Clean up and fix no_user
cover Clean up and fix no_user
From: Markus Armbruster <[email protected]>
In an ideal world, machines can be built by wiring devices together with configuration, not code. Unfortunately, that's not the world we live in right now. We still have quite a few devices that need to be wired up by code. If you try to device_add such a device, it'll fail in sometimes mysterious ways. If you're lucky, you get an unmysterious immediate crash.
We used to protect users from such badness by marking devices where device_add cannot possibly work "no-user", and refusing to device_add them. Anthony silently broke the protection in v1.1. He has rejected attempts to unbreak it with the argument that the protection makes it impossible to wire devices together with configuration, not code, and that the protection is being misused[*].
On the former, I disagree. The problem isn't protecting users from devices that cannot be wired up that way, it's devices that cannot be wired up that way.
On the latter, Anthony has a point: the purpose of the no-user flag isn't obvious, and some of its uses are suspect.
So, instead of just fixing the regression, this series first addresses that point. PATCH 1 clarifies the purpose of no-user. PATCH 2-9 clean up and document its use. PATCH 10 fixes the regression.
Markus Armbruster (10):
qdev: Replace no_user by cannot_instantiate_with_device_add_yet
sysbus: Set cannot_instantiate_with_device_add_yet
cpu: Document why cannot_instantiate_with_device_add_yet
apic: Document why cannot_instantiate_with_device_add_yet
pci-host: Consistently set cannot_instantiate_with_device_add_yet
ich9: Document why cannot_instantiate_with_device_add_yet piix3 piix4: Clean up use of cannot_instantiate_with_device_add_yet
vt82c686: Clean up use of cannot_instantiate_with_device_add_yet
isa: Clean up use of cannot_instantiate_with_device_add_yet
qdev: Do not let the user try to device_add when it cannot work
hw/acpi/piix4.c | 7 ++++++- hw/alpha/typhoon.c | 2 -- hw/arm/versatilepb.c | 1 - hw/audio/pcspk.c | 3 ++- hw/audio/pl041.c | 1 - hw/block/fdc.c | 1 - hw/core/sysbus.c | 7 +++++++ hw/display/pl110.c | 1 - hw/dma/pl080.c | 1 - hw/i2c/smbus_ich9.c | 6 +++++- hw/i386/kvm/clock.c | 1 - hw/i386/kvmvapic.c | 1 - hw/i386/pc.c | 1 - hw/ide/piix.c | 3 --- hw/ide/via.c | 1 - hw/input/pckbd.c | 1 - hw/input/vmmouse.c | 3 ++- hw/intc/apic_common.c | 6 +++++- hw/intc/arm_gic.c | 1 - hw/intc/arm_gic_common.c | 1 - hw/intc/arm_gic_kvm.c | 1 - hw/intc/i8259_common.c | 8 +++++++- hw/intc/ioapic_common.c | 1 - hw/intc/pl190.c | 1 - hw/isa/isa-bus.c | 1 - hw/isa/lpc_ich9.c | 7 +++++-- hw/isa/piix4.c | 6 +++++- hw/isa/vt82c686.c | 6 +++++- hw/mips/gt64xxx_pci.c | 5 +++++ hw/misc/arm_l2x0.c | 1 - hw/misc/vmport.c | 3 ++- hw/nvram/fw_cfg.c | 1 - hw/pci-bridge/dec.c | 5 +++++ hw/pci-host/apb.c | 5 +++++ hw/pci-host/bonito.c | 8 +++++--- hw/pci-host/grackle.c | 8 +++++--- hw/pci-host/piix.c | 19 +++++++++++++++---- hw/pci-host/ppce500.c | 5 +++++ hw/pci-host/prep.c | 7 +++++-- hw/pci-host/q35.c | 5 +++++ hw/pci-host/uninorth.c | 20 ++++++++++++++++++++ hw/pci-host/versatile.c | 5 +++++ hw/ppc/ppc4xx_pci.c | 5 +++++ hw/ppc/spapr_vio.c | 2 -- hw/s390x/ipl.c | 1 - hw/s390x/s390-virtio-bus.c | 2 -- hw/s390x/virtio-ccw.c | 2 -- hw/sd/pl181.c | 1 - hw/sh4/sh_pci.c | 5 +++++ hw/timer/arm_mptimer.c | 1 - hw/timer/hpet.c | 1 - hw/timer/i8254_common.c | 1 - hw/timer/m48t59.c | 1 - hw/timer/mc146818rtc.c | 1 - hw/timer/pl031.c | 1 - include/hw/qdev-core.h | 13 ++++++++++++- qdev-monitor.c | 7 ++++--- qom/cpu.c | 6 +++++- 58 files changed, 162 insertions(+), 65 deletions(-)
Signature · 1 line
1.8.1.4
PATCH 1/10 qdev: Replace no_user by cannot_instantiate_with_device_add_yet
From: Markus Armbruster <[email protected]>
In an ideal world, machines can be built by wiring devices together with configuration, not code. Unfortunately, that's not the world we live in right now. We still have quite a few devices that need to be wired up by code. If you try to device_add such a device, it'll fail in sometimes mysterious ways. If you're lucky, you get an unmysterious immediate crash.
To protect users from such badness, DeviceClass member no_user used to make device models unavailable with -device / device_add, but that regressed in commit 18b6dad. The device model is still omitted from help, but is available anyway.
Attempts to fix the regression have been rejected with the argument that the purpose of no_user isn't clear, and it's prone to misuse.
This commit clarifies no_user's purpose. Anthony suggested to rename it cannot_instantiate_with_device_add_yet_due_to_internal_bugs, which I shorten somewhat to keep checkpatch happy. While there, make it bool.
Every use of cannot_instantiate_with_device_add_yet gets a FIXME comment asking for rationale. The next few commits will clean them all up, either by providing a rationale, or by getting rid of the use.
With that done, the regression fix is hopefully acceptable.
--- hw/acpi/piix4.c | 2 +- hw/alpha/typhoon.c | 2 +- hw/arm/versatilepb.c | 2 +- hw/audio/pcspk.c | 2 +- hw/audio/pl041.c | 2 +- hw/block/fdc.c | 2 +- hw/display/pl110.c | 2 +- hw/dma/pl080.c | 2 +- hw/i2c/smbus_ich9.c | 2 +- hw/i386/kvm/clock.c | 2 +- hw/i386/kvmvapic.c | 2 +- hw/i386/pc.c | 2 +- hw/ide/piix.c | 6 +++--- hw/ide/via.c | 2 +- hw/input/pckbd.c | 2 +- hw/input/vmmouse.c | 2 +- hw/intc/apic_common.c | 2 +- hw/intc/arm_gic.c | 2 +- hw/intc/arm_gic_common.c | 2 +- hw/intc/arm_gic_kvm.c | 2 +- hw/intc/i8259_common.c | 2 +- hw/intc/ioapic_common.c | 2 +- hw/intc/pl190.c | 2 +- hw/isa/isa-bus.c | 2 +- hw/isa/lpc_ich9.c | 2 +- hw/isa/piix4.c | 2 +- hw/isa/vt82c686.c | 2 +- hw/misc/arm_l2x0.c | 2 +- hw/misc/vmport.c | 2 +- hw/nvram/fw_cfg.c | 2 +- hw/pci-host/bonito.c | 4 ++-- hw/pci-host/grackle.c | 4 ++-- hw/pci-host/piix.c | 8 ++++---- hw/pci-host/prep.c | 4 ++-- hw/ppc/spapr_vio.c | 2 +- hw/s390x/ipl.c | 2 +- hw/s390x/s390-virtio-bus.c | 2 +- hw/s390x/virtio-ccw.c | 2 +- hw/sd/pl181.c | 2 +- hw/timer/arm_mptimer.c | 2 +- hw/timer/hpet.c | 2 +- hw/timer/i8254_common.c | 2 +- hw/timer/m48t59.c | 2 +- hw/timer/mc146818rtc.c | 2 +- hw/timer/pl031.c | 2 +- include/hw/qdev-core.h | 13 ++++++++++++- qdev-monitor.c | 5 +++-- qom/cpu.c | 2 +- 48 files changed, 69 insertions(+), 57 deletions(-)
hw/acpi/piix4.c +1 -1
@@ -508,7 +508,7 @@ static void piix4_pm_class_init(ObjectClass *klass, void *data) k->revision =0x03 ; k->class_id =PCI_CLASS_BRIDGE_OTHER ; dc->desc ="PM" ; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_acpi; dc->props = piix4_pm_properties; }
hw/alpha/typhoon.c +1 -1
@@ -938,7 +938,7 @@ static void typhoon_pcihost_class_init(ObjectClass *klass, void *data)SysBusDeviceClass *k =SYS_BUS_DEVICE_CLASS (klass); k->init = typhoon_pcihost_init; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo typhoon_pcihost_info = {
hw/arm/versatilepb.c +1 -1
@@ -390,7 +390,7 @@ static void vpb_sic_class_init(ObjectClass *klass, void *data)SysBusDeviceClass *k =SYS_BUS_DEVICE_CLASS (klass); k->init = vpb_sic_init; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_vpb_sic; }
hw/audio/pcspk.c +1 -1
@@ -192,7 +192,7 @@ static void pcspk_class_initfn(ObjectClass *klass, void *data) dc->realize = pcspk_realizefn;set_bit (DEVICE_CATEGORY_SOUND , dc->categories); - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->props = pcspk_properties; }
hw/audio/pl041.c +1 -1
@@ -632,7 +632,7 @@ static void pl041_device_class_init(ObjectClass *klass, void *data) k->init = pl041_init;set_bit (DEVICE_CATEGORY_SOUND , dc->categories); - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->reset = pl041_device_reset; dc->vmsd = &vmstate_pl041; dc->props = pl041_device_properties;
hw/block/fdc.c +1 -1
@@ -2234,7 +2234,7 @@ static void isabus_fdc_class_init(ObjectClass *klass, void *data) dc->realize = isabus_fdc_realize; dc->fw_name ="fdc" ; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->reset = fdctrl_external_reset_isa; dc->vmsd = &vmstate_isa_fdc; dc->props = isa_fdc_properties;
hw/display/pl110.c +1 -1
@@ -496,7 +496,7 @@ static void pl110_class_init(ObjectClass *klass, void *data) k->init = pl110_initfn;set_bit (DEVICE_CATEGORY_DISPLAY , dc->categories); - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_pl110; }
hw/dma/pl080.c +1 -1
@@ -381,7 +381,7 @@ static void pl080_class_init(ObjectClass *oc, void *data) {DeviceClass *dc =DEVICE_CLASS (oc); - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_pl080; }
hw/i2c/smbus_ich9.c +1 -1
@@ -97,7 +97,7 @@ static void ich9_smb_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_INTEL_ICH9_6 ; k->revision =ICH9_A2_SMB_REVISION ; k->class_id =PCI_CLASS_SERIAL_SMBUS ; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_ich9_smbus; dc->desc ="ICH9 SMBUS Bridge" ; k->init = ich9_smbus_initfn;
