
+ *
+ * Authors:
+ *  Igor Mammedov <[email protected]>,
+ *
+ * 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 "hw/hotplug.h"
+#include "qemu/module.h"
+
+void hotplug_handler_plug(HotplugHandler *plug_handler,
+                          DeviceState *plugged_dev,
+                          Error **errp)
+{
+    HotplugHandlerClass *hdc = HOTPLUG_HANDLER_GET_CLASS(plug_handler);
Quoted 8 lines

hw/core/hotplug.c line 23

Quoted 4 lines
hw/core/hotplug.c new file +4 -0
+
+    if (hdc->plug) {
+        hdc->plug(plug_handler, plugged_dev, errp);
+    }

So does it mean anything to have a hotplug handler device that can't plug? Im thinking this API definition should be compusorly and throw-errors/assert rather than silent fail.

It doesn't have "fail" semantic but rather "handler doesn't need to do anything for plug".
And although in current usage we always have plug/unplug pair, not asserting would be more flexible from POV of generic API.

Quoted 77 lines
Quoted 50 lines
hw/core/hotplug.c new file +25 -0
+}
+
+void hotplug_handler_unplug(HotplugHandler *plug_handler,
+                            DeviceState *plugged_dev,
+                            Error **errp)
+{
+    HotplugHandlerClass *hdc = HOTPLUG_HANDLER_GET_CLASS(plug_handler);
+
+    if (hdc->unplug) {
+        hdc->unplug(plug_handler, plugged_dev, errp);
+    }
+}
+
+static const TypeInfo hotplug_handler_info = {
+    .name          = TYPE_HOTPLUG_HANDLER,
+    .parent        = TYPE_INTERFACE,
+    .class_size = sizeof(HotplugHandlerClass),
+};
+
+static void hotplug_handler_register_types(void)
+{
+    type_register_static(&hotplug_handler_info);
+}
+
+type_init(hotplug_handler_register_types)
include/hw/hotplug.h new file +18 -0
new file mode 100644
index 0000000..64ae6f7
@@ -0,0 +1,75 @@
+/*
+ * Hotplug handler interface.
+ *
+ * Copyright (c) 2013 Red Hat Inc.
+ *
+ * Authors:
+ *  Igor Mammedov <[email protected]>,
+ *
+ * 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 HOTPLUG_H
+#define HOTPLUG_H
+
+#include "qom/object.h"
+#include "qemu/typedefs.h"
+
+#define TYPE_HOTPLUG_HANDLER "hotplug-handler"

include/hw/hotplug.h line 28

Quoted 10 lines
include/hw/hotplug.h new file +10 -0
+
+#define HOTPLUG_HANDLER_CLASS(klass) \
+     OBJECT_CLASS_CHECK(HotplugHandlerClass, (klass), TYPE_HOTPLUG_HANDLER)
+#define HOTPLUG_HANDLER_GET_CLASS(obj) \
+     OBJECT_GET_CLASS(HotplugHandlerClass, (obj), TYPE_HOTPLUG_HANDLER)
+#define HOTPLUG_HANDLER(obj) \
+     INTERFACE_CHECK(HotplugHandler, (obj), TYPE_HOTPLUG_HANDLER)
+
+
+typedef struct HotplugHandler {

/* <private> */

Quoted 14 lines
include/hw/hotplug.h new file +14 -0
+    Object Parent;
+} HotplugHandler;
+
+/**
+ * hotplug_fn:
+ * @plug_handler: a device performing plug/uplug action
+ * @plugged_dev: a device that has been (un)plugged
+ * @errp: returns an error if this function fails
+ */
+typedef void (*hotplug_fn)(HotplugHandler *plug_handler,
+                           DeviceState *plugged_dev, Error **errp);
+
+/**
+ * HotplugDeviceClass:
Quoted 8 lines

include/hw/hotplug.h line 47

Quoted 5 lines
include/hw/hotplug.h new file +5 -0
+ *
+ * Interface to be implemented by a device performing
+ * hardware (un)plug functions.
+ *
+ * @parent: Opaque parent interface.

Not sure if it makes sense to publicly document the QOM parent first-field. It should not be used in code from this context.

I'd be glad to, but I think Andreas asked me once to do so for "parent" in DimmDevice class. I also no sure that is necessary to document this field in every child.

Quoted 31 lines

include/hw/hotplug.h line 51

Quoted 5 lines
include/hw/hotplug.h new file +4 -0
+ * @plug: plug callback.
+ * @unplug: unplug callback.
+ */
+typedef struct HotplugHandlerClass {

/* <private> */

include/hw/hotplug.h line 53

Quoted 2 lines
include/hw/hotplug.h new file +2 -0
+    InterfaceClass parent;
+

/* <public> */

Quoted 18 lines
include/hw/hotplug.h new file +18 -0
+    hotplug_fn plug;
+    hotplug_fn unplug;
+} HotplugHandlerClass;
+
+/**
+ * hotplug_handler_plug:
+ *
+ * Call #HotplugHandlerClass.plug callback of @plug_handler.
+ */
+void hotplug_handler_plug(HotplugHandler *plug_handler,
+                          DeviceState *plugged_dev,
+                          Error **errp);
+
+/**
+ * hotplug_handler_unplug:
+ *
+ * Call #HotplugHandlerClass.unplug callback of @plug_handler.
+ */
Quoted 8 lines
Quoted 8 lines
include/hw/hotplug.h new file +4 -1
+void hotplug_handler_unplug(HotplugHandler *plug_handler,
+                            DeviceState *plugged_dev,
+                            Error **errp);
+#endif
--

1.8.3.1

I'll ally the rest of changes on next respin, thanks for review!

#8 lore raw

PATCH v3 4/11 qdev: add to BusState "hotplug-handler" link

In reply to Igor Mammedov

It will allow to reuse field with different BUSes, reducing code duplication. Field is intended fot replacing 'hotplug_qdev' field in PCIBus and also will allow to avoid adding equivalent field to DimmBus with possiblitity to refactor other BUSes to use it instead of custom field.
In addition once all users of allow_hotplug field are converted to new API, link could replace allow_hotplug in qdev hotplug code.

---
 hw/core/qdev.c         | 4 ++++
 include/hw/qdev-core.h | 5 +++++
 2 files changed, 9 insertions(+)
hw/core/qdev.c +4 -0
index e374a93..25c2d2c 100644
@@ -32,6 +32,7 @@
 #include "qapi/visitor.h"
 #include "qapi/qmp/qjson.h"
 #include "monitor/monitor.h"
+#include "hw/hotplug.h"
 
 int qdev_hotplug = 0;
 static bool qdev_hot_added = false;
@@ -868,6 +869,9 @@ static void qbus_initfn(Object *obj)
     BusState *bus = BUS(obj);
 
     QTAILQ_INIT(&bus->children);
+    object_property_add_link(obj, QDEV_HOTPLUG_HANDLER_PROPERTY,
+                             TYPE_HOTPLUG_HANDLER,
+                             (Object **)&bus->hotplug_handler, NULL);
 }
 
 static char *default_bus_get_fw_dev_path(DeviceState *dev)
include/hw/qdev-core.h +5 -0
index f2043a6..684a5da 100644
@@ -8,6 +8,7 @@
 #include "qom/object.h"
 #include "hw/irq.h"
 #include "qapi/error.h"
+#include "hw/hotplug.h"
 
 enum {
     DEV_NVECTORS_UNSPECIFIED = -1,
@@ -169,14 +170,18 @@ typedef struct BusChild {
     QTAILQ_ENTRY(BusChild) sibling;
 } BusChild;
 
+#define QDEV_HOTPLUG_HANDLER_PROPERTY "hotplug-handler"
+
 /**
  * BusState:
+ * @hotplug_device: link to a hotplug device associated with bus.
  */
 struct BusState {
     Object obj;
     DeviceState *parent;
     const char *name;
     int allow_hotplug;
+    HotplugHandler *hotplug_handler;
     int max_index;
     QTAILQ_HEAD(ChildrenHead, BusChild) children;
     QLIST_ENTRY(BusState) sibling;
Signature · 1 line
1.8.3.1
#9 lore raw

Re: [Qemu-devel] [PATCH 04/11] qdev: add to BusState "hotplug-handler" link

Quoted 26 lines · Igor Mammedov

It will allow to reuse field with different BUSes, reducing code duplication. Field is intended fot replacing 'hotplug_qdev' field in PCIBus and also will allow to avoid adding equivalent field to DimmBus with possiblitity to refactor other BUSes to use it instead of custom field.
In addition once all users of allow_hotplug field are converted to new API, link could replace allow_hotplug in qdev hotplug code.

---
 hw/core/qdev.c         | 4 ++++
 include/hw/qdev-core.h | 5 +++++
 2 files changed, 9 insertions(+)
hw/core/qdev.c +1 -0
index e374a93..25c2d2c 100644
@@ -32,6 +32,7 @@
 #include "qapi/visitor.h"
 #include "qapi/qmp/qjson.h"
 #include "monitor/monitor.h"
+#include "hw/hotplug.h"
 
 int qdev_hotplug = 0;
 static bool qdev_hot_added = false;

hw/core/qdev.c static void qbus_initfn(Object *obj) · line 874

Quoted 6 lines
hw/core/qdev.c +3 -0
@@ -868,6 +869,9 @@ static void qbus_initfn(Object *obj)
     BusState *bus = BUS(obj);
 
     QTAILQ_INIT(&bus->children);
+    object_property_add_link(obj, QDEV_HOTPLUG_HANDLER_PROPERTY,
+                             TYPE_HOTPLUG_HANDLER,
+                             (Object **)&bus->hotplug_handler, NULL);

I think failure of this is fatal. When the patches go through, probably want to &error_abort.

Quoted 38 lines
hw/core/qdev.c +0 -0
 }
 
 static char *default_bus_get_fw_dev_path(DeviceState *dev)
include/hw/qdev-core.h +5 -1
index f2043a6..684a5da 100644
@@ -8,6 +8,7 @@
 #include "qom/object.h"
 #include "hw/irq.h"
 #include "qapi/error.h"
+#include "hw/hotplug.h"
 
 enum {
     DEV_NVECTORS_UNSPECIFIED = -1,
@@ -169,14 +170,18 @@ typedef struct BusChild {
     QTAILQ_ENTRY(BusChild) sibling;
 } BusChild;
 
+#define QDEV_HOTPLUG_HANDLER_PROPERTY "hotplug-handler"
+
 /**
  * BusState:
+ * @hotplug_device: link to a hotplug device associated with bus.
  */
 struct BusState {
     Object obj;
     DeviceState *parent;
     const char *name;
     int allow_hotplug;
+    HotplugHandler *hotplug_handler;
     int max_index;
     QTAILQ_HEAD(ChildrenHead, BusChild) children;
     QLIST_ENTRY(BusState) sibling;
--

1.8.3.1

#10 lore raw

Re: [Qemu-devel] [PATCH 04/11] qdev: add to BusState "hotplug-handler" link

Quoted 29 lines · On Sat, 14 Dec 2013 17:05:57 +1000 Peter Crosthwaite…
Quoted 28 lines · Igor Mammedov

It will allow to reuse field with different BUSes, reducing code duplication. Field is intended fot replacing 'hotplug_qdev' field in PCIBus and also will allow to avoid adding equivalent field to DimmBus with possiblitity to refactor other BUSes to use it instead of custom field.
In addition once all users of allow_hotplug field are converted to new API, link could replace allow_hotplug in qdev hotplug code.

---
 hw/core/qdev.c         | 4 ++++
 include/hw/qdev-core.h | 5 +++++
 2 files changed, 9 insertions(+)
hw/core/qdev.c +1 -0
index e374a93..25c2d2c 100644
@@ -32,6 +32,7 @@
 #include "qapi/visitor.h"
 #include "qapi/qmp/qjson.h"
 #include "monitor/monitor.h"
+#include "hw/hotplug.h"
 
 int qdev_hotplug = 0;
 static bool qdev_hot_added = false;
@@ -868,6 +869,9 @@ static void qbus_initfn(Object *obj)
     BusState *bus = BUS(obj);
 
Quoted 7 lines

hw/core/qdev.c static void qbus_initfn(Object *obj) · line 874

Quoted 4 lines
hw/core/qdev.c +3 -0
     QTAILQ_INIT(&bus->children);
+    object_property_add_link(obj, QDEV_HOTPLUG_HANDLER_PROPERTY,
+                             TYPE_HOTPLUG_HANDLER,
+                             (Object **)&bus->hotplug_handler, NULL);

I think failure of this is fatal. When the patches go through, probably want to &error_abort.

That's just another example of we don't care about error that gave birth to &error_abort topic. But I think is a drawback if error-less initfn. We have a lot of such usage without a way to report error, that's why "we don't care about error" was invented I guess.

But more to the point, bus could be created during hotplug, and we do not want to abort guest, instead hotplug operation should fail. In this case following setting of QDEV_HOTPLUG_HANDLER_PROPERTY will fail if object_property_add_link() failed before.

Quoted 39 lines
Quoted 39 lines
hw/core/qdev.c +0 -0
 }
 
 static char *default_bus_get_fw_dev_path(DeviceState *dev)
include/hw/qdev-core.h +5 -1
index f2043a6..684a5da 100644
@@ -8,6 +8,7 @@
 #include "qom/object.h"
 #include "hw/irq.h"
 #include "qapi/error.h"
+#include "hw/hotplug.h"
 
 enum {
     DEV_NVECTORS_UNSPECIFIED = -1,
@@ -169,14 +170,18 @@ typedef struct BusChild {
     QTAILQ_ENTRY(BusChild) sibling;
 } BusChild;
 
+#define QDEV_HOTPLUG_HANDLER_PROPERTY "hotplug-handler"
+
 /**
  * BusState:
+ * @hotplug_device: link to a hotplug device associated with bus.
  */
 struct BusState {
     Object obj;
     DeviceState *parent;
     const char *name;
     int allow_hotplug;
+    HotplugHandler *hotplug_handler;
     int max_index;
     QTAILQ_HEAD(ChildrenHead, BusChild) children;
     QLIST_ENTRY(BusState) sibling;
--

1.8.3.1

#11 lore raw

Re: [Qemu-devel] [PATCH 04/11] qdev: add to BusState "hotplug-handler" link

Quoted 3 lines · On Mon, 16 Dec 2013 16:26:25 +0100 Igor Mammedov <[email protected]> wrote:

On Sat, 14 Dec 2013 17:05:57 +1000
Peter Crosthwaite <[email protected]> wrote:

[...]

