hw/i386/pc: Remove deprecated 2.4 and 2.5 PC machines
v4 cover hw/i386/pc: Remove deprecated 2.4 and 2.5 PC machines
(series reviewed)
Since v3:
- Fixed 2 issues noticed by Thomas (floppy fallback, e1000)
Since v2:
- Removed qtest in test-x86-cpuid-compat.c
Since v1:
- Fixed issues noticed by Thomas
The versioned 'pc' and 'q35' machines up to 2.12 been marked as deprecated two releases ago, and are older than 6 years, so according to our support policy we can remove them.
This series only includes the 2.4 and 2.5 machines removal, as it is a big enough number of LoC removed. Rest will follow. Highlight is the legacy fw_cfg API removal :)
Philippe Mathieu-Daudé (19):
hw/i386/pc: Remove deprecated pc-q35-2.4 and pc-i440fx-2.4 machines hw/i386/pc: Remove PCMachineClass::broken_reserved_end field hw/i386/pc: Remove pc_compat_2_4[] array
target/i386/cpu: Remove X86CPU::check_cpuid field hw/core/machine: Remove hw_compat_2_4[] array hw/net/e1000: Remove unused E1000_FLAG_MAC flag hw/virtio/virtio-pci: Remove VIRTIO_PCI_FLAG_MIGRATE_EXTRA definition hw/virtio/virtio-pci: Remove VIRTIO_PCI_FLAG_DISABLE_PCIE definition hw/i386/pc: Remove deprecated pc-q35-2.5 and pc-i440fx-2.5 machines hw/i386/x86: Remove X86MachineClass::save_tsc_khz field hw/nvram/fw_cfg: Remove legacy FW_CFG_ORDER_OVERRIDE hw/core/machine: Remove hw_compat_2_5[] array hw/block/fdc-isa: Remove 'fallback' property hw/scsi/vmw_pvscsi: Remove PVSCSI_COMPAT_OLD_PCI_CONFIGURATION
definition
hw/scsi/vmw_pvscsi: Remove PVSCSI_COMPAT_DISABLE_PCIE_BIT definition hw/scsi/vmw_pvscsi: Convert DeviceRealize -> InstanceInit hw/net/vmxnet3: Remove VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS definition hw/net/vmxnet3: Remove VMXNET3_COMPAT_FLAG_DISABLE_PCIE definition hw/net/vmxnet3: Merge DeviceRealize in InstanceInit
include/hw/boards.h | 9 +-- include/hw/i386/pc.h | 7 -- include/hw/i386/x86.h | 5 -- include/hw/loader.h | 2 - include/hw/nvram/fw_cfg.h | 10 --- include/hw/virtio/virtio-pci.h | 8 -- target/i386/cpu.h | 1 - hw/block/fdc-isa.c | 4 +- hw/core/loader.c | 14 ---- hw/core/machine.c | 18 ----- hw/i386/pc.c | 42 ++--------- hw/i386/pc_piix.c | 26 ------- hw/i386/pc_q35.c | 26 ------- hw/i386/x86.c | 1 - hw/net/e1000.c | 93 ++++++++++-------------- hw/net/vmxnet3.c | 44 ++--------- hw/nvram/fw_cfg.c | 109 ++-------------------------- hw/scsi/vmw_pvscsi.c | 67 +++-------------- hw/virtio/virtio-pci.c | 11 +-- system/vl.c | 5 -- target/i386/cpu.c | 3 +- target/i386/machine.c | 5 +- tests/qtest/test-x86-cpuid-compat.c | 14 ---- 23 files changed, 77 insertions(+), 447 deletions(-)
Signature · 1 line
2.47.1
PATCH v4 1/19 hw/i386/pc: Remove deprecated pc-q35-2.4 and pc-i440fx-2.4 machines
These machines has been supported for a period of more than 6 years. According to our versioned machine support policy (see commit ce80c4fa6ff "docs: document special exception for machine type deprecation & removal") they can now be removed.
Remove the qtest in test-x86-cpuid-compat.c file.