hw/i386/kvm/clock.c +1 -1
@@ -114,7 +114,7 @@ static void kvmclock_class_init(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->realize = kvmclock_realize; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &kvmclock_vmsd; }
hw/i386/kvmvapic.c +1 -1
@@ -824,7 +824,7 @@ static void vapic_class_init(ObjectClass *klass, void *data) {DeviceClass *dc =DEVICE_CLASS (klass); - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->reset = vapic_reset; dc->vmsd = &vmstate_vapic; dc->realize = vapic_realize;
hw/i386/pc.c +1 -1
@@ -544,7 +544,7 @@ static void port92_class_initfn(ObjectClass *klass, void *data) {DeviceClass *dc =DEVICE_CLASS (klass); - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->realize = port92_realizefn; dc->reset = port92_reset; dc->vmsd = &vmstate_port92_isa;
hw/ide/piix.c +3 -3
@@ -248,7 +248,7 @@ static void piix3_ide_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_INTEL_82371SB_1 ; k->class_id =PCI_CLASS_STORAGE_IDE ;set_bit (DEVICE_CATEGORY_STORAGE , dc->categories); - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo piix3_ide_info = {
@@ -267,7 +267,7 @@ static void piix3_ide_xen_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_INTEL_82371SB_1 ; k->class_id =PCI_CLASS_STORAGE_IDE ;set_bit (DEVICE_CATEGORY_STORAGE , dc->categories); - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->unplug = pci_piix3_xen_ide_unplug; }
@@ -289,7 +289,7 @@ static void piix4_ide_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_INTEL_82371AB ; k->class_id =PCI_CLASS_STORAGE_IDE ;set_bit (DEVICE_CATEGORY_STORAGE , dc->categories); - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo piix4_ide_info = {
hw/ide/via.c +1 -1
@@ -225,7 +225,7 @@ static void via_ide_class_init(ObjectClass *klass, void *data) k->revision =0x06 ; k->class_id =PCI_CLASS_STORAGE_IDE ;set_bit (DEVICE_CATEGORY_STORAGE , dc->categories); - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo via_ide_info = {
hw/input/pckbd.c +1 -1
@@ -522,7 +522,7 @@ static void i8042_class_initfn(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->realize = i8042_realizefn; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_kbd_isa; }
hw/input/vmmouse.c +1 -1
@@ -282,7 +282,7 @@ static void vmmouse_class_initfn(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->realize = vmmouse_realizefn; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->reset = vmmouse_reset; dc->vmsd = &vmstate_vmmouse; dc->props = vmmouse_properties;
hw/intc/apic_common.c +1 -1
@@ -386,7 +386,7 @@ static void apic_common_class_init(ObjectClass *klass, void *data) dc->vmsd = &vmstate_apic_common; dc->reset = apic_reset_common; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->props = apic_properties_common; idc->init = apic_init_common; }
hw/intc/arm_gic.c +1 -1
@@ -704,7 +704,7 @@ static void arm_gic_class_init(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass);ARMGICClass *agc =ARM_GIC_CLASS (klass); - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ agc->parent_realize = dc->realize; dc->realize = arm_gic_realize; }
hw/intc/arm_gic_common.c +1 -1
@@ -156,7 +156,7 @@ static void arm_gic_common_class_init(ObjectClass *klass, void *data) dc->realize = arm_gic_common_realize; dc->props = arm_gic_common_properties; dc->vmsd = &vmstate_gic; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo arm_gic_common_type = {
hw/intc/arm_gic_kvm.c +1 -1
@@ -150,7 +150,7 @@ static void kvm_arm_gic_class_init(ObjectClass *klass, void *data) kgc->parent_reset = dc->reset; dc->realize = kvm_arm_gic_realize; dc->reset = kvm_arm_gic_reset; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo kvm_arm_gic_info = {
hw/intc/i8259_common.c +1 -1
@@ -135,7 +135,7 @@ static void pic_common_class_init(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->vmsd = &vmstate_pic_common; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->props = pic_properties_common; dc->realize = pic_common_realize; }
hw/intc/ioapic_common.c +1 -1
@@ -98,7 +98,7 @@ static void ioapic_common_class_init(ObjectClass *klass, void *data) dc->realize = ioapic_common_realize; dc->vmsd = &vmstate_ioapic_common; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo ioapic_common_type = {
hw/intc/pl190.c +1 -1
@@ -273,7 +273,7 @@ static void pl190_class_init(ObjectClass *klass, void *data)SysBusDeviceClass *k =SYS_BUS_DEVICE_CLASS (klass); k->init = pl190_init; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->reset = pl190_reset; dc->vmsd = &vmstate_pl190; }
hw/isa/isa-bus.c +1 -1
@@ -197,7 +197,7 @@ static void isabus_bridge_class_init(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->fw_name ="isa" ; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo isabus_bridge_info = {
hw/isa/lpc_ich9.c +1 -1
@@ -604,7 +604,7 @@ static void ich9_lpc_class_init(ObjectClass *klass, void *data) dc->reset = ich9_lpc_reset; k->init = ich9_lpc_initfn; dc->vmsd = &vmstate_ich9_lpc; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ k->config_write = ich9_lpc_config_write; dc->desc ="ICH9 LPC bridge" ; k->vendor_id =PCI_VENDOR_ID_INTEL ;
hw/isa/piix4.c +1 -1
@@ -113,7 +113,7 @@ static void piix4_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_INTEL_82371AB_0 ; k->class_id =PCI_CLASS_BRIDGE_ISA ; dc->desc ="ISA bridge" ; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_piix4; }
hw/isa/vt82c686.c +1 -1
@@ -480,7 +480,7 @@ static void via_class_init(ObjectClass *klass, void *data) k->class_id =PCI_CLASS_BRIDGE_ISA ; k->revision =0x40 ; dc->desc ="ISA bridge" ; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_via; }
hw/misc/arm_l2x0.c +1 -1
@@ -179,7 +179,7 @@ static void l2x0_class_init(ObjectClass *klass, void *data) k->init = l2x0_priv_init; dc->vmsd = &vmstate_l2x0; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->props = l2x0_properties; dc->reset = l2x0_priv_reset; }
hw/misc/vmport.c +1 -1
@@ -162,7 +162,7 @@ static void vmport_class_initfn(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->realize = vmport_realizefn; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo vmport_info = {
hw/nvram/fw_cfg.c +1 -1
@@ -576,7 +576,7 @@ static void fw_cfg_class_init(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->realize = fw_cfg_realize; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->reset = fw_cfg_reset; dc->vmsd = &vmstate_fw_cfg; dc->props = fw_cfg_properties;
hw/pci-host/bonito.c +2 -2
@@ -806,7 +806,7 @@ static void bonito_class_init(ObjectClass *klass, void *data) k->revision =0x01 ; k->class_id =PCI_CLASS_BRIDGE_HOST ; dc->desc ="Host bridge" ; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_bonito; }
@@ -823,7 +823,7 @@ static void bonito_pcihost_class_init(ObjectClass *klass, void *data)SysBusDeviceClass *k =SYS_BUS_DEVICE_CLASS (klass); k->init = bonito_pcihost_initfn; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo bonito_pcihost_info = {
hw/pci-host/grackle.c +2 -2
@@ -130,7 +130,7 @@ static void grackle_pci_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_MOTOROLA_MPC106 ; k->revision =0x00 ; k->class_id =PCI_CLASS_BRIDGE_HOST ; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo grackle_pci_info = {
@@ -146,7 +146,7 @@ static void pci_grackle_class_init(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); k->init = pci_grackle_init_device; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo grackle_pci_host_info = {
hw/pci-host/piix.c +4 -4
@@ -644,7 +644,7 @@ static void piix3_class_init(ObjectClass *klass, void *data) dc->desc ="ISA bridge" ; dc->vmsd = &vmstate_piix3; - dc->no_user =1 , + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ k->no_hotplug =1 ; k->init = piix3_initfn; k->config_write = piix3_write_config;
@@ -668,7 +668,7 @@ static void piix3_xen_class_init(ObjectClass *klass, void *data) dc->desc ="ISA bridge" ; dc->vmsd = &vmstate_piix3; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ k->no_hotplug =1 ; k->init = piix3_initfn; k->config_write = piix3_write_config_xen;
@@ -698,7 +698,7 @@ static void i440fx_class_init(ObjectClass *klass, void *data) k->revision =0x02 ; k->class_id =PCI_CLASS_BRIDGE_HOST ; dc->desc ="Host bridge" ; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_i440fx; }
@@ -730,7 +730,7 @@ static void i440fx_pcihost_class_init(ObjectClass *klass, void *data) hc->root_bus_path = i440fx_pcihost_root_bus_path; dc->realize = i440fx_pcihost_realize; dc->fw_name ="pci" ; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->props = i440fx_props; }
hw/pci-host/prep.c +2 -2
@@ -198,7 +198,7 @@ static void raven_class_init(ObjectClass *klass, void *data) k->class_id =PCI_CLASS_BRIDGE_HOST ; dc->desc ="PReP Host Bridge - Motorola Raven" ; dc->vmsd = &vmstate_raven; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo raven_info = {
