hw/i386/pc: Remove deprecated 2.6 and 2.7 PC machines
v3 cover hw/i386/pc: Remove deprecated 2.6 and 2.7 PC machines
Since v2:
- Addressed Mark review comments and added his R-b tags
The versioned 'pc' and 'q35' machines up to 2.12 been marked as deprecated two releases ago, and are older than 6 years, so according to our support policy we can remove them.
This series only includes the 2.6 and 2.7 machines removal, as it is a big enough number of LoC removed. Rest will follow.
Based-on: <[email protected]>
Philippe Mathieu-Daudé (19):
hw/i386/pc: Remove deprecated pc-q35-2.6 and pc-i440fx-2.6 machines hw/i386/pc: Remove PCMachineClass::legacy_cpu_hotplug field hw/mips/loongson3_virt: Prefer using fw_cfg_init_mem() hw/nvram/fw_cfg: Factor fw_cfg_init_mem_internal() out hw/i386/x86: Remove X86MachineClass::fwcfg_dma_enabled field hw/nvram/fw_cfg: Remove fw_cfg_io_properties::dma_enabled hw/i386/pc: Remove pc_compat_2_6[] array
hw/intc/apic: Remove APICCommonState::legacy_instance_id field hw/core/machine: Remove hw_compat_2_6[] array hw/virtio/virtio-mmio: Remove
VirtIOMMIOProxy::format_transport_address field
hw/i386/pc: Remove deprecated pc-q35-2.7 and pc-i440fx-2.7 machines hw/i386/pc: Remove pc_compat_2_7[] array
hw/audio/pcspk: Remove PCSpkState::migrate field hw/core/machine: Remove hw_compat_2_7[] array hw/i386/intel_iommu: Remove IntelIOMMUState::buggy_eim field hw/intc/ioapic: Remove IOAPICCommonState::version field hw/virtio/virtio-pci: Remove VirtIOPCIProxy::ignore_backend_features
field
hw/char/virtio-serial: Do not expose the 'emergency-write' property hw/virtio/virtio-pci: Remove VIRTIO_PCI_FLAG_PAGE_PER_VQ definition
hw/intc/ioapic_internal.h | 3 +- include/hw/acpi/cpu_hotplug.h | 3 - include/hw/boards.h | 6 - include/hw/i386/apic_internal.h | 1 - include/hw/i386/intel_iommu.h | 1 - include/hw/i386/pc.h | 9 -- include/hw/i386/x86.h | 2 - include/hw/virtio/virtio-mmio.h | 1 - include/hw/virtio/virtio-pci.h | 2 - include/hw/virtio/virtio-serial.h | 2 - hw/acpi/cpu_hotplug.c | 230 ------------------------------ hw/audio/pcspk.c | 10 -- hw/char/virtio-serial-bus.c | 9 +- hw/core/machine.c | 17 --- hw/display/virtio-vga.c | 10 -- hw/i386/acpi-build.c | 4 +- hw/i386/intel_iommu.c | 5 +- hw/i386/microvm.c | 3 - hw/i386/multiboot.c | 7 +- hw/i386/pc.c | 18 --- hw/i386/pc_piix.c | 23 --- hw/i386/pc_q35.c | 24 ---- hw/i386/x86-common.c | 3 +- hw/i386/x86.c | 2 - hw/intc/apic_common.c | 5 - hw/intc/ioapic.c | 18 +-- hw/intc/ioapic_common.c | 2 +- hw/mips/loongson3_virt.c | 2 +- hw/nvram/fw_cfg.c | 46 +++--- hw/virtio/virtio-mmio.c | 15 -- hw/virtio/virtio-pci.c | 12 +- 31 files changed, 37 insertions(+), 458 deletions(-)
Signature · 1 line
2.47.1
PATCH v3 1/19 hw/i386/pc: Remove deprecated pc-q35-2.6 and pc-i440fx-2.6 machines
These machines has been supported for a period of more than 6 years. According to our versioned machine support policy (see commit ce80c4fa6ff "docs: document special exception for machine type deprecation & removal") they can now be removed.
--- hw/i386/pc_piix.c | 14 -------------- hw/i386/pc_q35.c | 14 -------------- 2 files changed, 28 deletions(-)
hw/i386/pc_piix.c +0 -14
@@ -764,20 +764,6 @@ static void pc_i440fx_machine_2_7_options(MachineClass *m)DEFINE_I440FX_MACHINE (2 ,7 ); -static void pc_i440fx_machine_2_6_options (MachineClass *m) -{ -X86MachineClass *x86mc =X86_MACHINE_CLASS (m); -PCMachineClass *pcmc =PC_MACHINE_CLASS (m); - -pc_i440fx_machine_2_7_options (m); - pcmc->legacy_cpu_hotplug = true; - x86mc->fwcfg_dma_enabled = false; -compat_props_add (m->compat_props, hw_compat_2_6, hw_compat_2_6_len); -compat_props_add (m->compat_props, pc_compat_2_6, pc_compat_2_6_len); -} - -DEFINE_I440FX_MACHINE (2 ,6 ); -#ifdef CONFIG_ISAPC static void isapc_machine_options (MachineClass *m) {
hw/i386/pc_q35.c +0 -14
@@ -658,17 +658,3 @@ static void pc_q35_machine_2_7_options(MachineClass *m) }DEFINE_Q35_MACHINE (2 ,7 ); - -static void pc_q35_machine_2_6_options (MachineClass *m) -{ -X86MachineClass *x86mc =X86_MACHINE_CLASS (m); -PCMachineClass *pcmc =PC_MACHINE_CLASS (m); - -pc_q35_machine_2_7_options (m); - pcmc->legacy_cpu_hotplug = true; - x86mc->fwcfg_dma_enabled = false; -compat_props_add (m->compat_props, hw_compat_2_6, hw_compat_2_6_len); -compat_props_add (m->compat_props, pc_compat_2_6, pc_compat_2_6_len); -} - -DEFINE_Q35_MACHINE (2 ,6 );
Signature · 1 line
2.47.1
Re: [PATCH v3 01/19] hw/i386/pc: Remove deprecated pc-q35-2.6 and pc-i440fx-2.6 machines
Quoted 11 lines · Philippe Mathieu-Daudé
Date: Fri, 2 May 2025 20:56:33 +0200
From: Philippe Mathieu-Daudé <[email protected]>
Subject: [PATCH v3 01/19] hw/i386/pc: Remove deprecated pc-q35-2.6 and
pc-i440fx-2.6 machines
X-Mailer: git-send-email 2.47.1These machines has been supported for a period of more than 6 years. According to our versioned machine support policy (see commit
ce80c4fa6ff"docs: document special exception for machine type deprecation & removal") they can now be removed.
Quoted 6 lines
Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Mark Cave-Ayland <[email protected]>--- hw/i386/pc_piix.c | 14 -------------- hw/i386/pc_q35.c | 14 -------------- 2 files changed, 28 deletions(-)
PATCH v3 2/19 hw/i386/pc: Remove PCMachineClass::legacy_cpu_hotplug field
In reply to Philippe Mathieu-DaudéThe PCMachineClass::legacy_cpu_hotplug boolean was only used by the pc-q35-2.6 and pc-i440fx-2.6 machines, which got removed. Remove it and simplify build_dsdt(), removing build_legacy_cpu_hotplug_aml() altogether.
Note, this field was added by commit 679dd1a957d ("pc: use new CPU hotplug interface since 2.7 machine type"):
Quoted 11 lines
For compatibility reasons PC/Q35 will start with legacy CPU hotplug interface by default but with new CPU hotplug AML code since 2.7 machine type. That way legacy firmware that doesn't use QEMU generated ACPI tables will be able to continue using legacy CPU hotplug interface.
While new machine type, with firmware supporting QEMU provided ACPI tables, will generate new CPU hotplug AML, which will switch to new CPU hotplug interface when guest OS executes its
_INImethod on ACPI tables loading.
--- include/hw/acpi/cpu_hotplug.h | 3 - include/hw/i386/pc.h | 3 - hw/acpi/cpu_hotplug.c | 230 ---------------------------------- hw/i386/acpi-build.c | 4 +- 4 files changed, 1 insertion(+), 239 deletions(-)
include/hw/acpi/cpu_hotplug.h +0 -3
@@ -34,7 +34,4 @@ void legacy_acpi_cpu_hotplug_init(MemoryRegion *parent, Object *owner,void acpi_switch_to_modern_cphp (AcpiCpuHotplug *gpe_cpu,CPUHotplugState *cpuhp_state,uint16_t io_port); - -void build_legacy_cpu_hotplug_aml (Aml *ctx,MachineState *machine, -uint16_t io_base);#endif
include/hw/i386/pc.h +0 -3
@@ -110,9 +110,6 @@ struct PCMachineClass {bool enforce_amd_1tb_hole;bool isa_bios_alias; -/* generate legacy CPU hotplug AML */ -bool legacy_cpu_hotplug; -/* use PVH to load kernels that support this feature */ bool pvh_enabled;
hw/acpi/cpu_hotplug.c +0 -230
233 lines not on this page · raw message
hw/i386/acpi-build.c +1 -3
@@ -1465,9 +1465,7 @@ build_dsdt(GArray *table_data, BIOSLinker *linker, }aml_append (dsdt, scope); -if (pcmc->legacy_cpu_hotplug) { -build_legacy_cpu_hotplug_aml (dsdt, machine, pm->cpu_hp_io_base); - }else { + {CPUHotplugFeatures opts = { .acpi_1_compatible = true, .has_legacy_cphp = true, .smi_path = pm->smi_on_cpuhp ?"\\_SB.PCI0.SMI0.SMIC" :NULL ,
Signature · 1 line
2.47.1
Re: [PATCH v3 02/19] hw/i386/pc: Remove PCMachineClass::legacy_cpu_hotplug field
hw/i386/acpi-build.c build_dsdt(GArray *table_data, BIOSLinker *linker, · line 1468
Quoted 12 lines
hw/i386/acpi-build.c +1
-3@@ -1465,9 +1465,7 @@ build_dsdt(GArray *table_data, BIOSLinker *linker, }aml_append (dsdt, scope); -if (pcmc->legacy_cpu_hotplug) { -build_legacy_cpu_hotplug_aml (dsdt, machine, pm->cpu_hp_io_base); - }else { + {
Why leave the {} here? for clarity?
hw/i386/acpi-build.c build_dsdt(GArray *table_data, BIOSLinker *linker, · line 1471
Quoted 3 lines
hw/i386/acpi-build.c +0
-0CPUHotplugFeatures opts = { .acpi_1_compatible = true, .has_legacy_cphp = true, .smi_path = pm->smi_on_cpuhp ?"\\_SB.PCI0.SMI0.SMIC" :NULL ,
Otherwise,
Re: [PATCH v3 02/19] hw/i386/pc: Remove PCMachineClass::legacy_cpu_hotplug field
Quoted 8 lines · Zhao Liu
Quoted 8 lines
hw/i386/acpi-build.c +0
-0@@ -1465,9 +1465,7 @@ build_dsdt(GArray *table_data, BIOSLinker *linker, }aml_append (dsdt, scope);
Quoted 6 lines
hw/i386/acpi-build.c build_dsdt(GArray *table_data, BIOSLinker *linker, · line 1468
Quoted 4 lines
hw/i386/acpi-build.c +1
-3-if (pcmc->legacy_cpu_hotplug) { -build_legacy_cpu_hotplug_aml (dsdt, machine, pm->cpu_hp_io_base); - }else { + {Why leave the {} here? for clarity?