--- hw/i386/pc_piix.c | 13 ------------- hw/i386/pc_q35.c | 13 ------------- tests/qtest/test-x86-cpuid-compat.c | 14 -------------- 3 files changed, 40 deletions(-)
hw/i386/pc_piix.c +0 -13
@@ -791,19 +791,6 @@ static void pc_i440fx_machine_2_5_options(MachineClass *m)DEFINE_I440FX_MACHINE (2 ,5 ); -static void pc_i440fx_machine_2_4_options (MachineClass *m) -{ -PCMachineClass *pcmc =PC_MACHINE_CLASS (m); - -pc_i440fx_machine_2_5_options (m); - m->hw_version ="2.4.0" ; - pcmc->broken_reserved_end = true; -compat_props_add (m->compat_props, hw_compat_2_4, hw_compat_2_4_len); -compat_props_add (m->compat_props, pc_compat_2_4, pc_compat_2_4_len); -} - -DEFINE_I440FX_MACHINE (2 ,4 ); -#ifdef CONFIG_ISAPC static void isapc_machine_options (MachineClass *m) {
hw/i386/pc_q35.c +0 -13
@@ -685,16 +685,3 @@ static void pc_q35_machine_2_5_options(MachineClass *m) }DEFINE_Q35_MACHINE (2 ,5 ); - -static void pc_q35_machine_2_4_options (MachineClass *m) -{ -PCMachineClass *pcmc =PC_MACHINE_CLASS (m); - -pc_q35_machine_2_5_options (m); - m->hw_version ="2.4.0" ; - pcmc->broken_reserved_end = true; -compat_props_add (m->compat_props, hw_compat_2_4, hw_compat_2_4_len); -compat_props_add (m->compat_props, pc_compat_2_4, pc_compat_2_4_len); -} - -DEFINE_Q35_MACHINE (2 ,4 );
tests/qtest/test-x86-cpuid-compat.c +0 -14
@@ -365,20 +365,6 @@ int main(int argc, char **argv) "level", 10); } -/* -* xlevel doesn't have any feature that triggers auto-level -* code on old machine-types. Just check that the compat code -* is working correctly: -*/ -if (qtest_has_machine ("pc-i440fx-2.4" )) { -add_cpuid_test ("x86/cpuid/xlevel-compat/pc-i440fx-2.4/npt-off" , -"SandyBridge" ,NULL ,"pc-i440fx-2.4" , -"xlevel" ,0x80000008 ); -add_cpuid_test ("x86/cpuid/xlevel-compat/pc-i440fx-2.4/npt-on" , -"SandyBridge" ,"svm=on,npt=on" ,"pc-i440fx-2.4" , -"xlevel" ,0x80000008 ); - } -/* Test feature parsing */ add_feature_test("x86/cpuid/features/plus", "486", "+arat",
Signature · 1 line
2.47.1
Re: [PATCH v4 01/19] hw/i386/pc: Remove deprecated pc-q35-2.4 and pc-i440fx-2.4 machines
Quoted 11 lines · Philippe Mathieu-Daudé
These machines has been supported for a period of more than 6 years. According to our versioned machine support policy (see commit
ce80c4fa6ff"docs: document special exception for machine type deprecation & removal") they can now be removed.Remove the qtest in
test-x86-cpuid-compat.cfile.Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Thomas Huth <[email protected]>Reviewed-by Zhao Liu <[email protected]>
Quoted 76 lines
--- hw/i386/pc_piix.c | 13 ------------- hw/i386/pc_q35.c | 13 ------------- tests/qtest/test-x86-cpuid-compat.c | 14 -------------- 3 files changed, 40 deletions(-)hw/i386/pc_piix.c +0
-13@@ -791,19 +791,6 @@ static void pc_i440fx_machine_2_5_options(MachineClass *m)DEFINE_I440FX_MACHINE (2 ,5 ); -static void pc_i440fx_machine_2_4_options (MachineClass *m) -{ -PCMachineClass *pcmc =PC_MACHINE_CLASS (m); - -pc_i440fx_machine_2_5_options (m); - m->hw_version ="2.4.0" ; - pcmc->broken_reserved_end = true; -compat_props_add (m->compat_props, hw_compat_2_4, hw_compat_2_4_len); -compat_props_add (m->compat_props, pc_compat_2_4, pc_compat_2_4_len); -} - -DEFINE_I440FX_MACHINE (2 ,4 ); -#ifdef CONFIG_ISAPC static void isapc_machine_options (MachineClass *m) {hw/i386/pc_q35.c +0
