QOM, target/i386 patches for 2026-02-12
The following changes since commit cd5a79dc98e3087e7658e643bdbbb0baec77ac8a:
Merge tag 'for_upstream' of https://git.kernel.org/pub/scm/virt/kvm/mst/qemu into staging (2026-02-05 13:54:50 +0000)
are available in the Git repository at:
https://gitlab.com/bonzini/qemu.git tags/for-upstream
for you to fetch changes up to 28a29efe4342055b4631de96ba71fd577ae83e07:
target/i386/mshv: remove unused optimization of gva=>gpa translation (2026-02-10 14:41:08 +0100)
----------------------------------------------------------------
* hw/i386: Remove deprecated PC 2.6 and 2.7 machines
* i386/cpu: Fix incorrect initializer in Diamond Rapids definition
* qom: Clean up property release
* target/i386/kvm: set KVM_PMU_CAP_DISABLE if "-pmu" is configured
* target/i386/kvm: reset AMD and perfmon-v2 PMU registers during VM reset
* mshv: Cleanup
* target/i386: convert SEV-ES termination requests to guest panic events
----------------------------------------------------------------
Aidan Khoury (1):
i386/cpu: Fix incorrect initializer in Diamond Rapids definition
Chandan Somani (1):
qdev: Free property array on release
Dongli Zhang (5):
target/i386/kvm: set KVM_PMU_CAP_DISABLE if "-pmu" is configured target/i386/kvm: extract unrelated code out of kvm_x86_build_cpuid() target/i386/kvm: rename architectural PMU variables target/i386/kvm: reset AMD PMU registers during VM reset target/i386/kvm: support perfmon-v2 for reset
Igor Mammedov (1):
tests/acpi: Allow DSDT table change for x86 machines
Magnus Kulke (2):
accel/mshv: Remove remap overlapping mappings code target/i386/mshv: remove unused optimization of gva=>gpa translation
Marc-André Lureau (4):
qdev: make release_string() idempotent
qdev: make release_drive() idempotent
qdev: make release_tpm() idempotent
tests: add /qdev/free-properties test
Paolo Bonzini (1):
target/i386: convert SEV-ES termination requests to guest panic events
Philippe Mathieu-Daudé (22):
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/nvram/fw_cfg: Rename fw_cfg_init_mem() with '_nodma' suffix hw/mips/loongson3_virt: Prefer using fw_cfg_init_mem_nodma() hw/nvram/fw_cfg: Factor fw_cfg_init_mem_internal() out hw/nvram/fw_cfg: Rename fw_cfg_init_mem_wide() -> fw_cfg_init_mem_dma() hw/i386/x86: Remove X86MachineClass::fwcfg_dma_enabled field hw/i386/pc: Remove multiboot.bin
hw/i386: Assume fw_cfg DMA is always enabled
hw/i386: Remove linuxboot.bin
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
target/i386/cpu: Remove CPUX86State::full_cpuid_auto_level field 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/virtio/virtio-pci: Remove VirtIOPCIProxy::ignore_backend_features field hw/char/virtio-serial: Do not expose the 'emergency-write' property
Zhao Liu (4):
pc: Start with modern CPU hotplug interface by default
acpi: Remove legacy cpu hotplug utilities docs/specs/acpi_cpu_hotplug: Remove legacy cpu hotplug descriptions tests/acpi: Update DSDT tables for pc & q35 machines
docs/specs/acpi_cpu_hotplug.rst | 28 +- qapi/run-state.json | 23 +- include/hw/acpi/cpu.h | 1 - include/hw/acpi/ich9.h | 4 +- include/hw/acpi/piix4.h | 4 +- include/hw/core/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/nvram/fw_cfg.h | 9 +- include/hw/virtio/virtio-mmio.h | 1 - include/hw/virtio/virtio-pci.h | 1 - include/hw/virtio/virtio-serial.h | 2 - include/system/mshv_int.h | 22 +- pc-bios/optionrom/optionrom.h | 4 - target/i386/cpu.h | 19 +- accel/kvm/kvm-all.c | 1 + accel/mshv/mem.c | 406 ++-------------------- accel/mshv/mshv-all.c | 2 - backends/tpm/tpm_util.c | 1 + block/accounting.c | 1 - hw/acpi/acpi-cpu-hotplug-stub.c | 19 +- hw/acpi/cpu.c | 10 - hw/acpi/cpu_hotplug.c | 348 ------------------- hw/acpi/generic_event_device.c | 1 + hw/acpi/ich9.c | 61 +--- hw/acpi/piix4.c | 61 +--- hw/arm/virt.c | 2 +- hw/audio/pcspk.c | 10 - hw/char/virtio-serial-bus.c | 9 +- hw/core/machine.c | 17 - hw/core/qdev-properties-system.c | 1 + hw/core/qdev-properties.c | 24 +- hw/hppa/machine.c | 2 +- hw/i386/acpi-build.c | 7 +- hw/i386/fw_cfg.c | 16 +- hw/i386/intel_iommu.c | 5 +- hw/i386/microvm.c | 3 - hw/i386/multiboot.c | 7 +- hw/i386/pc.c | 25 +- hw/i386/pc_piix.c | 23 -- hw/i386/pc_q35.c | 24 -- hw/i386/x86-common.c | 8 +- hw/i386/x86.c | 2 - hw/input/stellaris_gamepad.c | 8 - hw/intc/apic_common.c | 5 - hw/intc/arm_gicv3_common.c | 8 - hw/intc/rx_icu.c | 8 - hw/loongarch/fw_cfg.c | 4 +- hw/loongarch/virt-acpi-build.c | 1 - hw/mips/loongson3_virt.c | 2 +- hw/misc/arm_sysctl.c | 2 - hw/misc/mps2-scc.c | 8 - hw/net/rocker/rocker.c | 1 - hw/nvram/fw_cfg.c | 22 +- hw/nvram/xlnx-efuse.c | 8 - hw/nvram/xlnx-versal-efuse-ctrl.c | 8 - hw/riscv/virt.c | 4 +- hw/virtio/virtio-iommu-pci.c | 8 - hw/virtio/virtio-mmio.c | 15 - hw/virtio/virtio-pci.c | 5 +- system/runstate.c | 4 + target/i386/cpu-system.c | 22 ++ target/i386/cpu.c | 119 ++++--- target/i386/kvm/kvm.c | 314 ++++++++++++++--- target/i386/mshv/mshv-cpu.c | 63 +--- tests/qtest/test-x86-cpuid-compat.c | 11 - tests/unit/test-qdev.c | 96 +++++ accel/mshv/trace-events | 5 - hw/acpi/meson.build | 2 +- pc-bios/meson.build | 2 - pc-bios/multiboot.bin | Bin 1024 -> 0 bytes pc-bios/optionrom/Makefile | 2 +- pc-bios/optionrom/linuxboot.S | 195 ----------- pc-bios/optionrom/multiboot.S | 232 ------------- pc-bios/optionrom/multiboot_dma.S | 234 ++++++++++++- roms/seabios-hppa | 2 +- tests/data/acpi/x86/pc/DSDT | Bin 8611 -> 8598 bytes tests/data/acpi/x86/pc/DSDT.acpierst | Bin 8522 -> 8509 bytes tests/data/acpi/x86/pc/DSDT.acpihmat | Bin 9936 -> 9923 bytes tests/data/acpi/x86/pc/DSDT.bridge | Bin 15482 -> 15469 bytes tests/data/acpi/x86/pc/DSDT.cphp | Bin 9075 -> 9062 bytes tests/data/acpi/x86/pc/DSDT.dimmpxm | Bin 10265 -> 10252 bytes tests/data/acpi/x86/pc/DSDT.hpbridge | Bin 8562 -> 8549 bytes tests/data/acpi/x86/pc/DSDT.hpbrroot | Bin 5100 -> 5087 bytes tests/data/acpi/x86/pc/DSDT.ipmikcs | Bin 8683 -> 8670 bytes tests/data/acpi/x86/pc/DSDT.memhp | Bin 9970 -> 9957 bytes tests/data/acpi/x86/pc/DSDT.nohpet | Bin 8469 -> 8456 bytes tests/data/acpi/x86/pc/DSDT.numamem | Bin 8617 -> 8604 bytes tests/data/acpi/x86/pc/DSDT.roothp | Bin 12404 -> 12391 bytes tests/data/acpi/x86/q35/DSDT | Bin 8440 -> 8427 bytes tests/data/acpi/x86/q35/DSDT.acpierst | Bin 8457 -> 8444 bytes tests/data/acpi/x86/q35/DSDT.acpihmat | Bin 9765 -> 9752 bytes tests/data/acpi/x86/q35/DSDT.acpihmat-generic-x | Bin 12650 -> 12637 bytes tests/data/acpi/x86/q35/DSDT.acpihmat-noinitiator | Bin 8719 -> 8706 bytes tests/data/acpi/x86/q35/DSDT.applesmc | Bin 8486 -> 8473 bytes tests/data/acpi/x86/q35/DSDT.bridge | Bin 12053 -> 12040 bytes tests/data/acpi/x86/q35/DSDT.core-count | Bin 12998 -> 12985 bytes tests/data/acpi/x86/q35/DSDT.core-count2 | Bin 33855 -> 33842 bytes tests/data/acpi/x86/q35/DSDT.cphp | Bin 8904 -> 8891 bytes tests/data/acpi/x86/q35/DSDT.cxl | Bin 13231 -> 13218 bytes tests/data/acpi/x86/q35/DSDT.dimmpxm | Bin 10094 -> 10081 bytes tests/data/acpi/x86/q35/DSDT.ipmibt | Bin 8515 -> 8502 bytes tests/data/acpi/x86/q35/DSDT.ipmismbus | Bin 8528 -> 8515 bytes tests/data/acpi/x86/q35/DSDT.ivrs | Bin 8457 -> 8444 bytes tests/data/acpi/x86/q35/DSDT.memhp | Bin 9799 -> 9786 bytes tests/data/acpi/x86/q35/DSDT.mmio64 | Bin 9570 -> 9557 bytes tests/data/acpi/x86/q35/DSDT.multi-bridge | Bin 13293 -> 13280 bytes tests/data/acpi/x86/q35/DSDT.noacpihp | Bin 8302 -> 8289 bytes tests/data/acpi/x86/q35/DSDT.nohpet | Bin 8298 -> 8285 bytes tests/data/acpi/x86/q35/DSDT.numamem | Bin 8446 -> 8433 bytes tests/data/acpi/x86/q35/DSDT.pvpanic-isa | Bin 8541 -> 8528 bytes tests/data/acpi/x86/q35/DSDT.thread-count | Bin 12998 -> 12985 bytes tests/data/acpi/x86/q35/DSDT.thread-count2 | Bin 33855 -> 33842 bytes tests/data/acpi/x86/q35/DSDT.tis.tpm12 | Bin 9046 -> 9033 bytes tests/data/acpi/x86/q35/DSDT.tis.tpm2 | Bin 9072 -> 9059 bytes tests/data/acpi/x86/q35/DSDT.type4-count | Bin 18674 -> 18661 bytes tests/data/acpi/x86/q35/DSDT.viot | Bin 14697 -> 14684 bytes tests/data/acpi/x86/q35/DSDT.xapic | Bin 35803 -> 35790 bytes tests/unit/meson.build | 1 + 121 files changed, 852 insertions(+), 1802 deletions(-) delete mode 100644 hw/acpi/cpu_hotplug.c create mode 100644 tests/unit/test-qdev.c delete mode 100644 pc-bios/multiboot.bin delete mode 100644 pc-bios/optionrom/linuxboot.S delete mode 100644 pc-bios/optionrom/multiboot.S
Signature · 1 line
2.52.0
[PULL 23/41] target/i386/cpu: Remove CPUX86State::full_cpuid_auto_level field
From: Philippe Mathieu-Daudé <[email protected]>
The CPUX86State::full_cpuid_auto_level boolean was only disabled for the pc-q35-2.7 and pc-i440fx-2.7 machines, which got removed. Being now always %true, we can remove it and simplify x86_cpu_expand_features().
--- target/i386/cpu.h | 3 -- target/i386/cpu.c | 117 ++++++++++++++++++++++------------------------ 2 files changed, 57 insertions(+), 63 deletions(-)
target/i386/cpu.h +0 -3
@@ -2425,9 +2425,6 @@ struct ArchCPU {/* Force to enable cpuid 0x1f */ bool force_cpuid_0x1f; -/* Enable auto level-increase for all CPUID leaves */ -bool full_cpuid_auto_level; -/* * Compatibility bits for old machine types (PC machine v6.0 and older). * Only advertise CPUID leaves defined by the vendor.
target/i386/cpu.c +57 -60
138 lines not on this page · raw message
Signature · 1 line
2.52.0
[PULL 30/41] target/i386/kvm: set KVM_PMU_CAP_DISABLE if "-pmu" is configured
In reply to Paolo BonziniFrom: Dongli Zhang <[email protected]>
Although AMD PERFCORE and PerfMonV2 are removed when "-pmu" is configured, there is no way to fully disable KVM AMD PMU virtualization. Neither "-cpu host,-pmu" nor "-cpu EPYC" achieves this.
As a result, the following message still appears in the VM dmesg:
[ 0.263615] Performance Events: AMD PMU driver.
However, the expected output should be:
[ 0.596381] Performance Events: PMU not available due to virtualization, using software events only. [ 0.600972] NMI watchdog: Perf NMI watchdog permanently disabled
This occurs because AMD does not use any CPUID bit to indicate PMU availability.
To address this, KVM_CAP_PMU_CAPABILITY is used to set KVM_PMU_CAP_DISABLE when "-pmu" is configured.
--- target/i386/kvm/kvm.c | 31 +++++++++++++++++++++++++++++++ 1 file changed, 31 insertions(+)
target/i386/kvm/kvm.c +31 -0
@@ -178,6 +178,8 @@ static int has_triple_fault_event;static bool has_msr_mcg_ext_ctl; +static int pmu_cap; +static struct kvm_cpuid2 *cpuid_cache;static struct kvm_cpuid2 *hv_cpuid_cache;static struct kvm_msr_list *kvm_feature_msrs;
@@ -2079,6 +2081,33 @@ full:int kvm_arch_pre_create_vcpu (CPUState *cpu,Error **errp) { +static bool first = true; +int ret; + +if (first) { + first = false; + +/* +* Since Linux v5.18, KVM provides a VM-level capability to easily +* disable PMUs; however, QEMU has been providing PMU property per +* CPU since v1.6. In order to accommodate both, have to configure +* the VM-level capability here. +* +* KVM_PMU_CAP_DISABLE doesn't change the PMU +* behavior on Intel platform because current "pmu" property works +* as expected. +*/ +if ((pmu_cap &KVM_PMU_CAP_DISABLE ) && !X86_CPU (cpu)->enable_pmu) { + ret =kvm_vm_enable_cap (kvm_state,KVM_CAP_PMU_CAPABILITY ,0 , +KVM_PMU_CAP_DISABLE ); +if (ret <0 ) { +error_setg_errno (errp, -ret, +"Failed to set KVM_PMU_CAP_DISABLE" ); +return ret; + } + } + } +if (is_tdx_vm ()) {return tdx_pre_create_vcpu (cpu, errp); }
@@ -3390,6 +3419,8 @@ int kvm_arch_init(MachineState *ms, KVMState *s) } } + pmu_cap =kvm_check_extension (s,KVM_CAP_PMU_CAPABILITY ); +return 0 ; }
Signature · 1 line
2.52.0
[PULL 10/41] hw/mips/loongson3_virt: Prefer using fw_cfg_init_mem_nodma()
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
fw_cfg_init_mem_wide() is prefered to initialize fw_cfg with DMA support. Without DMA, use fw_cfg_init_mem_nodma().
--- hw/mips/loongson3_virt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
hw/mips/loongson3_virt.c +1 -1
@@ -287,7 +287,7 @@ 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_nodma (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
2.52.0
[PULL 28/41] hw/char/virtio-serial: Do not expose the 'emergency-write' property
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
The 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
@@ -186,8 +186,6 @@ 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 ; }
@@ -1040,7 +1039,7 @@ 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); }
@@ -1156,8 +1155,6 @@ 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
2.52.0
Re: [PULL 28/41] hw/char/virtio-serial: Do not expose the 'emergency-write' property
Quoted 9 lines · Paolo Bonzini
From: Philippe Mathieu-Daudé <[email protected]>
The
VIRTIO_CONSOLE_F_EMERG_WRITEfeature 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_featuresmask is now unused, remove it.
Quoted 7 lines
Reviewed-by Mark Cave-Ayland <[email protected]>Reviewed-by Zhao Liu <[email protected]>Reviewed-by Igor Mammedov <[email protected]>Signed-off-by Philippe Mathieu-Daudé <[email protected]>Signed-off-by Zhao Liu <[email protected]>Signed-off-by Paolo Bonzini <[email protected]>
Hello,
This change appears to have introduced a Heap-buffer-overflow:
https://gitlab.com/qemu-project/qemu/-/issues/3303
-Alex
Quoted 63 lines
--- 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@@ -186,8 +186,6 @@ 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
-7@@ -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 ; }@@ -1040,7 +1039,7 @@ 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); }@@ -1156,8 +1155,6 @@ 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) --2.52.0
[PULL 16/41] hw/i386: Remove linuxboot.bin
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
All machines now use the linuxboot_dma.bin binary, so it's safe to remove the non-DMA version (linuxboot.bin).
--- hw/i386/pc.c | 3 +- pc-bios/meson.build | 1 - pc-bios/optionrom/Makefile | 2 +- pc-bios/optionrom/linuxboot.S | 195 ---------------------------------- 4 files changed, 2 insertions(+), 199 deletions(-) delete mode 100644 pc-bios/optionrom/linuxboot.S
hw/i386/pc.c +1 -2
@@ -669,8 +669,7 @@ void xen_load_linux(PCMachineState *pcms)x86_load_linux (x86ms, fw_cfg,PC_FW_DATA , pcmc->pvh_enabled);for (i =0 ; i < nb_option_roms; i++) { -assert (!strcmp (option_rom[i].name,"linuxboot.bin" ) || - !strcmp (option_rom[i].name,"linuxboot_dma.bin" ) || +assert (!strcmp (option_rom[i].name,"linuxboot_dma.bin" ) || !strcmp (option_rom[i].name,"pvh.bin" ) || !strcmp (option_rom[i].name,"multiboot_dma.bin" ));rom_add_option (option_rom[i].name, option_rom[i].bootindex);
pc-bios/meson.build +0 -1
@@ -63,7 +63,6 @@ blobs = [ 'efi-vmxnet3.rom', 'qemu-nsis.bmp', 'multiboot_dma.bin', - 'linuxboot.bin', 'linuxboot_dma.bin', 'kvmvapic.bin', 'pvh.bin',
pc-bios/optionrom/Makefile +1 -1
@@ -2,7 +2,7 @@ include config.mak SRC_DIR := $(TOPSRC_DIR)/pc-bios/optionrom VPATH = $(SRC_DIR) -all: multiboot_dma.bin linuxboot.bin linuxboot_dma.bin kvmvapic.bin pvh.bin +all: multiboot_dma.bin linuxboot_dma.bin kvmvapic.bin pvh.bin # Dummy command so that make thinks it has done something @true
pc-bios/optionrom/linuxboot.S deleted +0 -195
195 lines not on this page · raw message
Signature · 1 line
2.52.0
Re: [PULL 16/41] hw/i386: Remove linuxboot.bin
Quoted 12 lines · Paolo Bonzini
From: Philippe Mathieu-Daudé <[email protected]>
All machines now use the linuxboot_dma.bin binary, so it's safe to remove the non-DMA version (linuxboot.bin).
Suggested-by Thomas Huth <[email protected]>Reviewed-by Zhao Liu <[email protected]>Signed-off-by Philippe Mathieu-Daudé <[email protected]>Signed-off-by Zhao Liu <[email protected]>Reviewed-by Thomas Huth <[email protected]>Signed-off-by Paolo Bonzini <[email protected]>
Quoted 7 lines
--- hw/i386/pc.c | 3 +- pc-bios/meson.build | 1 - pc-bios/optionrom/Makefile | 2 +- pc-bios/optionrom/linuxboot.S | 195 ---------------------------------- 4 files changed, 2 insertions(+), 199 deletions(-) delete mode 100644 pc-bios/optionrom/linuxboot.S
The pre-built pc-bios/linuxboot_dma.bin still exists in the tree, shouldn't that have been deleted too ?