Quoted 8 lines
Quoted 8 lines

hw/core/qdev.c static void qbus_initfn(Object *obj) · line 874

Quoted 5 lines
     QTAILQ_INIT(&bus->children);
hw/core/qdev.c +3 -0
+    object_property_add_link(obj, QDEV_HOTPLUG_HANDLER_PROPERTY,
+                             TYPE_HOTPLUG_HANDLER,
+                             (Object **)&bus->hotplug_handler, NULL);

I think failure of this is fatal. When the patches go through, probably want to &error_abort.

[...]

Quoted 5 lines

But more to the point, bus could be created during hotplug, and we do not want to abort guest, instead hotplug operation should fail. In this case following setting of QDEV_HOTPLUG_HANDLER_PROPERTY will fail if object_property_add_link() failed before.

In following patches there is a couple places asking for being replaced with &error_abort, so if your patches get in before respin, I'll gladly use &error_abort there.

#12 lore raw

PATCH v3 5/11 qdev: add "hotpluggable" property to Device

In reply to Igor Mammedov

Currently it's possible to make PCIDevice not hotpluggable by using no_hotplug field of PCIDeviceClass. However it limits this only to PCI devices and prevents from generalizing hotplug code.

So add similar field to DeviceClass so it could be reused with other Devices and would allow to replace PCI specific hotplug callbacks with generic implementation.

In addition expose field as "hotpluggable" readonly property, to make it possible to get it via QOM interface.

Make DeviceClass hotpluggable by default as it was assumed before.

---
v4:
* s/hotplugable/hotpluggable/

v3:
* make DeviceClass hotpluggable by default
Since PCIDevice still uses internal no_hotlpug checks it shouldn't reggress. And follow up patch that converts PCIDevices to use "hotpluggable" property will take care about not hotpluggable PCI devices explicitly setting "hotpluggable" to false in their class_init().

* move generic hotplug checks from
"7/11 qdev:pci: refactor PCIDevice to use generic "hotplugable" property" to this patch

---
 hw/core/qdev.c         | 29 +++++++++++++++++++++++++++++
 include/hw/qdev-core.h |  3 +++
 2 files changed, 32 insertions(+)
hw/core/qdev.c +29 -0
index 25c2d2c..9418fea 100644
@@ -215,6 +215,12 @@ void qdev_unplug(DeviceState *dev, Error **errp)
     }
     assert(dc->unplug != NULL);
 
+    if (!dc->hotpluggable) {
+        error_set(errp, QERR_DEVICE_NO_HOTPLUG,
+                  object_get_typename(OBJECT(dev)));
+        return;
+    }
+
     qdev_hot_removed = true;
 
     if (dc->unplug(dev) < 0) {
@@ -679,6 +685,11 @@ static void device_set_realized(Object *obj, bool value, Error **err)
     DeviceClass *dc = DEVICE_GET_CLASS(dev);
     Error *local_err = NULL;
 
+    if (dev->hotplugged && !dc->hotpluggable) {
+        error_set(err, QERR_DEVICE_NO_HOTPLUG, object_get_typename(obj));
+        return;
+    }
+
     if (value && !dev->realized) {
         if (!obj->parent && local_err == NULL) {
             static int unattached_count;
@@ -719,6 +730,14 @@ static void device_set_realized(Object *obj, bool value, Error **err)
     dev->realized = value;
 }
 
+static bool device_get_hotpluggable(Object *obj, Error **err)
+{
+    DeviceClass *dc = DEVICE_GET_CLASS(obj);
+    DeviceState *dev = DEVICE(obj);
+
+    return dc->hotpluggable && dev->parent_bus->allow_hotplug;
+}
+
 static void device_initfn(Object *obj)
 {
     DeviceState *dev = DEVICE(obj);
@@ -736,6 +755,8 @@ static void device_initfn(Object *obj)
 
     object_property_add_bool(obj, "realized",
                              device_get_realized, device_set_realized, NULL);
+    object_property_add_bool(obj, "hotpluggable",
+                             device_get_hotpluggable, NULL, NULL);
 
     class = object_get_class(OBJECT(dev));
     do {
@@ -784,6 +805,14 @@ static void device_class_base_init(ObjectClass *class, void *data)
      * so do not propagate them to the subclasses.
      */
     klass->props = NULL;
+
+    /* by default all devices were considered as hotpluggable,
+     * so with intent to check it in generic qdev_unplug() /
+     * device_set_realized() functions make every device
+     * hotpluggable. Devices that shouldn't be hoplugable,
+     * should override it in their class_init()
+     */
+    klass->hotpluggable = true;
 }
 
 static void device_unparent(Object *obj)
include/hw/qdev-core.h +3 -0
index 684a5da..04bbef4 100644
@@ -50,6 +50,8 @@ struct VMStateDescription;
  * is changed to %true. Deprecated, new types inheriting directly from
  * TYPE_DEVICE should use @realize instead, new leaf types should consult
  * their respective parent type.
+ * @hotpluggable: booleean indicating if #DeviceClass is hotpluggable, available
+ * as readonly "hotpluggable" property of #DeviceState instance
  *
  * # Realization #
  * Devices are constructed in two stages,
@@ -99,6 +101,7 @@ typedef struct DeviceClass {
     const char *desc;
     Property *props;
     int no_user;
+    bool hotpluggable;
 
     /* callbacks */
     void (*reset)(DeviceState *dev);
Signature · 1 line
1.8.3.1
#13 lore raw

Re: [Qemu-devel] [PATCH 05/11] qdev: add "hotpluggable" property to Device

Quoted 106 lines · Igor Mammedov

Currently it's possible to make PCIDevice not hotpluggable by using no_hotplug field of PCIDeviceClass. However it limits this only to PCI devices and prevents from generalizing hotplug code.

So add similar field to DeviceClass so it could be reused with other Devices and would allow to replace PCI specific hotplug callbacks with generic implementation.

In addition expose field as "hotpluggable" readonly property, to make it possible to get it via QOM interface.

Make DeviceClass hotpluggable by default as it was assumed before.

---
v4:
* s/hotplugable/hotpluggable/

v3:
* make DeviceClass hotpluggable by default
Since PCIDevice still uses internal no_hotlpug checks it shouldn't reggress. And follow up patch that converts PCIDevices to use "hotpluggable" property will take care about not hotpluggable PCI devices explicitly setting "hotpluggable" to false in their class_init().

* move generic hotplug checks from
"7/11 qdev:pci: refactor PCIDevice to use generic "hotplugable" property" to this patch

---
 hw/core/qdev.c         | 29 +++++++++++++++++++++++++++++
 include/hw/qdev-core.h |  3 +++
 2 files changed, 32 insertions(+)
hw/core/qdev.c +29 -0
index 25c2d2c..9418fea 100644
@@ -215,6 +215,12 @@ void qdev_unplug(DeviceState *dev, Error **errp)
     }
     assert(dc->unplug != NULL);
 
+    if (!dc->hotpluggable) {
+        error_set(errp, QERR_DEVICE_NO_HOTPLUG,
+                  object_get_typename(OBJECT(dev)));
+        return;
+    }
+
     qdev_hot_removed = true;
 
     if (dc->unplug(dev) < 0) {
@@ -679,6 +685,11 @@ static void device_set_realized(Object *obj, bool value, Error **err)
     DeviceClass *dc = DEVICE_GET_CLASS(dev);
     Error *local_err = NULL;
 
+    if (dev->hotplugged && !dc->hotpluggable) {
+        error_set(err, QERR_DEVICE_NO_HOTPLUG, object_get_typename(obj));
+        return;
+    }
+
     if (value && !dev->realized) {
         if (!obj->parent && local_err == NULL) {
             static int unattached_count;
@@ -719,6 +730,14 @@ static void device_set_realized(Object *obj, bool value, Error **err)
     dev->realized = value;
 }
 
+static bool device_get_hotpluggable(Object *obj, Error **err)
+{
+    DeviceClass *dc = DEVICE_GET_CLASS(obj);
+    DeviceState *dev = DEVICE(obj);
+
+    return dc->hotpluggable && dev->parent_bus->allow_hotplug;
+}
+
 static void device_initfn(Object *obj)
 {
     DeviceState *dev = DEVICE(obj);
@@ -736,6 +755,8 @@ static void device_initfn(Object *obj)
 
     object_property_add_bool(obj, "realized",
                              device_get_realized, device_set_realized, NULL);
+    object_property_add_bool(obj, "hotpluggable",
+                             device_get_hotpluggable, NULL, NULL);
 
     class = object_get_class(OBJECT(dev));
     do {
@@ -784,6 +805,14 @@ static void device_class_base_init(ObjectClass *class, void *data)
      * so do not propagate them to the subclasses.
      */
     klass->props = NULL;
+
+    /* by default all devices were considered as hotpluggable,
+     * so with intent to check it in generic qdev_unplug() /
+     * device_set_realized() functions make every device
+     * hotpluggable. Devices that shouldn't be hoplugable,
+     * should override it in their class_init()
+     */
+    klass->hotpluggable = true;
 }
 
 static void device_unparent(Object *obj)
include/hw/qdev-core.h +0 -0
index 684a5da..04bbef4 100644

include/hw/qdev-core.h struct VMStateDescription; · line 53

Quoted 4 lines
include/hw/qdev-core.h +1 -0
@@ -50,6 +50,8 @@ struct VMStateDescription;
  * is changed to %true. Deprecated, new types inheriting directly from
  * TYPE_DEVICE should use @realize instead, new leaf types should consult
  * their respective parent type.
+ * @hotpluggable: booleean indicating if #DeviceClass is hotpluggable, available

"boolean". Although the fact that its bool is implicit just by saying:

"@hotpluggable: indicates if #DeviceClass is hotpluggable ...

Regards,
Peter

Quoted 16 lines
include/hw/qdev-core.h +2 -1
+ * as readonly "hotpluggable" property of #DeviceState instance
  *
  * # Realization #
  * Devices are constructed in two stages,
@@ -99,6 +101,7 @@ typedef struct DeviceClass {
     const char *desc;
     Property *props;
     int no_user;
+    bool hotpluggable;
 
     /* callbacks */
     void (*reset)(DeviceState *dev);
--

1.8.3.1

#14 lore raw

Re: [Qemu-devel] [PATCH 05/11] qdev: add "hotpluggable" property to Device

Quoted 114 lines · On Sat, 14 Dec 2013 17:10:35 +1000 Peter Crosthwaite…
Quoted 106 lines · Igor Mammedov

Currently it's possible to make PCIDevice not hotpluggable by using no_hotplug field of PCIDeviceClass. However it limits this only to PCI devices and prevents from generalizing hotplug code.

So add similar field to DeviceClass so it could be reused with other Devices and would allow to replace PCI specific hotplug callbacks with generic implementation.

In addition expose field as "hotpluggable" readonly property, to make it possible to get it via QOM interface.

Make DeviceClass hotpluggable by default as it was assumed before.

---
v4:
* s/hotplugable/hotpluggable/

v3:
* make DeviceClass hotpluggable by default
Since PCIDevice still uses internal no_hotlpug checks it shouldn't reggress. And follow up patch that converts PCIDevices to use "hotpluggable" property will take care about not hotpluggable PCI devices explicitly setting "hotpluggable" to false in their class_init().

* move generic hotplug checks from
"7/11 qdev:pci: refactor PCIDevice to use generic "hotplugable" property" to this patch

---
 hw/core/qdev.c         | 29 +++++++++++++++++++++++++++++
 include/hw/qdev-core.h |  3 +++
 2 files changed, 32 insertions(+)
hw/core/qdev.c +29 -0
index 25c2d2c..9418fea 100644
@@ -215,6 +215,12 @@ void qdev_unplug(DeviceState *dev, Error **errp)
     }
     assert(dc->unplug != NULL);
 
+    if (!dc->hotpluggable) {
+        error_set(errp, QERR_DEVICE_NO_HOTPLUG,
+                  object_get_typename(OBJECT(dev)));
+        return;
+    }
+
     qdev_hot_removed = true;
 
     if (dc->unplug(dev) < 0) {
@@ -679,6 +685,11 @@ static void device_set_realized(Object *obj, bool value, Error **err)
     DeviceClass *dc = DEVICE_GET_CLASS(dev);
     Error *local_err = NULL;
 
+    if (dev->hotplugged && !dc->hotpluggable) {
+        error_set(err, QERR_DEVICE_NO_HOTPLUG, object_get_typename(obj));
+        return;
+    }
+
     if (value && !dev->realized) {
         if (!obj->parent && local_err == NULL) {
             static int unattached_count;
@@ -719,6 +730,14 @@ static void device_set_realized(Object *obj, bool value, Error **err)
     dev->realized = value;
 }
 
+static bool device_get_hotpluggable(Object *obj, Error **err)
+{
+    DeviceClass *dc = DEVICE_GET_CLASS(obj);
+    DeviceState *dev = DEVICE(obj);
+
+    return dc->hotpluggable && dev->parent_bus->allow_hotplug;
+}
+
 static void device_initfn(Object *obj)
 {
     DeviceState *dev = DEVICE(obj);
@@ -736,6 +755,8 @@ static void device_initfn(Object *obj)
 
     object_property_add_bool(obj, "realized",
                              device_get_realized, device_set_realized, NULL);
+    object_property_add_bool(obj, "hotpluggable",
+                             device_get_hotpluggable, NULL, NULL);
 
     class = object_get_class(OBJECT(dev));
     do {
@@ -784,6 +805,14 @@ static void device_class_base_init(ObjectClass *class, void *data)
      * so do not propagate them to the subclasses.
      */
     klass->props = NULL;
+
+    /* by default all devices were considered as hotpluggable,
+     * so with intent to check it in generic qdev_unplug() /
+     * device_set_realized() functions make every device
+     * hotpluggable. Devices that shouldn't be hoplugable,
+     * should override it in their class_init()
+     */
+    klass->hotpluggable = true;
 }
 
 static void device_unparent(Object *obj)
include/hw/qdev-core.h +0 -0
index 684a5da..04bbef4 100644

include/hw/qdev-core.h struct VMStateDescription; · line 53

Quoted 4 lines
include/hw/qdev-core.h +1 -0
@@ -50,6 +50,8 @@ struct VMStateDescription;
  * is changed to %true. Deprecated, new types inheriting directly from
  * TYPE_DEVICE should use @realize instead, new leaf types should consult
  * their respective parent type.
+ * @hotpluggable: booleean indicating if #DeviceClass is hotpluggable, available

"boolean". Although the fact that its bool is implicit just by saying:

Quoted 1 line

"@hotpluggable: indicates if #DeviceClass is hotpluggable ...

sure

Quoted 20 lines

Regards,
Peter

Quoted 16 lines
include/hw/qdev-core.h +2 -1
+ * as readonly "hotpluggable" property of #DeviceState instance
  *
  * # Realization #
  * Devices are constructed in two stages,
@@ -99,6 +101,7 @@ typedef struct DeviceClass {
     const char *desc;
     Property *props;
     int no_user;
+    bool hotpluggable;
 
     /* callbacks */
     void (*reset)(DeviceState *dev);
--

1.8.3.1

#15 lore raw

Re: [Qemu-devel] [PATCH 05/11] qdev: add "hotpluggable" property to Device

In reply to Igor Mammedov
Quoted 8 lines · Igor Mammedov

Currently it's possible to make PCIDevice not hotpluggable by using no_hotplug field of PCIDeviceClass. However it limits this only to PCI devices and prevents from generalizing hotplug code.