Exactly, the diff-stat contains already more than 200 lines, too much to my taste. If we de-indent this block, per our coding style we also need to declare 'CPUHotplugFeatures opts' earlier. If you really want the braces removed, I'd to it in a follow-up cleanup patch.
Quoted 8 lines
hw/i386/acpi-build.c build_dsdt(GArray *table_data, BIOSLinker *linker, · line 1471
Quoted 4 lines
hw/i386/acpi-build.c +0
-0CPUHotplugFeatures opts = { .acpi_1_compatible = true, .has_legacy_cphp = true, .smi_path = pm->smi_on_cpuhp ?"\\_SB.PCI0.SMI0.SMIC" :NULL ,Otherwise,
Reviewed-by Zhao Liu <[email protected]>
Thanks!
PATCH v3 3/19 hw/mips/loongson3_virt: Prefer using fw_cfg_init_mem()
In reply to Philippe Mathieu-Daudéfw_cfg_init_mem_wide() is prefered to initialize fw_cfg with DMA support. Without DMA, use fw_cfg_init_mem().
--- hw/mips/loongson3_virt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
hw/mips/loongson3_virt.c +1 -1
@@ -286,7 +286,7 @@ static void fw_conf_init(void)FWCfgState *fw_cfg;hwaddr cfg_addr = virt_memmap[VIRT_FW_CFG ].base; - fw_cfg =fw_cfg_init_mem_wide (cfg_addr, cfg_addr +8 ,8 ,0 ,NULL ); + fw_cfg =fw_cfg_init_mem (cfg_addr, cfg_addr +8 ,8 );fw_cfg_add_i16 (fw_cfg,FW_CFG_NB_CPUS , (uint16_t )current_machine->smp.cpus);fw_cfg_add_i16 (fw_cfg,FW_CFG_MAX_CPUS , (uint16_t )current_machine->smp.max_cpus);fw_cfg_add_i64 (fw_cfg,FW_CFG_RAM_SIZE , loaderparams.ram_size);
Signature · 1 line
2.47.1
Re: [PATCH v3 03/19] hw/mips/loongson3_virt: Prefer using fw_cfg_init_mem()
Quoted 19 lines · Philippe Mathieu-Daudé
Date: Fri, 2 May 2025 20:56:35 +0200
From: Philippe Mathieu-Daudé <[email protected]>
Subject: [PATCH v3 03/19] hw/mips/loongson3_virt: Prefer using
fw_cfg_init_mem()
X-Mailer: git-send-email 2.47.1
fw_cfg_init_mem_wide()is prefered to initializefw_cfgwith DMA support. Without DMA, usefw_cfg_init_mem().Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- hw/mips/loongson3_virt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)hw/mips/loongson3_virt.c +0
-0
hw/mips/loongson3_virt.c static void fw_conf_init(void) · line 289
Quoted 5 lines
hw/mips/loongson3_virt.c +1
-1@@ -286,7 +286,7 @@ static void fw_conf_init(void)FWCfgState *fw_cfg;hwaddr cfg_addr = virt_memmap[VIRT_FW_CFG ].base; - fw_cfg =fw_cfg_init_mem_wide (cfg_addr, cfg_addr +8 ,8 ,0 ,NULL ); + fw_cfg =fw_cfg_init_mem (cfg_addr, cfg_addr +8 ,8 );
EMM, I'm not sure if I'm basing on the wrong code base... in which patch does fw_cfg_init_mem() accept the third argument?
Quoted 6 lines
hw/mips/loongson3_virt.c +0
-0fw_cfg_add_i16 (fw_cfg,FW_CFG_NB_CPUS , (uint16_t )current_machine->smp.cpus);fw_cfg_add_i16 (fw_cfg,FW_CFG_MAX_CPUS , (uint16_t )current_machine->smp.max_cpus);fw_cfg_add_i64 (fw_cfg,FW_CFG_RAM_SIZE , loaderparams.ram_size);Signature · 1 line
2.47.1
Re: [PATCH v3 03/19] hw/mips/loongson3_virt: Prefer using fw_cfg_init_mem()
Quoted 23 lines · Zhao Liu
Quoted 22 lines · Philippe Mathieu-Daudé
Date: Fri, 2 May 2025 20:56:35 +0200
From: Philippe Mathieu-Daudé <[email protected]>
Subject: [PATCH v3 03/19] hw/mips/loongson3_virt: Prefer using
fw_cfg_init_mem()
X-Mailer: git-send-email 2.47.1
fw_cfg_init_mem_wide()is prefered to initializefw_cfgwith DMA support. Without DMA, usefw_cfg_init_mem().Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- hw/mips/loongson3_virt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)hw/mips/loongson3_virt.c +0
-0@@ -286,7 +286,7 @@ static void fw_conf_init(void)FWCfgState *fw_cfg;hwaddr cfg_addr = virt_memmap[VIRT_FW_CFG ].base;
Quoted 5 lines
hw/mips/loongson3_virt.c static void fw_conf_init(void) · line 289
Quoted 2 lines
hw/mips/loongson3_virt.c +1
-1- fw_cfg =fw_cfg_init_mem_wide (cfg_addr, cfg_addr +8 ,8 ,0 ,NULL ); + fw_cfg =fw_cfg_init_mem (cfg_addr, cfg_addr +8 ,8 );EMM, I'm not sure if I'm basing on the wrong code base... in which patch does
fw_cfg_init_mem()accept the third argument?
Doh, I missed that, thanks :)
PATCH v3 4/19 hw/nvram/fw_cfg: Factor fw_cfg_init_mem_internal() out
In reply to Philippe Mathieu-DaudéFactor fw_cfg_init_mem_internal() out of fw_cfg_init_mem_wide(). In fw_cfg_init_mem_wide(), assert DMA arguments are provided. Callers without DMA have to use the fw_cfg_init_mem() helper.
--- hw/nvram/fw_cfg.c | 20 ++++++++++++++------ 1 file changed, 14 insertions(+), 6 deletions(-)
hw/nvram/fw_cfg.c +14 -6
@@ -1053,9 +1053,9 @@ FWCfgState *fw_cfg_init_io_dma(uint32_t iobase, uint32_t dma_iobase,return s; } -FWCfgState *fw_cfg_init_mem_wide (hwaddr ctl_addr, -hwaddr data_addr,uint32_t data_width, -hwaddr dma_addr,AddressSpace *dma_as) +static FWCfgState *fw_cfg_init_mem_internal (hwaddr ctl_addr, +hwaddr data_addr,uint32_t data_width, +hwaddr dma_addr,AddressSpace *dma_as) {DeviceState *dev;SysBusDevice *sbd;
@@ -1087,11 +1087,19 @@ FWCfgState *fw_cfg_init_mem_wide(hwaddr ctl_addr,return s; } +FWCfgState *fw_cfg_init_mem_wide (hwaddr ctl_addr, +hwaddr data_addr,uint32_t data_width, +hwaddr dma_addr,AddressSpace *dma_as) +{ +assert (dma_iobase && dma_as); +fw_cfg_init_mem_internal (ctl_addr, data_addr, data_addr, dma_addr, dma_as); +} +FWCfgState *fw_cfg_init_mem (hwaddr ctl_addr,hwaddr data_addr) { -return fw_cfg_init_mem_wide (ctl_addr, data_addr, - fw_cfg_data_mem_ops.valid.max_access_size, -0 ,NULL ); +return fw_cfg_init_mem_internal (ctl_addr, data_addr, + fw_cfg_data_mem_ops.valid.max_access_size, +0 ,NULL ); }
Signature · 1 line
2.47.1
Re: [PATCH v3 04/19] hw/nvram/fw_cfg: Factor fw_cfg_init_mem_internal() out
Quoted 33 lines · Philippe Mathieu-Daudé
Date: Fri, 2 May 2025 20:56:36 +0200
From: Philippe Mathieu-Daudé <[email protected]>
Subject: [PATCH v3 04/19] hw/nvram/fw_cfg: Factor
fw_cfg_init_mem_internal()out
X-Mailer: git-send-email 2.47.1Factor
fw_cfg_init_mem_internal()out offw_cfg_init_mem_wide(). Infw_cfg_init_mem_wide(), assert DMA arguments are provided. Callers without DMA have to use thefw_cfg_init_mem()helper.Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- hw/nvram/fw_cfg.c | 20 ++++++++++++++------ 1 file changed, 14 insertions(+), 6 deletions(-)hw/nvram/fw_cfg.c +3
-3@@ -1053,9 +1053,9 @@ FWCfgState *fw_cfg_init_io_dma(uint32_t iobase, uint32_t dma_iobase,return s; } -FWCfgState *fw_cfg_init_mem_wide (hwaddr ctl_addr, -hwaddr data_addr,uint32_t data_width, -hwaddr dma_addr,AddressSpace *dma_as) +static FWCfgState *fw_cfg_init_mem_internal (hwaddr ctl_addr, +hwaddr data_addr,uint32_t data_width, +hwaddr dma_addr,AddressSpace *dma_as) {DeviceState *dev;SysBusDevice *sbd;
hw/nvram/fw_cfg.c FWCfgState *fw_cfg_init_mem_wide(hwaddr ctl_addr, · line 1092
Quoted 6 lines
hw/nvram/fw_cfg.c +3
-0@@ -1087,11 +1087,19 @@ FWCfgState *fw_cfg_init_mem_wide(hwaddr ctl_addr,return s; } +FWCfgState *fw_cfg_init_mem_wide (hwaddr ctl_addr, +hwaddr data_addr,uint32_t data_width, +hwaddr dma_addr,AddressSpace *dma_as)
How about mentioning DMA in the name? fw_cfg_init_mem_dma.