Quoted 245 lines
hw/i386/pc.c +1
-2@@ -669,8 +669,7 @@ void xen_load_linux(PCMachineState *pcms)x86_load_linux (x86ms, fw_cfg,PC_FW_DATA , pcmc->pvh_enabled);for (i =0 ; i < nb_option_roms; i++) { -assert (!strcmp (option_rom[i].name,"linuxboot.bin" ) || - !strcmp (option_rom[i].name,"linuxboot_dma.bin" ) || +assert (!strcmp (option_rom[i].name,"linuxboot_dma.bin" ) || !strcmp (option_rom[i].name,"pvh.bin" ) || !strcmp (option_rom[i].name,"multiboot_dma.bin" ));rom_add_option (option_rom[i].name, option_rom[i].bootindex);pc-bios/meson.build +0
-1@@ -63,7 +63,6 @@ blobs = [ 'efi-vmxnet3.rom', 'qemu-nsis.bmp', 'multiboot_dma.bin', - 'linuxboot.bin', 'linuxboot_dma.bin', 'kvmvapic.bin', 'pvh.bin',pc-bios/optionrom/Makefile +1
-1@@ -2,7 +2,7 @@ include config.mak SRC_DIR := $(TOPSRC_DIR)/pc-bios/optionrom VPATH = $(SRC_DIR) -all: multiboot_dma.bin linuxboot.bin linuxboot_dma.bin kvmvapic.bin pvh.bin +all: multiboot_dma.bin linuxboot_dma.bin kvmvapic.bin pvh.bin # Dummy command so that make thinks it has done something @truepc-bios/optionrom/linuxboot.S deleted +0
-195@@ -1,195 +0,0 @@ -/* - * Linux Boot Option ROM - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, see <http://www.gnu.org/licenses/>. - * - * Copyright Novell Inc, 2009 - * Authors: Alexander Graf <[email protected]> - * - * Based on code in hw/pc.c. - */ - -#include "optionrom.h" - -#define BOOT_ROM_PRODUCT "Linux loader" - -BOOT_ROM_START - -run_linuxboot: - - cli - cld - - jmp copy_kernel -boot_kernel: - - read_fw FW_CFG_SETUP_ADDR - - mov %eax, %ebx - shr $4, %ebx - - /* All segments contain real_addr */ - mov %bx, %ds - mov %bx, %es - mov %bx, %fs - mov %bx, %gs - mov %bx, %ss - - /* CX = CS we want to jump to */ - add $0x20, %bx - mov %bx, %cx - - /* SP = cmdline_addr-real_addr-16 */ - read_fw FW_CFG_CMDLINE_ADDR - mov %eax, %ebx - read_fw FW_CFG_SETUP_ADDR - sub %eax, %ebx - sub $16, %ebx - mov %ebx, %esp - - /* Build indirect lret descriptor */ - pushw %cx /* CS */ - xor %ax, %ax - pushw %ax /* IP = 0 */ - - /* Clear registers */ - xor %eax, %eax - xor %ebx, %ebx - xor %ecx, %ecx - xor %edx, %edx - xor %edi, %edi - xor %ebp, %ebp - - /* Jump to Linux */ - lret - - -copy_kernel: - /* Read info block in low memory (0x10000 or 0x90000) */ - read_fw FW_CFG_SETUP_ADDR - shr $4, %eax - mov %eax, %es - xor %edi, %edi - read_fw_blob_addr32_edi(FW_CFG_SETUP) - - cmpw $0x203, %es:0x206 // if protocol >= 0x203 - jae 1f // have initrd_max - movl $0x37ffffff, %es:0x22c // else assume 0x37ffffff -1: - - /* Check if using kernel-specified initrd address */ - read_fw FW_CFG_INITRD_ADDR - mov %eax, %edi // (load_kernel wants it in %edi) - read_fw FW_CFG_INITRD_SIZE // find end of initrd - add %edi, %eax - xor %es:0x22c, %eax // if it matches es:0x22c - and $-4096, %eax // (apart from padding for page) - jz load_kernel // then initrd is not at top - // of memory - - /* pc.c placed the initrd at end of memory. Compute a better - * initrd address based on e801 data. - */ - mov $0xe801, %ax - xor %cx, %cx - xor %dx, %dx - int $0x15 - - /* Output could be in AX/BX or CX/DX */ - or %cx, %cx - jnz 1f - or %dx, %dx - jnz 1f - mov %ax, %cx - mov %bx, %dx -1: - - or %dx, %dx - jnz 2f - addw $1024, %cx /* add 1 MB */ - movzwl %cx, %edi - shll $10, %edi /* convert to bytes */ - jmp 3f - -2: - addw $16777216 >> 16, %dx /* add 16 MB */ - movzwl %dx, %edi - shll $16, %edi /* convert to bytes */ - -3: - read_fw FW_CFG_INITRD_SIZE - subl %eax, %edi - andl $-4096, %edi /* EDI = start of initrd */ - movl %edi, %es:0x218 /* put it in the header */ - -load_kernel: - /* We need to load the kernel into memory we can't access in 16 bit - mode, so let's get into 32 bit mode, write the kernel and jump - back again. */ - - /* Reserve space on the stack for our GDT descriptor. */ - mov %esp, %ebp - sub $16, %esp - - /* Now create the GDT descriptor */ - movw $((3 * 8) - 1), -16(%bp) - mov %cs, %eax - movzwl %ax, %eax - shl $4, %eax - addl $gdt, %eax - movl %eax, -14(%bp) - - /* And load the GDT */ - data32 lgdt -16(%bp) - mov %ebp, %esp - - /* Get us to protected mode now */ - mov $1, %eax - mov %eax, %cr0 - - /* So we can set ES to a 32-bit segment */ - mov $0x10, %eax - mov %eax, %es - - /* We're now running in 16-bit CS, but 32-bit ES! */ - - /* Load kernel and initrd */ - read_fw_blob_addr32_edi(FW_CFG_INITRD) - read_fw_blob_addr32(FW_CFG_KERNEL) - read_fw_blob_addr32(FW_CFG_CMDLINE) - - /* And now jump into Linux! */ - mov $0, %eax - mov %eax, %cr0 - - /* ES = CS */ - mov %cs, %ax - mov %ax, %es - - jmp boot_kernel - -/* Variables */ - -.align 4, 0 -gdt: - /* 0x00 */ -.byte 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 - - /* 0x08: code segment (base=0, limit=0xfffff, type=32bit code exec/read, DPL=0, 4k) */ -.byte 0xff, 0xff, 0x00, 0x00, 0x00, 0x9a, 0xcf, 0x00 - - /* 0x10: data segment (base=0, limit=0xfffff, type=32bit data read/write, DPL=0, 4k) */ -.byte 0xff, 0xff, 0x00, 0x00, 0x00, 0x92, 0xcf, 0x00 - -BOOT_ROM_ENDSignature · 1 line
2.52.0
With regards,
Daniel
Signature · 3 lines
|: https://berrange.com ~~ https://hachyderm.io/@berrange :| |: https://libvirt.org ~~ https://entangle-photo.org :| |: https://pixelfed.art/berrange ~~ https://fstop138.berrange.com :|
Re: [PULL 16/41] hw/i386: Remove linuxboot.bin
Quoted 22 lines · Daniel P. Berrangé
Quoted 12 lines · Paolo Bonzini
From: Philippe Mathieu-Daudé <[email protected]>
All machines now use the linuxboot_dma.bin binary, so it's safe to remove the non-DMA version (linuxboot.bin).
Suggested-by Thomas Huth <[email protected]>Reviewed-by Zhao Liu <[email protected]>Signed-off-by Philippe Mathieu-Daudé <[email protected]>Signed-off-by Zhao Liu <[email protected]>Reviewed-by Thomas Huth <[email protected]>Signed-off-by Paolo Bonzini <[email protected]>Quoted 7 lines
--- hw/i386/pc.c | 3 +- pc-bios/meson.build | 1 - pc-bios/optionrom/Makefile | 2 +- pc-bios/optionrom/linuxboot.S | 195 ---------------------------------- 4 files changed, 2 insertions(+), 199 deletions(-) delete mode 100644 pc-bios/optionrom/linuxboot.S
Quoted 2 lines
The pre-built pc-bios/linuxboot_dma.bin still exists in the tree, shouldn't that have been deleted too ?
I think you mean pc-bios/linuxboot.bin - yeah, that one should go as well.
Paolo
[PULL 02/41] hw/i386/pc: Remove deprecated pc-q35-2.6 and pc-i440fx-2.6 machines
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
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
@@ -732,20 +732,6 @@ 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_XEN static void xenfv_machine_4_2_options (MachineClass *m) {
hw/i386/pc_q35.c +0 -14
@@ -680,17 +680,3 @@ 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
2.52.0
[PULL 14/41] hw/i386/pc: Remove multiboot.bin
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
All PC machines now use the multiboot_dma.bin binary, we can remove the non-DMA version (multiboot.bin).
This doesn't change multiboot_dma binary file.
--- pc-bios/optionrom/optionrom.h | 4 - hw/i386/pc.c | 1 - pc-bios/meson.build | 1 - pc-bios/multiboot.bin | Bin 1024 -> 0 bytes pc-bios/optionrom/Makefile | 2 +- pc-bios/optionrom/multiboot.S | 232 ----------------------------- pc-bios/optionrom/multiboot_dma.S | 234 +++++++++++++++++++++++++++++- 7 files changed, 233 insertions(+), 241 deletions(-) delete mode 100644 pc-bios/multiboot.bin delete mode 100644 pc-bios/optionrom/multiboot.S
pc-bios/optionrom/optionrom.h +0 -4
@@ -117,16 +117,12 @@* * Clobbers: %eax, %edx, %es, %ecx, %edi and adresses %esp-20 to %esp */ -#ifdef USE_FW_CFG_DMA #define read_fw_blob_dma (var) \ read_fw var ## _SIZE; \ mov %eax, %ecx; \ read_fw var ## _ADDR; \ mov %eax, %edi ; \ read_fw_dma var ## _DATA, %ecx, %edi -#else -#define read_fw_blob_dma (var) read_fw_blob(var) -#endif #define read_fw_blob_pre (var) \ read_fw var ## _SIZE; \
hw/i386/pc.c +0 -1
@@ -672,7 +672,6 @@ void xen_load_linux(PCMachineState *pcms)assert (!strcmp (option_rom[i].name,"linuxboot.bin" ) || !strcmp (option_rom[i].name,"linuxboot_dma.bin" ) || !strcmp (option_rom[i].name,"pvh.bin" ) || - !strcmp (option_rom[i].name,"multiboot.bin" ) || !strcmp (option_rom[i].name,"multiboot_dma.bin" ));rom_add_option (option_rom[i].name, option_rom[i].bootindex); }
pc-bios/meson.build +0 -1
@@ -62,7 +62,6 @@ blobs = [ 'efi-e1000e.rom', 'efi-vmxnet3.rom', 'qemu-nsis.bmp', - 'multiboot.bin', 'multiboot_dma.bin', 'linuxboot.bin', 'linuxboot_dma.bin',
pc-bios/multiboot.bin deleted binary +0 -0
pc-bios/optionrom/Makefile +1 -1
@@ -2,7 +2,7 @@ include config.mak SRC_DIR := $(TOPSRC_DIR)/pc-bios/optionrom VPATH = $(SRC_DIR) -all: multiboot.bin multiboot_dma.bin linuxboot.bin linuxboot_dma.bin kvmvapic.bin pvh.bin +all: multiboot_dma.bin linuxboot.bin linuxboot_dma.bin kvmvapic.bin pvh.bin # Dummy command so that make thinks it has done something @true
pc-bios/optionrom/multiboot.S deleted +0 -232
232 lines not on this page · raw message
pc-bios/optionrom/multiboot_dma.S +232 -2
234 lines not on this page · raw message
Signature · 1 line
2.52.0
[PULL 04/41] pc: Start with modern CPU hotplug interface by default
In reply to Paolo BonziniFrom: Zhao Liu <[email protected]>
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 (in commit 679dd1a957df ("pc: use new CPU hotplug interface since 2.7 machine type")). In that way, legacy firmware that doesn't use QEMU generated ACPI tables was able to continue using legacy CPU hotplug interface.
While later machine types, with firmware supporting QEMU provided ACPI tables, generate new CPU hotplug AML, which will switch to new CPU hotplug interface when guest OS executes its _INI method on ACPI tables loading.
Since 2.6 machine type is now gone, and consider that the legacy BIOS (based on QEMU ACPI prior to v2.7) should be no longer in use, previous compatibility requirements are no longer necessary. So initialize 'modern' hotplug directly from the very beginning for PC/Q35 machines with cpu_hotplug_hw_init(), and drop _INIT method.
Additionally, remove the checks and settings around cpu_hotplug_legacy in cpuhp VMState (for piix4 & ich9), to eliminate the risk of segmentation faults, as gpe_cpu no longer has the opportunity to be initialized. This is safe because all hotplug now start with the modern way, and it's impossible to switch to legacy way at runtime (even the "cpu-hotplug-legacy" properties does not allow it either).
--- include/hw/acpi/cpu.h | 1 - hw/acpi/cpu.c | 10 ---------- hw/acpi/ich9.c | 22 +++------------------- hw/acpi/piix4.c | 21 +++------------------ hw/i386/acpi-build.c | 2 +- hw/loongarch/virt-acpi-build.c | 1 - 6 files changed, 7 insertions(+), 50 deletions(-)
include/hw/acpi/cpu.h +0 -1
@@ -54,7 +54,6 @@ void cpu_hotplug_hw_init(MemoryRegion *as, Object *owner,typedef struct CPUHotplugFeatures {bool acpi_1_compatible; -bool has_legacy_cphp;bool fw_unplugs_cpu;const char *smi_path; }CPUHotplugFeatures ;
hw/acpi/cpu.c +0 -10
@@ -408,16 +408,6 @@ void build_cpus_aml(Aml *table, MachineState *machine, CPUHotplugFeatures opts,aml_append (field,aml_reserved_field (4 *8 ));aml_append (field,aml_named_field (CPU_DATA ,32 ));aml_append (cpu_ctrl_dev, field); - -if (opts.has_legacy_cphp) { - method =aml_method ("_INI" ,0 ,AML_SERIALIZED ); -/* switch off legacy CPU hotplug HW and use new one, -* on reboot system is in new mode and writing 0 -* in CPU_SELECTOR selects BSP, which is NOP at -* the time _INI is called */ -aml_append (method,aml_store (zero,aml_name (CPU_SELECTOR ))); -aml_append (cpu_ctrl_dev, method); - } }aml_append (sb_scope, cpu_ctrl_dev);
hw/acpi/ich9.c +3 -19
@@ -183,26 +183,10 @@ static const VMStateDescription vmstate_tco_io_state = { } }; -static bool vmstate_test_use_cpuhp (void *opaque) -{ -ICH9LPCPMRegs *s = opaque; -return !s->cpu_hotplug_legacy; -} - -static int vmstate_cpuhp_pre_load (void *opaque) -{ -ICH9LPCPMRegs *s = opaque; -Object *obj =OBJECT (s->gpe_cpu.device); -object_property_set_bool (obj,"cpu-hotplug-legacy" , false, &error_abort); -return 0 ; -} -static const VMStateDescription vmstate_cpuhp_state = { .name ="ich9_pm/cpuhp" , .version_id =1 , .minimum_version_id =1 , - .needed = vmstate_test_use_cpuhp, - .pre_load = vmstate_cpuhp_pre_load, .fields = (const VMStateField []) {VMSTATE_CPU_HOTPLUG (cpuhp_state, ICH9LPCPMRegs),VMSTATE_END_OF_LIST ()
@@ -338,8 +322,8 @@ void ich9_pm_init(PCIDevice *lpc_pci, ICH9LPCPMRegs *pm, qemu_irq sci_irq) pm->powerdown_notifier.notify = pm_powerdown_req;qemu_register_powerdown_notifier (&pm->powerdown_notifier); -legacy_acpi_cpu_hotplug_init (pci_address_space_io (lpc_pci), -OBJECT (lpc_pci), &pm->gpe_cpu,ICH9_CPU_HOTPLUG_IO_BASE ); +cpu_hotplug_hw_init (pci_address_space_io (lpc_pci), +OBJECT (lpc_pci), &pm->cpuhp_state,ICH9_CPU_HOTPLUG_IO_BASE ); acpi_memory_hotplug_init(pci_address_space_io(lpc_pci), OBJECT(lpc_pci), &pm->acpi_memory_hotplug,
@@ -419,7 +403,7 @@ void ich9_pm_add_properties(Object *obj, ICH9LPCPMRegs *pm) {static const uint32_t gpe0_len =ICH9_PMIO_GPE0_LEN ; pm->acpi_memory_hotplug.is_enabled = true; - pm->cpu_hotplug_legacy = true; + pm->cpu_hotplug_legacy = false; pm->disable_s3 =0 ; pm->disable_s4 =0 ; pm->s4_val =2 ;
hw/acpi/piix4.c +3 -18
@@ -195,25 +195,10 @@ static const VMStateDescription vmstate_memhp_state = { } }; -static bool vmstate_test_use_cpuhp (void *opaque) -{ -PIIX4PMState *s = opaque; -return !s->cpu_hotplug_legacy; -} - -static int vmstate_cpuhp_pre_load (void *opaque) -{ -Object *obj =OBJECT (opaque); -object_property_set_bool (obj,"cpu-hotplug-legacy" , false, &error_abort); -return 0 ; -} -static const VMStateDescription vmstate_cpuhp_state = { .name ="piix4_pm/cpuhp" , .version_id =1 , .minimum_version_id =1 , - .needed = vmstate_test_use_cpuhp, - .pre_load = vmstate_cpuhp_pre_load, .fields = (const VMStateField []) {VMSTATE_CPU_HOTPLUG (cpuhp_state, PIIX4PMState),VMSTATE_END_OF_LIST ()
@@ -573,12 +558,12 @@ static void piix4_acpi_system_hot_add_init(MemoryRegion *parent,qbus_set_hotplug_handler (BUS (pci_get_bus (PCI_DEVICE (s))),OBJECT (s)); } - s->cpu_hotplug_legacy = true; + s->cpu_hotplug_legacy = false;object_property_add_bool (OBJECT (s),"cpu-hotplug-legacy" , piix4_get_cpu_hotplug_legacy, piix4_set_cpu_hotplug_legacy); -legacy_acpi_cpu_hotplug_init (parent,OBJECT (s), &s->gpe_cpu, -PIIX4_CPU_HOTPLUG_IO_BASE ); +cpu_hotplug_hw_init (parent,OBJECT (s), &s->cpuhp_state, +PIIX4_CPU_HOTPLUG_IO_BASE );if (s->acpi_memory_hotplug.is_enabled) { acpi_memory_hotplug_init(parent, OBJECT(s), &s->acpi_memory_hotplug,
hw/i386/acpi-build.c +1 -1
@@ -964,7 +964,7 @@ build_dsdt(GArray *table_data, BIOSLinker *linker,build_legacy_cpu_hotplug_aml (dsdt, machine, pm->cpu_hp_io_base); }else {CPUHotplugFeatures opts = { - .acpi_1_compatible = true, .has_legacy_cphp = true, + .acpi_1_compatible = true, .smi_path = pm->smi_on_cpuhp ?"\\_SB.PCI0.SMI0.SMIC" :NULL , .fw_unplugs_cpu = pm->smi_on_cpu_unplug, };
hw/loongarch/virt-acpi-build.c +0 -1
@@ -369,7 +369,6 @@ build_la_ged_aml(Aml *dsdt, MachineState *machine)if (event &ACPI_GED_CPU_HOTPLUG_EVT ) { opts.acpi_1_compatible = false; - opts.has_legacy_cphp = false; opts.fw_unplugs_cpu = false; opts.smi_path =NULL ;
Signature · 1 line
2.52.0
Re: [PULL 04/41] pc: Start with modern CPU hotplug interface by default
Quoted 21 lines · Paolo Bonzini
From: Zhao Liu <[email protected]>
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 (in commit
679dd1a957df("pc: use new CPU hotplug interface since 2.7 machine type")). In that way, legacy firmware that doesn't use QEMU generated ACPI tables was able to continue using legacy CPU hotplug interface.While later machine types, with firmware supporting QEMU provided ACPI tables, generate new CPU hotplug AML, which will switch to new CPU hotplug interface when guest OS executes its
_INImethod on ACPI tables loading.Since 2.6 machine type is now gone, and consider that the legacy BIOS (based on QEMU ACPI prior to v2.7) should be no longer in use, previous compatibility requirements are no longer necessary. So initialize 'modern' hotplug directly from the very beginning for PC/Q35 machines with
cpu_hotplug_hw_init(), and drop_INITmethod.