So add similar field to DeviceClass so it could be reused with other Devices and would allow to replace PCI specific hotplug callbacks with generic implementation.

Quoted 6 lines

In addition expose field as "hotpluggable" readonly property, to make it possible to get it via QOM interface.

Make DeviceClass hotpluggable by default as it was assumed before.

Pushing stuff to a base class in order to reuse code isn't the right way to model things.

hotplug is a bus concept. It doesn't apply widely to all possible devices.

Regards,

Anthony Liguori

Quoted 111 lines

---
v4:
* s/hotplugable/hotpluggable/

v3:
* make DeviceClass hotpluggable by default
Since PCIDevice still uses internal no_hotlpug checks it shouldn't reggress. And follow up patch that converts PCIDevices to use "hotpluggable" property will take care about not hotpluggable PCI devices explicitly setting "hotpluggable" to false in their class_init().

* move generic hotplug checks from
"7/11 qdev:pci: refactor PCIDevice to use generic "hotplugable" property" to this patch

---
 hw/core/qdev.c         | 29 +++++++++++++++++++++++++++++
 include/hw/qdev-core.h |  3 +++
 2 files changed, 32 insertions(+)
hw/core/qdev.c +29 -0
index 25c2d2c..9418fea 100644
@@ -215,6 +215,12 @@ void qdev_unplug(DeviceState *dev, Error **errp)
     }
     assert(dc->unplug != NULL);
 
+    if (!dc->hotpluggable) {
+        error_set(errp, QERR_DEVICE_NO_HOTPLUG,
+                  object_get_typename(OBJECT(dev)));
+        return;
+    }
+
     qdev_hot_removed = true;
 
     if (dc->unplug(dev) < 0) {
@@ -679,6 +685,11 @@ static void device_set_realized(Object *obj, bool value, Error **err)
     DeviceClass *dc = DEVICE_GET_CLASS(dev);
     Error *local_err = NULL;
 
+    if (dev->hotplugged && !dc->hotpluggable) {
+        error_set(err, QERR_DEVICE_NO_HOTPLUG, object_get_typename(obj));
+        return;
+    }
+
     if (value && !dev->realized) {
         if (!obj->parent && local_err == NULL) {
             static int unattached_count;
@@ -719,6 +730,14 @@ static void device_set_realized(Object *obj, bool value, Error **err)
     dev->realized = value;
 }
 
+static bool device_get_hotpluggable(Object *obj, Error **err)
+{
+    DeviceClass *dc = DEVICE_GET_CLASS(obj);
+    DeviceState *dev = DEVICE(obj);
+
+    return dc->hotpluggable && dev->parent_bus->allow_hotplug;
+}
+
 static void device_initfn(Object *obj)
 {
     DeviceState *dev = DEVICE(obj);
@@ -736,6 +755,8 @@ static void device_initfn(Object *obj)
 
     object_property_add_bool(obj, "realized",
                              device_get_realized, device_set_realized, NULL);
+    object_property_add_bool(obj, "hotpluggable",
+                             device_get_hotpluggable, NULL, NULL);
 
     class = object_get_class(OBJECT(dev));
     do {
@@ -784,6 +805,14 @@ static void device_class_base_init(ObjectClass *class, void *data)
      * so do not propagate them to the subclasses.
      */
     klass->props = NULL;
+
+    /* by default all devices were considered as hotpluggable,
+     * so with intent to check it in generic qdev_unplug() /
+     * device_set_realized() functions make every device
+     * hotpluggable. Devices that shouldn't be hoplugable,
+     * should override it in their class_init()
+     */
+    klass->hotpluggable = true;
 }
 
 static void device_unparent(Object *obj)
include/hw/qdev-core.h +3 -1
index 684a5da..04bbef4 100644
@@ -50,6 +50,8 @@ struct VMStateDescription;
  * is changed to %true. Deprecated, new types inheriting directly from
  * TYPE_DEVICE should use @realize instead, new leaf types should consult
  * their respective parent type.
+ * @hotpluggable: booleean indicating if #DeviceClass is hotpluggable, available
+ * as readonly "hotpluggable" property of #DeviceState instance
  *
  * # Realization #
  * Devices are constructed in two stages,
@@ -99,6 +101,7 @@ typedef struct DeviceClass {
     const char *desc;
     Property *props;
     int no_user;
+    bool hotpluggable;
 
     /* callbacks */
     void (*reset)(DeviceState *dev);
--

1.8.3.1

#16 lore raw

PATCH v3 6/11 hw/acpi: move typeinfo to the file end

In reply to Igor Mammedov

do so to avoid not necessary forward declarations and place typeinfo registration at the file end where it's usualy expected.

---
 hw/acpi/piix4.c | 80 ++++++++++++++++++++++++++++-----------------------------
 1 file changed, 40 insertions(+), 40 deletions(-)
hw/acpi/piix4.c +40 -40
index 93849c8..8084a60 100644
@@ -523,46 +523,6 @@ i2c_bus *piix4_pm_init(PCIBus *bus, int devfn, uint32_t smb_io_base,
     return s->smb.smbus;
 }
 
-static Property piix4_pm_properties[] = {
-    DEFINE_PROP_UINT32("smb_io_base", PIIX4PMState, smb_io_base, 0),
-    DEFINE_PROP_UINT8(ACPI_PM_PROP_S3_DISABLED, PIIX4PMState, disable_s3, 0),
-    DEFINE_PROP_UINT8(ACPI_PM_PROP_S4_DISABLED, PIIX4PMState, disable_s4, 0),
-    DEFINE_PROP_UINT8(ACPI_PM_PROP_S4_VAL, PIIX4PMState, s4_val, 2),
-    DEFINE_PROP_END_OF_LIST(),
-};
-
-static void piix4_pm_class_init(ObjectClass *klass, void *data)
-{
-    DeviceClass *dc = DEVICE_CLASS(klass);
-    PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
-
-    k->no_hotplug = 1;
-    k->init = piix4_pm_initfn;
-    k->config_write = pm_write_config;
-    k->vendor_id = PCI_VENDOR_ID_INTEL;
-    k->device_id = PCI_DEVICE_ID_INTEL_82371AB_3;
-    k->revision = 0x03;
-    k->class_id = PCI_CLASS_BRIDGE_OTHER;
-    dc->desc = "PM";
-    dc->no_user = 1;
-    dc->vmsd = &vmstate_acpi;
-    dc->props = piix4_pm_properties;
-}
-
-static const TypeInfo piix4_pm_info = {
-    .name          = TYPE_PIIX4_PM,
-    .parent        = TYPE_PCI_DEVICE,
-    .instance_size = sizeof(PIIX4PMState),
-    .class_init    = piix4_pm_class_init,
-};
-
-static void piix4_pm_register_types(void)
-{
-    type_register_static(&piix4_pm_info);
-}
-
-type_init(piix4_pm_register_types)
-
 static uint64_t gpe_readb(void *opaque, hwaddr addr, unsigned width)
 {
     PIIX4PMState *s = opaque;
@@ -772,3 +732,43 @@ static int piix4_device_hotplug(DeviceState *qdev, PCIDevice *dev,
 
     return 0;
 }
+
+static Property piix4_pm_properties[] = {
+    DEFINE_PROP_UINT32("smb_io_base", PIIX4PMState, smb_io_base, 0),
+    DEFINE_PROP_UINT8(ACPI_PM_PROP_S3_DISABLED, PIIX4PMState, disable_s3, 0),
+    DEFINE_PROP_UINT8(ACPI_PM_PROP_S4_DISABLED, PIIX4PMState, disable_s4, 0),
+    DEFINE_PROP_UINT8(ACPI_PM_PROP_S4_VAL, PIIX4PMState, s4_val, 2),
+    DEFINE_PROP_END_OF_LIST(),
+};
+
+static void piix4_pm_class_init(ObjectClass *klass, void *data)
+{
+    DeviceClass *dc = DEVICE_CLASS(klass);
+    PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
+
+    k->no_hotplug = 1;
+    k->init = piix4_pm_initfn;
+    k->config_write = pm_write_config;
+    k->vendor_id = PCI_VENDOR_ID_INTEL;
+    k->device_id = PCI_DEVICE_ID_INTEL_82371AB_3;
+    k->revision = 0x03;
+    k->class_id = PCI_CLASS_BRIDGE_OTHER;
+    dc->desc = "PM";
+    dc->no_user = 1;
+    dc->vmsd = &vmstate_acpi;
+    dc->props = piix4_pm_properties;
+}
+
+static const TypeInfo piix4_pm_info = {
+    .name          = TYPE_PIIX4_PM,
+    .parent        = TYPE_PCI_DEVICE,
+    .instance_size = sizeof(PIIX4PMState),
+    .class_init    = piix4_pm_class_init,
+};
+
+static void piix4_pm_register_types(void)
+{
+    type_register_static(&piix4_pm_info);
+}
+
+type_init(piix4_pm_register_types)
Signature · 1 line
1.8.3.1
#17 lore raw

PATCH v3 7/11 qdev:pci: refactor PCIDevice to use generic "hotpluggable" property

In reply to Igor Mammedov

Get rid of PCIDevice specific PCIDeviceClass.no_hotplug and use generic DeviceClass.hotpluggable field instead.

---
v2:
* move generic hotplug checks to
"qdev: add "hotpluggable" property to Device" patch
* s/hotplugable/hotpluggable/

---
 hw/acpi/piix4.c         | 10 +++++-----
 hw/display/cirrus_vga.c |  2 +-
 hw/display/qxl.c        |  2 +-
 hw/display/vga-pci.c    |  2 +-
 hw/display/vmware_vga.c |  2 +-
 hw/i386/acpi-build.c    |  6 +++---
 hw/ide/piix.c           |  4 ++--
 hw/isa/piix4.c          |  2 +-
 hw/pci-host/piix.c      |  6 +++---
 hw/pci/pci.c            | 11 +----------
 hw/usb/hcd-ehci-pci.c   |  2 +-
 hw/usb/hcd-ohci.c       |  2 +-
 hw/usb/hcd-uhci.c       |  2 +-
 hw/usb/hcd-xhci.c       |  2 +-
 include/hw/pci/pci.h    |  3 ---
 15 files changed, 23 insertions(+), 35 deletions(-)
hw/acpi/piix4.c +5 -5
index 8084a60..fff2126 100644
@@ -323,9 +323,9 @@ static void acpi_piix_eject_slot(PIIX4PMState *s, unsigned slots)
     QTAILQ_FOREACH_SAFE(kid, &bus->children, sibling, next) {
         DeviceState *qdev = kid->child;
         PCIDevice *dev = PCI_DEVICE(qdev);
-        PCIDeviceClass *pc = PCI_DEVICE_GET_CLASS(dev);
+        DeviceClass *dc = DEVICE_GET_CLASS(dev);
         if (PCI_SLOT(dev->devfn) == slot) {
-            if (pc->no_hotplug) {
+            if (!dc->hotpluggable) {
                 slot_free = false;
             } else {
                 object_unparent(OBJECT(qdev));
@@ -353,10 +353,10 @@ static void piix4_update_hotplug(PIIX4PMState *s)
     QTAILQ_FOREACH_SAFE(kid, &bus->children, sibling, next) {
         DeviceState *qdev = kid->child;
         PCIDevice *pdev = PCI_DEVICE(qdev);
-        PCIDeviceClass *pc = PCI_DEVICE_GET_CLASS(pdev);
+        DeviceClass *dc = DEVICE_GET_CLASS(qdev);
         int slot = PCI_SLOT(pdev->devfn);
 
-        if (pc->no_hotplug) {
+        if (!dc->hotpluggable) {
             s->pci0_hotplug_enable &= ~(1U << slot);
         }
 
@@ -746,7 +746,6 @@ static void piix4_pm_class_init(ObjectClass *klass, void *data)
     DeviceClass *dc = DEVICE_CLASS(klass);
     PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
 
-    k->no_hotplug = 1;
     k->init = piix4_pm_initfn;
     k->config_write = pm_write_config;
     k->vendor_id = PCI_VENDOR_ID_INTEL;
@@ -757,6 +756,7 @@ static void piix4_pm_class_init(ObjectClass *klass, void *data)
     dc->no_user = 1;
     dc->vmsd = &vmstate_acpi;
     dc->props = piix4_pm_properties;
+    dc->hotpluggable = false;
 }
 
 static const TypeInfo piix4_pm_info = {
hw/display/cirrus_vga.c +1 -1
index e4c345f..3a8fc0b 100644
@@ -2996,7 +2996,6 @@ static void cirrus_vga_class_init(ObjectClass *klass, void *data)
     DeviceClass *dc = DEVICE_CLASS(klass);
     PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
 
-    k->no_hotplug = 1;
     k->init = pci_cirrus_vga_initfn;
     k->romfile = VGABIOS_CIRRUS_FILENAME;
     k->vendor_id = PCI_VENDOR_ID_CIRRUS;
@@ -3006,6 +3005,7 @@ static void cirrus_vga_class_init(ObjectClass *klass, void *data)
     dc->desc = "Cirrus CLGD 54xx VGA";
     dc->vmsd = &vmstate_pci_cirrus_vga;
     dc->props = pci_vga_cirrus_properties;
+    dc->hotpluggable = false;
 }
 
 static const TypeInfo cirrus_vga_info = {
hw/display/qxl.c +1 -1
index efdefd6..abb45ce 100644
@@ -2289,7 +2289,6 @@ static void qxl_primary_class_init(ObjectClass *klass, void *data)
     DeviceClass *dc = DEVICE_CLASS(klass);
     PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
 
-    k->no_hotplug = 1;
     k->init = qxl_init_primary;
     k->romfile = "vgabios-qxl.bin";
     k->vendor_id = REDHAT_PCI_VENDOR_ID;
@@ -2300,6 +2299,7 @@ static void qxl_primary_class_init(ObjectClass *klass, void *data)
     dc->reset = qxl_reset_handler;
     dc->vmsd = &qxl_vmstate;
     dc->props = qxl_properties;
+    dc->hotpluggable = false;
 }
 
 static const TypeInfo qxl_primary_info = {
hw/display/vga-pci.c +1 -1
index b3a45c8..f74fc43 100644
@@ -190,7 +190,6 @@ static void vga_class_init(ObjectClass *klass, void *data)
     DeviceClass *dc = DEVICE_CLASS(klass);
     PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
 