-13@@ -685,16 +685,3 @@ static void pc_q35_machine_2_5_options(MachineClass *m) }DEFINE_Q35_MACHINE (2 ,5 ); - -static void pc_q35_machine_2_4_options (MachineClass *m) -{ -PCMachineClass *pcmc =PC_MACHINE_CLASS (m); - -pc_q35_machine_2_5_options (m); - m->hw_version ="2.4.0" ; - pcmc->broken_reserved_end = true; -compat_props_add (m->compat_props, hw_compat_2_4, hw_compat_2_4_len); -compat_props_add (m->compat_props, pc_compat_2_4, pc_compat_2_4_len); -} - -DEFINE_Q35_MACHINE (2 ,4 );tests/qtest/test-x86-cpuid-compat.c +0
-14@@ -365,20 +365,6 @@ int main(int argc, char **argv) "level", 10); } -/* -* xlevel doesn't have any feature that triggers auto-level -* code on old machine-types. Just check that the compat code -* is working correctly: -*/ -if (qtest_has_machine ("pc-i440fx-2.4" )) { -add_cpuid_test ("x86/cpuid/xlevel-compat/pc-i440fx-2.4/npt-off" , -"SandyBridge" ,NULL ,"pc-i440fx-2.4" , -"xlevel" ,0x80000008 ); -add_cpuid_test ("x86/cpuid/xlevel-compat/pc-i440fx-2.4/npt-on" , -"SandyBridge" ,"svm=on,npt=on" ,"pc-i440fx-2.4" , -"xlevel" ,0x80000008 ); - } -/* Test feature parsing */ add_feature_test("x86/cpuid/features/plus", "486", "+arat",
Re: [PATCH v4 01/19] hw/i386/pc: Remove deprecated pc-q35-2.4 and pc-i440fx-2.4 machines
In reply to Philippe Mathieu-DaudéQuoted 11 lines · On Mon, 12 May 2025 10:39:30 +0200 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.cfile.Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Thomas Huth <[email protected]>Reviewed-by Zhao Liu <[email protected]>
Quoted 76 lines
--- hw/i386/pc_piix.c | 13 ------------- hw/i386/pc_q35.c | 13 ------------- tests/qtest/test-x86-cpuid-compat.c | 14 -------------- 3 files changed, 40 deletions(-)hw/i386/pc_piix.c +0
-13@@ -791,19 +791,6 @@ static void pc_i440fx_machine_2_5_options(MachineClass *m)DEFINE_I440FX_MACHINE (2 ,5 ); -static void pc_i440fx_machine_2_4_options (MachineClass *m) -{ -PCMachineClass *pcmc =PC_MACHINE_CLASS (m); - -pc_i440fx_machine_2_5_options (m); - m->hw_version ="2.4.0" ; - pcmc->broken_reserved_end = true; -compat_props_add (m->compat_props, hw_compat_2_4, hw_compat_2_4_len); -compat_props_add (m->compat_props, pc_compat_2_4, pc_compat_2_4_len); -} - -DEFINE_I440FX_MACHINE (2 ,4 ); -#ifdef CONFIG_ISAPC static void isapc_machine_options (MachineClass *m) {hw/i386/pc_q35.c +0
-13@@ -685,16 +685,3 @@ static void pc_q35_machine_2_5_options(MachineClass *m) }DEFINE_Q35_MACHINE (2 ,5 ); - -static void pc_q35_machine_2_4_options (MachineClass *m) -{ -PCMachineClass *pcmc =PC_MACHINE_CLASS (m); - -pc_q35_machine_2_5_options (m); - m->hw_version ="2.4.0" ; - pcmc->broken_reserved_end = true; -compat_props_add (m->compat_props, hw_compat_2_4, hw_compat_2_4_len); -compat_props_add (m->compat_props, pc_compat_2_4, pc_compat_2_4_len); -} - -DEFINE_Q35_MACHINE (2 ,4 );tests/qtest/test-x86-cpuid-compat.c +0
-14@@ -365,20 +365,6 @@ int main(int argc, char **argv) "level", 10); } -/* -* xlevel doesn't have any feature that triggers auto-level -* code on old machine-types. Just check that the compat code -* is working correctly: -*/ -if (qtest_has_machine ("pc-i440fx-2.4" )) { -add_cpuid_test ("x86/cpuid/xlevel-compat/pc-i440fx-2.4/npt-off" , -"SandyBridge" ,NULL ,"pc-i440fx-2.4" , -"xlevel" ,0x80000008 ); -add_cpuid_test ("x86/cpuid/xlevel-compat/pc-i440fx-2.4/npt-on" , -"SandyBridge" ,"svm=on,npt=on" ,"pc-i440fx-2.4" , -"xlevel" ,0x80000008 ); - } -/* Test feature parsing */ add_feature_test("x86/cpuid/features/plus", "486", "+arat",
PATCH v4 2/19 hw/i386/pc: Remove PCMachineClass::broken_reserved_end field
In reply to Philippe Mathieu-DaudéThe PCMachineClass::broken_reserved_end field was only used by the pc-q35-2.4 and pc-i440fx-2.4 machines, which got removed. Remove it and simplify pc_memory_init().