Quoted 6 lines
Additionally, remove the checks and settings around
cpu_hotplug_legacyin cpuhp VMState (for piix4 & ich9), to eliminate the risk of segmentation faults, asgpe_cpuno longer has the opportunity to be initialized. This is safe because all hotplug now start with the modern way, and it's impossible to switch to legacy way at runtime (even the "cpu-hotplug-legacy" properties does not allow it either).
Hi. This commit seems (according to git bisect) to have broken "savevm" for the MIPS malta board (which uses piix4): it now segfaults when you run the command from the monitor. Repro:
$ qemu-img create -f qcow2 dummy.qcow2 32M
Build QEMU with '--target-list=mips-linux-user,mips64-linux-user,mipsel-linux-user,mipsn32-linux-user,mipsn32el-linux-user,mips-softmmu,mipsel-softmmu,mips64-softmmu,mips64el-softmmu'.
$ ./build/mips/qemu-system-mipsel -nographic -drive if=none,format=qcow2,file=dummy.qcow2 [Type "C-a c" to get the "(qemu)" monitor prompt)] (qemu) savevm foo
Backtrace from doing this under gdb:
#0 0x0000555555df7d4d in vmsd_can_compress (field=0x5555564f78a0 <__compound_literal.3>) at ../../migration/vmstate.c:339 #1 0x0000555555df7dbb in vmsd_desc_field_start
(vmsd=0x555556431ba0 <vmstate_cpuhp_state>, vmdesc=0x555556918690,
field=0x5555564f78a0 <__compound_literal.3>, i=0, max=1) at
../../migration/vmstate.c:362
#2 0x0000555555df85a7 in vmstate_save_state_v
(f=0x555556b5a0c0, vmsd=0x555556431ba0 <vmstate_cpuhp_state>,
opaque=0x555556c9aac0, vmdesc=0x555556918690, version_id=1,
errp=0x7fffffffc948) at ../../migration/vmstate.c:528
#3 0x0000555555df8032 in vmstate_save_state
(f=0x555556b5a0c0, vmsd=0x555556431ba0 <vmstate_cpuhp_state>,
opaque=0x555556c9aac0, vmdesc_id=0x555556918690, errp=0x7fffffffc948)
at ../../migration/vmstate.c:427
#4 0x0000555555df8f83 in vmstate_subsection_save
(f=0x555556b5a0c0, vmsd=0x555556431c40 <vmstate_acpi>,
opaque=0x555556c9aac0, vmdesc=0x555556918690, errp=0x7fffffffc948)
at ../../migration/vmstate.c:695
#5 0x0000555555df88a0 in vmstate_save_state_v
(f=0x555556b5a0c0, vmsd=0x555556431c40 <vmstate_acpi>,
opaque=0x555556c9aac0, vmdesc=0x555556918690, version_id=3,
errp=0x7fffffffc948) at ../../migration/vmstate.c:582
#6 0x0000555555df8032 in vmstate_save_state
(f=0x555556b5a0c0, vmsd=0x555556431c40 <vmstate_acpi>,opaque=0x555556c9aac0, vmdesc_id=0x555556918690, errp=0x7fffffffc948) at ../../migration/vmstate.c:427
#7 0x0000555555bd379e in vmstate_save (f=0x555556b5a0c0, se=0x555557090ff0, vmdesc=0x555556918690, errp=0x7fffffffc948)
at ../../migration/savevm.c:1054
#8 0x0000555555bd5133 in qemu_savevm_state_non_iterable
(f=0x555556b5a0c0, errp=0x7fffffffc948)
at ../../migration/savevm.c:1726#9 0x0000555555bd5226 in qemu_savevm_state_complete_precopy (s=0x5555567a1d00) at ../../migration/savevm.c:1753 #10 0x0000555555bd5615 in qemu_savevm_state (f=0x555556b5a0c0, errp=0x7fffffffcbf0) at ../../migration/savevm.c:1855 #11 0x0000555555bd8b83 in save_snapshot
(name=0x555557497150 "foo", overwrite=true, vmstate=0x0,
has_devices=false, devices=0x0, errp=0x7fffffffcbf0)
at ../../migration/savevm.c:3267#12 0x0000555555ba9a62 in hmp_savevm (mon=0x55555691d950, qdict=0x5555574973a0) at ../../migration/migration-hmp-cmds.c:481
The problem is that the vmstate_cpu_hotplug VMStateDescription that is being used has a NULL "fields" (it is the "stub" one). This used to be OK because the "piix4_pm/cpuhp" vmstate had a "needed" function that would return false for Malta, so we never tried to use the stub code.
What's the intention here ? Shouldn't we still have a "needed" function that returns "false" for the cases where we would be using the stub version of vmstate_cpu_hotplug ? Otherwise we would be changing the migration data. We can permit a compat break for MIPS boards, but the commit message suggests that this wasn't intended to be a compat break.
In this thread Fabiano suggests a fix which is probably OK if we want to take the "break migration compat" route:
https://lore.kernel.org/qemu-devel/CAFEAcA9BzSp9F8yv4eZv5eeUM4JXSy9T-QnRr_UbJZpNmhtHyw@mail.gmail.com/T/#m6f4a672ddfa8a0a34dfdf9b57bb04e529ab5c9a4
Quoted 164 lines
include/hw/acpi/cpu.h | 1 - hw/acpi/cpu.c | 10 ---------- hw/acpi/ich9.c | 22 +++------------------- hw/acpi/piix4.c | 21 +++------------------ hw/i386/acpi-build.c | 2 +- hw/loongarch/virt-acpi-build.c | 1 - 6 files changed, 7 insertions(+), 50 deletions(-)include/hw/acpi/cpu.h +0
-1@@ -54,7 +54,6 @@ void cpu_hotplug_hw_init(MemoryRegion *as, Object *owner,typedef struct CPUHotplugFeatures {bool acpi_1_compatible; -bool has_legacy_cphp;bool fw_unplugs_cpu;const char *smi_path; }CPUHotplugFeatures ;hw/acpi/cpu.c +0
-10@@ -408,16 +408,6 @@ void build_cpus_aml(Aml *table, MachineState *machine, CPUHotplugFeatures opts,aml_append (field,aml_reserved_field (4 *8 ));aml_append (field,aml_named_field (CPU_DATA ,32 ));aml_append (cpu_ctrl_dev, field); - -if (opts.has_legacy_cphp) { - method =aml_method ("_INI" ,0 ,AML_SERIALIZED ); -/* switch off legacy CPU hotplug HW and use new one, -* on reboot system is in new mode and writing 0 -* in CPU_SELECTOR selects BSP, which is NOP at -* the time _INI is called */ -aml_append (method,aml_store (zero,aml_name (CPU_SELECTOR ))); -aml_append (cpu_ctrl_dev, method); - } }aml_append (sb_scope, cpu_ctrl_dev);hw/acpi/ich9.c +3
-19@@ -183,26 +183,10 @@ static const VMStateDescription vmstate_tco_io_state = { } }; -static bool vmstate_test_use_cpuhp (void *opaque) -{ -ICH9LPCPMRegs *s = opaque; -return !s->cpu_hotplug_legacy; -} - -static int vmstate_cpuhp_pre_load (void *opaque) -{ -ICH9LPCPMRegs *s = opaque; -Object *obj =OBJECT (s->gpe_cpu.device); -object_property_set_bool (obj,"cpu-hotplug-legacy" , false, &error_abort); -return 0 ; -} -static const VMStateDescription vmstate_cpuhp_state = { .name ="ich9_pm/cpuhp" , .version_id =1 , .minimum_version_id =1 , - .needed = vmstate_test_use_cpuhp, - .pre_load = vmstate_cpuhp_pre_load, .fields = (const VMStateField []) {VMSTATE_CPU_HOTPLUG (cpuhp_state, ICH9LPCPMRegs),VMSTATE_END_OF_LIST ()@@ -338,8 +322,8 @@ void ich9_pm_init(PCIDevice *lpc_pci, ICH9LPCPMRegs *pm, qemu_irq sci_irq) pm->powerdown_notifier.notify = pm_powerdown_req;qemu_register_powerdown_notifier (&pm->powerdown_notifier); -legacy_acpi_cpu_hotplug_init (pci_address_space_io (lpc_pci), -OBJECT (lpc_pci), &pm->gpe_cpu,ICH9_CPU_HOTPLUG_IO_BASE ); +cpu_hotplug_hw_init (pci_address_space_io (lpc_pci), +OBJECT (lpc_pci), &pm->cpuhp_state,ICH9_CPU_HOTPLUG_IO_BASE ); acpi_memory_hotplug_init(pci_address_space_io(lpc_pci), OBJECT(lpc_pci), &pm->acpi_memory_hotplug,@@ -419,7 +403,7 @@ void ich9_pm_add_properties(Object *obj, ICH9LPCPMRegs *pm) {static const uint32_t gpe0_len =ICH9_PMIO_GPE0_LEN ; pm->acpi_memory_hotplug.is_enabled = true; - pm->cpu_hotplug_legacy = true; + pm->cpu_hotplug_legacy = false; pm->disable_s3 =0 ; pm->disable_s4 =0 ; pm->s4_val =2 ;hw/acpi/piix4.c +3
-18@@ -195,25 +195,10 @@ static const VMStateDescription vmstate_memhp_state = { } }; -static bool vmstate_test_use_cpuhp (void *opaque) -{ -PIIX4PMState *s = opaque; -return !s->cpu_hotplug_legacy; -} - -static int vmstate_cpuhp_pre_load (void *opaque) -{ -Object *obj =OBJECT (opaque); -object_property_set_bool (obj,"cpu-hotplug-legacy" , false, &error_abort); -return 0 ; -} -static const VMStateDescription vmstate_cpuhp_state = { .name ="piix4_pm/cpuhp" , .version_id =1 , .minimum_version_id =1 , - .needed = vmstate_test_use_cpuhp, - .pre_load = vmstate_cpuhp_pre_load, .fields = (const VMStateField []) {VMSTATE_CPU_HOTPLUG (cpuhp_state, PIIX4PMState),VMSTATE_END_OF_LIST ()@@ -573,12 +558,12 @@ static void piix4_acpi_system_hot_add_init(MemoryRegion *parent,qbus_set_hotplug_handler (BUS (pci_get_bus (PCI_DEVICE (s))),OBJECT (s)); } - s->cpu_hotplug_legacy = true; + s->cpu_hotplug_legacy = false;object_property_add_bool (OBJECT (s),"cpu-hotplug-legacy" , piix4_get_cpu_hotplug_legacy, piix4_set_cpu_hotplug_legacy); -legacy_acpi_cpu_hotplug_init (parent,OBJECT (s), &s->gpe_cpu, -PIIX4_CPU_HOTPLUG_IO_BASE ); +cpu_hotplug_hw_init (parent,OBJECT (s), &s->cpuhp_state, +PIIX4_CPU_HOTPLUG_IO_BASE );if (s->acpi_memory_hotplug.is_enabled) { acpi_memory_hotplug_init(parent, OBJECT(s), &s->acpi_memory_hotplug,hw/i386/acpi-build.c +1
-1@@ -964,7 +964,7 @@ build_dsdt(GArray *table_data, BIOSLinker *linker,build_legacy_cpu_hotplug_aml (dsdt, machine, pm->cpu_hp_io_base); }else {CPUHotplugFeatures opts = { - .acpi_1_compatible = true, .has_legacy_cphp = true, + .acpi_1_compatible = true, .smi_path = pm->smi_on_cpuhp ?"\\_SB.PCI0.SMI0.SMIC" :NULL , .fw_unplugs_cpu = pm->smi_on_cpu_unplug, };hw/loongarch/virt-acpi-build.c +0
-2@@ -369,7 +369,6 @@ build_la_ged_aml(Aml *dsdt, MachineState *machine)if (event &ACPI_GED_CPU_HOTPLUG_EVT ) { opts.acpi_1_compatible = false; - opts.has_legacy_cphp = false; opts.fw_unplugs_cpu = false; opts.smi_path =NULL ; --
Quoted 1 line
2.52.0
thanks
-- PMM
Re: [PULL 04/41] pc: Start with modern CPU hotplug interface by default
+Philippe & Aurelien
(And seeking Igor's input)
Quoted 6 lines
The problem is that the
vmstate_cpu_hotplugVMStateDescription that is being used has a NULL "fields" (it is the "stub" one). This used to be OK because the "piix4_pm/cpuhp" vmstate had a "needed" function that would return false for Malta, so we never tried to use the stub code.
Quoted 6 lines
What's the intention here ? Shouldn't we still have a "needed" function that returns "false" for the cases where we would be using the stub version of
vmstate_cpu_hotplug? Otherwise we would be changing the migration data. We can permit a compat break for MIPS boards, but the commit message suggests that this wasn't intended to be a compat break.
Thanks, this is my bad - I didn't realize it would break Malta. The "needed" method looks necessary, but previously making the "stub" one depend on cpu_hotplug_legacy=true also sounds a bit not proper.
Maybe in the needed() method we should also check whether the board/ machine supports CPU hotplug: check mc->has_hotpluggable_cpus. IIUC, Malta doesn't support this.
Quoted 3 lines
In this thread Fabiano suggests a fix which is probably OK if we want to take the "break migration compat" route:
https://lore.kernel.org/qemu-devel/CAFEAcA9BzSp9F8yv4eZv5eeUM4JXSy9T-QnRr_UbJZpNmhtHyw@mail.gmail.com/T/#m6f4a672ddfa8a0a34dfdf9b57bb04e529ab5c9a4
I think Fabiano's patch is a clean solution. Or I think there's another
option: we can bring needed() back but check mc->has_hotpluggable_cpus instead, like acpi-ged/cpuhp did ().
Igor, what do you think?
---
hw/acpi/ich9.c +8 -0
@@ -184,10 +184,18 @@ static const VMStateDescription vmstate_tco_io_state = { } }; +static bool cpuhp_needed (void *opaque) +{ +MachineClass *mc =MACHINE_GET_CLASS (qdev_get_machine ()); + +return mc->has_hotpluggable_cpus; +} +static const VMStateDescription vmstate_cpuhp_state = { .name ="ich9_pm/cpuhp" , .version_id =1 , .minimum_version_id =1 , + .needed = cpuhp_needed, .fields = (const VMStateField []) {VMSTATE_CPU_HOTPLUG (cpuhp_state, ICH9LPCPMRegs),VMSTATE_END_OF_LIST ()
hw/acpi/piix4.c +8 -0
@@ -195,10 +195,18 @@ static const VMStateDescription vmstate_memhp_state = { } }; +static bool cpuhp_needed (void *opaque) +{ +MachineClass *mc =MACHINE_GET_CLASS (qdev_get_machine ()); + +return mc->has_hotpluggable_cpus; +} +static const VMStateDescription vmstate_cpuhp_state = { .name ="piix4_pm/cpuhp" , .version_id =1 , .minimum_version_id =1 , + .needed = cpuhp_needed, .fields = (const VMStateField []) {VMSTATE_CPU_HOTPLUG (cpuhp_state, PIIX4PMState),VMSTATE_END_OF_LIST ()
Re: [PULL 04/41] pc: Start with modern CPU hotplug interface by default
Quoted 43 lines · Zhao Liu
I think Fabiano's patch is a clean solution. Or I think there's another
option: we can bringneeded()back but checkmc->has_hotpluggable_cpusinstead, like acpi-ged/cpuhp did ().Igor, what do you think?
---
hw/acpi/ich9.c +8
-0@@ -184,10 +184,18 @@ static const VMStateDescription vmstate_tco_io_state = { } }; +static bool cpuhp_needed (void *opaque) +{ +MachineClass *mc =MACHINE_GET_CLASS (qdev_get_machine ()); + +return mc->has_hotpluggable_cpus; +} +static const VMStateDescription vmstate_cpuhp_state = { .name ="ich9_pm/cpuhp" , .version_id =1 , .minimum_version_id =1 , + .needed = cpuhp_needed, .fields = (const VMStateField []) {VMSTATE_CPU_HOTPLUG (cpuhp_state, ICH9LPCPMRegs),VMSTATE_END_OF_LIST ()hw/acpi/piix4.c +5
-0@@ -195,10 +195,18 @@ static const VMStateDescription vmstate_memhp_state = { } }; +static bool cpuhp_needed (void *opaque) +{ +MachineClass *mc =MACHINE_GET_CLASS (qdev_get_machine ()); + +return mc->has_hotpluggable_cpus;
hw/acpi/piix4.c static const VMStateDescription vmstate_memhp_state = { · line 212
Quoted 10 lines
hw/acpi/piix4.c +3
-0+} +static const VMStateDescription vmstate_cpuhp_state = { .name ="piix4_pm/cpuhp" , .version_id =1 , .minimum_version_id =1 , + .needed = cpuhp_needed, .fields = (const VMStateField []) {VMSTATE_CPU_HOTPLUG (cpuhp_state, PIIX4PMState),VMSTATE_END_OF_LIST ()
I quite like this option, though I defer to people more expert on the piix4/hotplug code. This patch does fix the test failures I was seeing on MIPS.
-- PMM
Re: [PULL 04/41] pc: Start with modern CPU hotplug interface by default
In reply to Zhao LiuQuoted 30 lines · Zhao Liu
+Philippe & Aurelien
(And seeking Igor's input)
Quoted 6 lines
The problem is that the
vmstate_cpu_hotplugVMStateDescription that is being used has a NULL "fields" (it is the "stub" one). This used to be OK because the "piix4_pm/cpuhp" vmstate had a "needed" function that would return false for Malta, so we never tried to use the stub code.Quoted 6 lines
What's the intention here ? Shouldn't we still have a "needed" function that returns "false" for the cases where we would be using the stub version of
vmstate_cpu_hotplug? Otherwise we would be changing the migration data. We can permit a compat break for MIPS boards, but the commit message suggests that this wasn't intended to be a compat break.Thanks, this is my bad - I didn't realize it would break Malta. The "needed" method looks necessary, but previously making the "stub" one depend on cpu_hotplug_legacy=true also sounds a bit not proper.
Maybe in the
needed()method we should also check whether the board/ machine supports CPU hotplug: checkmc->has_hotpluggable_cpus. IIUC, Malta doesn't support this.Quoted 3 lines
In this thread Fabiano suggests a fix which is probably OK if we want to take the "break migration compat" route:
https://lore.kernel.org/qemu-devel/CAFEAcA9BzSp9F8yv4eZv5eeUM4JXSy9T-QnRr_UbJZpNmhtHyw@mail.gmail.com/T/#m6f4a672ddfa8a0a34dfdf9b57bb04e529ab5c9a4
Quoted 5 lines
I think Fabiano's patch is a clean solution. Or I think there's another
option: we can bringneeded()back but checkmc->has_hotpluggable_cpusinstead, like acpi-ged/cpuhp did ().Igor, what do you think?