-    k->no_hotplug = 1;
     k->init = pci_std_vga_initfn;
     k->romfile = "vgabios-stdvga.bin";
     k->vendor_id = PCI_VENDOR_ID_QEMU;
@@ -198,6 +197,7 @@ static void vga_class_init(ObjectClass *klass, void *data)
     k->class_id = PCI_CLASS_DISPLAY_VGA;
     dc->vmsd = &vmstate_vga_pci;
     dc->props = vga_pci_properties;
+    dc->hotpluggable = false;
     set_bit(DEVICE_CATEGORY_DISPLAY, dc->categories);
 }
 
hw/display/vmware_vga.c +1 -1
index aba292c..334e718 100644
@@ -1296,7 +1296,6 @@ static void vmsvga_class_init(ObjectClass *klass, void *data)
     DeviceClass *dc = DEVICE_CLASS(klass);
     PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
 
-    k->no_hotplug = 1;
     k->init = pci_vmsvga_initfn;
     k->romfile = "vgabios-vmware.bin";
     k->vendor_id = PCI_VENDOR_ID_VMWARE;
@@ -1307,6 +1306,7 @@ static void vmsvga_class_init(ObjectClass *klass, void *data)
     dc->reset = vmsvga_reset;
     dc->vmsd = &vmstate_vmware_vga;
     dc->props = vga_vmware_properties;
+    dc->hotpluggable = false;
     set_bit(DEVICE_CATEGORY_DISPLAY, dc->categories);
 }
 
hw/i386/acpi-build.c +3 -3
index befc39f..821589d 100644
@@ -186,16 +186,16 @@ static void acpi_get_hotplug_info(AcpiMiscInfo *misc)
            DIV_ROUND_UP(PCI_SLOT_MAX, BITS_PER_BYTE));
 
     for (i = 0; i < ARRAY_SIZE(bus->devices); ++i) {
-        PCIDeviceClass *pc;
+        DeviceClass *dc;
         PCIDevice *pdev = bus->devices[i];
 
         if (!pdev) {
             continue;
         }
 
-        pc = PCI_DEVICE_GET_CLASS(pdev);
+        dc = DEVICE_GET_CLASS(pdev);
 
-        if (pc->no_hotplug) {
+        if (!dc->hotpluggable) {
             int slot = PCI_SLOT(i);
 
             clear_bit(slot, misc->slot_hotplug_enable);
hw/ide/piix.c +2 -2
index ab36749..df127dd 100644
@@ -241,7 +241,6 @@ static void piix3_ide_class_init(ObjectClass *klass, void *data)
     DeviceClass *dc = DEVICE_CLASS(klass);
     PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
 
-    k->no_hotplug = 1;
     k->init = pci_piix_ide_initfn;
     k->exit = pci_piix_ide_exitfn;
     k->vendor_id = PCI_VENDOR_ID_INTEL;
@@ -249,6 +248,7 @@ static void piix3_ide_class_init(ObjectClass *klass, void *data)
     k->class_id = PCI_CLASS_STORAGE_IDE;
     set_bit(DEVICE_CATEGORY_STORAGE, dc->categories);
     dc->no_user = 1;
+    dc->hotpluggable = false;
 }
 
 static const TypeInfo piix3_ide_info = {
@@ -282,7 +282,6 @@ static void piix4_ide_class_init(ObjectClass *klass, void *data)
     DeviceClass *dc = DEVICE_CLASS(klass);
     PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
 
-    k->no_hotplug = 1;
     k->init = pci_piix_ide_initfn;
     k->exit = pci_piix_ide_exitfn;
     k->vendor_id = PCI_VENDOR_ID_INTEL;
@@ -290,6 +289,7 @@ static void piix4_ide_class_init(ObjectClass *klass, void *data)
     k->class_id = PCI_CLASS_STORAGE_IDE;
     set_bit(DEVICE_CATEGORY_STORAGE, dc->categories);
     dc->no_user = 1;
+    dc->hotpluggable = false;
 }
 
 static const TypeInfo piix4_ide_info = {
hw/isa/piix4.c +1 -1
index 1a1d451..53b15bc 100644
@@ -107,7 +107,6 @@ static void piix4_class_init(ObjectClass *klass, void *data)
     DeviceClass *dc = DEVICE_CLASS(klass);
     PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
 
-    k->no_hotplug = 1;
     k->init = piix4_initfn;
     k->vendor_id = PCI_VENDOR_ID_INTEL;
     k->device_id = PCI_DEVICE_ID_INTEL_82371AB_0;
@@ -115,6 +114,7 @@ static void piix4_class_init(ObjectClass *klass, void *data)
     dc->desc = "ISA bridge";
     dc->no_user = 1;
     dc->vmsd = &vmstate_piix4;
+    dc->hotpluggable = false;
 }
 
 static const TypeInfo piix4_info = {
hw/pci-host/piix.c +3 -3
index edc974e..4cdaabc 100644
@@ -654,7 +654,7 @@ static void piix3_class_init(ObjectClass *klass, void *data)
     dc->desc        = "ISA bridge";
     dc->vmsd        = &vmstate_piix3;
     dc->no_user     = 1,
-    k->no_hotplug   = 1;
+    dc->hotpluggable   = false;
     k->init         = piix3_initfn;
     k->config_write = piix3_write_config;
     k->vendor_id    = PCI_VENDOR_ID_INTEL;
@@ -678,7 +678,7 @@ static void piix3_xen_class_init(ObjectClass *klass, void *data)
     dc->desc        = "ISA bridge";
     dc->vmsd        = &vmstate_piix3;
     dc->no_user     = 1;
-    k->no_hotplug   = 1;
+    dc->hotpluggable   = false;
     k->init         = piix3_initfn;
     k->config_write = piix3_write_config_xen;
     k->vendor_id    = PCI_VENDOR_ID_INTEL;
@@ -699,7 +699,6 @@ static void i440fx_class_init(ObjectClass *klass, void *data)
     DeviceClass *dc = DEVICE_CLASS(klass);
     PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
 
-    k->no_hotplug = 1;
     k->init = i440fx_initfn;
     k->config_write = i440fx_write_config;
     k->vendor_id = PCI_VENDOR_ID_INTEL;
@@ -709,6 +708,7 @@ static void i440fx_class_init(ObjectClass *klass, void *data)
     dc->desc = "Host bridge";
     dc->no_user = 1;
     dc->vmsd = &vmstate_i440fx;
+    dc->hotpluggable   = false;
 }
 
 static const TypeInfo i440fx_info = {
hw/pci/pci.c +1 -10
index 49eca95..8a7a21b 100644
@@ -1735,11 +1735,7 @@ static int pci_qdev_init(DeviceState *qdev)
                                      pci_dev->devfn);
     if (pci_dev == NULL)
         return -1;
-    if (qdev->hotplugged && pc->no_hotplug) {
-        qerror_report(QERR_DEVICE_NO_HOTPLUG, object_get_typename(OBJECT(pci_dev)));
-        do_pci_unregister_device(pci_dev);
-        return -1;
-    }
+
     if (pc->init) {
         rc = pc->init(pci_dev);
         if (rc != 0) {
@@ -1774,12 +1770,7 @@ static int pci_qdev_init(DeviceState *qdev)
 static int pci_unplug_device(DeviceState *qdev)
 {
     PCIDevice *dev = PCI_DEVICE(qdev);
-    PCIDeviceClass *pc = PCI_DEVICE_GET_CLASS(dev);
 
-    if (pc->no_hotplug) {
-        qerror_report(QERR_DEVICE_NO_HOTPLUG, object_get_typename(OBJECT(dev)));
-        return -1;
-    }
     return dev->bus->hotplug(dev->bus->hotplug_qdev, dev,
                              PCI_HOTPLUG_DISABLED);
 }
hw/usb/hcd-ehci-pci.c +1 -1
index 0c98594..484a9bd 100644
@@ -123,7 +123,7 @@ static void ehci_class_init(ObjectClass *klass, void *data)
     k->init = usb_ehci_pci_initfn;
     k->class_id = PCI_CLASS_SERIAL_USB;
     k->config_write = usb_ehci_pci_write_config;
-    k->no_hotplug = 1;
+    dc->hotpluggable = false;
     dc->vmsd = &vmstate_ehci_pci;
     dc->props = ehci_pci_properties;
 }
hw/usb/hcd-ohci.c +1 -1
index e38cdeb..3d35058 100644
@@ -1993,10 +1993,10 @@ static void ohci_pci_class_init(ObjectClass *klass, void *data)
     k->vendor_id = PCI_VENDOR_ID_APPLE;
     k->device_id = PCI_DEVICE_ID_APPLE_IPID_USB;
     k->class_id = PCI_CLASS_SERIAL_USB;
-    k->no_hotplug = 1;
     set_bit(DEVICE_CATEGORY_USB, dc->categories);
     dc->desc = "Apple USB Controller";
     dc->props = ohci_pci_properties;
+    dc->hotpluggable = false;
 }
 
 static const TypeInfo ohci_pci_info = {
hw/usb/hcd-uhci.c +1 -1
index 238d1d2..ad814b5 100644
@@ -1318,7 +1318,7 @@ static void uhci_class_init(ObjectClass *klass, void *data)
     k->device_id = info->device_id;
     k->revision  = info->revision;
     k->class_id  = PCI_CLASS_SERIAL_USB;
-    k->no_hotplug = 1;
+    dc->hotpluggable = false;
     dc->vmsd = &vmstate_uhci;
     dc->props = uhci_properties;
     set_bit(DEVICE_CATEGORY_USB, dc->categories);
hw/usb/hcd-xhci.c +1 -1
index bafe085..0fa814e 100644
@@ -3798,6 +3798,7 @@ static void xhci_class_init(ObjectClass *klass, void *data)
     dc->vmsd    = &vmstate_xhci;
     dc->props   = xhci_properties;
     dc->reset   = xhci_reset;
+    dc->hotpluggable   = false;
     set_bit(DEVICE_CATEGORY_USB, dc->categories);
     k->init         = usb_xhci_initfn;
     k->vendor_id    = PCI_VENDOR_ID_NEC;
@@ -3805,7 +3806,6 @@ static void xhci_class_init(ObjectClass *klass, void *data)
     k->class_id     = PCI_CLASS_SERIAL_USB;
     k->revision     = 0x03;
     k->is_express   = 1;
-    k->no_hotplug   = 1;
 }
 
 static const TypeInfo xhci_info = {
include/hw/pci/pci.h +0 -3
index b783e68..6b79358 100644
@@ -201,9 +201,6 @@ typedef struct PCIDeviceClass {
     /* pcie stuff */
     int is_express;   /* is this device pci express? */
 
-    /* device isn't hot-pluggable */
-    int no_hotplug;
-
     /* rom bar */
     const char *romfile;
 } PCIDeviceClass;
Signature · 1 line
1.8.3.1
#18 lore raw

PATCH v3 8/11 acpi/piix4pm: convert ACPI PCI hotplug to use hotplug-handler API

In reply to Igor Mammedov

Split piix4_device_hotplug() into hotplug/unplug callbacks and register them as "hotplug-handler" interface implementation of PIIX4_PM device.

Replace pci_bus_hotplug() wiring with setting link on PCI BUS "hotplug-handler" property to PIIX4_PM device.

---
 hw/acpi/piix4.c | 73 ++++++++++++++++++++++++++++++++-------------------------
 1 file changed, 41 insertions(+), 32 deletions(-)
hw/acpi/piix4.c +41 -32
index fff2126..857d039 100644
@@ -30,6 +30,8 @@
 #include "hw/nvram/fw_cfg.h"
 #include "exec/address-spaces.h"
 #include "hw/acpi/piix4.h"
+#include "qapi/qmp/qerror.h"
+#include "hw/hotplug.h"
 
 //#define DEBUG
 
@@ -107,7 +109,7 @@ typedef struct PIIX4PMState {
     OBJECT_CHECK(PIIX4PMState, (obj), TYPE_PIIX4_PM)
 
 static void piix4_acpi_system_hot_add_init(MemoryRegion *parent,
-                                           PCIBus *bus, PIIX4PMState *s);
+                                           BusState *bus, PIIX4PMState *s);
 
 #define ACPI_ENABLE 0xf1
 #define ACPI_DISABLE 0xf0
@@ -478,7 +480,7 @@ static int piix4_pm_initfn(PCIDevice *dev)
     qemu_add_machine_init_done_notifier(&s->machine_ready);
     qemu_register_reset(piix4_reset, s);
 
-    piix4_acpi_system_hot_add_init(pci_address_space_io(dev), dev->bus, s);
+    piix4_acpi_system_hot_add_init(pci_address_space_io(dev), BUS(dev->bus), s);
 
     piix4_pm_add_propeties(s);
     return 0;
@@ -664,13 +666,11 @@ static void piix4_cpu_added_req(Notifier *n, void *opaque)
     piix4_cpu_hotplug_req(s, CPU(opaque), PLUG);
 }
 
-static int piix4_device_hotplug(DeviceState *qdev, PCIDevice *dev,
-                                PCIHotplugState state);
-
 static void piix4_acpi_system_hot_add_init(MemoryRegion *parent,
-                                           PCIBus *bus, PIIX4PMState *s)
+                                           BusState *bus, PIIX4PMState *s)
 {
     CPUState *cpu;
+    Error *local_error = NULL;
 
     memory_region_init_io(&s->io_gpe, OBJECT(s), &piix4_gpe_ops, s,
                           "acpi-gpe0", GPE_LEN);
@@ -680,7 +680,14 @@ static void piix4_acpi_system_hot_add_init(MemoryRegion *parent,
                           "acpi-pci-hotplug", PCI_HOTPLUG_SIZE);
     memory_region_add_subregion(parent, PCI_HOTPLUG_ADDR,
                                 &s->io_pci);
-    pci_bus_hotplug(bus, piix4_device_hotplug, DEVICE(s));
+    object_property_set_link(OBJECT(bus), OBJECT(s),
+                             QDEV_HOTPLUG_HANDLER_PROPERTY, &local_error);
+    if (error_is_set(&local_error)) {
+        qerror_report_err(local_error);
+        error_free(local_error);
+        abort();
+    }
+    bus->allow_hotplug = 1;
 
     CPU_FOREACH(cpu) {
         CPUClass *cc = CPU_GET_CLASS(cpu);
@@ -696,41 +703,36 @@ static void piix4_acpi_system_hot_add_init(MemoryRegion *parent,
     qemu_register_cpu_added_notifier(&s->cpu_added_notifier);
 }
 