@@ -215,7 +215,7 @@ static void raven_pcihost_class_init(ObjectClass *klass, void *data)set_bit (DEVICE_CATEGORY_BRIDGE , dc->categories); dc->realize = raven_pcihost_realizefn; dc->fw_name ="pci" ; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo raven_pcihost_info = {
hw/ppc/spapr_vio.c +1 -1
@@ -532,7 +532,7 @@ static void spapr_vio_bridge_class_init(ObjectClass *klass, void *data)SysBusDeviceClass *k =SYS_BUS_DEVICE_CLASS (klass); k->init = spapr_vio_bridge_init; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo spapr_vio_bridge_info = {
hw/s390x/ipl.c +1 -1
@@ -181,7 +181,7 @@ static void s390_ipl_class_init(ObjectClass *klass, void *data) k->init = s390_ipl_init; dc->props = s390_ipl_properties; dc->reset = s390_ipl_reset; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo s390_ipl_info = {
hw/s390x/s390-virtio-bus.c +1 -1
@@ -680,7 +680,7 @@ static void s390_virtio_bridge_class_init(ObjectClass *klass, void *data)SysBusDeviceClass *k =SYS_BUS_DEVICE_CLASS (klass); k->init = s390_virtio_bridge_init; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo s390_virtio_bridge_info = {
hw/s390x/virtio-ccw.c +1 -1
@@ -1278,7 +1278,7 @@ static void virtual_css_bridge_class_init(ObjectClass *klass, void *data)SysBusDeviceClass *k =SYS_BUS_DEVICE_CLASS (klass); k->init = virtual_css_bridge_init; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo virtual_css_bridge_info = {
hw/sd/pl181.c +1 -1
@@ -506,7 +506,7 @@ static void pl181_class_init(ObjectClass *klass, void *data) sdc->init = pl181_init; k->vmsd = &vmstate_pl181; k->reset = pl181_reset; - k->no_user =1 ; + k->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo pl181_info = {
hw/timer/arm_mptimer.c +1 -1
@@ -297,7 +297,7 @@ static void arm_mptimer_class_init(ObjectClass *klass, void *data) sbc->init = arm_mptimer_init; dc->vmsd = &vmstate_arm_mptimer; dc->reset = arm_mptimer_reset; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->props = arm_mptimer_properties; }
hw/timer/hpet.c +1 -1
@@ -751,7 +751,7 @@ static void hpet_device_class_init(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->realize = hpet_realize; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->reset = hpet_reset; dc->vmsd = &vmstate_hpet; dc->props = hpet_device_properties;
hw/timer/i8254_common.c +1 -1
@@ -282,7 +282,7 @@ static void pit_common_class_init(ObjectClass *klass, void *data) dc->realize = pit_common_realize; dc->vmsd = &vmstate_pit_common; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo pit_common_type = {
hw/timer/m48t59.c +1 -1
@@ -750,7 +750,7 @@ static void m48t59_isa_class_init(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->realize = m48t59_isa_realize; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->reset = m48t59_reset_isa; dc->props = m48t59_isa_properties; }
hw/timer/mc146818rtc.c +1 -1
@@ -906,7 +906,7 @@ static void rtc_class_initfn(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->realize = rtc_realizefn; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_rtc; dc->props = mc146818rtc_properties; }
hw/timer/pl031.c +1 -1
@@ -251,7 +251,7 @@ static void pl031_class_init(ObjectClass *klass, void *data)SysBusDeviceClass *k =SYS_BUS_DEVICE_CLASS (klass); k->init = pl031_init; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_pl031; }
include/hw/qdev-core.h +12 -1
@@ -97,7 +97,18 @@ typedef struct DeviceClass {const char *fw_name;const char *desc;Property *props; -int no_user; + +/* +* Shall we hide this device model from -device / device_add? +* All devices should support instantiation with device_add, and +* this flag should not exist. But we're not there, yet. Some +* devices fail to instantiate with cryptic error messages. +* Others instantiate, but don't work. Exposing users to such +* behavior would be cruel; this flag serves to protect them. It +* should never be set without a comment explaining why it is set. +* TODO remove once we're there +*/ +bool cannot_instantiate_with_device_add_yet;/* callbacks */ void (*reset)(DeviceState *dev);
qdev-monitor.c +3 -2
@@ -87,7 +87,7 @@ static void qdev_print_devinfo(DeviceClass *dc)if (dc->desc) {error_printf (", desc \"%s\"" , dc->desc); } -if (dc->no_user) { +if (dc->cannot_instantiate_with_device_add_yet) {error_printf (", no-user" ); }error_printf ("\n" );
@@ -127,7 +127,8 @@ static void qdev_print_devinfos(bool show_no_user)if ((i <DEVICE_CATEGORY_MAX ? !test_bit (i, dc->categories) : !bitmap_empty (dc->categories,DEVICE_CATEGORY_MAX )) - || (!show_no_user && dc->no_user)) { + || (!show_no_user + && dc->cannot_instantiate_with_device_add_yet)) {continue ; }if (!cat_printed) {
qom/cpu.c +1 -1
@@ -254,7 +254,7 @@ static void cpu_class_init(ObjectClass *klass, void *data) k->gdb_read_register = cpu_common_gdb_read_register; k->gdb_write_register = cpu_common_gdb_write_register; dc->realize = cpu_common_realizefn; - dc->no_user =1 ; + dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo cpu_type_info = {
Signature · 1 line
1.8.1.4
Re: [Qemu-devel] [PATCH 01/10] qdev: Replace no_user by cannot_instantiate_with_device_add_yet
Quoted 10 lines · On 17 October 2013 14:54, <[email protected]> wrote:
From: Markus Armbruster <[email protected]>
In an ideal world, machines can be built by wiring devices together with configuration, not code. Unfortunately, that's not the world we live in right now. We still have quite a few devices that need to be wired up by code. If you try to
device_addsuch a device, it'll fail in sometimes mysterious ways. If you're lucky, you get an unmysterious immediate crash.include/hw/qdev-core.h +1
-0+/*
include/hw/qdev-core.h typedef struct DeviceClass { · line 111
Quoted 10 lines
include/hw/qdev-core.h +10
-0+* Shall we hide this device model from -device / device_add? +* All devices should support instantiation with device_add, and +* this flag should not exist. But we're not there, yet. Some +* devices fail to instantiate with cryptic error messages. +* Others instantiate, but don't work. Exposing users to such +* behavior would be cruel; this flag serves to protect them. It +* should never be set without a comment explaining why it is set. +* TODO remove once we're there +*/ +bool cannot_instantiate_with_device_add_yet;
So reading this I'm still not entirely sure what the scope of this flag is intended to be. I can think of two possibilities:
(1) the minimal definition: this device would actually crash or cause QEMU to break if you created it with device_add
(2) a larger definition, which includes also devices which are completely useless if created with device_add because there's no way for the user to wire them up properly.
I think most sysbus devices are going to be in (2) but not (1), because they should be fine to create and initialize, but they'll just be sitting completely pointlessly totally disconnected from the machine model.
Definition (1) is a harder boundary and more straightforward to check against, but definition (2) is arguably a bit more useful for the end user.
-- PMM
Re: [Qemu-devel] [PATCH 01/10] qdev: Replace no_user by cannot_instantiate_with_device_add_yet
Quoted 31 lines · Peter Maydell
Quoted 10 lines · On 17 October 2013 14:54, <[email protected]> wrote:
From: Markus Armbruster <[email protected]>
In an ideal world, machines can be built by wiring devices together with configuration, not code. Unfortunately, that's not the world we live in right now. We still have quite a few devices that need to be wired up by code. If you try to
device_addsuch a device, it'll fail in sometimes mysterious ways. If you're lucky, you get an unmysterious immediate crash.include/hw/qdev-core.h +1
-0+/* include/hw/qdev-core.h typedef struct DeviceClass { · line 111
Quoted 10 lines
include/hw/qdev-core.h +10
-0+* Shall we hide this device model from -device / device_add? +* All devices should support instantiation with device_add, and +* this flag should not exist. But we're not there, yet. Some +* devices fail to instantiate with cryptic error messages. +* Others instantiate, but don't work. Exposing users to such +* behavior would be cruel; this flag serves to protect them. It +* should never be set without a comment explaining why it is set. +* TODO remove once we're there +*/ +bool cannot_instantiate_with_device_add_yet;So reading this I'm still not entirely sure what the scope of this flag is intended to be. I can think of two possibilities:
(1) the minimal definition: this device would actually crash or cause QEMU to break if you created it with
device_add
(2) a larger definition, which includes also devices which are completely useless if created withdevice_addbecause there's no way for the user to wire them up properly.