Hi, what's the plan here ? We need to fix this before 11.0 releases. Are you going to send your proposed ich9.c fix as a proper patch ?
thanks
-- PMM
Re: [PULL 04/41] pc: Start with modern CPU hotplug interface by default
Quoted 40 lines · Peter Maydell
Date: Sun, 29 Mar 2026 13:35:15 +0100
From: Peter Maydell <[email protected]>
Subject: Re: [PULL 04/41] pc: Start with modern CPU hotplug interface by
defaultQuoted 34 lines · Zhao Liu
+Philippe & Aurelien
(And seeking Igor's input)
Quoted 6 lines
The problem is that the
vmstate_cpu_hotplugVMStateDescription that is being used has a NULL "fields" (it is the "stub" one). This used to be OK because the "piix4_pm/cpuhp" vmstate had a "needed" function that would return false for Malta, so we never tried to use the stub code.Quoted 6 lines
What's the intention here ? Shouldn't we still have a "needed" function that returns "false" for the cases where we would be using the stub version of
vmstate_cpu_hotplug? Otherwise we would be changing the migration data. We can permit a compat break for MIPS boards, but the commit message suggests that this wasn't intended to be a compat break.Thanks, this is my bad - I didn't realize it would break Malta. The "needed" method looks necessary, but previously making the "stub" one depend on cpu_hotplug_legacy=true also sounds a bit not proper.
Maybe in the
needed()method we should also check whether the board/ machine supports CPU hotplug: checkmc->has_hotpluggable_cpus. IIUC, Malta doesn't support this.Quoted 3 lines
In this thread Fabiano suggests a fix which is probably OK if we want to take the "break migration compat" route:
https://lore.kernel.org/qemu-devel/CAFEAcA9BzSp9F8yv4eZv5eeUM4JXSy9T-QnRr_UbJZpNmhtHyw@mail.gmail.com/T/#m6f4a672ddfa8a0a34dfdf9b57bb04e529ab5c9a4I think Fabiano's patch is a clean solution. Or I think there's another
option: we can bringneeded()back but checkmc->has_hotpluggable_cpusinstead, like acpi-ged/cpuhp did ().
Quoted 5 lines
Quoted 1 line
Igor, what do you think?
Hi, what's the plan here ? We need to fix this before 11.0 releases. Are you going to send your proposed
ich9.cfix as a proper patch ?
Hi Peter,
Now I post a patch here:
https://lore.kernel.org/qemu-devel/[email protected]
Thanks,
Zhao
[PULL 22/41] hw/i386/pc: Remove pc_compat_2_7[] array
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
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.
--- include/hw/i386/pc.h | 3 --- hw/i386/pc.c | 10 ---------- 2 files changed, 13 deletions(-)
include/hw/i386/pc.h +0 -3
@@ -295,9 +295,6 @@ 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
@@ -256,16 +256,6 @@ 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
2.52.0
[PULL 35/41] qdev: Free property array on release
In reply to Paolo BonziniFrom: Chandan Somani <[email protected]>
Before this patch, users of the property array would free the array themselves in their cleanup functions. This causes inconsistencies where some users leak the array and some free them.
This patch makes it so that the property array's release function frees the property array (instead of just its elements). It fixes any leaks and requires less code.
DEFINE_PROP_ARRAY leakers that are fixed in this patch:
ebpf-rss_fds in hw/net/virtio-net.c rnmi_irqvec, rnmi_excpvec in hw/riscv/riscv_hart.c common.display_modes in hw/display/apple-gfx-mmio.m common.display_modes in hw/display/apple-gfx-pci.m
--- block/accounting.c | 1 - hw/core/qdev-properties.c | 21 ++++++++++----------- hw/input/stellaris_gamepad.c | 8 -------- hw/intc/arm_gicv3_common.c | 8 -------- hw/intc/rx_icu.c | 8 -------- hw/misc/arm_sysctl.c | 2 -- hw/misc/mps2-scc.c | 8 -------- hw/net/rocker/rocker.c | 1 - hw/nvram/xlnx-efuse.c | 8 -------- hw/nvram/xlnx-versal-efuse-ctrl.c | 8 -------- hw/virtio/virtio-iommu-pci.c | 8 -------- 11 files changed, 10 insertions(+), 71 deletions(-)
block/accounting.c +0 -1
@@ -73,7 +73,6 @@ bool block_acct_setup(BlockAcctStats *stats, enum OnOffAuto account_invalid, }block_acct_add_interval (stats, stats_intervals[i]); } -g_free (stats_intervals); }return true; }
hw/core/qdev-properties.c +10 -11
@@ -673,10 +673,8 @@ static Property array_elem_prop(Object *obj, const Property *parent_prop,/* * Object property release callback for array properties: We call the -* underlying element's property release hook for each element. -* -* Note that it is the responsibility of the individual device's deinit -* to free the array proper. +* underlying element's property release hook for each element and free the +* property array. */ static void release_prop_array (Object *obj,const char *name,void *opaque) {
@@ -686,15 +684,16 @@ static void release_prop_array(Object *obj, const char *name, void *opaque)char *elem = *arrayptr;int i; -if (!prop->arrayinfo->release) { -return ; +if (prop->arrayinfo->release) { +for (i =0 ; i < *alenptr; i++) { +Property elem_prop =array_elem_prop (obj, prop, name, elem); + prop->arrayinfo->release (obj,NULL , &elem_prop); + elem += prop->arrayfieldsize; + } } -for (i =0 ; i < *alenptr; i++) { -Property elem_prop =array_elem_prop (obj, prop, name, elem); - prop->arrayinfo->release (obj,NULL , &elem_prop); - elem += prop->arrayfieldsize; - } +g_clear_pointer (arrayptr, g_free); + *alenptr =0 ; }/*
hw/input/stellaris_gamepad.c +0 -8
@@ -63,13 +63,6 @@ static void stellaris_gamepad_realize(DeviceState *dev, Error **errp)qemu_input_handler_register (dev, &stellaris_gamepad_handler); } -static void stellaris_gamepad_finalize (Object *obj) -{ -StellarisGamepad *s =STELLARIS_GAMEPAD (obj); - -g_free (s->keycodes); -} -static void stellaris_gamepad_reset_enter (Object *obj,ResetType type) {StellarisGamepad *s =STELLARIS_GAMEPAD (obj);
@@ -98,7 +91,6 @@ static const TypeInfo stellaris_gamepad_info[] = { .name = TYPE_STELLARIS_GAMEPAD, .parent = TYPE_SYS_BUS_DEVICE, .instance_size = sizeof(StellarisGamepad), - .instance_finalize = stellaris_gamepad_finalize, .class_init = stellaris_gamepad_class_init, }, };
hw/intc/arm_gicv3_common.c +0 -8
@@ -488,13 +488,6 @@ static void arm_gicv3_common_realize(DeviceState *dev, Error **errp) s->itslist =g_ptr_array_new (); } -static void arm_gicv3_finalize (Object *obj) -{ -GICv3State *s =ARM_GICV3_COMMON (obj); - -g_free (s->redist_region_count); -} -static void arm_gicv3_common_reset_hold (Object *obj,ResetType type) {GICv3State *s =ARM_GICV3_COMMON (obj);
@@ -644,7 +637,6 @@ static const TypeInfo arm_gicv3_common_type = { .instance_size = sizeof(GICv3State), .class_size = sizeof(ARMGICv3CommonClass), .class_init = arm_gicv3_common_class_init, - .instance_finalize = arm_gicv3_finalize, .abstract = true, .interfaces = (const InterfaceInfo []) { {TYPE_ARM_LINUX_BOOT_IF },
hw/intc/rx_icu.c +0 -8
@@ -334,13 +334,6 @@ static void rxicu_init(Object *obj)sysbus_init_irq (d, &icu->_swi); } -static void rxicu_fini (Object *obj) -{ -RXICUState *icu =RX_ICU (obj); -g_free (icu->map); -g_free (icu->init_sense); -} -static const VMStateDescription vmstate_rxicu = { .name ="rx-icu" , .version_id =1 ,
@@ -382,7 +375,6 @@ static const TypeInfo rxicu_info = { .parent = TYPE_SYS_BUS_DEVICE, .instance_size = sizeof(RXICUState), .instance_init = rxicu_init, - .instance_finalize = rxicu_fini, .class_init = rxicu_class_init, };
hw/misc/arm_sysctl.c +0 -2
@@ -618,9 +618,7 @@ static void arm_sysctl_finalize(Object *obj) {arm_sysctl_state *s =ARM_SYSCTL (obj); -g_free (s->db_voltage);g_free (s->db_clock); -g_free (s->db_clock_reset); }static const Property arm_sysctl_properties[] = {
hw/misc/mps2-scc.c +0 -8
@@ -405,13 +405,6 @@ static void mps2_scc_realize(DeviceState *dev, Error **errp) s->oscclk =g_new0 (uint32_t, s->num_oscclk); } -static void mps2_scc_finalize (Object *obj) -{ -MPS2SCC *s =MPS2_SCC (obj); - -g_free (s->oscclk_reset); -} -static bool cfg7_needed (void *opaque) {MPS2SCC *s = opaque;
@@ -489,7 +482,6 @@ static const TypeInfo mps2_scc_info = { .parent = TYPE_SYS_BUS_DEVICE, .instance_size = sizeof(MPS2SCC), .instance_init = mps2_scc_init, - .instance_finalize = mps2_scc_finalize, .class_init = mps2_scc_class_init, };
hw/net/rocker/rocker.c +0 -1
@@ -1429,7 +1429,6 @@ static void pci_rocker_uninit(PCIDevice *dev)world_free (r->worlds[i]); } } -g_free (r->fp_ports_peers); }static void rocker_reset (DeviceState *dev)
hw/nvram/xlnx-efuse.c +0 -8
@@ -225,13 +225,6 @@ static void efuse_realize(DeviceState *dev, Error **errp) } } -static void efuse_finalize (Object *obj) -{ -XlnxEFuse *s =XLNX_EFUSE (obj); - -g_free (s->ro_bits); -} -static void efuse_prop_set_drive (Object *obj,Visitor *v,const char *name,void *opaque,Error **errp) {
@@ -289,7 +282,6 @@ static const TypeInfo efuse_info = { .name = TYPE_XLNX_EFUSE, .parent = TYPE_DEVICE, .instance_size = sizeof(XlnxEFuse), - .instance_finalize = efuse_finalize, .class_init = efuse_class_init, };
hw/nvram/xlnx-versal-efuse-ctrl.c +0 -8
@@ -724,13 +724,6 @@ static void efuse_ctrl_init(Object *obj)sysbus_init_irq (sbd, &s->irq_efuse_imr); } -static void efuse_ctrl_finalize (Object *obj) -{ -XlnxVersalEFuseCtrl *s =XLNX_VERSAL_EFUSE_CTRL (obj); - -g_free (s->extra_pg0_lock_spec); -} -static const VMStateDescription vmstate_efuse_ctrl = { .name =TYPE_XLNX_VERSAL_EFUSE_CTRL , .version_id =1 ,
@@ -767,7 +760,6 @@ static const TypeInfo efuse_ctrl_info = { .instance_size = sizeof(XlnxVersalEFuseCtrl), .class_init = efuse_ctrl_class_init, .instance_init = efuse_ctrl_init, - .instance_finalize = efuse_ctrl_finalize, };static void efuse_ctrl_register_types (void )
hw/virtio/virtio-iommu-pci.c +0 -8
@@ -94,18 +94,10 @@ static void virtio_iommu_pci_instance_init(Object *obj) TYPE_VIRTIO_IOMMU); } -static void virtio_iommu_pci_instance_finalize (Object *obj) -{ -VirtIOIOMMUPCI *dev =VIRTIO_IOMMU_PCI (obj); - -g_free (dev->vdev.prop_resv_regions); -} -static const VirtioPCIDeviceTypeInfo virtio_iommu_pci_info = { .generic_name =TYPE_VIRTIO_IOMMU_PCI , .instance_size =sizeof (VirtIOIOMMUPCI), .instance_init = virtio_iommu_pci_instance_init, - .instance_finalize = virtio_iommu_pci_instance_finalize, .class_init = virtio_iommu_pci_class_init, };
Signature · 1 line
2.52.0
[PULL 11/41] hw/nvram/fw_cfg: Factor fw_cfg_init_mem_internal() out
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
Factor 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 | 17 +++++++++++++---- 1 file changed, 13 insertions(+), 4 deletions(-)
hw/nvram/fw_cfg.c +13 -4
@@ -1054,9 +1054,9 @@ 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;
@@ -1088,10 +1088,19 @@ 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_addr && dma_as); +return fw_cfg_init_mem_internal (ctl_addr, data_addr, data_width, + dma_addr, dma_as); +} +FWCfgState *fw_cfg_init_mem_nodma (hwaddr ctl_addr,hwaddr data_addr,unsigned data_width) { -return fw_cfg_init_mem_wide (ctl_addr, data_addr, data_width,0 ,NULL ); +return fw_cfg_init_mem_internal (ctl_addr, data_addr, data_width,0 ,NULL ); }
Signature · 1 line
2.52.0
[PULL 01/41] target/i386: convert SEV-ES termination requests to guest panic events
In reply to Paolo BonziniThis produces a good error message instead of:
KVM: unknown exit reason 24
EAX=00000000 EBX=00000000 ECX=00000000 EDX=00a00f11
ESI=00000000 EDI=00000000 EBP=00000000 ESP=00000000
EIP=0000b004 EFL=00000002 [-------] CPL=0 II=0 A20=1 SMM=0 HLT=1
ES =0000 00000000 0000ffff 00009300
CS =f000 00800000 0000ffff 00009b00
SS =0000 00000000 0000ffff 00009300
DS =0000 00000000 0000ffff 00009300
FS =0000 00000000 0000ffff 00009300
GS =0000 00000000 0000ffff 00009300
LDT=0000 00000000 0000ffff 00008200
TR =0000 00000000 0000ffff 00008b00
GDT= 00000000 0000ffff
IDT= 00000000 0000ffff
CR0=60000010 CR2=00000000 CR3=00000000 CR4=00000000
DR0=0000000000000000 DR1=0000000000000000 DR2=0000000000000000 DR3=0000000000000000
DR6=00000000ffff0ff0 DR7=0000000000000400
EFER=0000000000000000
Code=00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 <00> 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
--- qapi/run-state.json | 23 +++++++++++++++++++++-- accel/kvm/kvm-all.c | 1 + system/runstate.c | 4 ++++ target/i386/cpu-system.c | 22 ++++++++++++++++++++++ 4 files changed, 48 insertions(+), 2 deletions(-)
qapi/run-state.json +21 -2
@@ -504,10 +504,12 @@ # # @tdx: tdx guest panic information type (Since: 10.1) # +# @sev: AMD SEV-ES guest termination information type (Since: 11.0) +# # Since: 2.9 ## { 'enum': 'GuestPanicInformationType', - 'data': [ 'hyper-v', 's390', 'tdx' ] } + 'data': [ 'hyper-v', 's390', 'tdx', 'sev' ] } ## # @GuestPanicInformation:
@@ -523,7 +525,8 @@ 'discriminator': 'type', 'data': {'hyper-v': 'GuestPanicInformationHyperV', 's390': 'GuestPanicInformationS390', - 'tdx' : 'GuestPanicInformationTdx'}} + 'tdx': 'GuestPanicInformationTdx', + 'sev': 'GuestPanicInformationSev' }} ## # @GuestPanicInformationHyperV:
@@ -625,6 +628,22 @@ 'message': 'str', '*gpa': 'uint64'}} +## +# @GuestPanicInformationSev: +# +# Information for AMD SEV-specific termination request (GHCB MSR +# contents) +# +# @set: The reason code set provided by the guest +# +# @code: The reason code provided by the guest +# +# Since: 11.0 +## +{'struct': 'GuestPanicInformationSev', + 'data': {'set': 'uint32', + 'code': 'uint32'}} + ## # @MEMORY_FAILURE: #
accel/kvm/kvm-all.c +1 -0
@@ -3300,6 +3300,7 @@ int kvm_cpu_exec(CPUState *cpu)qemu_system_reset_request (SHUTDOWN_CAUSE_GUEST_RESET ); ret =EXCP_INTERRUPT ;break ; +case KVM_SYSTEM_EVENT_SEV_TERM :case KVM_SYSTEM_EVENT_CRASH :kvm_cpu_synchronize_state (cpu);bql_lock ();
system/runstate.c +4 -0
@@ -669,6 +669,10 @@ void qemu_system_guest_panicked(GuestPanicInformation *info)"can be found at gpa page: 0x%" PRIx64"\n" , info->u.tdx.gpa); } + }else if (info->type ==GUEST_PANIC_INFORMATION_TYPE_SEV ) { +qemu_log_mask (LOG_GUEST_ERROR ,"SEV termination (reason set: %d code: %d)" , + info->u.sev.set, + info->u.sev.code); }qapi_free_GuestPanicInformation (info);
target/i386/cpu-system.c +22 -0
@@ -18,6 +18,8 @@*/ #include "qemu/osdep.h" +#include "qemu/error-report.h" +#include "system/kvm.h" #include "cpu.h" #include "qapi/error.h" #include "qapi/qapi-visit-run-state.h"
@@ -272,6 +274,26 @@ GuestPanicInformation *x86_cpu_get_crash_info(CPUState *cs)CPUX86State *env = &cpu->env;GuestPanicInformation *panic_info =NULL ; +#ifdef CONFIG_KVM +if (kvm_enabled ()) { +struct kvm_run *run = cs->kvm_run; + +if (run->exit_reason ==KVM_EXIT_SYSTEM_EVENT && + run->system_event.type ==KVM_SYSTEM_EVENT_SEV_TERM ) { + panic_info =g_new0 (GuestPanicInformation,1 ); + + panic_info->type =GUEST_PANIC_INFORMATION_TYPE_SEV ; +/* There should always be one data item, otherwise use zeroes. */ +if (run->system_event.ndata >0 ) { + panic_info->u.sev.set = (run->system_event.data[0 ] >>12 ) &0xf ; + panic_info->u.sev.code = (run->system_event.data[0 ] >>16 ) &0xff ; + }else { +warn_report ("Hypervisor did not provide any data for SEV-ES termination" ); + } + } + } else +#endif +if (hyperv_feat_enabled (cpu,HYPERV_FEAT_CRASH )) { panic_info =g_new0 (GuestPanicInformation,1 );
Signature · 1 line
2.52.0
[PULL 05/41] hw/i386/pc: Remove PCMachineClass::legacy_cpu_hotplug field
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
Now all PC & Q35 machiens are using modern hotplug from the beginning, and all legacy_cpu_hotplug flags keep false during runtime.
So it's safe to remove legacy_cpu_hotplug flags and related properties, with unused gpe_cpu field.