--- include/hw/i386/pc.h | 1 - hw/i386/pc.c | 13 +++++-------- 2 files changed, 5 insertions(+), 9 deletions(-)
include/hw/i386/pc.h +0 -1
@@ -107,7 +107,6 @@ struct PCMachineClass {/* RAM / address space compat: */ bool gigabyte_align;bool has_reserved_memory; -bool broken_reserved_end;bool enforce_amd_1tb_hole;bool isa_bios_alias;
hw/i386/pc.c +5 -8
@@ -999,14 +999,13 @@ void pc_memory_init(PCMachineState *pcms,if (machine->device_memory) {uint64_t *val =g_malloc (sizeof (*val)); -uint64_t res_mem_end = machine->device_memory->base; - -if (!pcmc->broken_reserved_end) { - res_mem_end +=memory_region_size (&machine->device_memory->mr); - } +uint64_t res_mem_end;if (pcms->cxl_devices_state.is_enabled) { res_mem_end = cxl_resv_end; + }else { + res_mem_end = machine->device_memory->base + +memory_region_size (&machine->device_memory->mr); } *val =cpu_to_le64 (ROUND_UP (res_mem_end,1 * GiB));fw_cfg_add_file (fw_cfg,"etc/reserved-memory-end" , val,sizeof (*val));
@@ -1044,9 +1043,7 @@ uint64_t pc_pci_hole64_start(void) hole64_start =pc_get_cxl_range_end (pcms); }else if (pcmc->has_reserved_memory && (ms->ram_size < ms->maxram_size)) {pc_get_device_memory_range (pcms, &hole64_start, &size); -if (!pcmc->broken_reserved_end) { - hole64_start += size; - } + hole64_start += size; }else { hole64_start =pc_above_4g_end (pcms); }
Signature · 1 line
2.47.1
Re: [PATCH v4 02/19] hw/i386/pc: Remove PCMachineClass::broken_reserved_end field
Quoted 8 lines · Philippe Mathieu-Daudé
The
PCMachineClass::broken_reserved_endfield was only used by the pc-q35-2.4 and pc-i440fx-2.4 machines, which got removed. Remove it and simplifypc_memory_init().Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Thomas Huth <[email protected]>Reviewed-by Zhao Liu <[email protected]>
Quoted 51 lines
--- include/hw/i386/pc.h | 1 - hw/i386/pc.c | 13 +++++-------- 2 files changed, 5 insertions(+), 9 deletions(-)include/hw/i386/pc.h +0
-1@@ -107,7 +107,6 @@ struct PCMachineClass {/* RAM / address space compat: */ bool gigabyte_align;bool has_reserved_memory; -bool broken_reserved_end;bool enforce_amd_1tb_hole;bool isa_bios_alias;hw/i386/pc.c +5
-8@@ -999,14 +999,13 @@ void pc_memory_init(PCMachineState *pcms,if (machine->device_memory) {uint64_t *val =g_malloc (sizeof (*val)); -uint64_t res_mem_end = machine->device_memory->base; - -if (!pcmc->broken_reserved_end) { - res_mem_end +=memory_region_size (&machine->device_memory->mr); - } +uint64_t res_mem_end;if (pcms->cxl_devices_state.is_enabled) { res_mem_end = cxl_resv_end; + }else { + res_mem_end = machine->device_memory->base + +memory_region_size (&machine->device_memory->mr); } *val =cpu_to_le64 (ROUND_UP (res_mem_end,1 * GiB));fw_cfg_add_file (fw_cfg,"etc/reserved-memory-end" , val,sizeof (*val));@@ -1044,9 +1043,7 @@ uint64_t pc_pci_hole64_start(void) hole64_start =pc_get_cxl_range_end (pcms); }else if (pcmc->has_reserved_memory && (ms->ram_size < ms->maxram_size)) {pc_get_device_memory_range (pcms, &hole64_start, &size); -if (!pcmc->broken_reserved_end) { - hole64_start += size; - } + hole64_start += size; }else { hole64_start =pc_above_4g_end (pcms); }
Re: [PATCH v4 02/19] hw/i386/pc: Remove PCMachineClass::broken_reserved_end field
In reply to Philippe Mathieu-DaudéQuoted 8 lines · On Mon, 12 May 2025 10:39:31 +0200 Philippe Mathieu-Daudé <[email protected]>…
The