hw/nvram/fw_cfg.c FWCfgState *fw_cfg_init_mem_wide(hwaddr ctl_addr, · line 1098
Quoted 6 lines
hw/nvram/fw_cfg.c +5
-0+{ +assert (dma_iobase && dma_as); +fw_cfg_init_mem_internal (ctl_addr, data_addr, data_addr, dma_addr, dma_as); +} +FWCfgState *fw_cfg_init_mem (hwaddr ctl_addr,hwaddr data_addr)
OK, now I'm sure fw_cfg_init_mem() only accepts 2 arguments. :-)
Quoted 13 lines
hw/nvram/fw_cfg.c +3
-3{ -return fw_cfg_init_mem_wide (ctl_addr, data_addr, - fw_cfg_data_mem_ops.valid.max_access_size, -0 ,NULL ); +return fw_cfg_init_mem_internal (ctl_addr, data_addr, + fw_cfg_data_mem_ops.valid.max_access_size, +0 ,NULL ); }Signature · 1 line
2.47.1
PATCH v3 5/19 hw/i386/x86: Remove X86MachineClass::fwcfg_dma_enabled field
In reply to Philippe Mathieu-DaudéThe X86MachineClass::fwcfg_dma_enabled boolean was only used by the pc-q35-2.6 and pc-i440fx-2.6 machines, which got removed. Remove it and simplify.
--- include/hw/i386/x86.h | 2 -- hw/i386/microvm.c | 3 --- hw/i386/multiboot.c | 7 +------ hw/i386/x86-common.c | 3 +-- hw/i386/x86.c | 2 -- 5 files changed, 2 insertions(+), 15 deletions(-)
include/hw/i386/x86.h +0 -2
@@ -29,8 +29,6 @@struct X86MachineClass {MachineClass parent; -/* use DMA capable linuxboot option rom */ -bool fwcfg_dma_enabled;/* CPU and apic information: */ bool apic_xrupt_override; };
hw/i386/microvm.c +0 -3
@@ -637,7 +637,6 @@ GlobalProperty microvm_properties[] = {static void microvm_class_init (ObjectClass *oc,const void *data) { -X86MachineClass *x86mc =X86_MACHINE_CLASS (oc);MicrovmMachineClass *mmc =MICROVM_MACHINE_CLASS (oc);MachineClass *mc =MACHINE_CLASS (oc);HotplugHandlerClass *hc =HOTPLUG_HANDLER_CLASS (oc);
@@ -671,8 +670,6 @@ static void microvm_class_init(ObjectClass *oc, const void *data) hc->unplug_request = microvm_device_unplug_request_cb; hc->unplug = microvm_device_unplug_cb; - x86mc->fwcfg_dma_enabled = true; - object_class_property_add(oc, MICROVM_MACHINE_RTC, "OnOffAuto", microvm_machine_get_rtc, microvm_machine_set_rtc,
hw/i386/multiboot.c +1 -6
@@ -153,7 +153,6 @@ int load_multiboot(X86MachineState *x86ms,int kernel_file_size, uint8_t *header) { - bool multiboot_dma_enabled = X86_MACHINE_GET_CLASS(x86ms)->fwcfg_dma_enabled; int i, is_multiboot = 0;uint32_t flags =0 ;uint32_t mh_entry_addr;
@@ -402,11 +401,7 @@ int load_multiboot(X86MachineState *x86ms,fw_cfg_add_bytes (fw_cfg,FW_CFG_INITRD_DATA , mb_bootinfo_data,sizeof (bootinfo)); -if (multiboot_dma_enabled) { - option_rom[nb_option_roms].name ="multiboot_dma.bin" ; - }else { - option_rom[nb_option_roms].name ="multiboot.bin" ; - } + option_rom[nb_option_roms].name ="multiboot_dma.bin" ; option_rom[nb_option_roms].bootindex =0 ; nb_option_roms++;
hw/i386/x86-common.c +1 -2
@@ -634,7 +634,6 @@ void x86_load_linux(X86MachineState *x86ms,int acpi_data_size, bool pvh_enabled) { - bool linuxboot_dma_enabled = X86_MACHINE_GET_CLASS(x86ms)->fwcfg_dma_enabled; uint16_t protocol;int setup_size, kernel_size, cmdline_size;int dtb_size, setup_data_offset;
@@ -993,7 +992,7 @@ void x86_load_linux(X86MachineState *x86ms, option_rom[nb_option_roms].bootindex =0 ; option_rom[nb_option_roms].name ="linuxboot.bin" ; -if (linuxboot_dma_enabled &&fw_cfg_dma_enabled (fw_cfg)) { +if (fw_cfg_dma_enabled (fw_cfg)) { option_rom[nb_option_roms].name ="linuxboot_dma.bin" ; } nb_option_roms++;
hw/i386/x86.c +0 -2
@@ -375,14 +375,12 @@ static void x86_machine_initfn(Object *obj)static void x86_machine_class_init (ObjectClass *oc,const void *data) {MachineClass *mc =MACHINE_CLASS (oc); -X86MachineClass *x86mc =X86_MACHINE_CLASS (oc);NMIClass *nc =NMI_CLASS (oc); mc->cpu_index_to_instance_props = x86_cpu_index_to_props; mc->get_default_cpu_node_id = x86_get_default_cpu_node_id; mc->possible_cpu_arch_ids = x86_possible_cpu_arch_ids; mc->kvm_type = x86_kvm_type; - x86mc->fwcfg_dma_enabled = true; nc->nmi_monitor_handler = x86_nmi; object_class_property_add(oc, X86_MACHINE_SMM, "OnOffAuto",
Signature · 1 line
2.47.1
Re: [PATCH v3 05/19] hw/i386/x86: Remove X86MachineClass::fwcfg_dma_enabled field
Quoted 10 lines · Philippe Mathieu-Daudé
Date: Fri, 2 May 2025 20:56:37 +0200
From: Philippe Mathieu-Daudé <[email protected]>
Subject: [PATCH v3 05/19] hw/i386/x86: Remove
X86MachineClass::fwcfg_dma_enabledfield
X-Mailer: git-send-email 2.47.1The
X86MachineClass::fwcfg_dma_enabledboolean was only used by the pc-q35-2.6 and pc-i440fx-2.6 machines, which got removed. Remove it and simplify.
Quoted 8 lines
Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- include/hw/i386/x86.h | 2 -- hw/i386/microvm.c | 3 --- hw/i386/multiboot.c | 7 +------ hw/i386/x86-common.c | 3 +-- hw/i386/x86.c | 2 -- 5 files changed, 2 insertions(+), 15 deletions(-)
Note for myself: fwcfg_dma_enabled is not a property.
PATCH v3 6/19 hw/nvram/fw_cfg: Remove fw_cfg_io_properties::dma_enabled
In reply to Philippe Mathieu-DaudéNow than all calls to fw_cfg_init_io_dma() pass DMA arguments, the 'dma_enabled' of the TYPE_FW_CFG_IO type is not used anymore. Remove it, simplifying fw_cfg_init_io_dma() and fw_cfg_io_realize().
Note, we can not remove the equivalent in fw_cfg_mem_properties[] because it is still used in HPPA and MIPS Loongson3 machines:
$ git grep -w fw_cfg_init_mem hw/hppa/machine.c:204: fw_cfg = fw_cfg_init_mem(addr, addr + 4); hw/mips/loongson3_virt.c:289: fw_cfg = fw_cfg_init_mem(cfg_addr, cfg_addr + 8, 8);
--- hw/nvram/fw_cfg.c | 26 ++++++++------------------ 1 file changed, 8 insertions(+), 18 deletions(-)
hw/nvram/fw_cfg.c +8 -18
@@ -1026,12 +1026,9 @@ FWCfgState *fw_cfg_init_io_dma(uint32_t iobase, uint32_t dma_iobase,FWCfgIoState *ios;FWCfgState *s;MemoryRegion *iomem =get_system_io (); -bool dma_requested = dma_iobase && dma_as; +assert (dma_iobase && dma_as); dev =qdev_new (TYPE_FW_CFG_IO ); -if (!dma_requested) { -qdev_prop_set_bit (dev,"dma_enabled" , false); - }object_property_add_child (OBJECT (qdev_get_machine ()),TYPE_FW_CFG ,OBJECT (dev));
@@ -1042,13 +1039,10 @@ FWCfgState *fw_cfg_init_io_dma(uint32_t iobase, uint32_t dma_iobase,memory_region_add_subregion (iomem, iobase, &ios->comb_iomem); s =FW_CFG (dev); - -if (s->dma_enabled) { -/* 64 bits for the address field */ - s->dma_as = dma_as; - s->dma_addr =0 ; -memory_region_add_subregion (iomem, dma_iobase, &s->dma_iomem); - } +/* 64 bits for the address field */ + s->dma_as = dma_as; + s->dma_addr =0 ; +memory_region_add_subregion (iomem, dma_iobase, &s->dma_iomem);return s; }
@@ -1185,8 +1179,6 @@ static void fw_cfg_file_slots_allocate(FWCfgState *s, Error **errp) }static const Property fw_cfg_io_properties[] = { -DEFINE_PROP_BOOL ("dma_enabled" , FWCfgIoState, parent_obj.dma_enabled, - true),DEFINE_PROP_UINT16 ("x-file-slots" , FWCfgIoState, parent_obj.file_slots,FW_CFG_FILE_SLOTS_DFLT ), };
@@ -1207,11 +1199,9 @@ static void fw_cfg_io_realize(DeviceState *dev, Error **errp)memory_region_init_io (&s->comb_iomem,OBJECT (s), &fw_cfg_comb_mem_ops,FW_CFG (s),"fwcfg" ,FW_CFG_CTL_SIZE ); -if (FW_CFG (s)->dma_enabled) { -memory_region_init_io (&FW_CFG (s)->dma_iomem,OBJECT (s), - &fw_cfg_dma_mem_ops,FW_CFG (s),"fwcfg.dma" , -sizeof (dma_addr_t)); - } +memory_region_init_io (&FW_CFG (s)->dma_iomem,OBJECT (s), + &fw_cfg_dma_mem_ops,FW_CFG (s),"fwcfg.dma" , +sizeof (dma_addr_t));fw_cfg_common_realize (dev, errp); }
Signature · 1 line
2.47.1
Re: [PATCH v3 06/19] hw/nvram/fw_cfg: Remove fw_cfg_io_properties::dma_enabled
Quoted 10 lines · Philippe Mathieu-Daudé
Date: Fri, 2 May 2025 20:56:38 +0200
From: Philippe Mathieu-Daudé <[email protected]>
Subject: [PATCH v3 06/19] hw/nvram/fw_cfg: Remove
fw_cfg_io_properties::dma_enabled
X-Mailer: git-send-email 2.47.1Now than all calls to
fw_cfg_init_io_dma()pass DMA arguments, the 'dma_enabled' of theTYPE_FW_CFG_IOtype is not used anymore. Remove it, simplifyingfw_cfg_init_io_dma()andfw_cfg_io_realize().