--- include/hw/acpi/ich9.h | 2 -- include/hw/acpi/piix4.h | 2 -- include/hw/i386/pc.h | 3 --- hw/acpi/ich9.c | 40 ++++------------------------------------ hw/acpi/piix4.c | 40 ++++------------------------------------ hw/i386/acpi-build.c | 4 +--- 6 files changed, 9 insertions(+), 82 deletions(-)
include/hw/acpi/ich9.h +0 -2
@@ -53,8 +53,6 @@ typedef struct ICH9LPCPMRegs {uint32_t pm_io_base;Notifier powerdown_notifier; -bool cpu_hotplug_legacy; -AcpiCpuHotplug gpe_cpu;CPUHotplugState cpuhp_state;bool keep_pci_slot_hpc;
include/hw/acpi/piix4.h +0 -2
@@ -63,8 +63,6 @@ struct PIIX4PMState {uint8_t disable_s4;uint8_t s4_val; -bool cpu_hotplug_legacy; -AcpiCpuHotplug gpe_cpu;CPUHotplugState cpuhp_state;MemHotplugState acpi_memory_hotplug;
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/ich9.c +4 -36
@@ -339,26 +339,6 @@ static void ich9_pm_get_gpe0_blk(Object *obj, Visitor *v, const char *name,visit_type_uint32 (v, name, &value, errp); } -static bool ich9_pm_get_cpu_hotplug_legacy (Object *obj,Error **errp) -{ -ICH9LPCState *s =ICH9_LPC_DEVICE (obj); - -return s->pm.cpu_hotplug_legacy; -} - -static void ich9_pm_set_cpu_hotplug_legacy (Object *obj,bool value, -Error **errp) -{ -ICH9LPCState *s =ICH9_LPC_DEVICE (obj); - -assert (!value); -if (s->pm.cpu_hotplug_legacy && value == false) { -acpi_switch_to_modern_cphp (&s->pm.gpe_cpu, &s->pm.cpuhp_state, -ICH9_CPU_HOTPLUG_IO_BASE ); - } - s->pm.cpu_hotplug_legacy = value; -} -static bool ich9_pm_get_enable_tco (Object *obj,Error **errp) {ICH9LPCState *s =ICH9_LPC_DEVICE (obj);
@@ -403,7 +383,6 @@ void ich9_pm_add_properties(Object *obj, ICH9LPCPMRegs *pm) {static const uint32_t gpe0_len =ICH9_PMIO_GPE0_LEN ; pm->acpi_memory_hotplug.is_enabled = true; - pm->cpu_hotplug_legacy = false; pm->disable_s3 =0 ; pm->disable_s4 =0 ; pm->s4_val =2 ;
@@ -422,9 +401,6 @@ void ich9_pm_add_properties(Object *obj, ICH9LPCPMRegs *pm) NULL, NULL, pm);object_property_add_uint32_ptr (obj,ACPI_PM_PROP_GPE0_BLK_LEN , &gpe0_len,OBJ_PROP_FLAG_READ ); -object_property_add_bool (obj,"cpu-hotplug-legacy" , - ich9_pm_get_cpu_hotplug_legacy, - ich9_pm_set_cpu_hotplug_legacy);object_property_add_uint8_ptr (obj,ACPI_PM_PROP_S3_DISABLED , &pm->disable_s3,OBJ_PROP_FLAG_READWRITE ); object_property_add_uint8_ptr(obj, ACPI_PM_PROP_S4_DISABLED,
@@ -477,11 +453,7 @@ void ich9_pm_device_plug_cb(HotplugHandler *hotplug_dev, DeviceState *dev, dev, errp); } }else if (object_dynamic_cast (OBJECT (dev),TYPE_CPU )) { -if (lpc->pm.cpu_hotplug_legacy) { -legacy_acpi_cpu_plug_cb (hotplug_dev, &lpc->pm.gpe_cpu, dev, errp); - }else { -acpi_cpu_plug_cb (hotplug_dev, &lpc->pm.cpuhp_state, dev, errp); - } +acpi_cpu_plug_cb (hotplug_dev, &lpc->pm.cpuhp_state, dev, errp); }else if (object_dynamic_cast (OBJECT (dev),TYPE_PCI_DEVICE )) {acpi_pcihp_device_plug_cb (hotplug_dev, &lpc->pm.acpi_pci_hotplug, dev, errp);
@@ -500,8 +472,7 @@ void ich9_pm_device_unplug_request_cb(HotplugHandler *hotplug_dev,acpi_memory_unplug_request_cb (hotplug_dev, &lpc->pm.acpi_memory_hotplug, dev, errp); - }else if (object_dynamic_cast (OBJECT (dev),TYPE_CPU ) && - !lpc->pm.cpu_hotplug_legacy) { + }else if (object_dynamic_cast (OBJECT (dev),TYPE_CPU )) {uint64_t negotiated = lpc->smi_negotiated_features; if (negotiated & BIT_ULL(ICH9_LPC_SMI_F_BROADCAST_BIT) &&
@@ -533,8 +504,7 @@ void ich9_pm_device_unplug_cb(HotplugHandler *hotplug_dev, DeviceState *dev,if (object_dynamic_cast (OBJECT (dev),TYPE_PC_DIMM )) {acpi_memory_unplug_cb (&lpc->pm.acpi_memory_hotplug, dev, errp); - }else if (object_dynamic_cast (OBJECT (dev),TYPE_CPU ) && - !lpc->pm.cpu_hotplug_legacy) { + }else if (object_dynamic_cast (OBJECT (dev),TYPE_CPU )) {acpi_cpu_unplug_cb (&lpc->pm.cpuhp_state, dev, errp); }else if (object_dynamic_cast (OBJECT (dev),TYPE_PCI_DEVICE )) { acpi_pcihp_device_unplug_cb(hotplug_dev, &lpc->pm.acpi_pci_hotplug,
@@ -556,7 +526,5 @@ void ich9_pm_ospm_status(AcpiDeviceIf *adev, ACPIOSTInfoList ***list)ICH9LPCState *s =ICH9_LPC_DEVICE (adev);acpi_memory_ospm_status (&s->pm.acpi_memory_hotplug, list); -if (!s->pm.cpu_hotplug_legacy) { -acpi_cpu_ospm_status (&s->pm.cpuhp_state, list); - } +acpi_cpu_ospm_status (&s->pm.cpuhp_state, list); }
hw/acpi/piix4.c +4 -36
@@ -336,11 +336,7 @@ static void piix4_device_plug_cb(HotplugHandler *hotplug_dev, }else if (object_dynamic_cast (OBJECT (dev),TYPE_PCI_DEVICE )) {acpi_pcihp_device_plug_cb (hotplug_dev, &s->acpi_pci_hotplug, dev, errp); }else if (object_dynamic_cast (OBJECT (dev),TYPE_CPU )) { -if (s->cpu_hotplug_legacy) { -legacy_acpi_cpu_plug_cb (hotplug_dev, &s->gpe_cpu, dev, errp); - }else { -acpi_cpu_plug_cb (hotplug_dev, &s->cpuhp_state, dev, errp); - } +acpi_cpu_plug_cb (hotplug_dev, &s->cpuhp_state, dev, errp); }else {g_assert_not_reached (); }
@@ -358,8 +354,7 @@ static void piix4_device_unplug_request_cb(HotplugHandler *hotplug_dev, }else if (object_dynamic_cast (OBJECT (dev),TYPE_PCI_DEVICE )) {acpi_pcihp_device_unplug_request_cb (hotplug_dev, &s->acpi_pci_hotplug, dev, errp); - }else if (object_dynamic_cast (OBJECT (dev),TYPE_CPU ) && - !s->cpu_hotplug_legacy) { + }else if (object_dynamic_cast (OBJECT (dev),TYPE_CPU )) {acpi_cpu_unplug_request_cb (hotplug_dev, &s->cpuhp_state, dev, errp); }else { error_setg(errp, "acpi: device unplug request for not supported device"
@@ -378,8 +373,7 @@ static void piix4_device_unplug_cb(HotplugHandler *hotplug_dev, }else if (object_dynamic_cast (OBJECT (dev),TYPE_PCI_DEVICE )) {acpi_pcihp_device_unplug_cb (hotplug_dev, &s->acpi_pci_hotplug, dev, errp); - }else if (object_dynamic_cast (OBJECT (dev),TYPE_CPU ) && - !s->cpu_hotplug_legacy) { + }else if (object_dynamic_cast (OBJECT (dev),TYPE_CPU )) {acpi_cpu_unplug_cb (&s->cpuhp_state, dev, errp); }else { error_setg(errp, "acpi: device unplug for not supported device"
@@ -523,26 +517,6 @@ static const MemoryRegionOps piix4_gpe_ops = { .endianness = DEVICE_LITTLE_ENDIAN, }; - -static bool piix4_get_cpu_hotplug_legacy (Object *obj,Error **errp) -{ -PIIX4PMState *s =PIIX4_PM (obj); - -return s->cpu_hotplug_legacy; -} - -static void piix4_set_cpu_hotplug_legacy (Object *obj,bool value,Error **errp) -{ -PIIX4PMState *s =PIIX4_PM (obj); - -assert (!value); -if (s->cpu_hotplug_legacy && value == false) { -acpi_switch_to_modern_cphp (&s->gpe_cpu, &s->cpuhp_state, -PIIX4_CPU_HOTPLUG_IO_BASE ); - } - s->cpu_hotplug_legacy = value; -} -static void piix4_acpi_system_hot_add_init (MemoryRegion *parent,PCIBus *bus,PIIX4PMState *s) {
@@ -558,10 +532,6 @@ static void piix4_acpi_system_hot_add_init(MemoryRegion *parent,qbus_set_hotplug_handler (BUS (pci_get_bus (PCI_DEVICE (s))),OBJECT (s)); } - s->cpu_hotplug_legacy = false; -object_property_add_bool (OBJECT (s),"cpu-hotplug-legacy" , - piix4_get_cpu_hotplug_legacy, - piix4_set_cpu_hotplug_legacy);cpu_hotplug_hw_init (parent,OBJECT (s), &s->cpuhp_state,PIIX4_CPU_HOTPLUG_IO_BASE );
@@ -576,9 +546,7 @@ static void piix4_ospm_status(AcpiDeviceIf *adev, ACPIOSTInfoList ***list)PIIX4PMState *s =PIIX4_PM (adev);acpi_memory_ospm_status (&s->acpi_memory_hotplug, list); -if (!s->cpu_hotplug_legacy) { -acpi_cpu_ospm_status (&s->cpuhp_state, list); - } +acpi_cpu_ospm_status (&s->cpuhp_state, list); }static void piix4_send_gpe (AcpiDeviceIf *adev,AcpiEventStatusBits ev)
hw/i386/acpi-build.c +1 -3
@@ -960,9 +960,7 @@ 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, .smi_path = pm->smi_on_cpuhp ?"\\_SB.PCI0.SMI0.SMIC" :NULL ,
Signature · 1 line
2.52.0
[PULL 09/41] hw/nvram/fw_cfg: Rename fw_cfg_init_mem() with '_nodma' suffix
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
Rename fw_cfg_init_mem() as fw_cfg_init_mem_nodma() to distinct with the DMA version (currently named fw_cfg_init_mem_wide).
--- include/hw/nvram/fw_cfg.h | 3 ++- hw/hppa/machine.c | 2 +- hw/nvram/fw_cfg.c | 7 +++---- 3 files changed, 6 insertions(+), 6 deletions(-)
include/hw/nvram/fw_cfg.h +2 -1
@@ -307,7 +307,8 @@ bool fw_cfg_add_file_from_generator(FWCfgState *s,FWCfgState *fw_cfg_init_io_dma (uint32_t iobase,uint32_t dma_iobase,AddressSpace *dma_as); -FWCfgState *fw_cfg_init_mem (hwaddr ctl_addr,hwaddr data_addr); +FWCfgState *fw_cfg_init_mem_nodma (hwaddr ctl_addr,hwaddr data_addr, +unsigned data_width);FWCfgState *fw_cfg_init_mem_wide (hwaddr ctl_addr,hwaddr data_addr,uint32_t data_width,hwaddr dma_addr,AddressSpace *dma_as);
hw/hppa/machine.c +1 -1
@@ -208,7 +208,7 @@ static FWCfgState *create_fw_cfg(MachineState *ms, PCIBus *pci_bus,int btlb_entries =HPPA_BTLB_ENTRIES (&cpu[0 ]->env);int len; - fw_cfg =fw_cfg_init_mem (addr, addr +4 ); + fw_cfg =fw_cfg_init_mem_nodma (addr, addr +4 ,1 );fw_cfg_add_i16 (fw_cfg,FW_CFG_NB_CPUS , ms->smp.cpus);fw_cfg_add_i16 (fw_cfg,FW_CFG_MAX_CPUS ,HPPA_MAX_CPUS );fw_cfg_add_i64 (fw_cfg,FW_CFG_RAM_SIZE , ms->ram_size);
hw/nvram/fw_cfg.c +3 -4
@@ -1088,11 +1088,10 @@ FWCfgState *fw_cfg_init_mem_wide(hwaddr ctl_addr,return s; } -FWCfgState *fw_cfg_init_mem (hwaddr ctl_addr,hwaddr data_addr) +FWCfgState *fw_cfg_init_mem_nodma (hwaddr ctl_addr,hwaddr data_addr, +unsigned data_width) { -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_wide (ctl_addr, data_addr, data_width,0 ,NULL ); }
Signature · 1 line
2.52.0
[PULL 20/41] hw/virtio/virtio-mmio: Remove VirtIOMMIOProxy::format_transport_address field
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
The 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
@@ -764,8 +764,6 @@ 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),
@@ -827,19 +825,6 @@ 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
2.52.0
[PULL 17/41] hw/i386/pc: Remove pc_compat_2_6[] array
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
The 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
@@ -298,9 +298,6 @@ 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
@@ -266,14 +266,6 @@ 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
2.52.0
[PULL 06/41] acpi: Remove legacy cpu hotplug utilities
In reply to Paolo BonziniFrom: Zhao Liu <[email protected]>
The cpu_hotplug.h and cpu_hotplug.c contain legacy cpu hotplug utilities. Now there's no use case of legacy cpu hotplug, so it's safe to drop legacy cpu hotplug support totally.
--- include/hw/acpi/ich9.h | 2 +- include/hw/acpi/piix4.h | 2 +- hw/acpi/acpi-cpu-hotplug-stub.c | 19 +- hw/acpi/cpu_hotplug.c | 348 -------------------------------- hw/acpi/generic_event_device.c | 1 + hw/acpi/ich9.c | 1 + hw/acpi/piix4.c | 2 +- hw/i386/acpi-build.c | 1 + hw/i386/pc.c | 3 +- hw/i386/x86-common.c | 1 - hw/acpi/meson.build | 2 +- 11 files changed, 10 insertions(+), 372 deletions(-) delete mode 100644 hw/acpi/cpu_hotplug.c
include/hw/acpi/ich9.h +1 -1
@@ -22,12 +22,12 @@#define HW_ACPI_ICH9_H #include "hw/acpi/acpi.h" -#include "hw/acpi/cpu_hotplug.h" #include "hw/acpi/cpu.h" #include "hw/acpi/pcihp.h" #include "hw/acpi/memory_hotplug.h" #include "hw/acpi/acpi_dev_interface.h" #include "hw/acpi/ich9_tco.h" +#include "hw/acpi/cpu.h" #define ACPI_PCIHP_ADDR_ICH9 0x0cc0
include/hw/acpi/piix4.h +1 -1
@@ -24,11 +24,11 @@#include "hw/pci/pci_device.h" #include "hw/acpi/acpi.h" -#include "hw/acpi/cpu_hotplug.h" #include "hw/acpi/memory_hotplug.h" #include "hw/acpi/pcihp.h" #include "hw/i2c/pm_smbus.h" #include "hw/isa/apm.h" +#include "hw/acpi/cpu.h" #define TYPE_PIIX4_PM "PIIX4_PM"OBJECT_DECLARE_SIMPLE_TYPE (PIIX4PMState,PIIX4_PM )
hw/acpi/acpi-cpu-hotplug-stub.c +1 -18
@@ -1,22 +1,10 @@#include "qemu/osdep.h" -#include "hw/acpi/cpu_hotplug.h" #include "migration/vmstate.h" - +#include "hw/acpi/cpu.h" /* Following stubs are all related to ACPI cpu hotplug */ const VMStateDescription vmstate_cpu_hotplug; -void acpi_switch_to_modern_cphp (AcpiCpuHotplug *gpe_cpu, -CPUHotplugState *cpuhp_state, -uint16_t io_port) -{ -} - -void legacy_acpi_cpu_hotplug_init (MemoryRegion *parent,Object *owner, -AcpiCpuHotplug *gpe_cpu,uint16_t base) -{ -} -void cpu_hotplug_hw_init (MemoryRegion *as,Object *owner,CPUHotplugState *state,hwaddr base_addr) {
@@ -31,11 +19,6 @@ void acpi_cpu_plug_cb(HotplugHandler *hotplug_dev, { } -void legacy_acpi_cpu_plug_cb (HotplugHandler *hotplug_dev, -AcpiCpuHotplug *g,DeviceState *dev,Error **errp) -{ -} -void acpi_cpu_unplug_cb (CPUHotplugState *cpu_st,DeviceState *dev,Error **errp) {
hw/acpi/cpu_hotplug.c deleted +0 -348
348 lines not on this page · raw message
hw/acpi/generic_event_device.c +1 -0
@@ -13,6 +13,7 @@#include "qapi/error.h" #include "hw/acpi/acpi.h" #include "hw/acpi/pcihp.h" +#include "hw/acpi/cpu.h" #include "hw/acpi/generic_event_device.h" #include "hw/pci/pci.h" #include "hw/core/irq.h"
hw/acpi/ich9.c +1 -0
@@ -40,6 +40,7 @@#include "hw/southbridge/ich9.h" #include "hw/mem/pc-dimm.h" #include "hw/mem/nvdimm.h" +#include "hw/acpi/pc-hotplug.h" static void ich9_pm_update_sci_fn (ACPIREGS *regs) {
hw/acpi/piix4.c +1 -1
@@ -33,7 +33,6 @@#include "system/xen.h" #include "qapi/error.h" #include "qemu/range.h" -#include "hw/acpi/cpu_hotplug.h" #include "hw/acpi/cpu.h" #include "hw/core/hotplug.h" #include "hw/mem/pc-dimm.h"
@@ -43,6 +42,7 @@#include "migration/vmstate.h" #include "hw/core/cpu.h" #include "qom/object.h" +#include "hw/acpi/pc-hotplug.h" #define GPE_BASE 0xafe0#define GPE_LEN 4
hw/i386/acpi-build.c +1 -0
@@ -35,6 +35,7 @@#include "hw/acpi/acpi-defs.h" #include "hw/acpi/acpi.h" #include "hw/acpi/cpu.h" +#include "hw/acpi/pc-hotplug.h" #include "hw/nvram/fw_cfg.h" #include "hw/acpi/bios-linker-loader.h" #include "hw/acpi/acpi_aml_interface.h"
hw/i386/pc.c +2 -1
@@ -48,7 +48,8 @@#include "hw/xen/xen.h" #include "qobject/qlist.h" #include "qemu/error-report.h" -#include "hw/acpi/cpu_hotplug.h" +#include "hw/acpi/acpi.h" +#include "hw/acpi/pc-hotplug.h" #include "acpi-build.h" #include "hw/mem/nvdimm.h" #include "hw/cxl/cxl_host.h"
hw/i386/x86-common.c +0 -1
@@ -36,7 +36,6 @@#include "hw/rtc/mc146818rtc.h" #include "target/i386/sev.h" -#include "hw/acpi/cpu_hotplug.h" #include "hw/core/irq.h" #include "hw/core/loader.h" #include "multiboot.h"
hw/acpi/meson.build +1 -1
@@ -6,7 +6,7 @@ acpi_ss.add(files( 'core.c', 'utils.c', )) -acpi_ss.add(when: 'CONFIG_ACPI_CPU_HOTPLUG', if_true: files('cpu.c', 'cpu_hotplug.c')) +acpi_ss.add(when: 'CONFIG_ACPI_CPU_HOTPLUG', if_true: files('cpu.c')) acpi_ss.add(when: 'CONFIG_ACPI_CPU_HOTPLUG', if_false: files('acpi-cpu-hotplug-stub.c')) acpi_ss.add(when: 'CONFIG_ACPI_MEMORY_HOTPLUG', if_true: files('memory_hotplug.c')) acpi_ss.add(when: 'CONFIG_ACPI_MEMORY_HOTPLUG', if_false: files('acpi-mem-hotplug-stub.c'))
Signature · 1 line
2.52.0
[PULL 19/41] hw/core/machine: Remove hw_compat_2_6[] array
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
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.
--- include/hw/core/boards.h | 3 --- hw/core/machine.c | 8 -------- 2 files changed, 11 deletions(-)
include/hw/core/boards.h +0 -3
@@ -885,7 +885,4 @@ 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
@@ -293,14 +293,6 @@ 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
2.52.0
[PULL 12/41] hw/nvram/fw_cfg: Rename fw_cfg_init_mem_wide() -> fw_cfg_init_mem_dma()
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
"wide" in fw_cfg_init_mem_wide() means "DMA support". Rename for clarity.