Quoted 8 lines
I think most sysbus devices are going to be in (2) but not (1), because they should be fine to create and initialize, but they'll just be sitting completely pointlessly totally disconnected from the machine model.
Definition (1) is a harder boundary and more straightforward to check against, but definition (2) is arguably a bit more useful for the end user.
I agree, and I'd like us to adopt definition (2). I tried to express this when I wrote "instantiate, but don't work". Care to suggest clearer language for this comment?
Regarding (2) being less straightforward to check against: I think we should try hard to make our cannot_instantiate_with_device_add_yet use correct (any device we mark that way is actually useless with device_add), but I view completeness (all the devices that are actually useless with -device are marked) as not quite that important.
PATCH 2/10 sysbus: Set cannot_instantiate_with_device_add_yet
In reply to [email protected]From: Markus Armbruster <[email protected]>
device_add plugs devices into suitable bus. For "real" buses, that actually connects the device. For sysbus, the connections need to be made separately, and device_add can't do that. The device would be left unconncected, and could not possibly work.
Many, but not all sysbus devices alreasy set cannot_instantiate_with_device_add_yet in their class init function.
Set it in their abstract base's class init function sysbus_device_class_init(), and remove the now redundant assignments from device class init functions.
--- hw/alpha/typhoon.c | 2 -- hw/arm/versatilepb.c | 1 - hw/audio/pl041.c | 1 - hw/core/sysbus.c | 7 +++++++ hw/display/pl110.c | 1 - hw/dma/pl080.c | 1 - hw/i386/kvm/clock.c | 1 - hw/i386/kvmvapic.c | 1 - hw/intc/arm_gic.c | 1 - hw/intc/arm_gic_common.c | 1 - hw/intc/arm_gic_kvm.c | 1 - hw/intc/ioapic_common.c | 1 - hw/intc/pl190.c | 1 - hw/isa/isa-bus.c | 1 - hw/misc/arm_l2x0.c | 1 - hw/nvram/fw_cfg.c | 1 - hw/pci-host/bonito.c | 2 -- hw/pci-host/grackle.c | 2 -- hw/pci-host/piix.c | 1 - hw/pci-host/prep.c | 1 - hw/ppc/spapr_vio.c | 2 -- hw/s390x/ipl.c | 1 - hw/s390x/s390-virtio-bus.c | 2 -- hw/s390x/virtio-ccw.c | 2 -- hw/sd/pl181.c | 1 - hw/timer/arm_mptimer.c | 1 - hw/timer/hpet.c | 1 - hw/timer/pl031.c | 1 - 28 files changed, 7 insertions(+), 33 deletions(-)
hw/alpha/typhoon.c +0 -2
@@ -934,11 +934,9 @@ static int typhoon_pcihost_init(SysBusDevice *dev)static void typhoon_pcihost_class_init (ObjectClass *klass,void *data) { -DeviceClass *dc =DEVICE_CLASS (klass);SysBusDeviceClass *k =SYS_BUS_DEVICE_CLASS (klass); k->init = typhoon_pcihost_init; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo typhoon_pcihost_info = {
hw/arm/versatilepb.c +0 -1
@@ -390,7 +390,6 @@ static void vpb_sic_class_init(ObjectClass *klass, void *data)SysBusDeviceClass *k =SYS_BUS_DEVICE_CLASS (klass); k->init = vpb_sic_init; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_vpb_sic; }
hw/audio/pl041.c +0 -1
@@ -632,7 +632,6 @@ static void pl041_device_class_init(ObjectClass *klass, void *data) k->init = pl041_init;set_bit (DEVICE_CATEGORY_SOUND , dc->categories); - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->reset = pl041_device_reset; dc->vmsd = &vmstate_pl041; dc->props = pl041_device_properties;
hw/core/sysbus.c +7 -0
@@ -257,6 +257,13 @@ static void sysbus_device_class_init(ObjectClass *klass, void *data)DeviceClass *k =DEVICE_CLASS (klass); k->init = sysbus_device_init; k->bus_type =TYPE_SYSTEM_BUS ; +/* +* device_add plugs devices into suitable bus. For "real" buses, +* that actually connects the device. For sysbus, the connections +* need to be made separately, and device_add can't do that. The +* device would be left unconncected, and could not possibly work. +*/ + k->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo sysbus_device_type_info = {
hw/display/pl110.c +0 -1
@@ -496,7 +496,6 @@ static void pl110_class_init(ObjectClass *klass, void *data) k->init = pl110_initfn;set_bit (DEVICE_CATEGORY_DISPLAY , dc->categories); - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_pl110; }
hw/dma/pl080.c +0 -1
@@ -381,7 +381,6 @@ static void pl080_class_init(ObjectClass *oc, void *data) {DeviceClass *dc =DEVICE_CLASS (oc); - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_pl080; }
hw/i386/kvm/clock.c +0 -1
@@ -114,7 +114,6 @@ static void kvmclock_class_init(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->realize = kvmclock_realize; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &kvmclock_vmsd; }
hw/i386/kvmvapic.c +0 -1
@@ -824,7 +824,6 @@ static void vapic_class_init(ObjectClass *klass, void *data) {DeviceClass *dc =DEVICE_CLASS (klass); - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->reset = vapic_reset; dc->vmsd = &vmstate_vapic; dc->realize = vapic_realize;
hw/intc/arm_gic.c +0 -1
@@ -704,7 +704,6 @@ static void arm_gic_class_init(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass);ARMGICClass *agc =ARM_GIC_CLASS (klass); - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ agc->parent_realize = dc->realize; dc->realize = arm_gic_realize; }
hw/intc/arm_gic_common.c +0 -1
@@ -156,7 +156,6 @@ static void arm_gic_common_class_init(ObjectClass *klass, void *data) dc->realize = arm_gic_common_realize; dc->props = arm_gic_common_properties; dc->vmsd = &vmstate_gic; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo arm_gic_common_type = {
hw/intc/arm_gic_kvm.c +0 -1
@@ -150,7 +150,6 @@ static void kvm_arm_gic_class_init(ObjectClass *klass, void *data) kgc->parent_reset = dc->reset; dc->realize = kvm_arm_gic_realize; dc->reset = kvm_arm_gic_reset; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo kvm_arm_gic_info = {
hw/intc/ioapic_common.c +0 -1
@@ -98,7 +98,6 @@ static void ioapic_common_class_init(ObjectClass *klass, void *data) dc->realize = ioapic_common_realize; dc->vmsd = &vmstate_ioapic_common; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo ioapic_common_type = {
hw/intc/pl190.c +0 -1
@@ -273,7 +273,6 @@ static void pl190_class_init(ObjectClass *klass, void *data)SysBusDeviceClass *k =SYS_BUS_DEVICE_CLASS (klass); k->init = pl190_init; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->reset = pl190_reset; dc->vmsd = &vmstate_pl190; }
hw/isa/isa-bus.c +0 -1
@@ -197,7 +197,6 @@ static void isabus_bridge_class_init(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->fw_name ="isa" ; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo isabus_bridge_info = {
hw/misc/arm_l2x0.c +0 -1
@@ -179,7 +179,6 @@ static void l2x0_class_init(ObjectClass *klass, void *data) k->init = l2x0_priv_init; dc->vmsd = &vmstate_l2x0; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->props = l2x0_properties; dc->reset = l2x0_priv_reset; }
hw/nvram/fw_cfg.c +0 -1
@@ -576,7 +576,6 @@ static void fw_cfg_class_init(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->realize = fw_cfg_realize; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->reset = fw_cfg_reset; dc->vmsd = &vmstate_fw_cfg; dc->props = fw_cfg_properties;
hw/pci-host/bonito.c +0 -2
@@ -819,11 +819,9 @@ static const TypeInfo bonito_info = {static void bonito_pcihost_class_init (ObjectClass *klass,void *data) { -DeviceClass *dc =DEVICE_CLASS (klass);SysBusDeviceClass *k =SYS_BUS_DEVICE_CLASS (klass); k->init = bonito_pcihost_initfn; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo bonito_pcihost_info = {
hw/pci-host/grackle.c +0 -2
@@ -143,10 +143,8 @@ static const TypeInfo grackle_pci_info = {static void pci_grackle_class_init (ObjectClass *klass,void *data) {SysBusDeviceClass *k =SYS_BUS_DEVICE_CLASS (klass); -DeviceClass *dc =DEVICE_CLASS (klass); k->init = pci_grackle_init_device; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo grackle_pci_host_info = {
hw/pci-host/piix.c +0 -1
@@ -730,7 +730,6 @@ static void i440fx_pcihost_class_init(ObjectClass *klass, void *data) hc->root_bus_path = i440fx_pcihost_root_bus_path; dc->realize = i440fx_pcihost_realize; dc->fw_name ="pci" ; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->props = i440fx_props; }