Quoted 6 lines
Note, we can not remove the equivalent in fw_cfg_mem_properties[] because it is still used in HPPA and MIPS Loongson3 machines:
$ git grep -w fw_cfg_init_mem hw/hppa/machine.c:204: fw_cfg = fw_cfg_init_mem(addr, addr + 4); hw/mips/loongson3_virt.c:289: fw_cfg = fw_cfg_init_mem(cfg_addr, cfg_addr + 8, 8);
The argument here is wrong. :-)
Quoted 4 lines
Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- hw/nvram/fw_cfg.c | 26 ++++++++------------------ 1 file changed, 8 insertions(+), 18 deletions(-)
PATCH v3 7/19 hw/i386/pc: Remove pc_compat_2_6[] array
In reply to Philippe Mathieu-DaudéThe pc_compat_2_6[] array was only used by the pc-q35-2.6 and pc-i440fx-2.6 machines, which got removed. Remove it.
--- include/hw/i386/pc.h | 3 --- hw/i386/pc.c | 8 -------- 2 files changed, 11 deletions(-)
include/hw/i386/pc.h +0 -3
@@ -292,9 +292,6 @@ extern const size_t pc_compat_2_8_len;extern GlobalProperty pc_compat_2_7[];extern const size_t pc_compat_2_7_len; -extern GlobalProperty pc_compat_2_6[]; -extern const size_t pc_compat_2_6_len; -#define DEFINE_PC_MACHINE (suffix, namestr, initfn, optsfn) \ static void pc_machine_##suffix##_class_init(ObjectClass *oc, \ const void *data) \
hw/i386/pc.c +0 -8
@@ -251,14 +251,6 @@ GlobalProperty pc_compat_2_7[] = { };const size_t pc_compat_2_7_len =G_N_ELEMENTS (pc_compat_2_7); -GlobalProperty pc_compat_2_6[] = { - {TYPE_X86_CPU ,"cpuid-0xb" ,"off" }, - {"vmxnet3" ,"romfile" ,"" }, - {TYPE_X86_CPU ,"fill-mtrr-mask" ,"off" }, - {"apic-common" ,"legacy-instance-id" ,"on" , } -}; -const size_t pc_compat_2_6_len =G_N_ELEMENTS (pc_compat_2_6); -/* * @PC_FW_DATA: * Size of the chunk of memory at the top of RAM for the BIOS ACPI tables
Signature · 1 line
2.47.1
Re: [PATCH v3 07/19] hw/i386/pc: Remove pc_compat_2_6[] array
Quoted 8 lines · Philippe Mathieu-Daudé
Date: Fri, 2 May 2025 20:56:39 +0200
From: Philippe Mathieu-Daudé <[email protected]>
Subject: [PATCH v3 07/19] hw/i386/pc: Remove pc_compat_2_6[] array
X-Mailer: git-send-email 2.47.1The pc_compat_2_6[] array was only used by the pc-q35-2.6 and pc-i440fx-2.6 machines, which got removed. Remove it.
Quoted 6 lines
Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Mark Cave-Ayland <[email protected]>--- include/hw/i386/pc.h | 3 --- hw/i386/pc.c | 8 -------- 2 files changed, 11 deletions(-)
PATCH v3 8/19 hw/intc/apic: Remove APICCommonState::legacy_instance_id field
In reply to Philippe Mathieu-DaudéThe APICCommonState::legacy_instance_id boolean was only set in the pc_compat_2_6[] array, via the 'legacy-instance-id=on' property. We removed all machines using that array, lets remove that property, simplifying apic_common_realize().
Because instance_id is initialized as initial_apic_id, we can not register vmstate_apic_common directly via dc->vmsd.
--- include/hw/i386/apic_internal.h | 1 - hw/intc/apic_common.c | 5 ----- 2 files changed, 6 deletions(-)
include/hw/i386/apic_internal.h +0 -1
@@ -188,7 +188,6 @@ struct APICCommonState {uint32_t vapic_control;DeviceState *vapic;hwaddr vapic_paddr;/* note: persistence via kvmvapic */ -bool legacy_instance_id;uint32_t extended_log_dest; };
hw/intc/apic_common.c +0 -5
@@ -294,9 +294,6 @@ static void apic_common_realize(DeviceState *dev, Error **errp) info->enable_tpr_reporting (s, true); } -if (s->legacy_instance_id) { - instance_id =VMSTATE_INSTANCE_ID_ANY ; - }vmstate_register_with_alias_id (NULL , instance_id, &vmstate_apic_common, s,-1 ,0 ,NULL );
@@ -412,8 +409,6 @@ static const Property apic_properties_common[] = {DEFINE_PROP_UINT8 ("version" , APICCommonState, version,0x14 ),DEFINE_PROP_BIT ("vapic" , APICCommonState, vapic_control,VAPIC_ENABLE_BIT , true), -DEFINE_PROP_BOOL ("legacy-instance-id" , APICCommonState, legacy_instance_id, - false), }; static void apic_common_get_id(Object *obj, Visitor *v, const char *name,
Signature · 1 line
2.47.1
PATCH v3 9/19 hw/core/machine: Remove hw_compat_2_6[] array
In reply to Philippe Mathieu-DaudéThe hw_compat_2_6[] array was only used by the pc-q35-2.6 and pc-i440fx-2.6 machines, which got removed. Remove it.
--- include/hw/boards.h | 3 --- hw/core/machine.c | 8 -------- 2 files changed, 11 deletions(-)
include/hw/boards.h +0 -3
@@ -841,7 +841,4 @@ extern const size_t hw_compat_2_8_len;extern GlobalProperty hw_compat_2_7[];extern const size_t hw_compat_2_7_len; -extern GlobalProperty hw_compat_2_6[]; -extern const size_t hw_compat_2_6_len; -#endif
hw/core/machine.c +0 -8
@@ -275,14 +275,6 @@ GlobalProperty hw_compat_2_7[] = { };const size_t hw_compat_2_7_len =G_N_ELEMENTS (hw_compat_2_7); -GlobalProperty hw_compat_2_6[] = { - {"virtio-mmio" ,"format_transport_address" ,"off" }, -/* Optional because not all virtio-pci devices support legacy mode */ - {"virtio-pci" ,"disable-modern" ,"on" , .optional = true }, - {"virtio-pci" ,"disable-legacy" ,"off" , .optional = true }, -}; -const size_t hw_compat_2_6_len =G_N_ELEMENTS (hw_compat_2_6); -MachineState *current_machine;static char *machine_get_kernel (Object *obj,Error **errp)
Signature · 1 line
2.47.1
Re: [PATCH v3 09/19] hw/core/machine: Remove hw_compat_2_6[] array
Quoted 9 lines · Philippe Mathieu-Daudé
The hw_compat_2_6[] array was only used by the pc-q35-2.6 and pc-i440fx-2.6 machines, which got removed. Remove it.
Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Mark Cave-Ayland <[email protected]>--- include/hw/boards.h | 3 --- hw/core/machine.c | 8 -------- 2 files changed, 11 deletions(-)
PATCH v3 10/19 hw/virtio/virtio-mmio: Remove VirtIOMMIOProxy::format_transport_address field
In reply to Philippe Mathieu-DaudéThe VirtIOMMIOProxy::format_transport_address boolean was only set in the hw_compat_2_6[] array, via the 'format_transport_address=off' property. We removed all machines using that array, lets remove that property, simplifying virtio_mmio_bus_get_dev_path().
--- include/hw/virtio/virtio-mmio.h | 1 - hw/virtio/virtio-mmio.c | 15 --------------- 2 files changed, 16 deletions(-)
include/hw/virtio/virtio-mmio.h +0 -1
@@ -66,7 +66,6 @@ struct VirtIOMMIOProxy {uint32_t guest_page_shift;/* virtio-bus */ VirtioBusState bus; -bool format_transport_address;/* Fields only used for non-legacy (v2) devices */ uint32_t guest_features[2 ];VirtIOMMIOQueue vqs[VIRTIO_QUEUE_MAX ];
hw/virtio/virtio-mmio.c +0 -15
@@ -752,8 +752,6 @@ static void virtio_mmio_pre_plugged(DeviceState *d, Error **errp)/* virtio-mmio device */ static const Property virtio_mmio_properties[] = { -DEFINE_PROP_BOOL ("format_transport_address" , VirtIOMMIOProxy, - format_transport_address, true),DEFINE_PROP_BOOL ("force-legacy" , VirtIOMMIOProxy, legacy, true),DEFINE_PROP_BIT ("ioeventfd" , VirtIOMMIOProxy, flags,VIRTIO_IOMMIO_FLAG_USE_IOEVENTFD_BIT , true),
@@ -815,19 +813,6 @@ static char *virtio_mmio_bus_get_dev_path(DeviceState *dev) virtio_mmio_proxy =VIRTIO_MMIO (virtio_mmio_bus->parent); proxy_path =qdev_get_dev_path (DEVICE (virtio_mmio_proxy)); -/* -* If @format_transport_address is false, then we just perform the same as -* virtio_bus_get_dev_path(): we delegate the address formatting for the -* device on the virtio-mmio bus to the bus that the virtio-mmio proxy -* (i.e., the device that implements the virtio-mmio bus) resides on. In -* this case the base address of the virtio-mmio transport will be -* invisible. -*/ -if (!virtio_mmio_proxy->format_transport_address) { -return proxy_path; - } - -/* Otherwise, we append the base address of the transport. */ section =memory_region_find (&virtio_mmio_proxy->iomem,0 ,0x200 );assert (section.mr);
Signature · 1 line
2.47.1
Re: [PATCH v3 10/19] hw/virtio/virtio-mmio: Remove VirtIOMMIOProxy::format_transport_address field
Quoted 11 lines · Philippe Mathieu-Daudé
The
VirtIOMMIOProxy::format_transport_addressboolean was only set in the hw_compat_2_6[] array, via the 'format_transport_address=off' property. We removed all machines using that array, lets remove that property, simplifyingvirtio_mmio_bus_get_dev_path().Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Mark Cave-Ayland <[email protected]>--- include/hw/virtio/virtio-mmio.h | 1 - hw/virtio/virtio-mmio.c | 15 --------------- 2 files changed, 16 deletions(-)
PATCH v3 11/19 hw/i386/pc: Remove deprecated pc-q35-2.7 and pc-i440fx-2.7 machines
In reply to Philippe Mathieu-DaudéThese machines has been supported for a period of more than 6 years. According to our versioned machine support policy (see commit ce80c4fa6ff "docs: document special exception for machine type deprecation & removal") they can now be removed.