--- include/hw/nvram/fw_cfg.h | 6 +++--- hw/arm/virt.c | 2 +- hw/loongarch/fw_cfg.c | 4 ++-- hw/nvram/fw_cfg.c | 6 +++--- hw/riscv/virt.c | 4 ++-- 5 files changed, 11 insertions(+), 11 deletions(-)
include/hw/nvram/fw_cfg.h +3 -3
@@ -309,9 +309,9 @@ FWCfgState *fw_cfg_init_io_dma(uint32_t iobase, uint32_t dma_iobase, AddressSpace *dma_as);FWCfgState *fw_cfg_init_mem_nodma (hwaddr ctl_addr,hwaddr data_addr,unsigned data_width); -FWCfgState *fw_cfg_init_mem_wide (hwaddr ctl_addr, -hwaddr data_addr,uint32_t data_width, -hwaddr dma_addr,AddressSpace *dma_as); +FWCfgState *fw_cfg_init_mem_dma (hwaddr ctl_addr, +hwaddr data_addr,uint32_t data_width, +hwaddr dma_addr,AddressSpace *dma_as);FWCfgState *fw_cfg_find (void );bool fw_cfg_dma_enabled (void *opaque);
hw/arm/virt.c +1 -1
@@ -1412,7 +1412,7 @@ static FWCfgState *create_fw_cfg(const VirtMachineState *vms, AddressSpace *as)FWCfgState *fw_cfg;char *nodename; - fw_cfg =fw_cfg_init_mem_wide (base +8 , base,8 , base +16 , as); + fw_cfg =fw_cfg_init_mem_dma (base +8 , base,8 , base +16 , as);fw_cfg_add_i16 (fw_cfg,FW_CFG_NB_CPUS , (uint16_t )ms->smp.cpus); nodename =g_strdup_printf ("/fw-cfg@%" PRIx64, base);
hw/loongarch/fw_cfg.c +2 -2
@@ -23,8 +23,8 @@ FWCfgState *virt_fw_cfg_init(ram_addr_t ram_size, MachineState *ms)int max_cpus = ms->smp.max_cpus;int smp_cpus = ms->smp.cpus; - fw_cfg =fw_cfg_init_mem_wide (VIRT_FWCFG_BASE +8 ,VIRT_FWCFG_BASE ,8 , -VIRT_FWCFG_BASE +16 , &address_space_memory); + fw_cfg =fw_cfg_init_mem_dma (VIRT_FWCFG_BASE +8 ,VIRT_FWCFG_BASE ,8 , +VIRT_FWCFG_BASE +16 , &address_space_memory);fw_cfg_add_i16 (fw_cfg,FW_CFG_MAX_CPUS , (uint16_t )max_cpus);fw_cfg_add_i64 (fw_cfg,FW_CFG_RAM_SIZE , (uint64_t )ram_size);fw_cfg_add_i16 (fw_cfg,FW_CFG_NB_CPUS , (uint16_t )smp_cpus);
hw/nvram/fw_cfg.c +3 -3
@@ -1088,9 +1088,9 @@ static FWCfgState *fw_cfg_init_mem_internal(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) +FWCfgState *fw_cfg_init_mem_dma (hwaddr ctl_addr, +hwaddr data_addr,uint32_t data_width, +hwaddr dma_addr,AddressSpace *dma_as) {assert (dma_addr && dma_as); return fw_cfg_init_mem_internal(ctl_addr, data_addr, data_width,
hw/riscv/virt.c +2 -2
@@ -1274,8 +1274,8 @@ static FWCfgState *create_fw_cfg(const MachineState *ms, hwaddr base) {FWCfgState *fw_cfg; - fw_cfg =fw_cfg_init_mem_wide (base +8 , base,8 , base +16 , - &address_space_memory); + fw_cfg =fw_cfg_init_mem_dma (base +8 , base,8 , base +16 , + &address_space_memory);fw_cfg_add_i16 (fw_cfg,FW_CFG_NB_CPUS , (uint16_t )ms->smp.cpus);return fw_cfg;
Signature · 1 line
2.52.0
[PULL 15/41] hw/i386: Assume fw_cfg DMA is always enabled
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
Now all calls of x86 machines to fw_cfg_init_io_dma() pass DMA arguments, so the FWCfgState (FWCfgIoState) created by x86 machines enables DMA by default.
Although other callers of fw_cfg_init_io_dma() besides x86 also pass DMA arguments to create DMA-enabled FwCfgIoState, the "dma_enabled" property of FwCfgIoState cannot yet be removed, because Sun4u and Sun4v still create DMA-disabled FwCfgIoState (bypass fw_cfg_init_io_dma()) in sun4uv_init() (hw/sparc64/sun4u.c).
Maybe reusing fw_cfg_init_io_dma() for them would be a better choice, or adding fw_cfg_init_io_nodma(). However, before that, first simplify the handling of FwCfgState in x86.
Considering that FwCfgIoState in x86 enables DMA by default, remove the handling for DMA-disabled cases and replace DMA checks with assertions to ensure that the default DMA-enabled setting is not broken.
Then 'linuxboot.bin' isn't used anymore, and it will be removed in the next commit.
--- hw/i386/fw_cfg.c | 16 ++++++++-------- hw/i386/x86-common.c | 6 ++---- 2 files changed, 10 insertions(+), 12 deletions(-)
hw/i386/fw_cfg.c +8 -8
@@ -215,18 +215,18 @@ void fw_cfg_build_feature_control(MachineState *ms, FWCfgState *fw_cfg)#ifdef CONFIG_ACPI void fw_cfg_add_acpi_dsdt (Aml *scope,FWCfgState *fw_cfg) { +uint8_t io_size; +Aml *dev =aml_device ("FWCF" ); +Aml *crs =aml_resource_template (); +/* * when using port i/o, the 8-bit data register *always* overlaps * with half of the 16-bit control register. Hence, the total size -* of the i/o region used is FW_CFG_CTL_SIZE; when using DMA, the -* DMA control register is located at FW_CFG_DMA_IO_BASE + 4 +* of the i/o region used is FW_CFG_CTL_SIZE; And the DMA control +* register is located at FW_CFG_DMA_IO_BASE + 4 */ -Object *obj =OBJECT (fw_cfg); -uint8_t io_size =object_property_get_bool (obj,"dma_enabled" ,NULL ) ? -ROUND_UP (FW_CFG_CTL_SIZE ,4 ) +sizeof (dma_addr_t) : -FW_CFG_CTL_SIZE ; -Aml *dev =aml_device ("FWCF" ); -Aml *crs =aml_resource_template (); +assert (fw_cfg_dma_enabled (fw_cfg)); + io_size =ROUND_UP (FW_CFG_CTL_SIZE ,4 ) +sizeof (dma_addr_t);aml_append (dev,aml_name_decl ("_HID" ,aml_string ("QEMU0002" )));
hw/i386/x86-common.c +2 -4
@@ -1002,10 +1002,8 @@ void x86_load_linux(X86MachineState *x86ms, } option_rom[nb_option_roms].bootindex =0 ; - option_rom[nb_option_roms].name ="linuxboot.bin" ; -if (fw_cfg_dma_enabled (fw_cfg)) { - option_rom[nb_option_roms].name ="linuxboot_dma.bin" ; - } +assert (fw_cfg_dma_enabled (fw_cfg)); + option_rom[nb_option_roms].name ="linuxboot_dma.bin" ; nb_option_roms++; }
Signature · 1 line
2.52.0
[PULL 29/41] i386/cpu: Fix incorrect initializer in Diamond Rapids definition
In reply to Paolo BonziniFrom: Aidan Khoury <[email protected]>
The new Diamond Rapids x86 cpu model definition that was added in 7a6dd8bde1 has an unexpected comma in the .features[FEAT_VMX_EXIT_CTLS] subobject initializer, causing the prior initialization to be overridden. For this reason VMX_VM_EXIT_SAVE_DEBUG_CONTROLS | VMX_VM_EXIT_HOST_ADDR_SPACE_SIZE is not included.
Fix this by replacing the comma with the missing bitwise OR to properly combine all the flags into a single bitmask value.
--- target/i386/cpu.c | 2 +- roms/seabios-hppa | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-)
target/i386/cpu.c +1 -1
@@ -5579,7 +5579,7 @@ static const X86CPUDefinition builtin_x86_defs[] = {MSR_VMX_EPT_INVVPID_ALL_CONTEXT | MSR_VMX_EPT_INVVPID_SINGLE_CONTEXT_NOGLOBALS, .features[FEAT_VMX_EXIT_CTLS ] = -VMX_VM_EXIT_SAVE_DEBUG_CONTROLS |VMX_VM_EXIT_HOST_ADDR_SPACE_SIZE , +VMX_VM_EXIT_SAVE_DEBUG_CONTROLS |VMX_VM_EXIT_HOST_ADDR_SPACE_SIZE |VMX_VM_EXIT_LOAD_IA32_PERF_GLOBAL_CTRL |VMX_VM_EXIT_ACK_INTR_ON_EXIT |VMX_VM_EXIT_SAVE_IA32_PAT | VMX_VM_EXIT_LOAD_IA32_PAT | VMX_VM_EXIT_SAVE_IA32_EFER |
roms/seabios-hppa +1 -1
@@ -1 +1 @@ -Subproject commit b876684ec394856a8cc2523c57b3dc784dc119d4 +Subproject commit 3391c580960febcb9fa8f686f9666adaa462c349
Signature · 1 line
2.52.0
[PULL 31/41] target/i386/kvm: extract unrelated code out of kvm_x86_build_cpuid()
In reply to Paolo BonziniFrom: Dongli Zhang <[email protected]>
The initialization of 'has_architectural_pmu_version', 'num_architectural_pmu_gp_counters', and 'num_architectural_pmu_fixed_counters' is unrelated to the process of building the CPUID.
Extract them out of kvm_x86_build_cpuid().
In addition, use cpuid_find_entry() instead of cpu_x86_cpuid(), because CPUID has already been filled at this stage.
--- target/i386/kvm/kvm.c | 62 ++++++++++++++++++++++++------------------- 1 file changed, 35 insertions(+), 27 deletions(-)
target/i386/kvm/kvm.c +35 -27
@@ -1985,33 +1985,6 @@ uint32_t kvm_x86_build_cpuid(CPUX86State *env, struct kvm_cpuid_entry2 *entries, } } -if (limit >=0x0a ) { -uint32_t eax, edx; - -cpu_x86_cpuid (env,0x0a ,0 , &eax, &unused, &unused, &edx); - - has_architectural_pmu_version = eax &0xff ; -if (has_architectural_pmu_version >0 ) { - num_architectural_pmu_gp_counters = (eax &0xff00 ) >>8 ; - -/* Shouldn't be more than 32, since that's the number of bits -* available in EBX to tell us _which_ counters are available. -* Play it safe. -*/ -if (num_architectural_pmu_gp_counters >MAX_GP_COUNTERS ) { - num_architectural_pmu_gp_counters =MAX_GP_COUNTERS ; - } - -if (has_architectural_pmu_version >1 ) { - num_architectural_pmu_fixed_counters = edx &0x1f ; - -if (num_architectural_pmu_fixed_counters >MAX_FIXED_COUNTERS ) { - num_architectural_pmu_fixed_counters =MAX_FIXED_COUNTERS ; - } - } - } - } -cpu_x86_cpuid (env,0x80000000 ,0 , &limit, &unused, &unused, &unused);for (i =0x80000000 ; i <= limit; i++) {
@@ -2115,6 +2088,39 @@ int kvm_arch_pre_create_vcpu(CPUState *cpu, Error **errp)return 0 ; } +static void kvm_init_pmu_info (struct kvm_cpuid2 *cpuid) +{ +struct kvm_cpuid_entry2 *c; + + c =cpuid_find_entry (cpuid,0xa ,0 ); + +if (!c) { +return ; + } + + has_architectural_pmu_version = c->eax &0xff ; +if (has_architectural_pmu_version >0 ) { + num_architectural_pmu_gp_counters = (c->eax &0xff00 ) >>8 ; + +/* +* Shouldn't be more than 32, since that's the number of bits +* available in EBX to tell us _which_ counters are available. +* Play it safe. +*/ +if (num_architectural_pmu_gp_counters >MAX_GP_COUNTERS ) { + num_architectural_pmu_gp_counters =MAX_GP_COUNTERS ; + } + +if (has_architectural_pmu_version >1 ) { + num_architectural_pmu_fixed_counters = c->edx &0x1f ; + +if (num_architectural_pmu_fixed_counters >MAX_FIXED_COUNTERS ) { + num_architectural_pmu_fixed_counters =MAX_FIXED_COUNTERS ; + } + } + } +} +int kvm_arch_init_vcpu (CPUState *cs) {struct {
@@ -2305,6 +2311,8 @@ int kvm_arch_init_vcpu(CPUState *cs) cpuid_i =kvm_x86_build_cpuid (env, cpuid_data.entries, cpuid_i); cpuid_data.cpuid.nent = cpuid_i; +kvm_init_pmu_info (&cpuid_data.cpuid); +if (x86_cpu_family (env->cpuid_version) >=6 && (env->features[FEAT_1_EDX ] & (CPUID_MCE |CPUID_MCA )) == (CPUID_MCE |CPUID_MCA )) {
Signature · 1 line
2.52.0
[PULL 36/41] qdev: make release_string() idempotent
In reply to Paolo BonziniFrom: Marc-André Lureau <[email protected]>
So it can eventually be called multiple times safely.
--- hw/core/qdev-properties.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
hw/core/qdev-properties.c +2 -1
@@ -551,7 +551,8 @@ const PropertyInfo qdev_prop_usize = {static void release_string (Object *obj,const char *name,void *opaque) {const Property *prop = opaque; -g_free (*(char **)object_field_prop_ptr (obj, prop)); + +g_clear_pointer ((char **)object_field_prop_ptr (obj, prop), g_free); } static void get_string(Object *obj, Visitor *v, const char *name,
Signature · 1 line
2.52.0
[PULL 39/41] tests: add /qdev/free-properties test
In reply to Paolo BonziniFrom: Marc-André Lureau <[email protected]>
Add a simple qdev test to check that allocated properties get freed with the object. This test exhibited array leaks before the fixes.
--- tests/unit/test-qdev.c | 96 ++++++++++++++++++++++++++++++++++++++++++ tests/unit/meson.build | 1 + 2 files changed, 97 insertions(+) create mode 100644 tests/unit/test-qdev.c
tests/unit/test-qdev.c new file +96 -0
@@ -0,0 +1,96 @@ +#include "qemu/osdep.h" +#include "hw/core/qdev-properties.h" +#include "qom/object.h" +#include "qapi/error.h" +#include "qapi/visitor.h" + + +#define TYPE_MY_DEV "my-dev" +typedef struct MyDev MyDev ; +DECLARE_INSTANCE_CHECKER (MyDev,STATIC_TYPE , + TYPE_MY_DEV) + +struct MyDev { +DeviceState parent_obj; + +uint32_t prop_u32; +char *prop_string; +uint32_t *prop_array_u32; +uint32_t prop_array_u32_nb; +}; + +static const Property my_dev_props[] = { +DEFINE_PROP_UINT32 ("u32" , MyDev, prop_u32,100 ), +DEFINE_PROP_STRING ("string" , MyDev, prop_string), +DEFINE_PROP_ARRAY ("array-u32" , MyDev, prop_array_u32_nb, prop_array_u32, + qdev_prop_uint32, uint32_t), +}; + +static void my_dev_class_init (ObjectClass *oc,const void *data) +{ +DeviceClass *dc =DEVICE_CLASS (oc); + + dc->realize =NULL ; +device_class_set_props (dc, my_dev_props); +} + +static const TypeInfo my_dev_type_info = { + .name =TYPE_MY_DEV , + .parent =TYPE_DEVICE , + .instance_size =sizeof (MyDev), + .class_init = my_dev_class_init, +}; + +/* +* Initialize a fake machine, being prepared for future tests. +* +* Realization of anonymous qdev (with no parent object) requires both +* the machine object and its "unattached" container to be at least present. +*/ +static void test_init_machine (void ) +{ +/* This is a fake machine - it doesn't need to be a machine object */ +Object *machine =object_property_add_new_container ( +object_get_root (),"machine" ); + +/* This container must exist for anonymous qdevs to realize() */ +object_property_add_new_container (machine,"unattached" ); +} + +static void test_qdev_free_properties (void ) +{ +MyDev *mt; + + mt =STATIC_TYPE (object_new (TYPE_MY_DEV )); +object_set_props (OBJECT (mt), &error_fatal, +"string" ,"something" , +"array-u32" ,"12,13" , +NULL ); +qdev_realize (DEVICE (mt),NULL , &error_fatal); + + g_assert_cmpuint(mt->prop_u32, ==, 100); + g_assert_cmpstr(mt->prop_string, ==, "something"); + g_assert_cmpuint(mt->prop_array_u32_nb, ==, 2); + g_assert_cmpuint(mt->prop_array_u32[0], ==, 12); + g_assert_cmpuint(mt->prop_array_u32[1], ==, 13); + +object_unparent (OBJECT (mt)); +object_unref (mt); +} + + +int main (int argc,char **argv) +{ +g_test_init (&argc, &argv,NULL ); + +module_call_init (MODULE_INIT_QOM ); +type_register_static (&my_dev_type_info); +test_init_machine (); + +g_test_add_func ("/qdev/free-properties" , + test_qdev_free_properties); + +g_test_run (); + +return 0 ; +}
tests/unit/meson.build +1 -0
@@ -156,6 +156,7 @@ if have_system 'test-qdev-global-props': [qom, hwcore] } endif + tests += {'test-qdev': [qom, hwcore]} endif if have_ga and host_os == 'linux'
Signature · 1 line
2.52.0
[PULL 13/41] hw/i386/x86: Remove X86MachineClass::fwcfg_dma_enabled field
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
The 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.
'multiboot.bin' isn't used anymore, we'll remove it in the next commit.
--- 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
@@ -30,8 +30,6 @@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
@@ -640,7 +640,6 @@ 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);
@@ -674,8 +673,6 @@ 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
@@ -645,7 +645,6 @@ 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;
@@ -1004,7 +1003,7 @@ 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
2.52.0
[PULL 03/41] tests/acpi: Allow DSDT table change for x86 machines
In reply to Paolo BonziniFrom: Igor Mammedov <[email protected]>
Before dropping legacy CPU hotplug code, mark and allow the affected ACPI tables, to avoid breaking ACPI table testing.
--- tests/qtest/bios-tables-test-allowed-diff.h | 42 +++++++++++++++++++++ 1 file changed, 42 insertions(+)
tests/qtest/bios-tables-test-allowed-diff.h +42 -0
@@ -1 +1,43 @@/* List of comma-separated changed AML files to ignore */ +"tests/data/acpi/x86/pc/DSDT" , +"tests/data/acpi/x86/pc/DSDT.bridge" , +"tests/data/acpi/x86/pc/DSDT.ipmikcs" , +"tests/data/acpi/x86/pc/DSDT.cphp" , +"tests/data/acpi/x86/pc/DSDT.numamem" , +"tests/data/acpi/x86/pc/DSDT.nohpet" , +"tests/data/acpi/x86/pc/DSDT.memhp" , +"tests/data/acpi/x86/pc/DSDT.dimmpxm" , +"tests/data/acpi/x86/pc/DSDT.acpihmat" , +"tests/data/acpi/x86/pc/DSDT.acpierst" , +"tests/data/acpi/x86/pc/DSDT.roothp" , +"tests/data/acpi/x86/pc/DSDT.hpbridge" , +"tests/data/acpi/x86/pc/DSDT.hpbrroot" , +"tests/data/acpi/x86/q35/DSDT" , +"tests/data/acpi/x86/q35/DSDT.tis.tpm2" , +"tests/data/acpi/x86/q35/DSDT.tis.tpm12" , +"tests/data/acpi/x86/q35/DSDT.bridge" , +"tests/data/acpi/x86/q35/DSDT.noacpihp" , +"tests/data/acpi/x86/q35/DSDT.multi-bridge" , +"tests/data/acpi/x86/q35/DSDT.ipmibt" , +"tests/data/acpi/x86/q35/DSDT.cphp" , +"tests/data/acpi/x86/q35/DSDT.numamem" , +"tests/data/acpi/x86/q35/DSDT.nohpet" , +"tests/data/acpi/x86/q35/DSDT.acpihmat-noinitiator" , +"tests/data/acpi/x86/q35/DSDT.acpihmat-generic-x" , +"tests/data/acpi/x86/q35/DSDT.memhp" , +"tests/data/acpi/x86/q35/DSDT.dimmpxm" , +"tests/data/acpi/x86/q35/DSDT.acpihmat" , +"tests/data/acpi/x86/q35/DSDT.mmio64" , +"tests/data/acpi/x86/q35/DSDT.acpierst" , +"tests/data/acpi/x86/q35/DSDT.applesmc" , +"tests/data/acpi/x86/q35/DSDT.pvpanic-isa" , +"tests/data/acpi/x86/q35/DSDT.ivrs" , +"tests/data/acpi/x86/q35/DSDT.type4-count" , +"tests/data/acpi/x86/q35/DSDT.core-count" , +"tests/data/acpi/x86/q35/DSDT.core-count2" , +"tests/data/acpi/x86/q35/DSDT.thread-count" , +"tests/data/acpi/x86/q35/DSDT.thread-count2" , +"tests/data/acpi/x86/q35/DSDT.viot" , +"tests/data/acpi/x86/q35/DSDT.cxl" , +"tests/data/acpi/x86/q35/DSDT.ipmismbus" , +"tests/data/acpi/x86/q35/DSDT.xapic",
Signature · 1 line
2.52.0
[PULL 34/41] target/i386/kvm: support perfmon-v2 for reset
In reply to Paolo BonziniFrom: Dongli Zhang <[email protected]>
Since perfmon-v2, the AMD PMU supports additional registers. This update includes get/put functionality for these extra registers.