-static void enable_device(PIIX4PMState *s, int slot)
+static void piix4_pci_device_hotplug_cb(HotplugHandler *hotplug_dev,
+                                        DeviceState *dev, Error **errp)
 {
-    s->ar.gpe.sts[0] |= PIIX4_PCI_HOTPLUG_STATUS;
-    s->pci0_slot_device_present |= (1U << slot);
-}
-
-static void disable_device(PIIX4PMState *s, int slot)
-{
-    s->ar.gpe.sts[0] |= PIIX4_PCI_HOTPLUG_STATUS;
-    s->pci0_status.down |= (1U << slot);
-}
-
-static int piix4_device_hotplug(DeviceState *qdev, PCIDevice *dev,
-				PCIHotplugState state)
-{
-    int slot = PCI_SLOT(dev->devfn);
-    PIIX4PMState *s = PIIX4_PM(qdev);
+    PCIDevice *pci_dev = PCI_DEVICE(dev);
+    int slot = PCI_SLOT(pci_dev->devfn);
+    PIIX4PMState *s = PIIX4_PM(hotplug_dev);
 
     /* Don't send event when device is enabled during qemu machine creation:
      * it is present on boot, no hotplug event is necessary. We do send an
      * event when the device is disabled later. */
-    if (state == PCI_COLDPLUG_ENABLED) {
+    if (!dev->hotplugged) {
         s->pci0_slot_device_present |= (1U << slot);
-        return 0;
-    }
-
-    if (state == PCI_HOTPLUG_ENABLED) {
-        enable_device(s, slot);
-    } else {
-        disable_device(s, slot);
+        return;
     }
 
+    s->ar.gpe.sts[0] |= PIIX4_PCI_HOTPLUG_STATUS;
+    s->pci0_slot_device_present |= (1U << slot);
     pm_update_sci(s);
+}
 
-    return 0;
+static void piix4_pci_device_hot_unplug_cb(HotplugHandler *hotplug_dev,
+                                           DeviceState *dev, Error **errp)
+{
+    PCIDevice *pci_dev = PCI_DEVICE(dev);
+    int slot = PCI_SLOT(pci_dev->devfn);
+    PIIX4PMState *s = PIIX4_PM(hotplug_dev);
+
+    s->ar.gpe.sts[0] |= PIIX4_PCI_HOTPLUG_STATUS;
+    s->pci0_status.down |= (1U << slot);
+    pm_update_sci(s);
 }
 
 static Property piix4_pm_properties[] = {
@@ -745,6 +747,7 @@ static void piix4_pm_class_init(ObjectClass *klass, void *data)
 {
     DeviceClass *dc = DEVICE_CLASS(klass);
     PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
+    HotplugHandlerClass *hc = HOTPLUG_HANDLER_CLASS(klass);
 
     k->init = piix4_pm_initfn;
     k->config_write = pm_write_config;
@@ -757,6 +760,8 @@ static void piix4_pm_class_init(ObjectClass *klass, void *data)
     dc->vmsd = &vmstate_acpi;
     dc->props = piix4_pm_properties;
     dc->hotpluggable = false;
+    hc->plug = piix4_pci_device_hotplug_cb;
+    hc->unplug = piix4_pci_device_hot_unplug_cb;
 }
 
 static const TypeInfo piix4_pm_info = {
@@ -764,6 +769,10 @@ static const TypeInfo piix4_pm_info = {
     .parent        = TYPE_PCI_DEVICE,
     .instance_size = sizeof(PIIX4PMState),
     .class_init    = piix4_pm_class_init,
+    .interfaces = (InterfaceInfo[]) {
+        { TYPE_HOTPLUG_HANDLER },
+        { }
+    }
 };
 
 static void piix4_pm_register_types(void)
Signature · 1 line
1.8.3.1
#19 lore raw

PATCH v3 9/11 pci/shpc: convert SHPC hotplug to use hotplug-handler API

In reply to Igor Mammedov

Split shpc_device_hotplug() into hotplug/unplug callbacks and register them as "hotplug-handler" interface implementation of PCI_BRIDGE_DEV device.

Replace pci_bus_hotplug() wiring with setting link on PCI BUS "hotplug-handler" property to PCI_BRIDGE_DEV device.

---
 hw/pci-bridge/pci_bridge_dev.c |   9 +++
 hw/pci/shpc.c                  | 133 ++++++++++++++++++++++++++---------------
 include/hw/pci/shpc.h          |   8 +++
 3 files changed, 103 insertions(+), 47 deletions(-)
hw/pci-bridge/pci_bridge_dev.c +9 -0
index 440e187..e68145c 100644
@@ -26,6 +26,7 @@
 #include "hw/pci/slotid_cap.h"
 #include "exec/memory.h"
 #include "hw/pci/pci_bus.h"
+#include "hw/hotplug.h"
 
 #define TYPE_PCI_BRIDGE_DEV "pci-bridge"
 #define PCI_BRIDGE_DEV(obj) \
@@ -136,6 +137,8 @@ static void pci_bridge_dev_class_init(ObjectClass *klass, void *data)
 {
     DeviceClass *dc = DEVICE_CLASS(klass);
     PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
+    HotplugHandlerClass *hc = HOTPLUG_HANDLER_CLASS(klass);
+
     k->init = pci_bridge_dev_initfn;
     k->exit = pci_bridge_dev_exitfn;
     k->config_write = pci_bridge_dev_write_config;
@@ -148,6 +151,8 @@ static void pci_bridge_dev_class_init(ObjectClass *klass, void *data)
     dc->props = pci_bridge_dev_properties;
     dc->vmsd = &pci_bridge_dev_vmstate;
     set_bit(DEVICE_CATEGORY_BRIDGE, dc->categories);
+    hc->plug = shpc_device_hotplug_cb;
+    hc->unplug = shpc_device_hot_unplug_cb;
 }
 
 static const TypeInfo pci_bridge_dev_info = {
@@ -155,6 +160,10 @@ static const TypeInfo pci_bridge_dev_info = {
     .parent        = TYPE_PCI_BRIDGE,
     .instance_size = sizeof(PCIBridgeDev),
     .class_init = pci_bridge_dev_class_init,
+    .interfaces = (InterfaceInfo[]) {
+        { TYPE_HOTPLUG_HANDLER },
+        { }
+    }
 };
 
 static void pci_bridge_dev_register(void)
hw/pci/shpc.c +86 -47

172 lines not on this page · raw message

include/hw/pci/shpc.h +8 -0
index 467911a..eef1a1a 100644
@@ -4,6 +4,8 @@
 #include "qemu-common.h"
 #include "exec/memory.h"
 #include "migration/vmstate.h"
+#include "qapi/error.h"
+#include "hw/hotplug.h"
 
 struct SHPCDevice {
     /* Capability offset in device's config space */
@@ -41,6 +43,12 @@ int shpc_init(PCIDevice *dev, PCIBus *sec_bus, MemoryRegion *bar, unsigned off);
 void shpc_cleanup(PCIDevice *dev, MemoryRegion *bar);
 void shpc_cap_write_config(PCIDevice *d, uint32_t addr, uint32_t val, int len);
 
+
+void shpc_device_hotplug_cb(HotplugHandler *hotplug_dev, DeviceState *dev,
+                            Error **errp);
+void shpc_device_hot_unplug_cb(HotplugHandler *hotplug_dev, DeviceState *dev,
+                               Error **errp);
+
 extern VMStateInfo shpc_vmstate_info;
 #define SHPC_VMSTATE(_field, _type) \
     VMSTATE_BUFFER_UNSAFE_INFO(_field, _type, 0, shpc_vmstate_info, 0)
Signature · 1 line
1.8.3.1
#20 lore raw

PATCH v3 10/11 pci/pcie: convert PCIE hotplug to use hotplug-handler API

In reply to Igor Mammedov

Split pcie_cap_slot_hotplug() into hotplug/unplug callbacks and register them as "hotplug-handler" interface implementation of PCIE_SLOT device.

Replace pci_bus_hotplug() wiring with setting link on PCI BUS "hotplug-handler" property to PCI_BRIDGE_DEV device.

---
 hw/pci/pcie.c         | 73 +++++++++++++++++++++++++++++++++------------------
 hw/pci/pcie_port.c    |  8 ++++++
 include/hw/pci/pcie.h |  5 ++++
 3 files changed, 61 insertions(+), 25 deletions(-)
hw/pci/pcie.c +48 -25
index ca60cf2..d4e978c 100644
@@ -26,6 +26,7 @@
 #include "hw/pci/pci_bus.h"
 #include "hw/pci/pcie_regs.h"
 #include "qemu/range.h"
+#include "qapi/qmp/qerror.h"
 
 //#define DEBUG_PCIE
 #ifdef DEBUG_PCIE
@@ -216,28 +217,20 @@ static void pcie_cap_slot_event(PCIDevice *dev, PCIExpressHotPlugEvent event)
     hotplug_event_notify(dev);
 }
 
-static int pcie_cap_slot_hotplug(DeviceState *qdev,
-                                 PCIDevice *pci_dev, PCIHotplugState state)
+static void pcie_cap_slot_hotplug_common(PCIDevice *hotplug_dev,
+                                         DeviceState *dev,
+                                         uint8_t **exp_cap, Error **errp)
 {
-    PCIDevice *d = PCI_DEVICE(qdev);
-    uint8_t *exp_cap = d->config + d->exp.exp_cap;
-    uint16_t sltsta = pci_get_word(exp_cap + PCI_EXP_SLTSTA);
-
-    /* Don't send event when device is enabled during qemu machine creation:
-     * it is present on boot, no hotplug event is necessary. We do send an
-     * event when the device is disabled later. */
-    if (state == PCI_COLDPLUG_ENABLED) {
-        pci_word_test_and_set_mask(exp_cap + PCI_EXP_SLTSTA,
-                                   PCI_EXP_SLTSTA_PDS);
-        return 0;
-    }
+    PCIDevice *pci_dev = PCI_DEVICE(dev);
+    *exp_cap = hotplug_dev->config + hotplug_dev->exp.exp_cap;
+    uint16_t sltsta = pci_get_word(*exp_cap + PCI_EXP_SLTSTA);
 
     PCIE_DEV_PRINTF(pci_dev, "hotplug state: %d\n", state);
     if (sltsta & PCI_EXP_SLTSTA_EIS) {
         /* the slot is electromechanically locked.
          * This error is propagated up to qdev and then to HMP/QMP.
          */
-        return -EBUSY;
+        error_setg_errno(errp, -EBUSY, "slot is electromechanically locked");
     }
 
     /* TODO: multifunction hot-plug.
@@ -245,18 +238,40 @@ static int pcie_cap_slot_hotplug(DeviceState *qdev,
      * hot plugged/unplugged.
      */
     assert(PCI_FUNC(pci_dev->devfn) == 0);
+}
 
-    if (state == PCI_HOTPLUG_ENABLED) {
+void pcie_cap_slot_hotplug_cb(HotplugHandler *hotplug_dev, DeviceState *dev,
+                              Error **errp)
+{
+    uint8_t *exp_cap;
+
+    pcie_cap_slot_hotplug_common(PCI_DEVICE(hotplug_dev), dev, &exp_cap, errp);
+
+    /* Don't send event when device is enabled during qemu machine creation:
+     * it is present on boot, no hotplug event is necessary. We do send an
+     * event when the device is disabled later. */
+    if (!dev->hotplugged) {
         pci_word_test_and_set_mask(exp_cap + PCI_EXP_SLTSTA,
                                    PCI_EXP_SLTSTA_PDS);
-        pcie_cap_slot_event(d, PCI_EXP_HP_EV_PDC);
-    } else {
-        object_unparent(OBJECT(pci_dev));
-        pci_word_test_and_clear_mask(exp_cap + PCI_EXP_SLTSTA,
-                                     PCI_EXP_SLTSTA_PDS);
-        pcie_cap_slot_event(d, PCI_EXP_HP_EV_PDC);
+        return;
     }
-    return 0;
+
+    pci_word_test_and_set_mask(exp_cap + PCI_EXP_SLTSTA,
+                               PCI_EXP_SLTSTA_PDS);
+    pcie_cap_slot_event(PCI_DEVICE(hotplug_dev), PCI_EXP_HP_EV_PDC);
+}
+
+void pcie_cap_slot_hot_unplug_cb(HotplugHandler *hotplug_dev, DeviceState *dev,
+                                 Error **errp)
+{
+    uint8_t *exp_cap;
+
+    pcie_cap_slot_hotplug_common(PCI_DEVICE(hotplug_dev), dev, &exp_cap, errp);
+
+    object_unparent(OBJECT(dev));
+    pci_word_test_and_clear_mask(exp_cap + PCI_EXP_SLTSTA,
+                                 PCI_EXP_SLTSTA_PDS);
+    pcie_cap_slot_event(PCI_DEVICE(hotplug_dev), PCI_EXP_HP_EV_PDC);
 }
 
 /* pci express slot for pci express root/downstream port
@@ -264,6 +279,8 @@ static int pcie_cap_slot_hotplug(DeviceState *qdev,
 void pcie_cap_slot_init(PCIDevice *dev, uint16_t slot)
 {
     uint32_t pos = dev->exp.exp_cap;
+    BusState *bus = BUS(pci_bridge_get_sec_bus(PCI_BRIDGE(dev)));
+    Error *local_error = NULL;
 
     pci_word_test_and_set_mask(dev->config + pos + PCI_EXP_FLAGS,
                                PCI_EXP_FLAGS_SLOT);
@@ -305,8 +322,14 @@ void pcie_cap_slot_init(PCIDevice *dev, uint16_t slot)
 
     dev->exp.hpev_notified = false;
 
-    pci_bus_hotplug(pci_bridge_get_sec_bus(PCI_BRIDGE(dev)),
-                    pcie_cap_slot_hotplug, &dev->qdev);
+    object_property_set_link(OBJECT(bus), OBJECT(dev),
+                             QDEV_HOTPLUG_HANDLER_PROPERTY, &local_error);
+    if (error_is_set(&local_error)) {
+        qerror_report_err(local_error);
+        error_free(local_error);
+        abort();
+    }
+    bus->allow_hotplug = 1;
 }
 
 void pcie_cap_slot_reset(PCIDevice *dev)
hw/pci/pcie_port.c +8 -0
index 2adb030..fa24877 100644
@@ -19,6 +19,7 @@
  */
 
 #include "hw/pci/pcie_port.h"
+#include "hw/hotplug.h"
 
 void pcie_port_init_reg(PCIDevice *d)
 {
@@ -149,8 +150,11 @@ static Property pcie_slot_props[] = {
 static void pcie_slot_class_init(ObjectClass *oc, void *data)
 {
     DeviceClass *dc = DEVICE_CLASS(oc);
+    HotplugHandlerClass *hc = HOTPLUG_HANDLER_CLASS(oc);
 
     dc->props = pcie_slot_props;
+    hc->plug = pcie_cap_slot_hotplug_cb;
+    hc->unplug = pcie_cap_slot_hot_unplug_cb;
 }
 
 static const TypeInfo pcie_slot_type_info = {
@@ -159,6 +163,10 @@ static const TypeInfo pcie_slot_type_info = {
     .instance_size = sizeof(PCIESlot),
     .abstract = true,
     .class_init = pcie_slot_class_init,
+    .interfaces = (InterfaceInfo[]) {
+        { TYPE_HOTPLUG_HANDLER },
+        { }
+    }
 };
 
 static void pcie_port_register_types(void)
include/hw/pci/pcie.h +5 -0
index 1966169..b0bf7e3 100644
@@ -25,6 +25,7 @@
 #include "hw/pci/pci_regs.h"
 #include "hw/pci/pcie_regs.h"
 #include "hw/pci/pcie_aer.h"
+#include "hw/hotplug.h"
 
 typedef enum {
     /* for attention and power indicator */
@@ -122,4 +123,8 @@ extern const VMStateDescription vmstate_pcie_device;
     .offset     = vmstate_offset_value(_state, _field, PCIDevice),   \
 }
 
+void pcie_cap_slot_hotplug_cb(HotplugHandler *hotplug_dev, DeviceState *dev,
+                              Error **errp);
+void pcie_cap_slot_hot_unplug_cb(HotplugHandler *hotplug_dev, DeviceState *dev,
+                                 Error **errp);
 #endif /* QEMU_PCIE_H */
Signature · 1 line
1.8.3.1
#21 lore raw

PATCH v3 11/11 hw/pci: switch to a generic hotplug handling for PCIDevice

In reply to Igor Mammedov

make qdev_unplug()/device_set_realized() to call hotplug handler's plug/unplug methods if available and remove not needed anymore hot(un)plug handling from PCIDevice.