--- hw/i386/pc_piix.c | 9 --------- hw/i386/pc_q35.c | 10 ---------- 2 files changed, 19 deletions(-)
hw/i386/pc_piix.c +0 -9
@@ -755,15 +755,6 @@ static void pc_i440fx_machine_2_8_options(MachineClass *m)DEFINE_I440FX_MACHINE (2 ,8 ); -static void pc_i440fx_machine_2_7_options (MachineClass *m) -{ -pc_i440fx_machine_2_8_options (m); -compat_props_add (m->compat_props, hw_compat_2_7, hw_compat_2_7_len); -compat_props_add (m->compat_props, pc_compat_2_7, pc_compat_2_7_len); -} - -DEFINE_I440FX_MACHINE (2 ,7 ); -#ifdef CONFIG_ISAPC static void isapc_machine_options (MachineClass *m) {
hw/i386/pc_q35.c +0 -10
@@ -648,13 +648,3 @@ static void pc_q35_machine_2_8_options(MachineClass *m) }DEFINE_Q35_MACHINE (2 ,8 ); - -static void pc_q35_machine_2_7_options (MachineClass *m) -{ -pc_q35_machine_2_8_options (m); - m->max_cpus =255 ; -compat_props_add (m->compat_props, hw_compat_2_7, hw_compat_2_7_len); -compat_props_add (m->compat_props, pc_compat_2_7, pc_compat_2_7_len); -} - -DEFINE_Q35_MACHINE (2 ,7 );
Signature · 1 line
2.47.1
Re: [PATCH v3 11/19] hw/i386/pc: Remove deprecated pc-q35-2.7 and pc-i440fx-2.7 machines
Quoted 40 lines · Philippe Mathieu-Daudé
These machines has been supported for a period of more than 6 years. According to our versioned machine support policy (see commit
ce80c4fa6ff"docs: document special exception for machine type deprecation & removal") they can now be removed.Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Mark Cave-Ayland <[email protected]>--- hw/i386/pc_piix.c | 9 --------- hw/i386/pc_q35.c | 10 ---------- 2 files changed, 19 deletions(-)hw/i386/pc_piix.c +0
-9@@ -755,15 +755,6 @@ static void pc_i440fx_machine_2_8_options(MachineClass *m)DEFINE_I440FX_MACHINE (2 ,8 ); -static void pc_i440fx_machine_2_7_options (MachineClass *m) -{ -pc_i440fx_machine_2_8_options (m); -compat_props_add (m->compat_props, hw_compat_2_7, hw_compat_2_7_len); -compat_props_add (m->compat_props, pc_compat_2_7, pc_compat_2_7_len); -} - -DEFINE_I440FX_MACHINE (2 ,7 ); -#ifdef CONFIG_ISAPC static void isapc_machine_options (MachineClass *m) {hw/i386/pc_q35.c +0
-0@@ -648,13 +648,3 @@ static void pc_q35_machine_2_8_options(MachineClass *m) }DEFINE_Q35_MACHINE (2 ,8 );
hw/i386/pc_q35.c static void pc_q35_machine_2_8_options(MachineClass *m) · line 660
Quoted 10 lines
hw/i386/pc_q35.c +0
-10- -static void pc_q35_machine_2_7_options (MachineClass *m) -{ -pc_q35_machine_2_8_options (m); - m->max_cpus =255 ; -compat_props_add (m->compat_props, hw_compat_2_7, hw_compat_2_7_len); -compat_props_add (m->compat_props, pc_compat_2_7, pc_compat_2_7_len); -} - -DEFINE_Q35_MACHINE (2 ,7 );
As noticed Thomas for 2.4, here I missed to squash:
-- >8 --
tests/qtest/test-x86-cpuid-compat.c +0 -11
@@ -348,11 +347,0 @@ int main(int argc, char **argv) -if (qtest_has_machine ("pc-i440fx-2.7" )) { -add_cpuid_test ("x86/cpuid/auto-level/pc-2.7" , -"486" ,"arat=on,avx512vbmi=on,xsaveopt=on" , -"pc-i440fx-2.7" ,"level" ,1 ); -add_cpuid_test ("x86/cpuid/auto-xlevel/pc-2.7" , -"486" ,"3dnow=on,sse4a=on,invtsc=on,npt=on,svm=on" , -"pc-i440fx-2.7" ,"xlevel" ,0 ); -add_cpuid_test ("x86/cpuid/auto-xlevel2/pc-2.7" , -"486" ,"xstore=on" ,"pc-i440fx-2.7" , -"xlevel2" ,0 ); - }
---
Thomas Huth
Reviewed-byRe: [PATCH v3 11/19] hw/i386/pc: Remove deprecated pc-q35-2.7 and pc-i440fx-2.7 machines
Quoted 59 lines · Philippe Mathieu-Daudé
Quoted 38 lines · Philippe Mathieu-Daudé
These machines has been supported for a period of more than 6 years. According to our versioned machine support policy (see commit
ce80c4fa6ff"docs: document special exception for machine type deprecation & removal") they can now be removed.Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Mark Cave-Ayland <[email protected]>---
hw/i386/pc_piix.c| 9 ---------
hw/i386/pc_q35.c | 10 ----------2 files changed, 19 deletions(-)hw/i386/pc_piix.c +0
-9@@ -755,15 +755,6 @@ static void pc_i440fx_machine_2_8_options(MachineClass *m) DEFINE_I440FX_MACHINE(2, 8); -static void pc_i440fx_machine_2_7_options (MachineClass *m) -{ - pc_i440fx_machine_2_8_options(m); - compat_props_add(m->compat_props, hw_compat_2_7, hw_compat_2_7_len); - compat_props_add(m->compat_props, pc_compat_2_7, pc_compat_2_7_len); -} - -DEFINE_I440FX_MACHINE (2 ,7 ); - #ifdef CONFIG_ISAPC static void isapc_machine_options(MachineClass *m) {hw/i386/pc_q35.c +0
-0@@ -648,13 +648,3 @@ static void pc_q35_machine_2_8_options(MachineClass *m) } DEFINE_Q35_MACHINE(2, 8);hw/i386/pc_q35.c static void pc_q35_machine_2_8_options(MachineClass *m) · line 659
Quoted 10 lines
hw/i386/pc_q35.c +0
-10- -static void pc_q35_machine_2_7_options (MachineClass *m) -{ - pc_q35_machine_2_8_options(m); - m->max_cpus = 255; - compat_props_add(m->compat_props, hw_compat_2_7, hw_compat_2_7_len); - compat_props_add(m->compat_props, pc_compat_2_7, pc_compat_2_7_len); -} - -DEFINE_Q35_MACHINE (2 ,7 );As noticed Thomas for 2.4, here I missed to squash:
-- >8 --
tests/qtest/test-x86-cpuid-compat.c +0
-1@@ -348,11 +347,0 @@ int main(int argc, char **argv) - if (qtest_has_machine("pc-i440fx-2.7")) {
tests/qtest/test-x86-cpuid-compat.c int main(int argc, char **argv) · line 358
Quoted 10 lines
tests/qtest/test-x86-cpuid-compat.c +0
-10- add_cpuid_test("x86/cpuid/auto-level/pc-2.7", - "486", "arat=on,avx512vbmi=on,xsaveopt=on", - "pc-i440fx-2.7", "level", 1); - add_cpuid_test("x86/cpuid/auto-xlevel/pc-2.7", - "486", "3dnow=on,sse4a=on,invtsc=on,npt=on,svm=on", - "pc-i440fx-2.7", "xlevel", 0); - add_cpuid_test("x86/cpuid/auto-xlevel2/pc-2.7", - "486", "xstore=on", "pc-i440fx-2.7", - "xlevel2", 0); - }
With that hunk added:
PATCH v3 12/19 hw/i386/pc: Remove pc_compat_2_7[] array
In reply to Philippe Mathieu-DaudéThe pc_compat_2_7[] array was only used by the pc-q35-2.7 and pc-i440fx-2.7 machines, which got removed. Remove it.
--- include/hw/i386/pc.h | 3 --- hw/i386/pc.c | 10 ---------- 2 files changed, 13 deletions(-)
include/hw/i386/pc.h +0 -3
@@ -289,9 +289,6 @@ extern const size_t pc_compat_2_9_len;extern GlobalProperty pc_compat_2_8[];extern const size_t pc_compat_2_8_len; -extern GlobalProperty pc_compat_2_7[]; -extern const size_t pc_compat_2_7_len; -#define DEFINE_PC_MACHINE (suffix, namestr, initfn, optsfn) \ static void pc_machine_##suffix##_class_init(ObjectClass *oc, \ const void *data) \
hw/i386/pc.c +0 -10
@@ -241,16 +241,6 @@ GlobalProperty pc_compat_2_8[] = { };const size_t pc_compat_2_8_len =G_N_ELEMENTS (pc_compat_2_8); -GlobalProperty pc_compat_2_7[] = { - {TYPE_X86_CPU ,"l3-cache" ,"off" }, - {TYPE_X86_CPU ,"full-cpuid-auto-level" ,"off" }, - {"Opteron_G3" "-" TYPE_X86_CPU ,"family" ,"15" }, - {"Opteron_G3" "-" TYPE_X86_CPU ,"model" ,"6" }, - {"Opteron_G3" "-" TYPE_X86_CPU ,"stepping" ,"1" }, - {"isa-pcspk" ,"migrate" ,"off" }, -}; -const size_t pc_compat_2_7_len =G_N_ELEMENTS (pc_compat_2_7); -/* * @PC_FW_DATA: * Size of the chunk of memory at the top of RAM for the BIOS ACPI tables
Signature · 1 line
2.47.1
Re: [PATCH v3 12/19] hw/i386/pc: Remove pc_compat_2_7[] array
Quoted 31 lines · Philippe Mathieu-Daudé
The pc_compat_2_7[] array was only used by the pc-q35-2.7 and pc-i440fx-2.7 machines, which got removed. Remove it.
Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Mark Cave-Ayland <[email protected]>--- include/hw/i386/pc.h | 3 --- hw/i386/pc.c | 10 ---------- 2 files changed, 13 deletions(-)include/hw/i386/pc.h +0
-3@@ -289,9 +289,6 @@ extern const size_t pc_compat_2_9_len;extern GlobalProperty pc_compat_2_8[];extern const size_t pc_compat_2_8_len; -extern GlobalProperty pc_compat_2_7[]; -extern const size_t pc_compat_2_7_len; -#define DEFINE_PC_MACHINE (suffix, namestr, initfn, optsfn) \ static void pc_machine_##suffix##_class_init(ObjectClass *oc, \ const void *data) \hw/i386/pc.c +0
-0@@ -241,16 +241,6 @@ GlobalProperty pc_compat_2_8[] = { };const size_t pc_compat_2_8_len =G_N_ELEMENTS (pc_compat_2_8);
hw/i386/pc.c GlobalProperty pc_compat_2_8[] = { · line 252
Quoted 10 lines
hw/i386/pc.c +0
-9-GlobalProperty pc_compat_2_7[] = { - {TYPE_X86_CPU ,"l3-cache" ,"off" }, - {TYPE_X86_CPU ,"full-cpuid-auto-level" ,"off" }, - {"Opteron_G3" "-" TYPE_X86_CPU ,"family" ,"15" }, - {"Opteron_G3" "-" TYPE_X86_CPU ,"model" ,"6" }, - {"Opteron_G3" "-" TYPE_X86_CPU ,"stepping" ,"1" }, - {"isa-pcspk" ,"migrate" ,"off" }, -}; -const size_t pc_compat_2_7_len =G_N_ELEMENTS (pc_compat_2_7);
I'd really appreciate if you could provide clean-up patches for TYPE_X86_CPU, too. Otherwise I'm pretty sure we'll forget that there is some clean up possibility here.
Anyway, for this patch here:
Re: [PATCH v3 12/19] hw/i386/pc: Remove pc_compat_2_7[] array
Hi Thomas,
Quoted 34 lines · Thomas Huth
Quoted 33 lines · Philippe Mathieu-Daudé
The pc_compat_2_7[] array was only used by the pc-q35-2.7 and pc-i440fx-2.7 machines, which got removed. Remove it.
Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Mark Cave-Ayland <[email protected]>---
include/hw/i386/pc.h| 3 ---
hw/i386/pc.c | 10 ----------2 files changed, 13 deletions(-)include/hw/i386/pc.h +0
-3@@ -289,9 +289,6 @@ extern const size_t pc_compat_2_9_len; extern GlobalProperty pc_compat_2_8[]; extern const size_t pc_compat_2_8_len; -extern GlobalProperty pc_compat_2_7[]; -extern const size_t pc_compat_2_7_len; - #define DEFINE_PC_MACHINE(suffix, namestr, initfn, optsfn) \ static void pc_machine_##suffix##_class_init(ObjectClass *oc, \ const void *data) \hw/i386/pc.c +0
-3@@ -241,16 +241,6 @@ GlobalProperty pc_compat_2_8[] = { }; const size_t pc_compat_2_8_len = G_N_ELEMENTS(pc_compat_2_8); -GlobalProperty pc_compat_2_7[] = { - { TYPE_X86_CPU, "l3-cache", "off" }, - { TYPE_X86_CPU, "full-cpuid-auto-level", "off" },
Quoted 8 lines
hw/i386/pc.c GlobalProperty pc_compat_2_8[] = { · line 252
Quoted 6 lines
hw/i386/pc.c +0
-6- { "Opteron_G3" "-" TYPE_X86_CPU, "family", "15" }, - { "Opteron_G3" "-" TYPE_X86_CPU, "model", "6" }, - { "Opteron_G3" "-" TYPE_X86_CPU, "stepping", "1" }, - { "isa-pcspk", "migrate", "off" }, -}; -const size_t pc_compat_2_7_len =G_N_ELEMENTS (pc_compat_2_7);I'd really appreciate if you could provide clean-up patches for
TYPE_X86_CPU, too. Otherwise I'm pretty sure we'll forget that there is some clean up possibility here.
Well TBH it is too exhausting to keep rebasing these patches without feedback from maintainers. I'll respin a v4 with Zhao and your comments addressed but without touching the TYPE_X86_CPU properties. If maintainers prefer to remove dead code in one go -- something I certainly understand from a maintainer PoV -- I'll let someone else do it, taking over my series.
Quoted 2 lines
Anyway, for this patch here:
Reviewed-by Thomas Huth <[email protected]>
Thank you for your support with these series.
Regards,
Phil.
Re: [PATCH v3 12/19] hw/i386/pc: Remove pc_compat_2_7[] array
Quoted 46 lines · Philippe Mathieu-Daudé
Hi Thomas,
Quoted 34 lines · Thomas Huth
Quoted 33 lines · Philippe Mathieu-Daudé
The pc_compat_2_7[] array was only used by the pc-q35-2.7 and pc-i440fx-2.7 machines, which got removed. Remove it.
Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Mark Cave-Ayland <[email protected]>---
include/hw/i386/pc.h| 3 ---
hw/i386/pc.c | 10 ----------2 files changed, 13 deletions(-)include/hw/i386/pc.h +0
-3@@ -289,9 +289,6 @@ extern const size_t pc_compat_2_9_len; extern GlobalProperty pc_compat_2_8[]; extern const size_t pc_compat_2_8_len; -extern GlobalProperty pc_compat_2_7[]; -extern const size_t pc_compat_2_7_len; - #define DEFINE_PC_MACHINE(suffix, namestr, initfn, optsfn) \ static void pc_machine_##suffix##_class_init(ObjectClass *oc, \ const void *data) \hw/i386/pc.c +0
-3@@ -241,16 +241,6 @@ GlobalProperty pc_compat_2_8[] = { }; const size_t pc_compat_2_8_len = G_N_ELEMENTS(pc_compat_2_8); -GlobalProperty pc_compat_2_7[] = { - { TYPE_X86_CPU, "l3-cache", "off" }, - { TYPE_X86_CPU, "full-cpuid-auto-level", "off" },Quoted 8 lines
hw/i386/pc.c GlobalProperty pc_compat_2_8[] = { · line 252
Quoted 6 lines
hw/i386/pc.c +0
-6- { "Opteron_G3" "-" TYPE_X86_CPU, "family", "15" }, - { "Opteron_G3" "-" TYPE_X86_CPU, "model", "6" }, - { "Opteron_G3" "-" TYPE_X86_CPU, "stepping", "1" }, - { "isa-pcspk", "migrate", "off" }, -}; -const size_t pc_compat_2_7_len =G_N_ELEMENTS (pc_compat_2_7);I'd really appreciate if you could provide clean-up patches for
TYPE_X86_CPU, too. Otherwise I'm pretty sure we'll forget that there is some clean up possibility here.
Quoted 2 lines
Well TBH it is too exhausting to keep rebasing these patches without feedback from maintainers.
Ok, fair, if it's too much right now, it can of course also be done by someone else later.
Quoted 5 lines
I'll respin a v4 with Zhao and your comments
addressed but without touching theTYPE_X86_CPUproperties. If maintainers prefer to remove dead code in one go -- something I certainly understand from a maintainer PoV -- I'll let someone else do it, taking over my series.
Since the x86 maintainers seem to be busy with other stuff right now ... I'm currently preparing a machine type cleanup pull request that contains Daniels patches to automatically disable the old machine types, and my own patches to remove the old s390x machine types. If nobody objects, I could also add your patches (at least the reviewed ones) there. Michael? Paolo?
Thomas
Re: [PATCH v3 12/19] hw/i386/pc: Remove pc_compat_2_7[] array
In reply to Philippe Mathieu-DaudéQuoted 52 lines · Philippe Mathieu-Daudé
Date: Thu, 8 May 2025 12:40:35 +0200
From: Philippe Mathieu-Daudé <[email protected]>
Subject: Re: [PATCH v3 12/19] hw/i386/pc: Remove pc_compat_2_7[] arrayHi Thomas,
Quoted 36 lines · Thomas Huth
Quoted 35 lines · Philippe Mathieu-Daudé
The pc_compat_2_7[] array was only used by the pc-q35-2.7 and pc-i440fx-2.7 machines, which got removed. Remove it.
Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Mark Cave-Ayland <[email protected]>---
include/hw/i386/pc.h| 3 ---
hw/i386/pc.c | 10 ----------2 files changed, 13 deletions(-)include/hw/i386/pc.h +0
-3@@ -289,9 +289,6 @@ extern const size_t pc_compat_2_9_len; extern GlobalProperty pc_compat_2_8[]; extern const size_t pc_compat_2_8_len; -extern GlobalProperty pc_compat_2_7[]; -extern const size_t pc_compat_2_7_len; - #define DEFINE_PC_MACHINE(suffix, namestr, initfn, optsfn) \ static void pc_machine_##suffix##_class_init(ObjectClass *oc, \ const void *data) \hw/i386/pc.c +0
-5@@ -241,16 +241,6 @@ GlobalProperty pc_compat_2_8[] = { }; const size_t pc_compat_2_8_len = G_N_ELEMENTS(pc_compat_2_8); -GlobalProperty pc_compat_2_7[] = { - { TYPE_X86_CPU, "l3-cache", "off" }, - { TYPE_X86_CPU, "full-cpuid-auto-level", "off" }, - { "Opteron_G3" "-" TYPE_X86_CPU, "family", "15" }, - { "Opteron_G3" "-" TYPE_X86_CPU, "model", "6" },Quoted 8 lines
hw/i386/pc.c GlobalProperty pc_compat_2_8[] = { · line 252
Quoted 4 lines
hw/i386/pc.c +0
-4- { "Opteron_G3" "-" TYPE_X86_CPU, "stepping", "1" }, - { "isa-pcspk", "migrate", "off" }, -}; -const size_t pc_compat_2_7_len =G_N_ELEMENTS (pc_compat_2_7);I'd really appreciate if you could provide clean-up patches for
TYPE_X86_CPU, too. Otherwise I'm pretty sure we'll forget that there is some clean up possibility here.