Similar to the implementation in KVM:
- MSR_CORE_PERF_GLOBAL_STATUS and MSR_AMD64_PERF_CNTR_GLOBAL_STATUS both use env->msr_global_status.
- MSR_CORE_PERF_GLOBAL_CTRL and MSR_AMD64_PERF_CNTR_GLOBAL_CTL both use env->msr_global_ctrl.
- MSR_CORE_PERF_GLOBAL_OVF_CTRL and MSR_AMD64_PERF_CNTR_GLOBAL_STATUS_CLR both use env->msr_global_ovf_ctrl.
No changes are needed for vmstate_msr_architectural_pmu or pmu_enable_needed().
--- target/i386/cpu.h | 4 ++++ target/i386/kvm/kvm.c | 48 +++++++++++++++++++++++++++++++++++-------- 2 files changed, 43 insertions(+), 9 deletions(-)
target/i386/cpu.h +4 -0
@@ -506,6 +506,10 @@ typedef enum X86Seg {#define MSR_CORE_PERF_GLOBAL_CTRL 0x38f#define MSR_CORE_PERF_GLOBAL_OVF_CTRL 0x390 +#define MSR_AMD64_PERF_CNTR_GLOBAL_STATUS 0xc0000300 +#define MSR_AMD64_PERF_CNTR_GLOBAL_CTL 0xc0000301 +#define MSR_AMD64_PERF_CNTR_GLOBAL_STATUS_CLR 0xc0000302 +#define MSR_K7_EVNTSEL0 0xc0010000#define MSR_K7_PERFCTR0 0xc0010004#define MSR_F15H_PERF_CTL0 0xc0010200
target/i386/kvm/kvm.c +39 -9
95 lines not on this page · raw message
Signature · 1 line
2.52.0
[PULL 18/41] hw/intc/apic: Remove APICCommonState::legacy_instance_id field
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
The 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
@@ -187,7 +187,6 @@ 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
@@ -276,9 +276,6 @@ 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 );
@@ -395,8 +392,6 @@ 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
2.52.0
[PULL 27/41] hw/virtio/virtio-pci: Remove VirtIOPCIProxy::ignore_backend_features field
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
The 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
@@ -150,7 +150,6 @@ struct VirtIOPCIProxy {uint16_t last_pcie_cap_offset;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
@@ -2040,8 +2040,7 @@ 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) {
@@ -2443,8 +2442,6 @@ 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
2.52.0
[PULL 33/41] target/i386/kvm: reset AMD PMU registers during VM reset
In reply to Paolo BonziniFrom: Dongli Zhang <[email protected]>
QEMU uses the kvm_get_msrs() function to save Intel PMU registers from KVM and kvm_put_msrs() to restore them to KVM. However, there is no support for AMD PMU registers. Currently, pmu_version and num_pmu_gp_counters are initialized based on cpuid(0xa), which does not apply to AMD processors. For AMD CPUs, prior to PerfMonV2, the number of general-purpose registers is determined based on the CPU version.
To address this issue, we need to add support for AMD PMU registers. Without this support, the following problems can arise:
1. If the VM is reset (e.g., via QEMU system_reset or VM kdump/kexec) while running "perf top", the PMU registers are not disabled properly.
2. Despite x86_cpu_reset() resetting many registers to zero, kvm_put_msrs() does not handle AMD PMU registers, causing some PMU events to remain enabled in KVM.
3. The KVM kvm_pmc_speculative_in_use() function consistently returns true, preventing the reclamation of these events. Consequently, the kvm_pmc->perf_event remains active.
4. After a reboot, the VM kernel may report the following error:
[ 0.092011] Performance Events: Fam17h+ core perfctr, Broken BIOS detected, complain to your hardware vendor. [ 0.092023] [Firmware Bug]: the BIOS has corrupted hw-PMU resources (MSR c0010200 is 530076)
5. In the worst case, the active kvm_pmc->perf_event may inject unknown NMIs randomly into the VM kernel:
[...] Uhhuh. NMI received for unknown reason 30 on CPU 0.
To resolve these issues, we propose resetting AMD PMU registers during the VM reset process.
--- target/i386/cpu.h | 12 +++ target/i386/kvm/kvm.c | 168 +++++++++++++++++++++++++++++++++++++++++- 2 files changed, 176 insertions(+), 4 deletions(-)
target/i386/cpu.h +12 -0
@@ -506,6 +506,14 @@ typedef enum X86Seg {#define MSR_CORE_PERF_GLOBAL_CTRL 0x38f#define MSR_CORE_PERF_GLOBAL_OVF_CTRL 0x390 +#define MSR_K7_EVNTSEL0 0xc0010000 +#define MSR_K7_PERFCTR0 0xc0010004 +#define MSR_F15H_PERF_CTL0 0xc0010200 +#define MSR_F15H_PERF_CTR0 0xc0010201 + +#define AMD64_NUM_COUNTERS 4 +#define AMD64_NUM_COUNTERS_CORE 6 +#define MSR_MC0_CTL 0x400#define MSR_MC0_STATUS 0x401#define MSR_MC0_ADDR 0x402
@@ -1737,6 +1745,10 @@ typedef struct {#endif #define MAX_FIXED_COUNTERS 3 +/* +* This formula is based on Intel's MSR. The current size also meets AMD's +* needs. +*/ #define MAX_GP_COUNTERS (MSR_IA32_PERF_STATUS - MSR_P6_EVNTSEL0)#define NB_OPMASK_REGS 8
target/i386/kvm/kvm.c +164 -4
216 lines not on this page · raw message
Signature · 1 line
2.52.0
[PULL 38/41] qdev: make release_tpm() idempotent
In reply to Paolo BonziniFrom: Marc-André Lureau <[email protected]>
So it can eventually be called multiple times safely.
--- backends/tpm/tpm_util.c | 1 + 1 file changed, 1 insertion(+)
backends/tpm/tpm_util.c +1 -0
@@ -72,6 +72,7 @@ static void release_tpm(Object *obj, const char *name, void *opaque)if (*be) {tpm_backend_reset (*be); + *be =NULL ; } }
Signature · 1 line
2.52.0
[PULL 41/41] target/i386/mshv: remove unused optimization of gva=>gpa translation
In reply to Paolo BonziniFrom: Magnus Kulke <[email protected]>
The removed parameters are remnants of a prior attempt to optimize gva=>gpa translation. Currently there is only one call site and it's not using it. So we can remove it as dead code.
--- target/i386/mshv/mshv-cpu.c | 20 ++++++++------------ 1 file changed, 8 insertions(+), 12 deletions(-)
target/i386/mshv/mshv-cpu.c +8 -12
@@ -1336,23 +1336,19 @@ static int read_memory(const CPUState *cpu, uint64_t initial_gva,return 0 ; } -static int write_memory (const CPUState *cpu,uint64_t initial_gva, -uint64_t initial_gpa,uint64_t gva,const uint8_t *data, +static int write_memory (const CPUState *cpu,uint64_t gva,const uint8_t *data,size_t len) {int ret;uint64_t gpa, flags; -if (gva == initial_gva) { - gpa = initial_gpa; - }else { - flags =HV_TRANSLATE_GVA_VALIDATE_WRITE ; - ret =translate_gva (cpu, gva, &gpa, flags); -if (ret <0 ) { -error_report ("failed to translate gva to gpa" ); -return -1 ; - } + flags =HV_TRANSLATE_GVA_VALIDATE_WRITE ; + ret =translate_gva (cpu, gva, &gpa, flags); +if (ret <0 ) { +error_report ("failed to translate gva to gpa" ); +return -1 ; } + ret =mshv_guest_mem_write (gpa, data, len, false);if (ret !=MEMTX_OK ) {error_report ("failed to write to mmio" );
@@ -1407,7 +1403,7 @@ static int handle_pio_str_read(CPUState *cpu,for (size_t i =0 ; i < repeat; i++) {pio_read (port, data, len, false); - ret =write_memory (cpu,0 ,0 , dst, data, len); + ret =write_memory (cpu, dst, data, len);if (ret <0 ) {error_report ("Failed to write memory" );return -1 ;
Signature · 1 line
2.52.0
[PULL 24/41] hw/audio/pcspk: Remove PCSpkState::migrate field
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
The 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
@@ -57,7 +57,6 @@ struct PCSpkState {unsigned int play_pos;uint8_t data_on;uint8_t dummy_refresh_clock; -bool migrate; };static const char *s_spk ="pcspk" ;
@@ -202,18 +201,10 @@ static void pcspk_realizefn(DeviceState *dev, Error **errp) } } -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),
@@ -224,7 +215,6 @@ static const VMStateDescription vmstate_spk = {static const Property pcspk_properties[] = {DEFINE_AUDIO_PROPERTIES (PCSpkState, audio_be),DEFINE_PROP_UINT32 ("iobase" , PCSpkState, iobase,0x61 ), -DEFINE_PROP_BOOL ("migrate" , PCSpkState, migrate, true), DEFINE_PROP_LINK("pit", PCSpkState, pit, TYPE_PIT_COMMON, PITCommonState *), };
Signature · 1 line
2.52.0
[PULL 08/41] tests/acpi: Update DSDT tables for pc & q35 machines
In reply to Paolo BonziniFrom: Zhao Liu <[email protected]>
Now the legacy cpu hotplug way has gone away, and there's no _INIT method in DSDT table for modern cpu hotplug support.
Update DSDT tables for pc machine, and_INIT methods are removed from DSDT tables:
- Method (_INI, 0, Serialized) // _INI: Initialize
- {
- CSEL = Zero
- }
--- tests/qtest/bios-tables-test-allowed-diff.h | 42 ------------------ tests/data/acpi/x86/pc/DSDT | Bin 8611 -> 8598 bytes tests/data/acpi/x86/pc/DSDT.acpierst | Bin 8522 -> 8509 bytes tests/data/acpi/x86/pc/DSDT.acpihmat | Bin 9936 -> 9923 bytes tests/data/acpi/x86/pc/DSDT.bridge | Bin 15482 -> 15469 bytes tests/data/acpi/x86/pc/DSDT.cphp | Bin 9075 -> 9062 bytes tests/data/acpi/x86/pc/DSDT.dimmpxm | Bin 10265 -> 10252 bytes tests/data/acpi/x86/pc/DSDT.hpbridge | Bin 8562 -> 8549 bytes tests/data/acpi/x86/pc/DSDT.hpbrroot | Bin 5100 -> 5087 bytes tests/data/acpi/x86/pc/DSDT.ipmikcs | Bin 8683 -> 8670 bytes tests/data/acpi/x86/pc/DSDT.memhp | Bin 9970 -> 9957 bytes tests/data/acpi/x86/pc/DSDT.nohpet | Bin 8469 -> 8456 bytes tests/data/acpi/x86/pc/DSDT.numamem | Bin 8617 -> 8604 bytes tests/data/acpi/x86/pc/DSDT.roothp | Bin 12404 -> 12391 bytes tests/data/acpi/x86/q35/DSDT | Bin 8440 -> 8427 bytes tests/data/acpi/x86/q35/DSDT.acpierst | Bin 8457 -> 8444 bytes tests/data/acpi/x86/q35/DSDT.acpihmat | Bin 9765 -> 9752 bytes .../data/acpi/x86/q35/DSDT.acpihmat-generic-x | Bin 12650 -> 12637 bytes .../acpi/x86/q35/DSDT.acpihmat-noinitiator | Bin 8719 -> 8706 bytes tests/data/acpi/x86/q35/DSDT.applesmc | Bin 8486 -> 8473 bytes tests/data/acpi/x86/q35/DSDT.bridge | Bin 12053 -> 12040 bytes tests/data/acpi/x86/q35/DSDT.core-count | Bin 12998 -> 12985 bytes tests/data/acpi/x86/q35/DSDT.core-count2 | Bin 33855 -> 33842 bytes tests/data/acpi/x86/q35/DSDT.cphp | Bin 8904 -> 8891 bytes tests/data/acpi/x86/q35/DSDT.cxl | Bin 13231 -> 13218 bytes tests/data/acpi/x86/q35/DSDT.dimmpxm | Bin 10094 -> 10081 bytes tests/data/acpi/x86/q35/DSDT.ipmibt | Bin 8515 -> 8502 bytes tests/data/acpi/x86/q35/DSDT.ipmismbus | Bin 8528 -> 8515 bytes tests/data/acpi/x86/q35/DSDT.ivrs | Bin 8457 -> 8444 bytes tests/data/acpi/x86/q35/DSDT.memhp | Bin 9799 -> 9786 bytes tests/data/acpi/x86/q35/DSDT.mmio64 | Bin 9570 -> 9557 bytes tests/data/acpi/x86/q35/DSDT.multi-bridge | Bin 13293 -> 13280 bytes tests/data/acpi/x86/q35/DSDT.noacpihp | Bin 8302 -> 8289 bytes tests/data/acpi/x86/q35/DSDT.nohpet | Bin 8298 -> 8285 bytes tests/data/acpi/x86/q35/DSDT.numamem | Bin 8446 -> 8433 bytes tests/data/acpi/x86/q35/DSDT.pvpanic-isa | Bin 8541 -> 8528 bytes tests/data/acpi/x86/q35/DSDT.thread-count | Bin 12998 -> 12985 bytes tests/data/acpi/x86/q35/DSDT.thread-count2 | Bin 33855 -> 33842 bytes tests/data/acpi/x86/q35/DSDT.tis.tpm12 | Bin 9046 -> 9033 bytes tests/data/acpi/x86/q35/DSDT.tis.tpm2 | Bin 9072 -> 9059 bytes tests/data/acpi/x86/q35/DSDT.type4-count | Bin 18674 -> 18661 bytes tests/data/acpi/x86/q35/DSDT.viot | Bin 14697 -> 14684 bytes tests/data/acpi/x86/q35/DSDT.xapic | Bin 35803 -> 35790 bytes 43 files changed, 42 deletions(-)
tests/qtest/bios-tables-test-allowed-diff.h +0 -42
@@ -1,43 +1 @@/* List of comma-separated changed AML files to ignore */ -"tests/data/acpi/x86/pc/DSDT" , -"tests/data/acpi/x86/pc/DSDT.bridge" , -"tests/data/acpi/x86/pc/DSDT.ipmikcs" , -"tests/data/acpi/x86/pc/DSDT.cphp" , -"tests/data/acpi/x86/pc/DSDT.numamem" , -"tests/data/acpi/x86/pc/DSDT.nohpet" , -"tests/data/acpi/x86/pc/DSDT.memhp" , -"tests/data/acpi/x86/pc/DSDT.dimmpxm" , -"tests/data/acpi/x86/pc/DSDT.acpihmat" , -"tests/data/acpi/x86/pc/DSDT.acpierst" , -"tests/data/acpi/x86/pc/DSDT.roothp" , -"tests/data/acpi/x86/pc/DSDT.hpbridge" , -"tests/data/acpi/x86/pc/DSDT.hpbrroot" , -"tests/data/acpi/x86/q35/DSDT" , -"tests/data/acpi/x86/q35/DSDT.tis.tpm2" , -"tests/data/acpi/x86/q35/DSDT.tis.tpm12" , -"tests/data/acpi/x86/q35/DSDT.bridge" , -"tests/data/acpi/x86/q35/DSDT.noacpihp" , -"tests/data/acpi/x86/q35/DSDT.multi-bridge" , -"tests/data/acpi/x86/q35/DSDT.ipmibt" , -"tests/data/acpi/x86/q35/DSDT.cphp" , -"tests/data/acpi/x86/q35/DSDT.numamem" , -"tests/data/acpi/x86/q35/DSDT.nohpet" , -"tests/data/acpi/x86/q35/DSDT.acpihmat-noinitiator" , -"tests/data/acpi/x86/q35/DSDT.acpihmat-generic-x" , -"tests/data/acpi/x86/q35/DSDT.memhp" , -"tests/data/acpi/x86/q35/DSDT.dimmpxm" , -"tests/data/acpi/x86/q35/DSDT.acpihmat" , -"tests/data/acpi/x86/q35/DSDT.mmio64" , -"tests/data/acpi/x86/q35/DSDT.acpierst" , -"tests/data/acpi/x86/q35/DSDT.applesmc" , -"tests/data/acpi/x86/q35/DSDT.pvpanic-isa" , -"tests/data/acpi/x86/q35/DSDT.ivrs" , -"tests/data/acpi/x86/q35/DSDT.type4-count" , -"tests/data/acpi/x86/q35/DSDT.core-count" , -"tests/data/acpi/x86/q35/DSDT.core-count2" , -"tests/data/acpi/x86/q35/DSDT.thread-count" , -"tests/data/acpi/x86/q35/DSDT.thread-count2" , -"tests/data/acpi/x86/q35/DSDT.viot" , -"tests/data/acpi/x86/q35/DSDT.cxl" , -"tests/data/acpi/x86/q35/DSDT.ipmismbus" , -"tests/data/acpi/x86/q35/DSDT.xapic",
tests/data/acpi/x86/pc/DSDT binary +0 -0
tests/data/acpi/x86/pc/DSDT.acpierst binary +0 -0
tests/data/acpi/x86/pc/DSDT.acpihmat binary +0 -0
tests/data/acpi/x86/pc/DSDT.bridge binary +0 -0
tests/data/acpi/x86/pc/DSDT.cphp binary +0 -0
tests/data/acpi/x86/pc/DSDT.dimmpxm binary +0 -0
tests/data/acpi/x86/pc/DSDT.hpbridge binary +0 -0
tests/data/acpi/x86/pc/DSDT.hpbrroot binary +0 -0
tests/data/acpi/x86/pc/DSDT.ipmikcs binary +0 -0
tests/data/acpi/x86/pc/DSDT.memhp binary +0 -0
tests/data/acpi/x86/pc/DSDT.nohpet binary +0 -0
tests/data/acpi/x86/pc/DSDT.numamem binary +0 -0
tests/data/acpi/x86/pc/DSDT.roothp binary +0 -0
tests/data/acpi/x86/q35/DSDT binary +0 -0
tests/data/acpi/x86/q35/DSDT.acpierst binary +0 -0
tests/data/acpi/x86/q35/DSDT.acpihmat binary +0 -0
tests/data/acpi/x86/q35/DSDT.acpihmat-generic-x binary +0 -0
tests/data/acpi/x86/q35/DSDT.acpihmat-noinitiator binary +0 -0
tests/data/acpi/x86/q35/DSDT.applesmc binary +0 -0
tests/data/acpi/x86/q35/DSDT.bridge binary +0 -0
tests/data/acpi/x86/q35/DSDT.core-count binary +0 -0
tests/data/acpi/x86/q35/DSDT.core-count2 binary +0 -0
tests/data/acpi/x86/q35/DSDT.cphp binary +0 -0
tests/data/acpi/x86/q35/DSDT.cxl binary +0 -0
tests/data/acpi/x86/q35/DSDT.dimmpxm binary +0 -0
tests/data/acpi/x86/q35/DSDT.ipmibt binary +0 -0
tests/data/acpi/x86/q35/DSDT.ipmismbus binary +0 -0
tests/data/acpi/x86/q35/DSDT.ivrs binary +0 -0
tests/data/acpi/x86/q35/DSDT.memhp binary +0 -0
tests/data/acpi/x86/q35/DSDT.mmio64 binary +0 -0
tests/data/acpi/x86/q35/DSDT.multi-bridge binary +0 -0
tests/data/acpi/x86/q35/DSDT.noacpihp binary +0 -0
tests/data/acpi/x86/q35/DSDT.nohpet binary +0 -0
tests/data/acpi/x86/q35/DSDT.numamem binary +0 -0
tests/data/acpi/x86/q35/DSDT.pvpanic-isa binary +0 -0
tests/data/acpi/x86/q35/DSDT.thread-count binary +0 -0
tests/data/acpi/x86/q35/DSDT.thread-count2 binary +0 -0
tests/data/acpi/x86/q35/DSDT.tis.tpm12 binary +0 -0
tests/data/acpi/x86/q35/DSDT.tis.tpm2 binary +0 -0
tests/data/acpi/x86/q35/DSDT.type4-count binary +0 -0
tests/data/acpi/x86/q35/DSDT.viot binary +0 -0
tests/data/acpi/x86/q35/DSDT.xapic binary +0 -0
Signature · 1 line
2.52.0
[PULL 32/41] target/i386/kvm: rename architectural PMU variables
In reply to Paolo BonziniFrom: Dongli Zhang <[email protected]>
AMD does not have what is commonly referred to as an architectural PMU. Therefore, we need to rename the following variables to be applicable for both Intel and AMD:
- has_architectural_pmu_version
- num_architectural_pmu_gp_counters
- num_architectural_pmu_fixed_counters
For Intel processors, the meaning of pmu_version remains unchanged.