hw/pci-host/prep.c +0 -1
@@ -215,7 +215,6 @@ static void raven_pcihost_class_init(ObjectClass *klass, void *data)set_bit (DEVICE_CATEGORY_BRIDGE , dc->categories); dc->realize = raven_pcihost_realizefn; dc->fw_name ="pci" ; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo raven_pcihost_info = {
hw/ppc/spapr_vio.c +0 -2
@@ -528,11 +528,9 @@ static int spapr_vio_bridge_init(SysBusDevice *dev)static void spapr_vio_bridge_class_init (ObjectClass *klass,void *data) { -DeviceClass *dc =DEVICE_CLASS (klass);SysBusDeviceClass *k =SYS_BUS_DEVICE_CLASS (klass); k->init = spapr_vio_bridge_init; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo spapr_vio_bridge_info = {
hw/s390x/ipl.c +0 -1
@@ -181,7 +181,6 @@ static void s390_ipl_class_init(ObjectClass *klass, void *data) k->init = s390_ipl_init; dc->props = s390_ipl_properties; dc->reset = s390_ipl_reset; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo s390_ipl_info = {
hw/s390x/s390-virtio-bus.c +0 -2
@@ -676,11 +676,9 @@ static int s390_virtio_bridge_init(SysBusDevice *dev)static void s390_virtio_bridge_class_init (ObjectClass *klass,void *data) { -DeviceClass *dc =DEVICE_CLASS (klass);SysBusDeviceClass *k =SYS_BUS_DEVICE_CLASS (klass); k->init = s390_virtio_bridge_init; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo s390_virtio_bridge_info = {
hw/s390x/virtio-ccw.c +0 -2
@@ -1274,11 +1274,9 @@ static int virtual_css_bridge_init(SysBusDevice *dev)static void virtual_css_bridge_class_init (ObjectClass *klass,void *data) { -DeviceClass *dc =DEVICE_CLASS (klass);SysBusDeviceClass *k =SYS_BUS_DEVICE_CLASS (klass); k->init = virtual_css_bridge_init; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo virtual_css_bridge_info = {
hw/sd/pl181.c +0 -1
@@ -506,7 +506,6 @@ static void pl181_class_init(ObjectClass *klass, void *data) sdc->init = pl181_init; k->vmsd = &vmstate_pl181; k->reset = pl181_reset; - k->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo pl181_info = {
hw/timer/arm_mptimer.c +0 -1
@@ -297,7 +297,6 @@ static void arm_mptimer_class_init(ObjectClass *klass, void *data) sbc->init = arm_mptimer_init; dc->vmsd = &vmstate_arm_mptimer; dc->reset = arm_mptimer_reset; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->props = arm_mptimer_properties; }
hw/timer/hpet.c +0 -1
@@ -751,7 +751,6 @@ static void hpet_device_class_init(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->realize = hpet_realize; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->reset = hpet_reset; dc->vmsd = &vmstate_hpet; dc->props = hpet_device_properties;
hw/timer/pl031.c +0 -1
@@ -251,7 +251,6 @@ static void pl031_class_init(ObjectClass *klass, void *data)SysBusDeviceClass *k =SYS_BUS_DEVICE_CLASS (klass); k->init = pl031_init; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_pl031; }
Signature · 1 line
1.8.1.4
Re: [Qemu-devel] [PATCH 02/10] sysbus: Set cannot_instantiate_with_device_add_yet
Quoted 6 lines · On 17 October 2013 14:54, <[email protected]> wrote:
From: Markus Armbruster <[email protected]>
device_addplugs devices into suitable bus. For "real" buses, that actually connects the device. For sysbus, the connections need to be made separately, anddevice_addcan't do that. The device would be left unconncected, and could not possibly work.
"unconnected"
Quoted 2 lines
Many, but not all sysbus devices alreasy set
"already"
Quoted 5 lines
cannot_instantiate_with_device_add_yetin their class init function.Set it in their abstract base's class init function
sysbus_device_class_init(), and remove the now redundant assignments from device class init functions.
So I think this change is probably OK (but see my comments on patch 1 about what our definition of the flag is supposed to be). But I'd like to see a list of the devices which this patch makes no-user which previously weren't. Then I could eyeball the list and check whether there's anything in it which shouldn't be.
thanks
-- PMM
Re: [Qemu-devel] [PATCH 02/10] sysbus: Set cannot_instantiate_with_device_add_yet
Quoted 9 lines · Peter Maydell
Quoted 6 lines · On 17 October 2013 14:54, <[email protected]> wrote:
From: Markus Armbruster <[email protected]>
device_addplugs devices into suitable bus. For "real" buses, that actually connects the device. For sysbus, the connections need to be made separately, anddevice_addcan't do that. The device would be left unconncected, and could not possibly work."unconnected"
Will fix.
Quoted 3 lines
Quoted 1 line
Many, but not all sysbus devices alreasy set
"already"
Will fix.
Quoted 11 lines
Quoted 5 lines
cannot_instantiate_with_device_add_yetin their class init function.Set it in their abstract base's class init function
sysbus_device_class_init(), and remove the now redundant assignments from device class init functions.So I think this change is probably OK (but see my comments on patch 1 about what our definition of the flag is supposed to be). But I'd like to see a list of the devices which this patch makes no-user which previously weren't. Then I could eyeball the list and check whether there's anything in it which shouldn't be.
I'll include that list in v2.
Thanks!
PATCH 3/10 cpu: Document why cannot_instantiate_with_device_add_yet
In reply to [email protected]From: Markus Armbruster <[email protected]>
--- qom/cpu.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
qom/cpu.c +5 -1
@@ -254,7 +254,11 @@ static void cpu_class_init(ObjectClass *klass, void *data) k->gdb_read_register = cpu_common_gdb_read_register; k->gdb_write_register = cpu_common_gdb_write_register; dc->realize = cpu_common_realizefn; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ +/* +* Reason: CPUs still need special care by board code: wiring up +* IRQs, adding reset handlers, halting non-first CPUS, ... +*/ + dc->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo cpu_type_info = {
Signature · 1 line
1.8.1.4
Re: [Qemu-devel] [PATCH 03/10] cpu: Document why cannot_instantiate_with_device_add_yet
Quoted 4 lines · On 17 October 2013 14:54, <[email protected]> wrote:
From: Markus Armbruster <[email protected]>
Signed-off-by Markus Armbruster <[email protected]>
-- PMM
PATCH 4/10 apic: Document why cannot_instantiate_with_device_add_yet
In reply to [email protected]From: Markus Armbruster <[email protected]>
--- hw/intc/apic_common.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
hw/intc/apic_common.c +5 -1
@@ -386,9 +386,13 @@ static void apic_common_class_init(ObjectClass *klass, void *data) dc->vmsd = &vmstate_apic_common; dc->reset = apic_reset_common; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->props = apic_properties_common; idc->init = apic_init_common; +/* +* Reason: APIC and CPU need to be wired up by +* x86_cpu_apic_create() +*/ + dc->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo apic_common_type = {
Signature · 1 line
1.8.1.4
Re: [Qemu-devel] [PATCH 04/10] apic: Document why cannot_instantiate_with_device_add_yet
Quoted 4 lines · On 17 October 2013 14:54, <[email protected]> wrote:
From: Markus Armbruster <[email protected]>
Signed-off-by Markus Armbruster <[email protected]>
-- PMM
PATCH 5/10 pci-host: Consistently set cannot_instantiate_with_device_add_yet
In reply to [email protected]From: Markus Armbruster <[email protected]>
Many PCI host bridges consist of a sysbus device and a PCI device. You need both for the thing to work. Arguably, these bridges should be modelled as a single, composite devices instead of pairs of seemingly independent devices you can only use together, but we're not there, yet.
Since the sysbus part can't be instantiated with device_add, yet, permitting it with the PCI part is useless. We shouldn't offer useless options to the user, so let's set cannot_instantiate_with_device_add_yet for them.
It's already set for Bonito, grackle, i440FX, and raven. Document why.
Set it for the others: dec-21154, e500-host-bridge, gt64120_pci, pbm-pci, ppc4xx-host-bridge, sh_pci_host, u3-agp, uni-north-agp, uni-north-internal-pci, uni-north-pci, and versatile_pci_host.