In case if hotplug handler is not available, revert to the legacy hotplug method.

---
 hw/core/qdev.c           | 17 +++++++++++++----
 hw/pci/pci.c             | 29 +----------------------------
 include/hw/pci/pci.h     | 10 ----------
 include/hw/pci/pci_bus.h |  2 --
 4 files changed, 14 insertions(+), 44 deletions(-)
hw/core/qdev.c +13 -4
index 9418fea..7711c76 100644
@@ -213,7 +213,6 @@ void qdev_unplug(DeviceState *dev, Error **errp)
         error_set(errp, QERR_BUS_NO_HOTPLUG, dev->parent_bus->name);
         return;
     }
-    assert(dc->unplug != NULL);
 
     if (!dc->hotpluggable) {
         error_set(errp, QERR_DEVICE_NO_HOTPLUG,
@@ -223,9 +222,13 @@ void qdev_unplug(DeviceState *dev, Error **errp)
 
     qdev_hot_removed = true;
 
-    if (dc->unplug(dev) < 0) {
-        error_set(errp, QERR_UNDEFINED_ERROR);
-        return;
+    if (dev->parent_bus && dev->parent_bus->hotplug_handler) {
+        hotplug_handler_unplug(dev->parent_bus->hotplug_handler, dev, errp);
+    } else {
+        assert(dc->unplug != NULL);
+        if (dc->unplug(dev) < 0) { /* legacy handler */
+            error_set(errp, QERR_UNDEFINED_ERROR);
+        }
     }
 }
 
@@ -705,6 +708,12 @@ static void device_set_realized(Object *obj, bool value, Error **err)
             dc->realize(dev, &local_err);
         }
 
+        if (dev->parent_bus && dev->parent_bus->hotplug_handler &&
+            local_err == NULL) {
+            hotplug_handler_plug(dev->parent_bus->hotplug_handler,
+                                 dev, &local_err);
+        }
+
         if (qdev_get_vmsd(dev) && local_err == NULL) {
             vmstate_register_with_alias_id(dev, -1, qdev_get_vmsd(dev), dev,
                                            dev->instance_id_alias,
hw/pci/pci.c +1 -28
index 8a7a21b..6642e4c 100644
@@ -35,6 +35,7 @@
 #include "hw/pci/msi.h"
 #include "hw/pci/msix.h"
 #include "exec/address-spaces.h"
+#include "hw/hotplug.h"
 
 //#define DEBUG_PCI
 #ifdef DEBUG_PCI
@@ -348,13 +349,6 @@ void pci_bus_irqs(PCIBus *bus, pci_set_irq_fn set_irq, pci_map_irq_fn map_irq,
     bus->irq_count = g_malloc0(nirq * sizeof(bus->irq_count[0]));
 }
 
-void pci_bus_hotplug(PCIBus *bus, pci_hotplug_fn hotplug, DeviceState *qdev)
-{
-    bus->qbus.allow_hotplug = 1;
-    bus->hotplug = hotplug;
-    bus->hotplug_qdev = qdev;
-}
-
 PCIBus *pci_register_bus(DeviceState *parent, const char *name,
                          pci_set_irq_fn set_irq, pci_map_irq_fn map_irq,
                          void *irq_opaque,
@@ -1752,29 +1746,9 @@ static int pci_qdev_init(DeviceState *qdev)
     }
     pci_add_option_rom(pci_dev, is_default_rom);
 
-    if (bus->hotplug) {
-        /* Let buses differentiate between hotplug and when device is
-         * enabled during qemu machine creation. */
-        rc = bus->hotplug(bus->hotplug_qdev, pci_dev,
-                          qdev->hotplugged ? PCI_HOTPLUG_ENABLED:
-                          PCI_COLDPLUG_ENABLED);
-        if (rc != 0) {
-            int r = pci_unregister_device(&pci_dev->qdev);
-            assert(!r);
-            return rc;
-        }
-    }
     return 0;
 }
 
-static int pci_unplug_device(DeviceState *qdev)
-{
-    PCIDevice *dev = PCI_DEVICE(qdev);
-
-    return dev->bus->hotplug(dev->bus->hotplug_qdev, dev,
-                             PCI_HOTPLUG_DISABLED);
-}
-
 PCIDevice *pci_create_multifunction(PCIBus *bus, int devfn, bool multifunction,
                                     const char *name)
 {
@@ -2245,7 +2219,6 @@ static void pci_device_class_init(ObjectClass *klass, void *data)
 {
     DeviceClass *k = DEVICE_CLASS(klass);
     k->init = pci_qdev_init;
-    k->unplug = pci_unplug_device;
     k->exit = pci_unregister_device;
     k->bus_type = TYPE_PCI_BUS;
     k->props = pci_props;
include/hw/pci/pci.h +0 -10
index 6b79358..dd45e04 100644
@@ -327,15 +327,6 @@ typedef void (*pci_set_irq_fn)(void *opaque, int irq_num, int level);
 typedef int (*pci_map_irq_fn)(PCIDevice *pci_dev, int irq_num);
 typedef PCIINTxRoute (*pci_route_irq_fn)(void *opaque, int pin);
 
-typedef enum {
-    PCI_HOTPLUG_DISABLED,
-    PCI_HOTPLUG_ENABLED,
-    PCI_COLDPLUG_ENABLED,
-} PCIHotplugState;
-
-typedef int (*pci_hotplug_fn)(DeviceState *qdev, PCIDevice *pci_dev,
-                              PCIHotplugState state);
-
 #define TYPE_PCI_BUS "PCI"
 #define PCI_BUS(obj) OBJECT_CHECK(PCIBus, (obj), TYPE_PCI_BUS)
 #define TYPE_PCIE_BUS "PCIE"
@@ -354,7 +345,6 @@ PCIBus *pci_bus_new(DeviceState *parent, const char *name,
 void pci_bus_irqs(PCIBus *bus, pci_set_irq_fn set_irq, pci_map_irq_fn map_irq,
                   void *irq_opaque, int nirq);
 int pci_bus_get_irq_level(PCIBus *bus, int irq_num);
-void pci_bus_hotplug(PCIBus *bus, pci_hotplug_fn hotplug, DeviceState *dev);
 /* 0 <= pin <= 3 0 = INTA, 1 = INTB, 2 = INTC, 3 = INTD */
 int pci_swizzle_map_irq_fn(PCIDevice *pci_dev, int pin);
 PCIBus *pci_register_bus(DeviceState *parent, const char *name,
include/hw/pci/pci_bus.h +0 -2
index 9df1788..fabaeee 100644
@@ -16,8 +16,6 @@ struct PCIBus {
     pci_set_irq_fn set_irq;
     pci_map_irq_fn map_irq;
     pci_route_irq_fn route_intx_to_irq;
-    pci_hotplug_fn hotplug;
-    DeviceState *hotplug_qdev;
     void *irq_opaque;
     PCIDevice *devices[PCI_SLOT_MAX * PCI_FUNC_MAX];
     PCIDevice *parent_dev;
Signature · 1 line
1.8.3.1
#22 lore raw
In reply to Igor Mammedov
Quoted 24 lines · Igor Mammedov

changes since v2:
* s/hotplugable/hotpluggable/
* move hotplug check to an earlier patch:
"qdev: add "hotpluggable" property to Device"
--
Refactor PCI specific hotplug API to a more generic/reusable one. Model it after SCSI-BUS like hotplug API replacing single hotplug callback with hotplug/hot_unplug pair of callbacks as suggested by Paolo.
Difference between SCSI-BUS and this approach is that the former is BUS centric while the latter is device centred. Which is evolved from the fact that hotplug callbacks used by ACPI/SHPC/PCIE are implemented by devices rather than by bus and bus serves only as a proxy to forward event to hotplug device.
Memory hotplug also exposes tha same usage pattern hence an attempt to generalize hotplug API.

Refactoring also simplifies wiring of a hotplug device with a bus, all it needs is to set "hotplug-device" link on bus, which would potentially allow to do it from configuration file, there is not need to setup hotplug device callbacks on bus synce it can get them via HOTPLUG_DEVICE API of "hotplug-device" target.

Quoted 3 lines

In addition device centred hotplug API may be used by bus-less hotplug implementations as well if it's decided to use link<foo...> instead of bus.

I'm having a hard time parsing this description.

Sharing hot plug code is a good thing. Making hotplug a qdev-level concept seems like a bad thing to me.

The series is a net add of code so I don't think we're winning anything by generalizing here.

Is there a use-case this enables that isn't possible today?

Regards,

Anthony Liguori

Quoted 61 lines

Patches 8-11 are should be merged as one and are split only for simplifying review (they compile fine but PCI hotplug is broken until the last patch is applyed).

git tree for testing:
https://github.com/imammedo/qemu/commits/hotplug_dev_inf_v3

tested only ACPI and PCIE hotplug.

Hervé Poussineau (1):
qom: detect bad reentrance during object_class_foreach

Igor Mammedov (9):
define hotplug interface
qdev: add to BusState "hotplug-handler" link
qdev: add "hotpluggable" property to Device
hw/acpi: move typeinfo to the file end
qdev:pci: refactor PCIDevice to use generic "hotpluggable" property acpi/piix4pm: convert ACPI PCI hotplug to use hotplug-handler API pci/shpc: convert SHPC hotplug to use hotplug-handler API pci/pcie: convert PCIE hotplug to use hotplug-handler API hw/pci: switch to a generic hotplug handling for PCIDevice

Paolo Bonzini (1):
qom: do not register interface "types" in the type table

 hw/acpi/piix4.c                | 151 ++++++++++++++++++++++-------------------
 hw/core/Makefile.objs          |   1 +
 hw/core/hotplug.c              |  48 +++++++++++++
 hw/core/qdev.c                 |  50 ++++++++++++--
 hw/display/cirrus_vga.c        |   2 +-
 hw/display/qxl.c               |   2 +-
 hw/display/vga-pci.c           |   2 +-
 hw/display/vmware_vga.c        |   2 +-
 hw/i386/acpi-build.c           |   6 +-
 hw/ide/piix.c                  |   4 +-
 hw/isa/piix4.c                 |   2 +-
 hw/pci-bridge/pci_bridge_dev.c |   9 +++
 hw/pci-host/piix.c             |   6 +-
 hw/pci/pci.c                   |  40 +----------
 hw/pci/pcie.c                  |  73 +++++++++++++-------
 hw/pci/pcie_port.c             |   8 +++
 hw/pci/shpc.c                  | 133 +++++++++++++++++++++++-------------
 hw/usb/hcd-ehci-pci.c          |   2 +-
 hw/usb/hcd-ohci.c              |   2 +-
 hw/usb/hcd-uhci.c              |   2 +-
 hw/usb/hcd-xhci.c              |   2 +-
 include/hw/hotplug.h           |  75 ++++++++++++++++++++
 include/hw/pci/pci.h           |  13 ----
 include/hw/pci/pci_bus.h       |   2 -
 include/hw/pci/pcie.h          |   5 ++
 include/hw/pci/shpc.h          |   8 +++
 include/hw/qdev-core.h         |   8 +++
 qom/object.c                   |  17 ++++-
 28 files changed, 455 insertions(+), 220 deletions(-)
 create mode 100644 hw/core/hotplug.c
 create mode 100644 include/hw/hotplug.h
Signature · 1 line
1.8.3.1
#23 lore raw
Quoted 26 lines · Anthony Liguori
Quoted 24 lines · Igor Mammedov

changes since v2:
* s/hotplugable/hotpluggable/
* move hotplug check to an earlier patch:
"qdev: add "hotpluggable" property to Device"
--
Refactor PCI specific hotplug API to a more generic/reusable one. Model it after SCSI-BUS like hotplug API replacing single hotplug callback with hotplug/hot_unplug pair of callbacks as suggested by Paolo.
Difference between SCSI-BUS and this approach is that the former is BUS centric while the latter is device centred. Which is evolved from the fact that hotplug callbacks used by ACPI/SHPC/PCIE are implemented by devices rather than by bus and bus serves only as a proxy to forward event to hotplug device.
Memory hotplug also exposes tha same usage pattern hence an attempt to generalize hotplug API.

Refactoring also simplifies wiring of a hotplug device with a bus, all it needs is to set "hotplug-device" link on bus, which would potentially allow to do it from configuration file, there is not need to setup hotplug device callbacks on bus synce it can get them via HOTPLUG_DEVICE API of "hotplug-device" target.

Quoted 9 lines
Quoted 3 lines

In addition device centred hotplug API may be used by bus-less hotplug implementations as well if it's decided to use link<foo...> instead of bus.

I'm having a hard time parsing this description.

Sharing hot plug code is a good thing. Making hotplug a qdev-level concept seems like a bad thing to me.

So hotplug seems sane as a bus level concept to me. The problem is busses are a collection of general devices so the bus concept is lost when you move from the container to the contained. Are we missing an absraction layer here - TYPE_BUS_DEVICE?