PCMachineClass::broken_reserved_endfield was only used by the pc-q35-2.4 and pc-i440fx-2.4 machines, which got removed. Remove it and simplifypc_memory_init().Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Thomas Huth <[email protected]>Reviewed-by Zhao Liu <[email protected]>
Quoted 51 lines
--- include/hw/i386/pc.h | 1 - hw/i386/pc.c | 13 +++++-------- 2 files changed, 5 insertions(+), 9 deletions(-)include/hw/i386/pc.h +0
-1@@ -107,7 +107,6 @@ struct PCMachineClass {/* RAM / address space compat: */ bool gigabyte_align;bool has_reserved_memory; -bool broken_reserved_end;bool enforce_amd_1tb_hole;bool isa_bios_alias;hw/i386/pc.c +5
-8@@ -999,14 +999,13 @@ void pc_memory_init(PCMachineState *pcms,if (machine->device_memory) {uint64_t *val =g_malloc (sizeof (*val)); -uint64_t res_mem_end = machine->device_memory->base; - -if (!pcmc->broken_reserved_end) { - res_mem_end +=memory_region_size (&machine->device_memory->mr); - } +uint64_t res_mem_end;if (pcms->cxl_devices_state.is_enabled) { res_mem_end = cxl_resv_end; + }else { + res_mem_end = machine->device_memory->base + +memory_region_size (&machine->device_memory->mr); } *val =cpu_to_le64 (ROUND_UP (res_mem_end,1 * GiB));fw_cfg_add_file (fw_cfg,"etc/reserved-memory-end" , val,sizeof (*val));@@ -1044,9 +1043,7 @@ uint64_t pc_pci_hole64_start(void) hole64_start =pc_get_cxl_range_end (pcms); }else if (pcmc->has_reserved_memory && (ms->ram_size < ms->maxram_size)) {pc_get_device_memory_range (pcms, &hole64_start, &size); -if (!pcmc->broken_reserved_end) { - hole64_start += size; - } + hole64_start += size; }else { hole64_start =pc_above_4g_end (pcms); }
PATCH v4 3/19 hw/i386/pc: Remove pc_compat_2_4[] array
In reply to Philippe Mathieu-DaudéThe pc_compat_2_4[] array was only used by the pc-q35-2.4 and pc-i440fx-2.4 machines, which got removed. Remove it.
--- include/hw/i386/pc.h | 3 --- hw/i386/pc.c | 19 ------------------- 2 files changed, 22 deletions(-)
include/hw/i386/pc.h +0 -3
@@ -301,9 +301,6 @@ extern const size_t pc_compat_2_6_len;extern GlobalProperty pc_compat_2_5[];extern const size_t pc_compat_2_5_len; -extern GlobalProperty pc_compat_2_4[]; -extern const size_t pc_compat_2_4_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 -19
@@ -262,25 +262,6 @@ const size_t pc_compat_2_6_len = G_N_ELEMENTS(pc_compat_2_6);GlobalProperty pc_compat_2_5[] = {};const size_t pc_compat_2_5_len =G_N_ELEMENTS (pc_compat_2_5); -GlobalProperty pc_compat_2_4[] = { - PC_CPU_MODEL_IDS("2.4.0") - {"Haswell-" TYPE_X86_CPU ,"abm" ,"off" }, - {"Haswell-noTSX-" TYPE_X86_CPU ,"abm" ,"off" }, - {"Broadwell-" TYPE_X86_CPU ,"abm" ,"off" }, - {"Broadwell-noTSX-" TYPE_X86_CPU ,"abm" ,"off" }, - {"host" "-" TYPE_X86_CPU ,"host-cache-info" ,"on" }, - {TYPE_X86_CPU ,"check" ,"off" }, - {"qemu64" "-" TYPE_X86_CPU ,"sse4a" ,"on" }, - {"qemu64" "-" TYPE_X86_CPU ,"abm" ,"on" }, - {"qemu64" "-" TYPE_X86_CPU ,"popcnt" ,"on" }, - {"qemu32" "-" TYPE_X86_CPU ,"popcnt" ,"on" }, - {"Opteron_G2" "-" TYPE_X86_CPU ,"rdtscp" ,"on" }, - {"Opteron_G3" "-" TYPE_X86_CPU ,"rdtscp" ,"on" }, - {"Opteron_G4" "-" TYPE_X86_CPU ,"rdtscp" ,"on" }, - {"Opteron_G5" "-" TYPE_X86_CPU ,"rdtscp" ,"on" , } -}; -const size_t pc_compat_2_4_len =G_N_ELEMENTS (pc_compat_2_4); -/* * @PC_FW_DATA: * Size of the chunk of memory at the top of RAM for the BIOS ACPI tables
Signature · 1 line
2.47.1
Re: [PATCH v4 03/19] hw/i386/pc: Remove pc_compat_2_4[] array
Quoted 6 lines · Philippe Mathieu-Daudé
The pc_compat_2_4[] array was only used by the pc-q35-2.4 and pc-i440fx-2.4 machines, which got removed. Remove it.
Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Zhao Liu <[email protected]>
Quoted 49 lines
--- include/hw/i386/pc.h | 3 --- hw/i386/pc.c | 19 ------------------- 2 files changed, 22 deletions(-)include/hw/i386/pc.h +0
-3@@ -301,9 +301,6 @@ extern const size_t pc_compat_2_6_len;extern GlobalProperty pc_compat_2_5[];extern const size_t pc_compat_2_5_len; -extern GlobalProperty pc_compat_2_4[]; -extern const size_t pc_compat_2_4_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
-19@@ -262,25 +262,6 @@ const size_t pc_compat_2_6_len = G_N_ELEMENTS(pc_compat_2_6);GlobalProperty pc_compat_2_5[] = {};const size_t pc_compat_2_5_len =G_N_ELEMENTS (pc_compat_2_5); -GlobalProperty pc_compat_2_4[] = { - PC_CPU_MODEL_IDS("2.4.0") - {"Haswell-" TYPE_X86_CPU ,"abm" ,"off" }, - {"Haswell-noTSX-" TYPE_X86_CPU ,"abm" ,"off" }, - {"Broadwell-" TYPE_X86_CPU ,"abm" ,"off" }, - {"Broadwell-noTSX-" TYPE_X86_CPU ,"abm" ,"off" }, - {"host" "-" TYPE_X86_CPU ,"host-cache-info" ,"on" }, - {TYPE_X86_CPU ,"check" ,"off" }, - {"qemu64" "-" TYPE_X86_CPU ,"sse4a" ,"on" }, - {"qemu64" "-" TYPE_X86_CPU ,"abm" ,"on" }, - {"qemu64" "-" TYPE_X86_CPU ,"popcnt" ,"on" }, - {"qemu32" "-" TYPE_X86_CPU ,"popcnt" ,"on" }, - {"Opteron_G2" "-" TYPE_X86_CPU ,"rdtscp" ,"on" }, - {"Opteron_G3" "-" TYPE_X86_CPU ,"rdtscp" ,"on" }, - {"Opteron_G4" "-" TYPE_X86_CPU ,"rdtscp" ,"on" }, - {"Opteron_G5" "-" TYPE_X86_CPU ,"rdtscp" ,"on" , } -}; -const size_t pc_compat_2_4_len =G_N_ELEMENTS (pc_compat_2_4); -/* * @PC_FW_DATA: * Size of the chunk of memory at the top of RAM for the BIOS ACPI tables
Re: [PATCH v4 03/19] hw/i386/pc: Remove pc_compat_2_4[] array
In reply to Philippe Mathieu-DaudéQuoted 6 lines · On Mon, 12 May 2025 10:39:32 +0200 Philippe Mathieu-Daudé <[email protected]>…
The pc_compat_2_4[] array was only used by the pc-q35-2.4 and pc-i440fx-2.4 machines, which got removed. Remove it.
Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Zhao Liu <[email protected]>
Quoted 49 lines
--- include/hw/i386/pc.h | 3 --- hw/i386/pc.c | 19 ------------------- 2 files changed, 22 deletions(-)include/hw/i386/pc.h +0
-3@@ -301,9 +301,6 @@ extern const size_t pc_compat_2_6_len;extern GlobalProperty pc_compat_2_5[];extern const size_t pc_compat_2_5_len; -extern GlobalProperty pc_compat_2_4[]; -extern const size_t pc_compat_2_4_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
-19@@ -262,25 +262,6 @@ const size_t pc_compat_2_6_len = G_N_ELEMENTS(pc_compat_2_6);GlobalProperty pc_compat_2_5[] = {};const size_t pc_compat_2_5_len =G_N_ELEMENTS (pc_compat_2_5); -GlobalProperty pc_compat_2_4[] = { - PC_CPU_MODEL_IDS("2.4.0") - {"Haswell-" TYPE_X86_CPU ,"abm" ,"off" }, - {"Haswell-noTSX-" TYPE_X86_CPU ,"abm" ,"off" }, - {"Broadwell-" TYPE_X86_CPU ,"abm" ,"off" }, - {"Broadwell-noTSX-" TYPE_X86_CPU ,"abm" ,"off" }, - {"host" "-" TYPE_X86_CPU ,"host-cache-info" ,"on" }, - {TYPE_X86_CPU ,"check" ,"off" }, - {"qemu64" "-" TYPE_X86_CPU ,"sse4a" ,"on" }, - {"qemu64" "-" TYPE_X86_CPU ,"abm" ,"on" }, - {"qemu64" "-" TYPE_X86_CPU ,"popcnt" ,"on" }, - {"qemu32" "-" TYPE_X86_CPU ,"popcnt" ,"on" }, - {"Opteron_G2" "-" TYPE_X86_CPU ,"rdtscp" ,"on" }, - {"Opteron_G3" "-" TYPE_X86_CPU ,"rdtscp" ,"on" }, - {"Opteron_G4" "-" TYPE_X86_CPU ,"rdtscp" ,"on" }, - {"Opteron_G5" "-" TYPE_X86_CPU ,"rdtscp" ,"on" , } -}; -const size_t pc_compat_2_4_len =G_N_ELEMENTS (pc_compat_2_4); -/* * @PC_FW_DATA: * Size of the chunk of memory at the top of RAM for the BIOS ACPI tables
PATCH v4 4/19 target/i386/cpu: Remove X86CPU::check_cpuid field
In reply to Philippe Mathieu-DaudéThe X86CPU::check_cpuid boolean was only set in the pc_compat_2_4[] array, via the 'check=off' property. We removed all machines using that array, lets remove that CPU property and simplify x86_cpu_realizefn().
--- target/i386/cpu.h | 1 - target/i386/cpu.c | 3 +-- 2 files changed, 1 insertion(+), 3 deletions(-)
target/i386/cpu.h +0 -1
@@ -2164,7 +2164,6 @@ struct ArchCPU {uint8_t hyperv_ver_id_sb;uint32_t hyperv_ver_id_sn; -bool check_cpuid;bool enforce_cpuid;/* * Force features to be enabled even if the host doesn't support them.
target/i386/cpu.c +1 -2
@@ -8174,7 +8174,7 @@ static void x86_cpu_realizefn(DeviceState *dev, Error **errp) } } -if (x86_cpu_filter_features (cpu, cpu->check_cpuid || cpu->enforce_cpuid)) { +if (x86_cpu_filter_features (cpu, cpu->enforce_cpuid)) {if (cpu->enforce_cpuid) { error_setg(&local_err, accel_uses_host_cpuid() ?
@@ -8813,7 +8813,6 @@ static const Property x86_cpu_properties[] = {DEFINE_PROP_UINT8 ("hv-version-id-sbranch" ,X86CPU , hyperv_ver_id_sb,0 ),DEFINE_PROP_UINT32 ("hv-version-id-snumber" ,X86CPU , hyperv_ver_id_sn,0 ), -DEFINE_PROP_BOOL ("check" ,X86CPU , check_cpuid, true),DEFINE_PROP_BOOL ("enforce" ,X86CPU , enforce_cpuid, false),DEFINE_PROP_BOOL ("x-force-features" ,X86CPU , force_features, false), DEFINE_PROP_BOOL("kvm", X86CPU, expose_kvm, true),
Signature · 1 line
2.47.1
Re: [PATCH v4 04/19] target/i386/cpu: Remove X86CPU::check_cpuid field
Quoted 4 lines · Philippe Mathieu-Daudé
The
X86CPU::check_cpuidboolean was only set in the pc_compat_2_4[] array, via the 'check=off' property. We removed all machines using that array, lets remove that CPU property and simplifyx86_cpu_realizefn().
No.
We cannot do this. Because it changes the behavior of QEMU.
'check_cpuid' is true by default while 'enforce_cpuid' is false. So that QEMU emits warnings in x86_cpu_filter_features() by default when user requests unsupported CPU features. If remove "check" property and the internal 'check_cpuid', QEMU will not do it unless user sets enforce_cpuid explicitly.