For AMD processors:
pmu_version == 1 corresponds to versions before AMD PerfMonV2. pmu_version == 2 corresponds to AMD PerfMonV2.
--- target/i386/kvm/kvm.c | 49 ++++++++++++++++++++++++------------------- 1 file changed, 28 insertions(+), 21 deletions(-)
target/i386/kvm/kvm.c +28 -21
@@ -166,9 +166,16 @@ static bool has_msr_perf_capabs;static bool has_msr_pkrs;static bool has_msr_hwcr; -static uint32_t has_architectural_pmu_version; -static uint32_t num_architectural_pmu_gp_counters; -static uint32_t num_architectural_pmu_fixed_counters; +/* +* For Intel processors, the meaning is the architectural PMU version +* number. +* +* For AMD processors: 1 corresponds to the prior versions, and 2 +* corresponds to AMD PerfMonV2. +*/ +static uint32_t pmu_version; +static uint32_t num_pmu_gp_counters; +static uint32_t num_pmu_fixed_counters;static int has_xsave2;static int has_xcrs;
@@ -2098,24 +2105,24 @@ static void kvm_init_pmu_info(struct kvm_cpuid2 *cpuid)return ; } - has_architectural_pmu_version = c->eax &0xff ; -if (has_architectural_pmu_version >0 ) { - num_architectural_pmu_gp_counters = (c->eax &0xff00 ) >>8 ; + pmu_version = c->eax &0xff ; +if (pmu_version >0 ) { + num_pmu_gp_counters = (c->eax &0xff00 ) >>8 ;/* * Shouldn't be more than 32, since that's the number of bits * available in EBX to tell us _which_ counters are available. * Play it safe. */ -if (num_architectural_pmu_gp_counters >MAX_GP_COUNTERS ) { - num_architectural_pmu_gp_counters =MAX_GP_COUNTERS ; +if (num_pmu_gp_counters >MAX_GP_COUNTERS ) { + num_pmu_gp_counters =MAX_GP_COUNTERS ; } -if (has_architectural_pmu_version >1 ) { - num_architectural_pmu_fixed_counters = c->edx &0x1f ; +if (pmu_version >1 ) { + num_pmu_fixed_counters = c->edx &0x1f ; -if (num_architectural_pmu_fixed_counters >MAX_FIXED_COUNTERS ) { - num_architectural_pmu_fixed_counters =MAX_FIXED_COUNTERS ; +if (num_pmu_fixed_counters >MAX_FIXED_COUNTERS ) { + num_pmu_fixed_counters =MAX_FIXED_COUNTERS ; } } }
@@ -4086,25 +4093,25 @@ static int kvm_put_msrs(X86CPU *cpu, KvmPutState level)kvm_msr_entry_add (cpu,MSR_KVM_POLL_CONTROL , env->poll_control_msr); } -if (has_architectural_pmu_version >0 ) { -if (has_architectural_pmu_version >1 ) { +if (pmu_version >0 ) { +if (pmu_version >1 ) {/* Stop the counter. */ kvm_msr_entry_add (cpu,MSR_CORE_PERF_FIXED_CTR_CTRL ,0 );kvm_msr_entry_add (cpu,MSR_CORE_PERF_GLOBAL_CTRL ,0 ); }/* Set the counter values. */ -for (i =0 ; i < num_architectural_pmu_fixed_counters; i++) { +for (i =0 ; i < num_pmu_fixed_counters; i++) {kvm_msr_entry_add (cpu,MSR_CORE_PERF_FIXED_CTR0 + i, env->msr_fixed_counters[i]); } -for (i =0 ; i < num_architectural_pmu_gp_counters; i++) { +for (i =0 ; i < num_pmu_gp_counters; i++) {kvm_msr_entry_add (cpu,MSR_P6_PERFCTR0 + i, env->msr_gp_counters[i]);kvm_msr_entry_add (cpu,MSR_P6_EVNTSEL0 + i, env->msr_gp_evtsel[i]); } -if (has_architectural_pmu_version >1 ) { +if (pmu_version >1 ) {kvm_msr_entry_add (cpu,MSR_CORE_PERF_GLOBAL_STATUS , env->msr_global_status); kvm_msr_entry_add(cpu, MSR_CORE_PERF_GLOBAL_OVF_CTRL,
@@ -4621,17 +4628,17 @@ static int kvm_get_msrs(X86CPU *cpu)if (env->features[FEAT_KVM ] &CPUID_KVM_POLL_CONTROL ) {kvm_msr_entry_add (cpu,MSR_KVM_POLL_CONTROL ,1 ); } -if (has_architectural_pmu_version >0 ) { -if (has_architectural_pmu_version >1 ) { +if (pmu_version >0 ) { +if (pmu_version >1 ) {kvm_msr_entry_add (cpu,MSR_CORE_PERF_FIXED_CTR_CTRL ,0 );kvm_msr_entry_add (cpu,MSR_CORE_PERF_GLOBAL_CTRL ,0 );kvm_msr_entry_add (cpu,MSR_CORE_PERF_GLOBAL_STATUS ,0 );kvm_msr_entry_add (cpu,MSR_CORE_PERF_GLOBAL_OVF_CTRL ,0 ); } -for (i =0 ; i < num_architectural_pmu_fixed_counters; i++) { +for (i =0 ; i < num_pmu_fixed_counters; i++) {kvm_msr_entry_add (cpu,MSR_CORE_PERF_FIXED_CTR0 + i,0 ); } -for (i =0 ; i < num_architectural_pmu_gp_counters; i++) { +for (i =0 ; i < num_pmu_gp_counters; i++) {kvm_msr_entry_add (cpu,MSR_P6_PERFCTR0 + i,0 );kvm_msr_entry_add (cpu,MSR_P6_EVNTSEL0 + i,0 ); }
Signature · 1 line
2.52.0
[PULL 07/41] docs/specs/acpi_cpu_hotplug: Remove legacy cpu hotplug descriptions
In reply to Paolo BonziniFrom: Zhao Liu <[email protected]>
Legacy cpu hotplug has been removed totally and machines start with modern cpu hotplug interface directly.
Therefore, update the documentation to describe current QEMU cpu hotplug logic.
--- docs/specs/acpi_cpu_hotplug.rst | 28 +++------------------------- 1 file changed, 3 insertions(+), 25 deletions(-)
docs/specs/acpi_cpu_hotplug.rst +3 -25
@@ -8,22 +8,6 @@ ACPI BIOS GPE.2 handler is dedicated for notifying OS about CPU hot-add and hot-remove events. -Legacy ACPI CPU hotplug interface registers -------------------------------------------- - -CPU present bitmap for: - -- ICH9-LPC (IO port 0x0cd8-0xcf7, 1-byte access) -- PIIX-PM (IO port 0xaf00-0xaf1f, 1-byte access) -- One bit per CPU. Bit position reflects corresponding CPU APIC ID. Read-only. -- The first DWORD in bitmap is used in write mode to switch from legacy - to modern CPU hotplug interface, write 0 into it to do switch. - -QEMU sets corresponding CPU bit on hot-add event and issues SCI -with GPE.2 event set. CPU present map is read by ACPI BIOS GPE.2 handler -to notify OS about CPU hot-add events. CPU hot-remove isn't supported. - - Modern ACPI CPU hotplug interface registers -------------------------------------------
@@ -189,20 +173,14 @@ Typical usecases (x86) Detecting and enabling modern CPU hotplug interface ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ -QEMU starts with legacy CPU hotplug interface enabled. Detecting and -switching to modern interface is based on the 2 legacy CPU hotplug features: +QEMU starts with modern CPU hotplug interface enabled. Use following steps to +detect modern CPU hotplug interface: -#. Writes into CPU bitmap are ignored. -#. CPU bitmap always has bit #0 set, corresponding to boot CPU. - -Use following steps to detect and enable modern CPU hotplug interface: - -#. Store 0x0 to the 'CPU selector' register, attempting to switch to modern mode #. Store 0x0 to the 'CPU selector' register, to ensure valid selector value #. Store 0x0 to the 'Command field' register #. Read the 'Command data 2' register. If read value is 0x0, the modern interface is enabled. - Otherwise legacy or no CPU hotplug interface available + Otherwise no CPU hotplug interface available Get a cpu with pending event ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Signature · 1 line
2.52.0
[PULL 40/41] accel/mshv: Remove remap overlapping mappings code
In reply to Paolo BonziniFrom: Magnus Kulke <[email protected]>
This change removes userland code that worked around a restriction in the mshv driver in the 6.18 kernel: regions from userland couldn't be mapped to multiple regions in the kernel. We maintained a shadow mapping table in qemu and used a heuristic to swap in a requested region in case of UNMAPPED_GPA exits.
However, this heuristic wasn't reliable in all cases, since HyperV behaviour is not 100% reliable across versions. HyperV itself doesn't prohibit to map regions at multiple places into the guest, so the restriction has been removed in the mshv driver.
Hence we can remove the remapping code. Effectively this will mandate a 6.19 kernel, if the workload attempt to map e.g. BIOS to multiple reagions. I still think it's the right call to remove this logic:
- The workaround only seems to work reliably with a certain revision of HyperV as a nested hypervisor.
- We expect Direct Virtualization (L1VH) to be the main platform for the mshv accelerator, which also requires a 6.19 kernel
This reverts commit efc4093358511a58846a409b965213aa1bb9f31a.
--- include/system/mshv_int.h | 22 +- accel/mshv/mem.c | 406 +++--------------------------------- accel/mshv/mshv-all.c | 2 - target/i386/mshv/mshv-cpu.c | 43 ---- accel/mshv/trace-events | 5 - 5 files changed, 30 insertions(+), 448 deletions(-)
include/system/mshv_int.h +2 -20
@@ -16,8 +16,6 @@#define MSHV_MSR_ENTRIES_COUNT 64 -#define MSHV_MAX_MEM_SLOTS 32 -typedef struct hyperv_message hv_message ; typedef struct MshvHvCallArgs {
@@ -42,12 +40,6 @@ typedef struct MshvAddressSpace {AddressSpace *as; } MshvAddressSpace; -typedef struct MshvMemorySlotManager { -size_t n_slots; -GList *slots; -QemuMutex mutex; -}MshvMemorySlotManager ; -struct MshvState {AccelState parent_obj;int vm;
@@ -56,7 +48,6 @@ struct MshvState {int nr_as;MshvAddressSpace *as;int fd; -MshvMemorySlotManager msm; }; typedef struct MshvMsiControl {
@@ -87,12 +78,6 @@ typedef enum MshvVmExit { MshvVmExitSpecial =2 , } MshvVmExit; -typedef enum MshvRemapResult { - MshvRemapOk =0 , - MshvRemapNoMapping =1 , - MshvRemapNoOverlap =2 , -}MshvRemapResult ; -void mshv_init_mmio_emu (void );int mshv_create_vcpu (int vm_fd,uint8_t vp_index,int *cpu_fd);void mshv_remove_vcpu (int vm_fd,int cpu_fd);
@@ -116,22 +101,19 @@ int mshv_hvcall(int fd, const struct mshv_root_hvcall *args);#endif /* memory */ -typedef struct MshvMemorySlot { +typedef struct MshvMemoryRegion {uint64_t guest_phys_addr;uint64_t memory_size;uint64_t userspace_addr;bool readonly; -bool mapped; -}MshvMemorySlot ; +}MshvMemoryRegion ; -MshvRemapResult mshv_remap_overlap_region (int vm_fd,uint64_t gpa);int mshv_guest_mem_read (uint64_t gpa,uint8_t *data,uintptr_t size,bool is_secure_mode,bool instruction_fetch);int mshv_guest_mem_write (uint64_t gpa,const uint8_t *data,uintptr_t size,bool is_secure_mode);void mshv_set_phys_mem (MshvMemoryListener *mml,MemoryRegionSection *section,bool add); -void mshv_init_memory_slot_manager (MshvState *mshv_state);/* msr */ typedef struct MshvMsrEntry {
accel/mshv/mem.c +28 -378
470 lines not on this page · raw message
accel/mshv/mshv-all.c +0 -2
@@ -437,8 +437,6 @@ static int mshv_init(AccelState *as, MachineState *ms)mshv_init_msicontrol (); -mshv_init_memory_slot_manager (s); - ret =create_vm (mshv_fd, &vm_fd);if (ret <0 ) {close (mshv_fd);
target/i386/mshv/mshv-cpu.c +0 -43
55 lines not on this page · raw message
accel/mshv/trace-events +0 -5
@@ -26,8 +26,3 @@ mshv_map_memory(uint64_t userspace_addr, uint64_t gpa, uint64_t size) "\tu_a=0x% mshv_unmap_memory(uint64_t userspace_addr, uint64_t gpa, uint64_t size) "\tu_a=0x%" PRIx64 " gpa=0x%010" PRIx64 " size=0x%08" PRIx64 mshv_set_phys_mem(bool add, const char *name, uint64_t gpa) "\tadd=%d name=%s gpa=0x%010" PRIx64 mshv_handle_mmio(uint64_t gva, uint64_t gpa, uint64_t size, uint8_t access_type) "\tgva=0x%" PRIx64 " gpa=0x%010" PRIx64 " size=0x%" PRIx64 " access_type=%d" - -mshv_found_slot(uint64_t userspace_addr, uint64_t gpa, uint64_t size) "\tu_a=0x%" PRIx64 " gpa=0x%010" PRIx64 " size=0x%08" PRIx64 -mshv_skip_unset_mem(uint64_t userspace_addr, uint64_t gpa, uint64_t size) "\tu_a=0x%" PRIx64 " gpa=0x%010" PRIx64 " size=0x%08" PRIx64 -mshv_remap_attempt(uint64_t userspace_addr, uint64_t gpa, uint64_t size) "\tu_a=0x%" PRIx64 " gpa=0x%010" PRIx64 " size=0x%08" PRIx64 -mshv_find_slot_by_gpa(uint64_t gpa) "\tgpa=0x%010" PRIx64
Signature · 1 line
2.52.0
[PULL 26/41] hw/i386/intel_iommu: Remove IntelIOMMUState::buggy_eim field
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
The 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
@@ -313,7 +313,6 @@ 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 pasid;/* Whether to support PASID */
hw/i386/intel_iommu.c +2 -3
@@ -4180,7 +4180,6 @@ 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),
@@ -5539,11 +5538,11 @@ 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
2.52.0
[PULL 25/41] hw/core/machine: Remove hw_compat_2_7[] array
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
The 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/core/boards.h | 3 --- hw/core/machine.c | 9 --------- 2 files changed, 12 deletions(-)
include/hw/core/boards.h +0 -3
@@ -882,7 +882,4 @@ 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
@@ -284,15 +284,6 @@ 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
2.52.0
[PULL 21/41] hw/i386/pc: Remove deprecated pc-q35-2.7 and pc-i440fx-2.7 machines
In reply to Paolo BonziniFrom: Philippe Mathieu-Daudé <[email protected]>
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. Remove the qtest in test-x86-cpuid-compat.c file.
--- hw/i386/pc_piix.c | 9 --------- hw/i386/pc_q35.c | 10 ---------- tests/qtest/test-x86-cpuid-compat.c | 11 ----------- 3 files changed, 30 deletions(-)
hw/i386/pc_piix.c +0 -9
@@ -723,15 +723,6 @@ 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_XEN static void xenfv_machine_4_2_options (MachineClass *m) {
hw/i386/pc_q35.c +0 -10
@@ -670,13 +670,3 @@ 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 );
tests/qtest/test-x86-cpuid-compat.c +0 -11
@@ -345,17 +345,6 @@ int main(int argc, char **argv)/* Check compatibility of old machine-types that didn't * auto-increase level/xlevel/xlevel2: */ -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 ); - }if (qtest_has_machine ("pc-i440fx-2.9" )) { add_cpuid_test("x86/cpuid/auto-level7/pc-i440fx-2.9/off", "Conroe", NULL, "pc-i440fx-2.9",
Signature · 1 line
2.52.0
[PULL 37/41] qdev: make release_drive() idempotent
In reply to Paolo BonziniFrom: Marc-André Lureau <[email protected]>
So it can eventually be called multiple times safely.
--- hw/core/qdev-properties-system.c | 1 + 1 file changed, 1 insertion(+)
hw/core/qdev-properties-system.c +1 -0
@@ -232,6 +232,7 @@ static void release_drive(Object *obj, const char *name, void *opaque)if (*ptr) {blockdev_auto_del (*ptr);blk_detach_dev (*ptr, dev); + *ptr =NULL ; } }
Signature · 1 line
2.52.0
Quoted 22 lines · Paolo Bonzini
The following changes since commit
cd5a79dc98e3087e7658e643bdbbb0baec77ac8a:Merge tag '
for_upstream' of https://git.kernel.org/pub/scm/virt/kvm/mst/qemu into staging (2026-02-05 13:54:50 +0000)are available in the Git repository at:
https://gitlab.com/bonzini/qemu.git tags/for-upstream
for you to fetch changes up to
28a29efe4342055b4631de96ba71fd577ae83e07:target/i386/mshv: remove unused optimization of gva=>gpa translation (2026-02-10 14:41:08 +0100)
----------------------------------------------------------------
* hw/i386: Remove deprecated PC 2.6 and 2.7 machines
* i386/cpu: Fix incorrect initializer in Diamond Rapids definition
* qom: Clean up property release
* target/i386/kvm: setKVM_PMU_CAP_DISABLEif "-pmu" is configured
* target/i386/kvm: reset AMD and perfmon-v2 PMU registers during VM reset
* mshv: Cleanup
* target/i386: convert SEV-ES termination requests to guest panic events
Quoted 1 line
----------------------------------------------------------------
Hi; the commit for this patch:
Quoted 2 lines
Aidan Khoury (1):
i386/cpu: Fix incorrect initializer in Diamond Rapids definition
includes this change to the roms/seabios-hppa submodule:
Quoted 1 line
roms/seabios-hppa | 2 +-
which is not mentioned in the commit message and is not related to the change. This seems likely to have been introduced by accident during a patchseries rebase: could you remove it from the commit and resend, please ?
thanks
-- PMM