--- hw/mips/gt64xxx_pci.c | 5 +++++ hw/pci-bridge/dec.c | 5 +++++ hw/pci-host/apb.c | 5 +++++ hw/pci-host/bonito.c | 6 +++++- hw/pci-host/grackle.c | 6 +++++- hw/pci-host/piix.c | 6 +++++- hw/pci-host/ppce500.c | 5 +++++ hw/pci-host/prep.c | 6 +++++- hw/pci-host/q35.c | 5 +++++ hw/pci-host/uninorth.c | 20 ++++++++++++++++++++ hw/pci-host/versatile.c | 5 +++++ hw/ppc/ppc4xx_pci.c | 5 +++++ hw/sh4/sh_pci.c | 5 +++++ 13 files changed, 80 insertions(+), 4 deletions(-)
hw/mips/gt64xxx_pci.c +5 -0
@@ -1157,6 +1157,11 @@ static void gt64120_pci_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_MARVELL_GT6412X ; k->revision =0x10 ; k->class_id =PCI_CLASS_BRIDGE_HOST ; +/* +* PCI-facing part of the host bridge, not usable without the +* host-facing part, which can't be device_add'ed, yet. +*/ + k->parent_class.cannot_instantiate_with_device_add_yet = true; }static const TypeInfo gt64120_pci_info = {
hw/pci-bridge/dec.c +5 -0
@@ -123,6 +123,11 @@ static void dec_21154_pci_host_class_init(ObjectClass *klass, void *data) k->revision =0x02 ; k->class_id =PCI_CLASS_BRIDGE_PCI ; k->is_bridge =1 ; +/* +* PCI-facing part of the host bridge, not usable without the +* host-facing part, which can't be device_add'ed, yet. +*/ + k->parent_class.cannot_instantiate_with_device_add_yet = true; }static const TypeInfo dec_21154_pci_host_info = {
hw/pci-host/apb.c +5 -0
@@ -521,6 +521,11 @@ static void pbm_pci_host_class_init(ObjectClass *klass, void *data) k->vendor_id =PCI_VENDOR_ID_SUN ; k->device_id =PCI_DEVICE_ID_SUN_SABRE ; k->class_id =PCI_CLASS_BRIDGE_HOST ; +/* +* PCI-facing part of the host bridge, not usable without the +* host-facing part, which can't be device_add'ed, yet. +*/ + k->parent_class.cannot_instantiate_with_device_add_yet = true; }static const TypeInfo pbm_pci_host_info = {
hw/pci-host/bonito.c +5 -1
@@ -806,8 +806,12 @@ static void bonito_class_init(ObjectClass *klass, void *data) k->revision =0x01 ; k->class_id =PCI_CLASS_BRIDGE_HOST ; dc->desc ="Host bridge" ; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_bonito; +/* +* PCI-facing part of the host bridge, not usable without the +* host-facing part, which can't be device_add'ed, yet. +*/ + dc->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo bonito_info = {
hw/pci-host/grackle.c +5 -1
@@ -130,7 +130,11 @@ static void grackle_pci_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_MOTOROLA_MPC106 ; k->revision =0x00 ; k->class_id =PCI_CLASS_BRIDGE_HOST ; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ +/* +* PCI-facing part of the host bridge, not usable without the +* host-facing part, which can't be device_add'ed, yet. +*/ + dc->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo grackle_pci_info = {
hw/pci-host/piix.c +5 -1
@@ -698,8 +698,12 @@ static void i440fx_class_init(ObjectClass *klass, void *data) k->revision =0x02 ; k->class_id =PCI_CLASS_BRIDGE_HOST ; dc->desc ="Host bridge" ; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_i440fx; +/* +* PCI-facing part of the host bridge, not usable without the +* host-facing part, which can't be device_add'ed, yet. +*/ + dc->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo i440fx_info = {
hw/pci-host/ppce500.c +5 -0
@@ -387,6 +387,11 @@ static void e500_host_bridge_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_MPC8533E ; k->class_id =PCI_CLASS_PROCESSOR_POWERPC ; dc->desc ="Host bridge" ; +/* +* PCI-facing part of the host bridge, not usable without the +* host-facing part, which can't be device_add'ed, yet. +*/ + dc->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo e500_host_bridge_info = {
hw/pci-host/prep.c +5 -1
@@ -198,7 +198,11 @@ static void raven_class_init(ObjectClass *klass, void *data) k->class_id =PCI_CLASS_BRIDGE_HOST ; dc->desc ="PReP Host Bridge - Motorola Raven" ; dc->vmsd = &vmstate_raven; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ +/* +* PCI-facing part of the host bridge, not usable without the +* host-facing part, which can't be device_add'ed, yet. +*/ + dc->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo raven_info = {
hw/pci-host/q35.c +5 -0
@@ -390,6 +390,11 @@ static void mch_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_INTEL_Q35_MCH ; k->revision =MCH_HOST_BRIDGE_REVISION_DEFAULT ; k->class_id =PCI_CLASS_BRIDGE_HOST ; +/* +* PCI-facing part of the host bridge, not usable without the +* host-facing part, which can't be device_add'ed, yet. +*/ + dc->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo mch_info = {
hw/pci-host/uninorth.c +20 -0
@@ -357,6 +357,11 @@ static void unin_main_pci_host_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_APPLE_UNI_N_PCI ; k->revision =0x00 ; k->class_id =PCI_CLASS_BRIDGE_HOST ; +/* +* PCI-facing part of the host bridge, not usable without the +* host-facing part, which can't be device_add'ed, yet. +*/ + k->parent_class.cannot_instantiate_with_device_add_yet = true; }static const TypeInfo unin_main_pci_host_info = {
@@ -375,6 +380,11 @@ static void u3_agp_pci_host_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_APPLE_U3_AGP ; k->revision =0x00 ; k->class_id =PCI_CLASS_BRIDGE_HOST ; +/* +* PCI-facing part of the host bridge, not usable without the +* host-facing part, which can't be device_add'ed, yet. +*/ + k->parent_class.cannot_instantiate_with_device_add_yet = true; }static const TypeInfo u3_agp_pci_host_info = {
@@ -393,6 +403,11 @@ static void unin_agp_pci_host_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_APPLE_UNI_N_AGP ; k->revision =0x00 ; k->class_id =PCI_CLASS_BRIDGE_HOST ; +/* +* PCI-facing part of the host bridge, not usable without the +* host-facing part, which can't be device_add'ed, yet. +*/ + k->parent_class.cannot_instantiate_with_device_add_yet = true; }static const TypeInfo unin_agp_pci_host_info = {
@@ -411,6 +426,11 @@ static void unin_internal_pci_host_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_APPLE_UNI_N_I_PCI ; k->revision =0x00 ; k->class_id =PCI_CLASS_BRIDGE_HOST ; +/* +* PCI-facing part of the host bridge, not usable without the +* host-facing part, which can't be device_add'ed, yet. +*/ + k->parent_class.cannot_instantiate_with_device_add_yet = true; }static const TypeInfo unin_internal_pci_host_info = {
hw/pci-host/versatile.c +5 -0
@@ -472,6 +472,11 @@ static void versatile_pci_host_class_init(ObjectClass *klass, void *data) k->vendor_id =PCI_VENDOR_ID_XILINX ; k->device_id =PCI_DEVICE_ID_XILINX_XC2VP30 ; k->class_id =PCI_CLASS_PROCESSOR_CO ; +/* +* PCI-facing part of the host bridge, not usable without the +* host-facing part, which can't be device_add'ed, yet. +*/ + k->parent_class.cannot_instantiate_with_device_add_yet = true; }static const TypeInfo versatile_pci_host_info = {
hw/ppc/ppc4xx_pci.c +5 -0
@@ -380,6 +380,11 @@ static void ppc4xx_host_bridge_class_init(ObjectClass *klass, void *data) k->vendor_id =PCI_VENDOR_ID_IBM ; k->device_id =PCI_DEVICE_ID_IBM_440GX ; k->class_id =PCI_CLASS_BRIDGE_OTHER ; +/* +* PCI-facing part of the host bridge, not usable without the +* host-facing part, which can't be device_add'ed, yet. +*/ + dc->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo ppc4xx_host_bridge_info = {
hw/sh4/sh_pci.c +5 -0
@@ -166,6 +166,11 @@ static void sh_pci_host_class_init(ObjectClass *klass, void *data) k->init = sh_pci_host_init; k->vendor_id =PCI_VENDOR_ID_HITACHI ; k->device_id =PCI_DEVICE_ID_HITACHI_SH7751R ; +/* +* PCI-facing part of the host bridge, not usable without the +* host-facing part, which can't be device_add'ed, yet. +*/ + k->parent_class.cannot_instantiate_with_device_add_yet = true; }static const TypeInfo sh_pci_host_info = {
Signature · 1 line
1.8.1.4
Re: [Qemu-devel] [PATCH 05/10] pci-host: Consistently set cannot_instantiate_with_device_add_yet
Quoted 7 lines · On 17 October 2013 14:54, <[email protected]> wrote:
From: Markus Armbruster <[email protected]>
Many PCI host bridges consist of a sysbus device and a PCI device. You need both for the thing to work. Arguably, these bridges should be modelled as a single, composite devices instead of pairs of seemingly independent devices you can only use together, but we're not there, yet.
I disagree here -- we should be using the modularity that our device model provides, not arbitrarily squashing things together into single objects just because we've foolishly exposed to the end user direct command line access to "create any random object whatsoever even if it doesn't make sense".
Quoted 4 lines
Since the sysbus part can't be instantiated with
device_add, yet, permitting it with the PCI part is useless. We shouldn't offer useless options to the user, so let's setcannot_instantiate_with_device_add_yetfor them.
It doesn't make sense to allow the user to create the on-PCI-bus representation of the host controller anyway even if they could device_add the host controller proper: creating the host controller will always automatically create the on-PCI-bus part.
hw/mips/gt64xxx_pci.c static void gt64120_pci_class_init(ObjectClass *klass, void *data) · line 1164
Quoted 11 lines
hw/mips/gt64xxx_pci.c +5
-0@@ -1157,6 +1157,11 @@ static void gt64120_pci_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_MARVELL_GT6412X ; k->revision =0x10 ; k->class_id =PCI_CLASS_BRIDGE_HOST ; +/* +* PCI-facing part of the host bridge, not usable without the +* host-facing part, which can't be device_add'ed, yet. +*/ + k->parent_class.cannot_instantiate_with_device_add_yet = true;
Please don't directly access parent_class -- you should be using the proper QOM cast macros to get the DeviceClass pointer.
thanks
-- PMM
Re: [Qemu-devel] [PATCH 05/10] pci-host: Consistently set cannot_instantiate_with_device_add_yet
Quoted 9 lines · Peter Maydell
Quoted 7 lines · On 17 October 2013 14:54, <[email protected]> wrote:
From: Markus Armbruster <[email protected]>
Many PCI host bridges consist of a sysbus device and a PCI device. You need both for the thing to work. Arguably, these bridges should be modelled as a single, composite devices instead of pairs of seemingly independent devices you can only use together, but we're not there, yet.