Quoted 6 lines
Well TBH it is too exhausting to keep rebasing these patches without feedback from maintainers. I'll respin a v4 with Zhao and your comments addressed but without touching the
TYPE_X86_CPUproperties. If maintainers prefer to remove dead code in one go -- something I certainly understand from a maintainer PoV -- I'll let someone else do it, taking over my series.
Hi Philippe, I think I could volunteer help you to revisit the history of these properties (they're also too old for me :-)), and help identify if these properties should be removed or at least list the potential issues. Hopefully I can do this.
Thanks,
Zhao
PATCH v3 13/19 hw/audio/pcspk: Remove PCSpkState::migrate field
In reply to Philippe Mathieu-DaudéThe PCSpkState::migrate boolean was only set in the pc_compat_2_7[] array, via the 'migrate=off' property. We removed all machines using that array, lets remove that property, simplifying vmstate_spk[].
--- hw/audio/pcspk.c | 10 ---------- 1 file changed, 10 deletions(-)
hw/audio/pcspk.c +0 -10
@@ -56,7 +56,6 @@ struct PCSpkState {unsigned int play_pos;uint8_t data_on;uint8_t dummy_refresh_clock; -bool migrate; };static const char *s_spk ="pcspk" ;
@@ -196,18 +195,10 @@ static void pcspk_realizefn(DeviceState *dev, Error **errp) pcspk_state = s; } -static bool migrate_needed (void *opaque) -{ -PCSpkState *s = opaque; - -return s->migrate; -} -static const VMStateDescription vmstate_spk = { .name ="pcspk" , .version_id =1 , .minimum_version_id =1 , - .needed = migrate_needed, .fields = (const VMStateField []) {VMSTATE_UINT8 (data_on, PCSpkState),VMSTATE_UINT8 (dummy_refresh_clock, PCSpkState),
@@ -218,7 +209,6 @@ static const VMStateDescription vmstate_spk = {static const Property pcspk_properties[] = {DEFINE_AUDIO_PROPERTIES (PCSpkState, card),DEFINE_PROP_UINT32 ("iobase" , PCSpkState, iobase,0x61 ), -DEFINE_PROP_BOOL ("migrate" , PCSpkState, migrate, true), };static void pcspk_class_initfn (ObjectClass *klass,const void *data)
Signature · 1 line
2.47.1
Re: [PATCH v3 13/19] hw/audio/pcspk: Remove PCSpkState::migrate field
Quoted 10 lines · Philippe Mathieu-Daudé
The
PCSpkState::migrateboolean was only set in the pc_compat_2_7[] array, via the 'migrate=off' property. We removed all machines using that array, lets remove that property, simplifying vmstate_spk[].Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Mark Cave-Ayland <[email protected]>--- hw/audio/pcspk.c | 10 ---------- 1 file changed, 10 deletions(-)
PATCH v3 14/19 hw/core/machine: Remove hw_compat_2_7[] array
In reply to Philippe Mathieu-DaudéThe hw_compat_2_7[] array was only used by the pc-q35-2.7 and pc-i440fx-2.7 machines, which got removed. Remove it.
--- include/hw/boards.h | 3 --- hw/core/machine.c | 9 --------- 2 files changed, 12 deletions(-)
include/hw/boards.h +0 -3
@@ -838,7 +838,4 @@ extern const size_t hw_compat_2_9_len;extern GlobalProperty hw_compat_2_8[];extern const size_t hw_compat_2_8_len; -extern GlobalProperty hw_compat_2_7[]; -extern const size_t hw_compat_2_7_len; -#endif
hw/core/machine.c +0 -9
@@ -266,15 +266,6 @@ GlobalProperty hw_compat_2_8[] = { };const size_t hw_compat_2_8_len =G_N_ELEMENTS (hw_compat_2_8); -GlobalProperty hw_compat_2_7[] = { - {"virtio-pci" ,"page-per-vq" ,"on" }, - {"virtio-serial-device" ,"emergency-write" ,"off" }, - {"ioapic" ,"version" ,"0x11" }, - {"intel-iommu" ,"x-buggy-eim" ,"true" }, - {"virtio-pci" ,"x-ignore-backend-features" ,"on" }, -}; -const size_t hw_compat_2_7_len =G_N_ELEMENTS (hw_compat_2_7); -MachineState *current_machine;static char *machine_get_kernel (Object *obj,Error **errp)
Signature · 1 line
2.47.1
PATCH v3 15/19 hw/i386/intel_iommu: Remove IntelIOMMUState::buggy_eim field
In reply to Philippe Mathieu-DaudéThe IntelIOMMUState::buggy_eim boolean was only set in the hw_compat_2_7[] array, via the 'x-buggy-eim=true' property. We removed all machines using that array, lets remove that property, simplifying vtd_decide_config().
--- include/hw/i386/intel_iommu.h | 1 - hw/i386/intel_iommu.c | 5 ++--- 2 files changed, 2 insertions(+), 4 deletions(-)
include/hw/i386/intel_iommu.h +0 -1
@@ -303,7 +303,6 @@ struct IntelIOMMUState {uint32_t intr_size;/* Number of IR table entries */ bool intr_eime;/* Extended interrupt mode enabled */ OnOffAuto intr_eim;/* Toggle for EIM cabability */ -bool buggy_eim;/* Force buggy EIM unless eim=off */ uint8_t aw_bits;/* Host/IOVA address width (in bits) */ bool dma_drain;/* Whether DMA r/w draining enabled */ bool dma_translation;/* Whether DMA translation supported */
hw/i386/intel_iommu.c +2 -3
@@ -3823,7 +3823,6 @@ static const Property vtd_properties[] = {DEFINE_PROP_UINT32 ("version" , IntelIOMMUState, version,0 ),DEFINE_PROP_ON_OFF_AUTO ("eim" , IntelIOMMUState, intr_eim,ON_OFF_AUTO_AUTO ), -DEFINE_PROP_BOOL ("x-buggy-eim" , IntelIOMMUState, buggy_eim, false),DEFINE_PROP_UINT8 ("aw-bits" , IntelIOMMUState, aw_bits,VTD_HOST_ADDRESS_WIDTH ), DEFINE_PROP_BOOL("caching-mode", IntelIOMMUState, caching_mode, FALSE),
@@ -4731,11 +4730,11 @@ static bool vtd_decide_config(IntelIOMMUState *s, Error **errp) }if (s->intr_eim ==ON_OFF_AUTO_AUTO ) { - s->intr_eim = (kvm_irqchip_in_kernel () || s->buggy_eim) + s->intr_eim =kvm_irqchip_in_kernel () &&x86_iommu_ir_supported (x86_iommu) ?ON_OFF_AUTO_ON :ON_OFF_AUTO_OFF ; } -if (s->intr_eim ==ON_OFF_AUTO_ON && !s->buggy_eim) { +if (s->intr_eim ==ON_OFF_AUTO_ON ) {if (kvm_irqchip_is_split () && !kvm_enable_x2apic ()) {error_setg (errp,"eim=on requires support on the KVM side" "(X2APIC_API, first shipped in v4.7)" );
Signature · 1 line
2.47.1
PATCH v3 16/19 hw/intc/ioapic: Remove IOAPICCommonState::version field
In reply to Philippe Mathieu-DaudéThe IOAPICCommonState::version integer was only set in the hw_compat_2_7[] array, via the 'version=0x11' property. We removed all machines using that array, lets remove that property, simplify by only using the default version (defined as IOAPIC_VER_DEF).
For the record, this field was introduced in commit 20fd4b7b6d9 ("x86: ioapic: add support for explicit EOI"):
Quoted 14 lines
Some old Linux kernels (upstream before v4.0), or any released RHEL kernels has problem in sending APIC EOI when IR is enabled. Meanwhile, many of them only support explicit EOI for IOAPIC, which is only introduced in IOAPIC version
0x20. This patch provide a way to boost QEMU IOAPIC to version0x20, in order for QEMU to correctly receive EOI messages.Without boosting IOAPIC version to
0x20, kernels before commit d32932d ("x86/irq: Convert IOAPIC to use hierarchical irqdomain interfaces") will have trouble enabling both IR and level-triggered interrupt devices (like e1000).To upgrade IOAPIC to version
0x20, we need to specify:
Quoted 8 lines
-global ioapic.version=0x20To be compatible with old systems,
0x11will still be the default IOAPIC version. Here0x11and0x20are the only versions to be supported.One thing to mention: this patch only applies to emulated IOAPIC. It does not affect kernel IOAPIC behavior.