Quoted 5 lines

The series is a net add of code so I don't think we're winning anything by generalizing here.

Is there a use-case this enables that isn't possible today?

Has some potential embedded applications is we want to use it for power-down/up emulation. Exactly what such a sysbus hotplug handler would look like is an open question.

Regards,
Peter

Quoted 66 lines

Regards,

Anthony Liguori

Quoted 61 lines

Patches 8-11 are should be merged as one and are split only for simplifying review (they compile fine but PCI hotplug is broken until the last patch is applyed).

git tree for testing:
https://github.com/imammedo/qemu/commits/hotplug_dev_inf_v3

tested only ACPI and PCIE hotplug.

Hervé Poussineau (1):
qom: detect bad reentrance during object_class_foreach

Igor Mammedov (9):
define hotplug interface
qdev: add to BusState "hotplug-handler" link
qdev: add "hotpluggable" property to Device
hw/acpi: move typeinfo to the file end
qdev:pci: refactor PCIDevice to use generic "hotpluggable" property acpi/piix4pm: convert ACPI PCI hotplug to use hotplug-handler API pci/shpc: convert SHPC hotplug to use hotplug-handler API pci/pcie: convert PCIE hotplug to use hotplug-handler API hw/pci: switch to a generic hotplug handling for PCIDevice

Paolo Bonzini (1):
qom: do not register interface "types" in the type table

 hw/acpi/piix4.c                | 151 ++++++++++++++++++++++-------------------
 hw/core/Makefile.objs          |   1 +
 hw/core/hotplug.c              |  48 +++++++++++++
 hw/core/qdev.c                 |  50 ++++++++++++--
 hw/display/cirrus_vga.c        |   2 +-
 hw/display/qxl.c               |   2 +-
 hw/display/vga-pci.c           |   2 +-
 hw/display/vmware_vga.c        |   2 +-
 hw/i386/acpi-build.c           |   6 +-
 hw/ide/piix.c                  |   4 +-
 hw/isa/piix4.c                 |   2 +-
 hw/pci-bridge/pci_bridge_dev.c |   9 +++
 hw/pci-host/piix.c             |   6 +-
 hw/pci/pci.c                   |  40 +----------
 hw/pci/pcie.c                  |  73 +++++++++++++-------
 hw/pci/pcie_port.c             |   8 +++
 hw/pci/shpc.c                  | 133 +++++++++++++++++++++++-------------
 hw/usb/hcd-ehci-pci.c          |   2 +-
 hw/usb/hcd-ohci.c              |   2 +-
 hw/usb/hcd-uhci.c              |   2 +-
 hw/usb/hcd-xhci.c              |   2 +-
 include/hw/hotplug.h           |  75 ++++++++++++++++++++
 include/hw/pci/pci.h           |  13 ----
 include/hw/pci/pci_bus.h       |   2 -
 include/hw/pci/pcie.h          |   5 ++
 include/hw/pci/shpc.h          |   8 +++
 include/hw/qdev-core.h         |   8 +++
 qom/object.c                   |  17 ++++-
 28 files changed, 455 insertions(+), 220 deletions(-)
 create mode 100644 hw/core/hotplug.c
 create mode 100644 include/hw/hotplug.h

--
1.8.3.1

#24 lore raw
In reply to Anthony Liguori
Quoted 26 lines · On Mon, 16 Dec 2013 15:26:37 -0800 Anthony Liguori <[email protected]>…
Quoted 24 lines · Igor Mammedov

changes since v2:
* s/hotplugable/hotpluggable/
* move hotplug check to an earlier patch:
"qdev: add "hotpluggable" property to Device"
--
Refactor PCI specific hotplug API to a more generic/reusable one. Model it after SCSI-BUS like hotplug API replacing single hotplug callback with hotplug/hot_unplug pair of callbacks as suggested by Paolo.
Difference between SCSI-BUS and this approach is that the former is BUS centric while the latter is device centred. Which is evolved from the fact that hotplug callbacks used by ACPI/SHPC/PCIE are implemented by devices rather than by bus and bus serves only as a proxy to forward event to hotplug device.
Memory hotplug also exposes tha same usage pattern hence an attempt to generalize hotplug API.

Refactoring also simplifies wiring of a hotplug device with a bus, all it needs is to set "hotplug-device" link on bus, which would potentially allow to do it from configuration file, there is not need to setup hotplug device callbacks on bus synce it can get them via HOTPLUG_DEVICE API of "hotplug-device" target.

Quoted 8 lines
Quoted 3 lines

In addition device centred hotplug API may be used by bus-less hotplug implementations as well if it's decided to use link<foo...> instead of bus.

I'm having a hard time parsing this description.

Sharing hot plug code is a good thing. Making hotplug a qdev-level concept seems like a bad thing to me.

we could add TYPE_BUS_DEVICE as Peter suggests but let me explain why I've chosen DEVICE for it.
All current hotplug code depends on hotpluggable bus but periodically there were suggestions to use link<>s for hotplug instead of bus. Making hotplug the bus device only concept will not allow to move in direction of hotplug with links which are/were supposed to replace buses in future.
As side effect series lays ground work for link based hotplug, it's not complete but a step towards it. It still doesn't allow link based hotplug, since purpose was to have unified hotplug interface across PCI/SCSI/virtio devices/buses.

Maybe we need TYPE_HOTPLUGGABLE_DEVICE so it could be abstracted from bus as well and eventually it could be used with bussless devices?

Quoted 3 lines

The series is a net add of code so I don't think we're winning anything by generalizing here.

it will allow to remove different implementations of hotplug mechanism in SCSI/virtio and maybe USB buses in addition to converted PCI hotplug.

Quoted 1 line

Is there a use-case this enables that isn't possible today?

It was prompted by memory hotplug, alternative was that it can re-implement hotplug mechanism in its own way or duplicating code from scsi or pci buses/devices. Neither alternatives look nice when there is a common usage pattern.

Quoted 65 lines

Regards,

Anthony Liguori

Quoted 61 lines

Patches 8-11 are should be merged as one and are split only for simplifying review (they compile fine but PCI hotplug is broken until the last patch is applyed).

git tree for testing:
https://github.com/imammedo/qemu/commits/hotplug_dev_inf_v3

tested only ACPI and PCIE hotplug.

Hervé Poussineau (1):
qom: detect bad reentrance during object_class_foreach

Igor Mammedov (9):
define hotplug interface
qdev: add to BusState "hotplug-handler" link
qdev: add "hotpluggable" property to Device
hw/acpi: move typeinfo to the file end
qdev:pci: refactor PCIDevice to use generic "hotpluggable" property acpi/piix4pm: convert ACPI PCI hotplug to use hotplug-handler API pci/shpc: convert SHPC hotplug to use hotplug-handler API pci/pcie: convert PCIE hotplug to use hotplug-handler API hw/pci: switch to a generic hotplug handling for PCIDevice

Paolo Bonzini (1):
qom: do not register interface "types" in the type table

 hw/acpi/piix4.c                | 151 ++++++++++++++++++++++-------------------
 hw/core/Makefile.objs          |   1 +
 hw/core/hotplug.c              |  48 +++++++++++++
 hw/core/qdev.c                 |  50 ++++++++++++--
 hw/display/cirrus_vga.c        |   2 +-
 hw/display/qxl.c               |   2 +-
 hw/display/vga-pci.c           |   2 +-
 hw/display/vmware_vga.c        |   2 +-
 hw/i386/acpi-build.c           |   6 +-
 hw/ide/piix.c                  |   4 +-
 hw/isa/piix4.c                 |   2 +-
 hw/pci-bridge/pci_bridge_dev.c |   9 +++
 hw/pci-host/piix.c             |   6 +-
 hw/pci/pci.c                   |  40 +----------
 hw/pci/pcie.c                  |  73 +++++++++++++-------
 hw/pci/pcie_port.c             |   8 +++
 hw/pci/shpc.c                  | 133 +++++++++++++++++++++++-------------
 hw/usb/hcd-ehci-pci.c          |   2 +-
 hw/usb/hcd-ohci.c              |   2 +-
 hw/usb/hcd-uhci.c              |   2 +-
 hw/usb/hcd-xhci.c              |   2 +-
 include/hw/hotplug.h           |  75 ++++++++++++++++++++
 include/hw/pci/pci.h           |  13 ----
 include/hw/pci/pci_bus.h       |   2 -
 include/hw/pci/pcie.h          |   5 ++
 include/hw/pci/shpc.h          |   8 +++
 include/hw/qdev-core.h         |   8 +++
 qom/object.c                   |  17 ++++-
 28 files changed, 455 insertions(+), 220 deletions(-)
 create mode 100644 hw/core/hotplug.c
 create mode 100644 include/hw/hotplug.h
Signature · 1 line
1.8.3.1
#25 lore raw
In reply to Anthony Liguori

Il 17/12/2013 00:26, Anthony Liguori ha scritto:

Quoted 2 lines

Sharing hot plug code is a good thing. Making hotplug a qdev-level concept seems like a bad thing to me.

Can you explain what you mean?

Quoted 2 lines

The series is a net add of code so I don't think we're winning anything by generalizing here.

Any generalization that's used just once will be a net add of code (and this code will be reused by SCSI and x86 memory hotplug at least; perhaps x86 CPU hotplug too).

Any generalization that requires some boilerplate code will be a net add of code, too. QEMU being written in C, we unfortunately cannot avoid that.

So I don't think that lines of code are a good metric.

Paolo

Quoted 67 lines

Is there a use-case this enables that isn't possible today?

Regards,

Anthony Liguori

Quoted 61 lines

Patches 8-11 are should be merged as one and are split only for simplifying review (they compile fine but PCI hotplug is broken until the last patch is applyed).

git tree for testing:
https://github.com/imammedo/qemu/commits/hotplug_dev_inf_v3

tested only ACPI and PCIE hotplug.

Hervé Poussineau (1):
qom: detect bad reentrance during object_class_foreach

Igor Mammedov (9):
define hotplug interface
qdev: add to BusState "hotplug-handler" link
qdev: add "hotpluggable" property to Device
hw/acpi: move typeinfo to the file end
qdev:pci: refactor PCIDevice to use generic "hotpluggable" property acpi/piix4pm: convert ACPI PCI hotplug to use hotplug-handler API pci/shpc: convert SHPC hotplug to use hotplug-handler API pci/pcie: convert PCIE hotplug to use hotplug-handler API hw/pci: switch to a generic hotplug handling for PCIDevice

Paolo Bonzini (1):
qom: do not register interface "types" in the type table

 hw/acpi/piix4.c                | 151 ++++++++++++++++++++++-------------------
 hw/core/Makefile.objs          |   1 +
 hw/core/hotplug.c              |  48 +++++++++++++
 hw/core/qdev.c                 |  50 ++++++++++++--
 hw/display/cirrus_vga.c        |   2 +-
 hw/display/qxl.c               |   2 +-
 hw/display/vga-pci.c           |   2 +-
 hw/display/vmware_vga.c        |   2 +-
 hw/i386/acpi-build.c           |   6 +-
 hw/ide/piix.c                  |   4 +-
 hw/isa/piix4.c                 |   2 +-
 hw/pci-bridge/pci_bridge_dev.c |   9 +++
 hw/pci-host/piix.c             |   6 +-
 hw/pci/pci.c                   |  40 +----------
 hw/pci/pcie.c                  |  73 +++++++++++++-------
 hw/pci/pcie_port.c             |   8 +++
 hw/pci/shpc.c                  | 133 +++++++++++++++++++++++-------------
 hw/usb/hcd-ehci-pci.c          |   2 +-
 hw/usb/hcd-ohci.c              |   2 +-
 hw/usb/hcd-uhci.c              |   2 +-
 hw/usb/hcd-xhci.c              |   2 +-
 include/hw/hotplug.h           |  75 ++++++++++++++++++++
 include/hw/pci/pci.h           |  13 ----
 include/hw/pci/pci_bus.h       |   2 -
 include/hw/pci/pcie.h          |   5 ++
 include/hw/pci/shpc.h          |   8 +++
 include/hw/qdev-core.h         |   8 +++
 qom/object.c                   |  17 ++++-
 28 files changed, 455 insertions(+), 220 deletions(-)
 create mode 100644 hw/core/hotplug.c
 create mode 100644 include/hw/hotplug.h
Signature · 1 line
1.8.3.1
#26 lore raw
Quoted 5 lines · Paolo Bonzini

Il 17/12/2013 00:26, Anthony Liguori ha scritto:

Quoted 2 lines

Sharing hot plug code is a good thing. Making hotplug a qdev-level concept seems like a bad thing to me.

Can you explain what you mean?

The question is whether "hotpluggable" as a property applies to all devices or not.

But hotplug is strictly a bus level concept. It's a sequence of events that correspond to what happens when you add a new device to a bus after power on.

Quoted 6 lines
Quoted 2 lines

The series is a net add of code so I don't think we're winning anything by generalizing here.

Any generalization that's used just once will be a net add of code (and this code will be reused by SCSI and x86 memory hotplug at least; perhaps x86 CPU hotplug too).

The question is whether there can be code sharing without touching the base class. You could certainly have a HotpluggableBusState and then a HotpluggableDeviceState.

Interfaces would be another option too.

Quoted 4 lines

Any generalization that requires some boilerplate code will be a net add of code, too. QEMU being written in C, we unfortunately cannot avoid that.

So I don't think that lines of code are a good metric.

The general concern is about polluting widely used base classes. It's better if we can avoid adding things to DeviceState and Object whenever possible.

Regards,

Anthony Liguori

Quoted 70 lines

Paolo

Quoted 67 lines

Is there a use-case this enables that isn't possible today?

Regards,

Anthony Liguori

Quoted 61 lines

Patches 8-11 are should be merged as one and are split only for simplifying review (they compile fine but PCI hotplug is broken until the last patch is applyed).

git tree for testing:
https://github.com/imammedo/qemu/commits/hotplug_dev_inf_v3

tested only ACPI and PCIE hotplug.