Quoted 5 lines
I disagree here -- we should be using the modularity that our device model provides, not arbitrarily squashing things together into single objects just because we've foolishly exposed to the end user direct command line access to "create any random object whatsoever even if it doesn't make sense".
I'm afraid I didn't express myself clearly. I'm not advocating *squashing* these components together. I'm saying that if A and B can only be used wired together, there should be a C composed of A, B and the necessary wiring, and that C is what actually gets put on the board by configuration.
Quoted 9 lines
Quoted 4 lines
Since the sysbus part can't be instantiated with
device_add, yet, permitting it with the PCI part is useless. We shouldn't offer useless options to the user, so let's setcannot_instantiate_with_device_add_yetfor them.It doesn't make sense to allow the user to create the on-PCI-bus representation of the host controller anyway even if they could
device_addthe host controller proper: creating the host controller will always automatically create the on-PCI-bus part.
Technically, a device_add i440FX-pcihost doesn't automatically create i440FX *now*.
I suspect we're arguing only about what exact kind of crazy device_add of the PCI-facing part of the PCI host bridge is. Assuming we actually agree it's crazy in *today's* state of things, does it matter what kind of crazy it is? If it doesn't matter, perhaps you could give me a hint on how to rephrase the commit message.
Quoted 7 lines
Quoted 7 lines
hw/mips/gt64xxx_pci.c +0
-0@@ -1157,6 +1157,11 @@ static voidgt64120_pci_class_init (ObjectClass *klass, void *data) k->device_id = PCI_DEVICE_ID_MARVELL_GT6412X; k->revision =0x10 ; k->class_id =PCI_CLASS_BRIDGE_HOST ;
Quoted 8 lines
hw/mips/gt64xxx_pci.c static void gt64120_pci_class_init(ObjectClass *klass, void *data) · line 1164
Quoted 5 lines
hw/mips/gt64xxx_pci.c +5
-0+/* +* PCI-facing part of the host bridge, not usable without the +* host-facing part, which can't be device_add'ed, yet. +*/ + k->parent_class.cannot_instantiate_with_device_add_yet = true;Please don't directly access
parent_class-- you should be using the proper QOM cast macros to get the DeviceClass pointer.
Will fix, thanks!
PATCH 6/10 ich9: Document why cannot_instantiate_with_device_add_yet
In reply to [email protected]From: Markus Armbruster <[email protected]>
An ICH9 southbridge contains several PCI devices, some of them with multiple functions. We model each function as a separate qdev. Two of them need some special wiring set up in pc_q35_init() to work: the LPC controller at 00:1f.0, and the SMBus controller at 00:1f.3.
--- hw/i2c/smbus_ich9.c | 6 +++++- hw/isa/lpc_ich9.c | 7 +++++-- 2 files changed, 10 insertions(+), 3 deletions(-)
hw/i2c/smbus_ich9.c +5 -1
@@ -97,11 +97,15 @@ static void ich9_smb_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_INTEL_ICH9_6 ; k->revision =ICH9_A2_SMB_REVISION ; k->class_id =PCI_CLASS_SERIAL_SMBUS ; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_ich9_smbus; dc->desc ="ICH9 SMBUS Bridge" ; k->init = ich9_smbus_initfn; k->config_write = ich9_smbus_write_config; +/* +* Reason: part of ICH9 southbridge, needs to be wired up by +* pc_q35_init() +*/ + dc->cannot_instantiate_with_device_add_yet = true; }i2c_bus *ich9_smb_init (PCIBus *bus,int devfn,uint32_t smb_io_base)
hw/isa/lpc_ich9.c +5 -2
@@ -604,14 +604,17 @@ static void ich9_lpc_class_init(ObjectClass *klass, void *data) dc->reset = ich9_lpc_reset; k->init = ich9_lpc_initfn; dc->vmsd = &vmstate_ich9_lpc; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ k->config_write = ich9_lpc_config_write; dc->desc ="ICH9 LPC bridge" ; k->vendor_id =PCI_VENDOR_ID_INTEL ; k->device_id =PCI_DEVICE_ID_INTEL_ICH9_8 ; k->revision =ICH9_A2_LPC_REVISION ; k->class_id =PCI_CLASS_BRIDGE_ISA ; - +/* +* Reason: part of ICH9 southbridge, needs to be wired up by +* pc_q35_init() +*/ + dc->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo ich9_lpc_info = {
Signature · 1 line
1.8.1.4
Re: [Qemu-devel] [PATCH 06/10] ich9: Document why cannot_instantiate_with_device_add_yet
Quoted 8 lines · On 17 October 2013 14:54, <[email protected]> wrote:
From: Markus Armbruster <[email protected]>
An ICH9 southbridge contains several PCI devices, some of them with multiple functions. We model each function as a separate qdev. Two of them need some special wiring set up in
pc_q35_init()to work: the LPC controller at 00:1f.0, and the SMBus controller at 00:1f.3.Signed-off-by Markus Armbruster <[email protected]>
-- PMM
PATCH 7/10 piix3 piix4: Clean up use of cannot_instantiate_with_device_add_yet
In reply to [email protected]From: Markus Armbruster <[email protected]>
A PIIX3/PIIX4 southbridge has multiple functions. We model each function as a separate qdev. Two of them need some special wiring set up in pc_init1() or mips_malta_init() to work: the ISA bridge at 01.0, and the SMBus controller at 01.3.
The IDE controller at 01.1 (piix3-ide, piix3-ide-xen, piix4-ide) has always had cannot_instantiate_with_device_add_yet set, but there is no obvious reason why device_add could not work for them. Drop it.
--- hw/acpi/piix4.c | 7 ++++++- hw/ide/piix.c | 3 --- hw/isa/piix4.c | 6 +++++- hw/pci-host/piix.c | 12 ++++++++++-- 4 files changed, 21 insertions(+), 7 deletions(-)
hw/acpi/piix4.c +6 -1
@@ -508,9 +508,14 @@ static void piix4_pm_class_init(ObjectClass *klass, void *data) k->revision =0x03 ; k->class_id =PCI_CLASS_BRIDGE_OTHER ; dc->desc ="PM" ; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ + dc->cannot_instantiate_with_device_add_yet = true; dc->vmsd = &vmstate_acpi; dc->props = piix4_pm_properties; +/* +* Reason: part of PIIX4 southbridge, needs to be wired up, +* e.g. by mips_malta_init() +*/ + dc->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo piix4_pm_info = {
hw/ide/piix.c +0 -3
@@ -248,7 +248,6 @@ static void piix3_ide_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_INTEL_82371SB_1 ; k->class_id =PCI_CLASS_STORAGE_IDE ;set_bit (DEVICE_CATEGORY_STORAGE , dc->categories); - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo piix3_ide_info = {
@@ -267,7 +266,6 @@ static void piix3_ide_xen_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_INTEL_82371SB_1 ; k->class_id =PCI_CLASS_STORAGE_IDE ;set_bit (DEVICE_CATEGORY_STORAGE , dc->categories); - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->unplug = pci_piix3_xen_ide_unplug; }
@@ -289,7 +287,6 @@ static void piix4_ide_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_INTEL_82371AB ; k->class_id =PCI_CLASS_STORAGE_IDE ;set_bit (DEVICE_CATEGORY_STORAGE , dc->categories); - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo piix4_ide_info = {
hw/isa/piix4.c +5 -1
@@ -113,8 +113,12 @@ static void piix4_class_init(ObjectClass *klass, void *data) k->device_id =PCI_DEVICE_ID_INTEL_82371AB_0 ; k->class_id =PCI_CLASS_BRIDGE_ISA ; dc->desc ="ISA bridge" ; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_piix4; +/* +* Reason: part of PIIX4 southbridge, needs to be wired up, +* e.g. by mips_malta_init() +*/ + dc->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo piix4_info = {
hw/pci-host/piix.c +10 -2
@@ -644,7 +644,6 @@ static void piix3_class_init(ObjectClass *klass, void *data) dc->desc ="ISA bridge" ; dc->vmsd = &vmstate_piix3; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ k->no_hotplug =1 ; k->init = piix3_initfn; k->config_write = piix3_write_config;
@@ -652,6 +651,11 @@ static void piix3_class_init(ObjectClass *klass, void *data)/* 82371SB PIIX3 PCI-to-ISA bridge (Step A1) */ k->device_id =PCI_DEVICE_ID_INTEL_82371SB_0 ; k->class_id =PCI_CLASS_BRIDGE_ISA ; +/* +* Reason: part of PIIX3 southbridge, needs to be wired up by +* pc_piix.c's pc_init1() +*/ + dc->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo piix3_info = {
@@ -668,7 +672,6 @@ static void piix3_xen_class_init(ObjectClass *klass, void *data) dc->desc ="ISA bridge" ; dc->vmsd = &vmstate_piix3; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ k->no_hotplug =1 ; k->init = piix3_initfn; k->config_write = piix3_write_config_xen;
@@ -676,6 +679,11 @@ static void piix3_xen_class_init(ObjectClass *klass, void *data)/* 82371SB PIIX3 PCI-to-ISA bridge (Step A1) */ k->device_id =PCI_DEVICE_ID_INTEL_82371SB_0 ; k->class_id =PCI_CLASS_BRIDGE_ISA ; +/* +* Reason: part of PIIX3 southbridge, needs to be wired up by +* pc_piix.c's pc_init1() +*/ + dc->cannot_instantiate_with_device_add_yet = true; };static const TypeInfo piix3_xen_info = {
Signature · 1 line
1.8.1.4
PATCH 8/10 vt82c686: Clean up use of cannot_instantiate_with_device_add_yet
In reply to [email protected]From: Markus Armbruster <[email protected]>
A VT82C686B southbridge has multiple functions. We model each function as a separate qdev. One of them need some special wiring set up in mips_fulong2e_init() to work: the ISA bridge at 05.0.