--- hw/intc/ioapic_internal.h | 3 +-- hw/intc/ioapic.c | 18 ++---------------- hw/intc/ioapic_common.c | 2 +- 3 files changed, 4 insertions(+), 19 deletions(-)
hw/intc/ioapic_internal.h +1 -2
@@ -82,7 +82,7 @@#define IOAPIC_ID_MASK 0xf#define IOAPIC_VER_ENTRIES_SHIFT 16 - +#define IOAPIC_VER_DEF 0x20#define TYPE_IOAPIC_COMMON "ioapic-common"OBJECT_DECLARE_TYPE (IOAPICCommonState, IOAPICCommonClass,IOAPIC_COMMON )
@@ -104,7 +104,6 @@ struct IOAPICCommonState {uint32_t irr;uint64_t ioredtbl[IOAPIC_NUM_PINS ];Notifier machine_done; -uint8_t version;uint64_t irq_count[IOAPIC_NUM_PINS ];int irq_level[IOAPIC_NUM_PINS ];int irq_eoi[IOAPIC_NUM_PINS ];
hw/intc/ioapic.c +2 -16
@@ -315,7 +315,7 @@ ioapic_mem_read(void *opaque, hwaddr addr, unsigned int size) val = s->id <<IOAPIC_ID_SHIFT ;break ;case IOAPIC_REG_VER : - val = s->version | + val =IOAPIC_VER_DEF | ((IOAPIC_NUM_PINS -1 ) <<IOAPIC_VER_ENTRIES_SHIFT );break ;default :
@@ -411,8 +411,7 @@ ioapic_mem_write(void *opaque, hwaddr addr, uint64_t val, }break ;case IOAPIC_EOI : -/* Explicit EOI is only supported for IOAPIC version 0x20 */ -if (size !=4 || s->version !=0x20 ) { +if (size !=4 ) {break ; }ioapic_eoi_broadcast (val);
@@ -444,18 +443,10 @@ static void ioapic_machine_done_notify(Notifier *notifier, void *data)#endif } -#define IOAPIC_VER_DEF 0x20 -static void ioapic_realize (DeviceState *dev,Error **errp) {IOAPICCommonState *s =IOAPIC_COMMON (dev); -if (s->version !=0x11 && s->version !=0x20 ) { -error_setg (errp,"IOAPIC only supports version 0x11 or 0x20 " -"(default: 0x%x)." ,IOAPIC_VER_DEF ); -return ; - } -memory_region_init_io (&s->io_memory,OBJECT (s), &ioapic_io_ops, s,"ioapic" ,0x1000 );
@@ -476,10 +467,6 @@ static void ioapic_unrealize(DeviceState *dev)timer_free (s->delayed_ioapic_service_timer); } -static const Property ioapic_properties[] = { -DEFINE_PROP_UINT8 ("version" , IOAPICCommonState, version,IOAPIC_VER_DEF ), -}; -static void ioapic_class_init (ObjectClass *klass,const void *data) {IOAPICCommonClass *k =IOAPIC_COMMON_CLASS (klass);
@@ -493,7 +480,6 @@ static void ioapic_class_init(ObjectClass *klass, const void *data)*/ k->post_load = ioapic_update_kvm_routes;device_class_set_legacy_reset (dc, ioapic_reset_common); -device_class_set_props (dc, ioapic_properties); }static const TypeInfo ioapic_info = {
hw/intc/ioapic_common.c +1 -1
@@ -83,7 +83,7 @@ static void ioapic_print_redtbl(GString *buf, IOAPICCommonState *s)int i;g_string_append_printf (buf,"ioapic0: ver=0x%x id=0x%02x sel=0x%02x" , - s->version, s->id, s->ioregsel); +IOAPIC_VER_DEF , s->id, s->ioregsel);if (s->ioregsel) {g_string_append_printf (buf," (redir[%u])\n" , (s->ioregsel -IOAPIC_REG_REDTBL_BASE ) >>1 );
Signature · 1 line
2.47.1
PATCH v3 17/19 hw/virtio/virtio-pci: Remove VirtIOPCIProxy::ignore_backend_features field
In reply to Philippe Mathieu-DaudéThe VirtIOPCIProxy::ignore_backend_features boolean was only set in the hw_compat_2_7[] array, via the 'x-ignore-backend-features=on' property. We removed all machines using that array, lets remove that property, simplify by only using the default version.
--- include/hw/virtio/virtio-pci.h | 1 - hw/virtio/virtio-pci.c | 5 +---- 2 files changed, 1 insertion(+), 5 deletions(-)
include/hw/virtio/virtio-pci.h +0 -1
@@ -149,7 +149,6 @@ struct VirtIOPCIProxy {int config_cap;uint32_t flags;bool disable_modern; -bool ignore_backend_features;OnOffAuto disable_legacy;/* Transitional device id */ uint16_t trans_devid;
hw/virtio/virtio-pci.c +1 -4
@@ -1965,8 +1965,7 @@ static void virtio_pci_device_plugged(DeviceState *d, Error **errp)* Virtio capabilities present without * VIRTIO_F_VERSION_1 confuses guests */ -if (!proxy->ignore_backend_features && - !virtio_has_feature (vdev->host_features,VIRTIO_F_VERSION_1 )) { +if (!virtio_has_feature (vdev->host_features,VIRTIO_F_VERSION_1 )) {virtio_pci_disable_modern (proxy);if (!legacy) {
@@ -2351,8 +2350,6 @@ static const Property virtio_pci_properties[] = { VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY_BIT, false),DEFINE_PROP_BIT ("page-per-vq" , VirtIOPCIProxy, flags,VIRTIO_PCI_FLAG_PAGE_PER_VQ_BIT , false), -DEFINE_PROP_BOOL ("x-ignore-backend-features" , VirtIOPCIProxy, - ignore_backend_features, false),DEFINE_PROP_BIT ("ats" , VirtIOPCIProxy, flags,VIRTIO_PCI_FLAG_ATS_BIT , false), DEFINE_PROP_BIT("x-ats-page-aligned", VirtIOPCIProxy, flags,
Signature · 1 line
2.47.1
PATCH v3 18/19 hw/char/virtio-serial: Do not expose the 'emergency-write' property
In reply to Philippe Mathieu-DaudéThe VIRTIO_CONSOLE_F_EMERG_WRITE feature bit was only set in the hw_compat_2_7[] array, via the 'emergency-write=off' property. We removed all machines using that array, lets remove that property. All instances have this feature bit set and it can not be disabled. VirtIOSerial::host_features mask is now unused, remove it.
--- include/hw/virtio/virtio-serial.h | 2 -- hw/char/virtio-serial-bus.c | 9 +++------ 2 files changed, 3 insertions(+), 8 deletions(-)
include/hw/virtio/virtio-serial.h +0 -2
@@ -185,8 +185,6 @@ struct VirtIOSerial {struct VirtIOSerialPostLoad *post_load;virtio_serial_conf serial; - -uint64_t host_features; };/* Interface to the virtio-serial bus */
hw/char/virtio-serial-bus.c +3 -6
@@ -557,7 +557,7 @@ static uint64_t get_features(VirtIODevice *vdev, uint64_t features, vser =VIRTIO_SERIAL (vdev); - features |= vser->host_features; + features |=BIT_ULL (VIRTIO_CONSOLE_F_EMERG_WRITE );if (vser->bus.max_nr_ports >1 ) {virtio_add_feature (&features,VIRTIO_CONSOLE_F_MULTIPORT ); }
@@ -587,8 +587,7 @@ static void set_config(VirtIODevice *vdev, const uint8_t *config_data)VirtIOSerialPortClass *vsc;uint8_t emerg_wr_lo; -if (!virtio_has_feature (vser->host_features, -VIRTIO_CONSOLE_F_EMERG_WRITE ) || !config->emerg_wr) { +if (!config->emerg_wr) {return ; }
@@ -1039,7 +1038,7 @@ static void virtio_serial_device_realize(DeviceState *dev, Error **errp)return ; } -if (!virtio_has_feature (vser->host_features, +if (!virtio_has_feature (vdev->host_features,VIRTIO_CONSOLE_F_EMERG_WRITE )) { config_size = offsetof(struct virtio_console_config , emerg_wr); }
@@ -1155,8 +1154,6 @@ static const VMStateDescription vmstate_virtio_console = {static const Property virtio_serial_properties[] = {DEFINE_PROP_UINT32 ("max_ports" , VirtIOSerial, serial.max_virtserial_ports,31 ), -DEFINE_PROP_BIT64 ("emergency-write" , VirtIOSerial, host_features, -VIRTIO_CONSOLE_F_EMERG_WRITE , true), };static void virtio_serial_class_init (ObjectClass *klass,const void *data)
Signature · 1 line
2.47.1
PATCH v3 19/19 hw/virtio/virtio-pci: Remove VIRTIO_PCI_FLAG_PAGE_PER_VQ definition
In reply to Philippe Mathieu-DaudéVIRTIO_PCI_FLAG_PAGE_PER_VQ was only used by the hw_compat_2_7[] array, via the 'page-per-vq=on' property. We removed all machines using that array, lets remove all the code around VIRTIO_PCI_FLAG_PAGE_PER_VQ (see commit 9a4c0e220d8 for similar VIRTIO_PCI_FLAG_* enum removal).
--- include/hw/virtio/virtio-pci.h | 1 - hw/display/virtio-vga.c | 10 ---------- hw/virtio/virtio-pci.c | 7 +------ 3 files changed, 1 insertion(+), 17 deletions(-)
include/hw/virtio/virtio-pci.h +0 -1
@@ -33,7 +33,6 @@ enum {VIRTIO_PCI_FLAG_BUS_MASTER_BUG_MIGRATION_BIT ,VIRTIO_PCI_FLAG_USE_IOEVENTFD_BIT ,VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY_BIT , -VIRTIO_PCI_FLAG_PAGE_PER_VQ_BIT ,VIRTIO_PCI_FLAG_ATS_BIT ,VIRTIO_PCI_FLAG_INIT_DEVERR_BIT , VIRTIO_PCI_FLAG_INIT_LNKCTL_BIT,
hw/display/virtio-vga.c +0 -10
@@ -141,16 +141,6 @@ static void virtio_vga_base_realize(VirtIOPCIProxy *vpci_dev, Error **errp) VIRTIO_GPU_SHM_ID_HOST_VISIBLE); } -if (!(vpci_dev->flags &VIRTIO_PCI_FLAG_PAGE_PER_VQ )) { -/* -* with page-per-vq=off there is no padding space we can use -* for the stdvga registers. Make the common and isr regions -* smaller then. -*/ - vpci_dev->common.size /=2 ; - vpci_dev->isr.size /=2 ; - } - offset =memory_region_size (&vpci_dev->modern_bar); offset -= vpci_dev->notify.size; vpci_dev->notify.offset = offset;
hw/virtio/virtio-pci.c +1 -6
@@ -314,12 +314,9 @@ static bool virtio_pci_ioeventfd_enabled(DeviceState *d)return (proxy->flags &VIRTIO_PCI_FLAG_USE_IOEVENTFD ) !=0 ; } -#define QEMU_VIRTIO_PCI_QUEUE_MEM_MULT 0x1000 -static inline int virtio_pci_queue_mem_mult (struct VirtIOPCIProxy *proxy) { -return (proxy->flags &VIRTIO_PCI_FLAG_PAGE_PER_VQ ) ? -QEMU_VIRTIO_PCI_QUEUE_MEM_MULT :4 ; +return 4 ; } static int virtio_pci_ioeventfd_assign(DeviceState *d, EventNotifier *notifier,
@@ -2348,8 +2345,6 @@ static const Property virtio_pci_properties[] = { VIRTIO_PCI_FLAG_BUS_MASTER_BUG_MIGRATION_BIT, false),DEFINE_PROP_BIT ("modern-pio-notify" , VirtIOPCIProxy, flags,VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY_BIT , false), -DEFINE_PROP_BIT ("page-per-vq" , VirtIOPCIProxy, flags, -VIRTIO_PCI_FLAG_PAGE_PER_VQ_BIT , false),DEFINE_PROP_BIT ("ats" , VirtIOPCIProxy, flags,VIRTIO_PCI_FLAG_ATS_BIT , false), DEFINE_PROP_BIT("x-ats-page-aligned", VirtIOPCIProxy, flags,
Signature · 1 line
2.47.1