Hervé Poussineau (1):
qom: detect bad reentrance during object_class_foreach

Igor Mammedov (9):
define hotplug interface
qdev: add to BusState "hotplug-handler" link
qdev: add "hotpluggable" property to Device
hw/acpi: move typeinfo to the file end
qdev:pci: refactor PCIDevice to use generic "hotpluggable" property acpi/piix4pm: convert ACPI PCI hotplug to use hotplug-handler API pci/shpc: convert SHPC hotplug to use hotplug-handler API pci/pcie: convert PCIE hotplug to use hotplug-handler API hw/pci: switch to a generic hotplug handling for PCIDevice

Paolo Bonzini (1):
qom: do not register interface "types" in the type table

 hw/acpi/piix4.c                | 151 ++++++++++++++++++++++-------------------
 hw/core/Makefile.objs          |   1 +
 hw/core/hotplug.c              |  48 +++++++++++++
 hw/core/qdev.c                 |  50 ++++++++++++--
 hw/display/cirrus_vga.c        |   2 +-
 hw/display/qxl.c               |   2 +-
 hw/display/vga-pci.c           |   2 +-
 hw/display/vmware_vga.c        |   2 +-
 hw/i386/acpi-build.c           |   6 +-
 hw/ide/piix.c                  |   4 +-
 hw/isa/piix4.c                 |   2 +-
 hw/pci-bridge/pci_bridge_dev.c |   9 +++
 hw/pci-host/piix.c             |   6 +-
 hw/pci/pci.c                   |  40 +----------
 hw/pci/pcie.c                  |  73 +++++++++++++-------
 hw/pci/pcie_port.c             |   8 +++
 hw/pci/shpc.c                  | 133 +++++++++++++++++++++++-------------
 hw/usb/hcd-ehci-pci.c          |   2 +-
 hw/usb/hcd-ohci.c              |   2 +-
 hw/usb/hcd-uhci.c              |   2 +-
 hw/usb/hcd-xhci.c              |   2 +-
 include/hw/hotplug.h           |  75 ++++++++++++++++++++
 include/hw/pci/pci.h           |  13 ----
 include/hw/pci/pci_bus.h       |   2 -
 include/hw/pci/pcie.h          |   5 ++
 include/hw/pci/shpc.h          |   8 +++
 include/hw/qdev-core.h         |   8 +++
 qom/object.c                   |  17 ++++-
 28 files changed, 455 insertions(+), 220 deletions(-)
 create mode 100644 hw/core/hotplug.c
 create mode 100644 include/hw/hotplug.h

--
1.8.3.1

#27 lore raw

Il 17/12/2013 20:38, Anthony Liguori ha scritto:

Quoted 5 lines · Paolo Bonzini
Quoted 4 lines

Il 17/12/2013 00:26, Anthony Liguori ha scritto:

Quoted 2 lines

Sharing hot plug code is a good thing. Making hotplug a qdev-level concept seems like a bad thing to me.

Quoted 8 lines
Quoted 1 line

Can you explain what you mean?

The question is whether "hotpluggable" as a property applies to all devices or not.

But hotplug is strictly a bus level concept. It's a sequence of events that correspond to what happens when you add a new device to a bus after power on.

Hotplugging a device is a special case of plugging a device. If a bus or device only supports cold-plug, that can be done using "bc->allow_hotplug = false" or "dc->hotpluggable = false".

Igor's interface applies just as well to the case of plugging a device at startup; I think separating the two makes little sense. And once you have cold-plug and hot-plug in qdev core, it makes sense to add unplug as well. Also because we already have surprise removal in qdev core (that's unparent) and we have some kind of unplug request support (device_del/dc->unplug).

One possibility that remains is to put cold/hot-plug in a "BusDevice" class rather than in the core qdev:

Device
BusDevice <-- can be cold/hot-plugged

but this adds more complication. For example, the same CPU can be hotpluggable or not depending on the board model, should the superclass be Device or BusDevice. And if we ever have multi-CPU targets, with the "core" CPU not hotpluggable and additional hotpluggable ones (e.g. for GPUs) what would be the superclass of the common CPU superclass?

Quoted 5 lines

The question is whether there can be code sharing without touching the base class. You could certainly have a HotpluggableBusState and then a HotpluggableDeviceState.

Interfaces would be another option too.

Interfaces are fine, but the question is who finds them and calls them.
In this case, the discovery mechanism is a link property, and the
calling mechanism is an explicit hook in the "realized" property.

If we had aspect-oriented programming, we would be marking join points instead of writing "if (dev->foo) bar(dev->foo)" conditionals. But the idea is the same.

Quoted 3 lines

The general concern is about polluting widely used base classes. It's better if we can avoid adding things to DeviceState and Object whenever possible.

I agree. At the same time we should make base classes as small as possible, but not smaller than that.

Paolo

#28 lore raw
Quoted 10 lines · On Wed, 18 Dec 2013 11:36:52 +0100 Paolo Bonzini <[email protected]> wrote:

Il 17/12/2013 20:38, Anthony Liguori ha scritto:

Quoted 9 lines · Paolo Bonzini
Quoted 5 lines

Il 17/12/2013 00:26, Anthony Liguori ha scritto:

Quoted 2 lines

Sharing hot plug code is a good thing. Making hotplug a qdev-level concept seems like a bad thing to me.

Can you explain what you mean?

The question is whether "hotpluggable" as a property applies to all devices or not.

I think Andreas asked me to provide "hotpluggable" property to distinguish hotpluggable vs not hotpluggable DimmDevice via qom interface.

Quoted 8 lines
Quoted 4 lines

But hotplug is strictly a bus level concept. It's a sequence of events that correspond to what happens when you add a new device to a bus after power on.

Hotplugging a device is a special case of plugging a device. If a bus or device only supports cold-plug, that can be done using "bc->allow_hotplug = false" or "dc->hotpluggable = false".

Do we need per instance ability to set "hotpluggable" property? For example board might want to mark some CPUs as not hotpluggable.

Quoted 24 lines

Igor's interface applies just as well to the case of plugging a device at startup; I think separating the two makes little sense. And once you have cold-plug and hot-plug in qdev core, it makes sense to add unplug as well. Also because we already have surprise removal in qdev core (that's unparent) and we have some kind of unplug request support (device_del/dc->unplug).

One possibility that remains is to put cold/hot-plug in a "BusDevice" class rather than in the core qdev:

Device
BusDevice <-- can be cold/hot-plugged

but this adds more complication. For example, the same CPU can be hotpluggable or not depending on the board model, should the superclass be Device or BusDevice. And if we ever have multi-CPU targets, with the "core" CPU not hotpluggable and additional hotpluggable ones (e.g. for GPUs) what would be the superclass of the common CPU superclass?

Quoted 4 lines

The question is whether there can be code sharing without touching the base class. You could certainly have a HotpluggableBusState and then a HotpluggableDeviceState.

Quoted 5 lines
Quoted 1 line

Interfaces would be another option too.

Interfaces are fine, but the question is who finds them and calls them.
In this case, the discovery mechanism is a link property, and the
calling mechanism is an explicit hook in the "realized" property.

If we don't need per instance "hotpluggable" state and we can call interfaces from generic qdev/device code, then we would need at first only TYPE_HOTPLUGGABLE_BUS_DEVICE_IF and later for link<> based hotplug we could add just TYPE_HOTPLUGGABLE_DEVICE_IF. Difference would be in the way they get access to hotplug device link, former one will use bus for it and second some other way.

Quoted 14 lines

If we had aspect-oriented programming, we would be marking join points instead of writing "if (dev->foo) bar(dev->foo)" conditionals. But the idea is the same.

Quoted 3 lines

The general concern is about polluting widely used base classes. It's better if we can avoid adding things to DeviceState and Object whenever possible.

I agree. At the same time we should make base classes as small as possible, but not smaller than that.

Paolo

#29 lore raw

Il 18/12/2013 16:48, Igor Mammedov ha scritto:

Quoted 5 lines
Quoted 3 lines

Hotplugging a device is a special case of plugging a device. If a bus or device only supports cold-plug, that can be done using "bc->allow_hotplug = false" or "dc->hotpluggable = false".

Do we need per instance ability to set "hotpluggable" property? For example board might want to mark some CPUs as not hotpluggable.

I think such fine-grained control could be handled from dc->unplug. Let's not do anything more than we need, until we need it.

Right now, all we need to model is the fact that a device X can act as hotplug controller for multiple buses, X is not the parent of those buses, and we need to tell those buses about X. This is hotplug-handler in BusState.

We also like to distinguish devices that only support -device (or are even only board-instantiatable) from devices that support device_add. This is dc->hotpluggable. It is not absolutely necessary, but it makes sense if "plugging" gets a more central place in qdev. This is the case after you add hotplug-handler.

Quoted 6 lines
Quoted 5 lines
Quoted 1 line

Interfaces would be another option too.

Interfaces are fine, but the question is who finds them and calls them. In this case, the discovery mechanism is a link property, and the calling mechanism is an explicit hook in the "realized" property.

Quoted 6 lines

If we don't need per instance "hotpluggable" state and we can call interfaces from generic qdev/device code, then we would need at first only TYPE_HOTPLUGGABLE_BUS_DEVICE_IF and later for link<> based hotplug we could add just TYPE_HOTPLUGGABLE_DEVICE_IF. Difference would be in the way they get access to hotplug device link, former one will use bus for it and second some other way.

I think this is overengineered. What you have is flexible and decently clean. I don't think there's any need to go from the device to the bus and from there optionally to the handler. Most of the time you couldn't do anything really in the bus, you would have to call some sort of bus-specific callback (like SCSIBusInfo). It is then equivalent to set the bus's parent device as a (hot)plug handler and go straight from the device to the handler, as in your patches.

Paolo

#30 lore raw
Quoted 9 lines · On Wed, 18 Dec 2013 16:59:02 +0100 Paolo Bonzini <[email protected]> wrote:

Il 18/12/2013 16:48, Igor Mammedov ha scritto:

Quoted 5 lines
Quoted 3 lines

Hotplugging a device is a special case of plugging a device. If a bus or device only supports cold-plug, that can be done using "bc->allow_hotplug = false" or "dc->hotpluggable = false".

Do we need per instance ability to set "hotpluggable" property? For example board might want to mark some CPUs as not hotpluggable.

I think such fine-grained control could be handled from dc->unplug. Let's not do anything more than we need, until we need it.

I had in mind a lightweight conversion of initial memory to DimmDevices for i386 target and disabling its unplug until initial memory unplug would be properly handled by board and SeaBIOS.

Quoted 2 lines

Paolo

#31 lore raw
In reply to Igor Mammedov
Quoted 21 lines · Igor Mammedov
Quoted 10 lines · On Wed, 18 Dec 2013 11:36:52 +0100 Paolo Bonzini <[email protected]> wrote:

Il 17/12/2013 20:38, Anthony Liguori ha scritto:

Quoted 9 lines · Paolo Bonzini
Quoted 5 lines

Il 17/12/2013 00:26, Anthony Liguori ha scritto:

Quoted 2 lines

Sharing hot plug code is a good thing. Making hotplug a qdev-level concept seems like a bad thing to me.

Can you explain what you mean?

The question is whether "hotpluggable" as a property applies to all devices or not.

I think Andreas asked me to provide "hotpluggable" property to distinguish hotpluggable vs not hotpluggable DimmDevice via qom interface.

Quoted 5 lines
Quoted 4 lines

But hotplug is strictly a bus level concept. It's a sequence of events that correspond to what happens when you add a new device to a bus after power on.

Quoted 5 lines
Quoted 3 lines

Hotplugging a device is a special case of plugging a device. If a bus or device only supports cold-plug, that can be done using "bc->allow_hotplug = false" or "dc->hotpluggable = false".

Do we need per instance ability to set "hotpluggable" property? For example board might want to mark some CPUs as not hotpluggable.

It could be useful.
In real life same device can be on-board or on a plugin card. But it's not a must, we survived without this so far.

So maybe start not supporting it, add later?

Quoted 50 lines
Quoted 24 lines

Igor's interface applies just as well to the case of plugging a device at startup; I think separating the two makes little sense. And once you have cold-plug and hot-plug in qdev core, it makes sense to add unplug as well. Also because we already have surprise removal in qdev core (that's unparent) and we have some kind of unplug request support (device_del/dc->unplug).

One possibility that remains is to put cold/hot-plug in a "BusDevice" class rather than in the core qdev:

Device
BusDevice <-- can be cold/hot-plugged

but this adds more complication. For example, the same CPU can be hotpluggable or not depending on the board model, should the superclass be Device or BusDevice. And if we ever have multi-CPU targets, with the "core" CPU not hotpluggable and additional hotpluggable ones (e.g. for GPUs) what would be the superclass of the common CPU superclass?

Quoted 4 lines

The question is whether there can be code sharing without touching the base class. You could certainly have a HotpluggableBusState and then a HotpluggableDeviceState.

Quoted 5 lines
Quoted 1 line

Interfaces would be another option too.

Interfaces are fine, but the question is who finds them and calls them.
In this case, the discovery mechanism is a link property, and the
calling mechanism is an explicit hook in the "realized" property.

If we don't need per instance "hotpluggable" state and we can call interfaces from generic qdev/device code, then we would need at first only TYPE_HOTPLUGGABLE_BUS_DEVICE_IF and later for link<> based hotplug we could add just TYPE_HOTPLUGGABLE_DEVICE_IF. Difference would be in the way they get access to hotplug device link, former one will use bus for it and second some other way.

Quoted 14 lines

If we had aspect-oriented programming, we would be marking join points instead of writing "if (dev->foo) bar(dev->foo)" conditionals. But the idea is the same.

Quoted 3 lines

The general concern is about polluting widely used base classes. It's better if we can avoid adding things to DeviceState and Object whenever possible.

I agree. At the same time we should make base classes as small as possible, but not smaller than that.

Paolo

#32 lore raw
Quoted 25 lines · On Wed, 18 Dec 2013 18:26:07 +0200 "Michael S. Tsirkin" <[email protected]> wrote:
Quoted 24 lines · Igor Mammedov
Quoted 10 lines · On Wed, 18 Dec 2013 11:36:52 +0100 Paolo Bonzini <[email protected]> wrote:

Il 17/12/2013 20:38, Anthony Liguori ha scritto:

Quoted 9 lines · Paolo Bonzini
Quoted 5 lines

Il 17/12/2013 00:26, Anthony Liguori ha scritto:

Quoted 2 lines

Sharing hot plug code is a good thing. Making hotplug a qdev-level concept seems like a bad thing to me.

Can you explain what you mean?

The question is whether "hotpluggable" as a property applies to all devices or not.

I think Andreas asked me to provide "hotpluggable" property to distinguish hotpluggable vs not hotpluggable DimmDevice via qom interface.

Quoted 8 lines
Quoted 4 lines

But hotplug is strictly a bus level concept. It's a sequence of events that correspond to what happens when you add a new device to a bus after power on.

Hotplugging a device is a special case of plugging a device. If a bus or device only supports cold-plug, that can be done using "bc->allow_hotplug = false" or "dc->hotpluggable = false".

Quoted 9 lines
Quoted 2 lines

Do we need per instance ability to set "hotpluggable" property? For example board might want to mark some CPUs as not hotpluggable.

It could be useful.
In real life same device can be on-board or on a plugin card. But it's not a must, we survived without this so far.

So maybe start not supporting it, add later?

Yes, that's surely could be done later

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

Paolo