The IDE controller at 05.1 (via-ide) has always had cannot_instantiate_with_device_add_yet set, but there is no obvious reason why device_add could not work for them. Drop it.
--- hw/ide/via.c | 1 - hw/isa/vt82c686.c | 6 +++++- 2 files changed, 5 insertions(+), 2 deletions(-)
hw/ide/via.c +0 -1
@@ -225,7 +225,6 @@ static void via_ide_class_init(ObjectClass *klass, void *data) k->revision =0x06 ; k->class_id =PCI_CLASS_STORAGE_IDE ;set_bit (DEVICE_CATEGORY_STORAGE , dc->categories); - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo via_ide_info = {
hw/isa/vt82c686.c +5 -1
@@ -480,8 +480,12 @@ static void via_class_init(ObjectClass *klass, void *data) k->class_id =PCI_CLASS_BRIDGE_ISA ; k->revision =0x40 ; dc->desc ="ISA bridge" ; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_via; +/* +* Reason: part of VIA VT82C686 southbridge, needs to be wired up, +* e.g. by mips_fulong2e_init() +*/ + dc->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo via_info = {
Signature · 1 line
1.8.1.4
Re: [Qemu-devel] [PATCH 08/10] vt82c686: Clean up use of cannot_instantiate_with_device_add_yet
Quoted 11 lines · On 17 October 2013 14:54, <[email protected]> wrote:
From: Markus Armbruster <[email protected]>
A VT82C686B southbridge has multiple functions. We model each function as a separate qdev. One of them need some special wiring set up in
mips_fulong2e_init()to work: the ISA bridge at 05.0.The IDE controller at 05.1 (via-ide) has always had
cannot_instantiate_with_device_add_yetset, but there is no obvious reason whydevice_addcould not work for them. Drop it.Signed-off-by Markus Armbruster <[email protected]>
-- PMM
PATCH 9/10 isa: Clean up use of cannot_instantiate_with_device_add_yet
In reply to [email protected]From: Markus Armbruster <[email protected]>
Drop it when there's no obvious reason why device_add could not work. Else keep and document why.
* isa-fdc, port92, i8042, m48t59_isa, mc146818rtc, isa-pit, kvm-pit:
drop (from the last two by dropping it from their abstract base pit-common)
* pcspk: keep because of pointer property pit, and because realize sets global pcspk_state
* vmmouse: keep because of pointer property ps2_mouse
* vmport: keep because realize sets global port_state
* isa-i8259, kvm-i8259: keep (in their abstract base pic-common), because the PICs need additional wiring: its IRQ input lines are set up by board code, and the wiring of the slave to the master is hard-coded in device model code.
--- hw/audio/pcspk.c | 3 ++- hw/block/fdc.c | 1 - hw/i386/pc.c | 1 - hw/input/pckbd.c | 1 - hw/input/vmmouse.c | 3 ++- hw/intc/i8259_common.c | 8 +++++++- hw/misc/vmport.c | 3 ++- hw/timer/i8254_common.c | 1 - hw/timer/m48t59.c | 1 - hw/timer/mc146818rtc.c | 1 - 10 files changed, 13 insertions(+), 10 deletions(-)
hw/audio/pcspk.c +2 -1
@@ -192,8 +192,9 @@ static void pcspk_class_initfn(ObjectClass *klass, void *data) dc->realize = pcspk_realizefn;set_bit (DEVICE_CATEGORY_SOUND , dc->categories); - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->props = pcspk_properties; +/* Reason: pointer property "pit", realize sets global pcspk_state */ + dc->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo pcspk_info = {
hw/block/fdc.c +0 -1
@@ -2234,7 +2234,6 @@ static void isabus_fdc_class_init(ObjectClass *klass, void *data) dc->realize = isabus_fdc_realize; dc->fw_name ="fdc" ; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->reset = fdctrl_external_reset_isa; dc->vmsd = &vmstate_isa_fdc; dc->props = isa_fdc_properties;
hw/i386/pc.c +0 -1
@@ -544,7 +544,6 @@ static void port92_class_initfn(ObjectClass *klass, void *data) {DeviceClass *dc =DEVICE_CLASS (klass); - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->realize = port92_realizefn; dc->reset = port92_reset; dc->vmsd = &vmstate_port92_isa;
hw/input/pckbd.c +0 -1
@@ -522,7 +522,6 @@ static void i8042_class_initfn(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->realize = i8042_realizefn; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_kbd_isa; }
hw/input/vmmouse.c +2 -1
@@ -282,10 +282,11 @@ static void vmmouse_class_initfn(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->realize = vmmouse_realizefn; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->reset = vmmouse_reset; dc->vmsd = &vmstate_vmmouse; dc->props = vmmouse_properties; +/* Reason: pointer property "ps2_mouse" */ + dc->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo vmmouse_info = {
hw/intc/i8259_common.c +7 -1
@@ -135,9 +135,15 @@ static void pic_common_class_init(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->vmsd = &vmstate_pic_common; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->props = pic_properties_common; dc->realize = pic_common_realize; +/* +* Reason: unlike ordinary ISA devices, the PICs need additional +* wiring: its IRQ input lines are set up by board code, and the +* wiring of the slave to the master is hard-coded in device model +* code. +*/ + dc->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo pic_common_type = {
hw/misc/vmport.c +2 -1
@@ -162,7 +162,8 @@ static void vmport_class_initfn(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->realize = vmport_realizefn; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ +/* Reason: realize sets global port_state */ + dc->cannot_instantiate_with_device_add_yet = true; }static const TypeInfo vmport_info = {
hw/timer/i8254_common.c +0 -1
@@ -282,7 +282,6 @@ static void pit_common_class_init(ObjectClass *klass, void *data) dc->realize = pit_common_realize; dc->vmsd = &vmstate_pit_common; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ }static const TypeInfo pit_common_type = {
hw/timer/m48t59.c +0 -1
@@ -750,7 +750,6 @@ static void m48t59_isa_class_init(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->realize = m48t59_isa_realize; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->reset = m48t59_reset_isa; dc->props = m48t59_isa_properties; }
hw/timer/mc146818rtc.c +0 -1
@@ -906,7 +906,6 @@ static void rtc_class_initfn(ObjectClass *klass, void *data)DeviceClass *dc =DEVICE_CLASS (klass); dc->realize = rtc_realizefn; - dc->cannot_instantiate_with_device_add_yet = true;/* FIXME explain why */ dc->vmsd = &vmstate_rtc; dc->props = mc146818rtc_properties; }
Signature · 1 line
1.8.1.4
Re: [Qemu-devel] [PATCH 09/10] isa: Clean up use of cannot_instantiate_with_device_add_yet
Quoted 8 lines · On 17 October 2013 14:55, <[email protected]> wrote:
From: Markus Armbruster <[email protected]>
Drop it when there's no obvious reason why
device_addcould not work. Else keep and document why.* isa-fdc, port92, i8042,
m48t59_isa, mc146818rtc, isa-pit, kvm-pit:
drop (from the last two by dropping it from their abstract base pit-common)
port92 needs its a20_out qemu_irq line wiring up, doesn't it?
the pit devices have an output IRQ line that needs wiring up.
-- PMM
Re: [Qemu-devel] [PATCH 09/10] isa: Clean up use of cannot_instantiate_with_device_add_yet
Quoted 6 lines · Peter Maydell
Quoted 5 lines · On 17 October 2013 14:55, <[email protected]> wrote:
From: Markus Armbruster <[email protected]>
Drop it when there's no obvious reason why
device_addcould not work. Else keep and document why.
Quoted 7 lines
Quoted 3 lines
* isa-fdc, port92, i8042,
m48t59_isa, mc146818rtc, isa-pit, kvm-pit:
drop (from the last two by dropping it from their abstract base pit-common)port92 needs its
a20_outqemu_irqline wiring up, doesn't it?the pit devices have an output IRQ line that needs wiring up.
Good points; I'll have a second look at these two. Thanks!
PATCH 10/10 qdev: Do not let the user try to device_add when it cannot work
In reply to [email protected]From: Markus Armbruster <[email protected]>
Such devices have always been unavailable and omitted from the list of available devices shown by device_add help. Until commit 18b6dad silently broke the former, setting up nasty traps for unwary users, like this one:
$ qemu-system-x86_64 -nodefaults -monitor stdio -display none
QEMU 1.6.50 monitor - type 'help' for more information
(qemu) device_add apic
Segmentation fault (core dumped)I call that a regression. Fix it.
--- qdev-monitor.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
qdev-monitor.c +1 -1
@@ -477,7 +477,7 @@ DeviceState *qdev_device_add(QemuOpts *opts) } } -if (!obj) { +if (!obj ||DEVICE_CLASS (obj)->cannot_instantiate_with_device_add_yet) {qerror_report (QERR_INVALID_PARAMETER_VALUE ,"driver" ,"device type" );return NULL ; }
Signature · 1 line
1.8.1.4