Quoted 41 lines
Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Zhao Liu <[email protected]>--- target/i386/cpu.h | 1 - target/i386/cpu.c | 3 +-- 2 files changed, 1 insertion(+), 3 deletions(-)target/i386/cpu.h +0
-1@@ -2164,7 +2164,6 @@ struct ArchCPU {uint8_t hyperv_ver_id_sb;uint32_t hyperv_ver_id_sn; -bool check_cpuid;bool enforce_cpuid;/* * Force features to be enabled even if the host doesn't support them. target/i386/cpu.c +1
-2@@ -8174,7 +8174,7 @@ static void x86_cpu_realizefn(DeviceState *dev, Error **errp) } } -if (x86_cpu_filter_features (cpu, cpu->check_cpuid || cpu->enforce_cpuid)) { +if (x86_cpu_filter_features (cpu, cpu->enforce_cpuid)) {if (cpu->enforce_cpuid) { error_setg(&local_err, accel_uses_host_cpuid() ?@@ -8813,7 +8813,6 @@ static const Property x86_cpu_properties[] = {DEFINE_PROP_UINT8 ("hv-version-id-sbranch" ,X86CPU , hyperv_ver_id_sb,0 ),DEFINE_PROP_UINT32 ("hv-version-id-snumber" ,X86CPU , hyperv_ver_id_sn,0 ), -DEFINE_PROP_BOOL ("check" ,X86CPU , check_cpuid, true),DEFINE_PROP_BOOL ("enforce" ,X86CPU , enforce_cpuid, false),DEFINE_PROP_BOOL ("x-force-features" ,X86CPU , force_features, false), DEFINE_PROP_BOOL("kvm", X86CPU, expose_kvm, true),
Re: [PATCH v4 04/19] target/i386/cpu: Remove X86CPU::check_cpuid field
Quoted 13 lines · Xiaoyao Li
Date: Wed, 28 May 2025 10:09:56 +0800
From: Xiaoyao Li <[email protected]>
Subject: Re: [PATCH v4 04/19] target/i386/cpu: RemoveX86CPU::check_cpuid
fieldQuoted 4 lines · Philippe Mathieu-Daudé
The
X86CPU::check_cpuidboolean was only set in the pc_compat_2_4[] array, via the 'check=off' property. We removed all machines using that array, lets remove that CPU property and simplifyx86_cpu_realizefn().No.
Quoted 7 lines
We cannot do this. Because it changes the behavior of QEMU.
'
check_cpuid' is true by default while 'enforce_cpuid' is false. So that QEMU emits warnings inx86_cpu_filter_features()by default when user requests unsupported CPU features. If remove "check" property and the internal 'check_cpuid', QEMU will not do it unless user setsenforce_cpuidexplicitly.
One option would be to have x86_cpu_filter_features() unconditionally turn on verbose and print warnings, but some people might want to turn off these warning prints, I don't know if anyone would, but it would be possible.
The other option is still to keep the “check” property.
IMO, the latter option is the better way to reduce Philippe's burden.
Regards,
Zhao
Re: [PATCH v4 04/19] target/i386/cpu: Remove X86CPU::check_cpuid field
Quoted 22 lines · On Wed, 28 May 2025 13:23:49 +0800 Zhao Liu <[email protected]> wrote:
Quoted 13 lines · Xiaoyao Li
Date: Wed, 28 May 2025 10:09:56 +0800
From: Xiaoyao Li <[email protected]>
Subject: Re: [PATCH v4 04/19] target/i386/cpu: RemoveX86CPU::check_cpuid
fieldQuoted 4 lines · Philippe Mathieu-Daudé
The
X86CPU::check_cpuidboolean was only set in the pc_compat_2_4[] array, via the 'check=off' property. We removed all machines using that array, lets remove that CPU property and simplifyx86_cpu_realizefn().No.
Quoted 7 lines
We cannot do this. Because it changes the behavior of QEMU.
'
check_cpuid' is true by default while 'enforce_cpuid' is false. So that QEMU emits warnings inx86_cpu_filter_features()by default when user requests unsupported CPU features. If remove "check" property and the internal 'check_cpuid', QEMU will not do it unless user setsenforce_cpuidexplicitly.
Quoted 8 lines
One option would be to have
x86_cpu_filter_features()unconditionally turn on verbose and print warnings, but some people might want to turn off these warning prints, I don't know if anyone would, but it would be possible.The other option is still to keep the “check” property.
IMO, the latter option is the better way to reduce Philippe's burden.
we essentially loose warnings by default when some features aren't available, qemu still continues to run though.
Given that Daniel acked it from libvirt side, libvirt doesn't care about warnings (it does its has its own cpu model calculation). Likely other mgmt do not care about it either, and if they do they probably doing something wrong and should use QMP to get that data.
That leaves us with human users, for that case I'd say one should use enforce_cpuid if feature availability matters.
so +1 to removal
Quoted 5 lines
Regards,
Zhao
Re: [PATCH v4 04/19] target/i386/cpu: Remove X86CPU::check_cpuid field
Quoted 37 lines · Igor Mammedov
Quoted 22 lines · On Wed, 28 May 2025 13:23:49 +0800 Zhao Liu <[email protected]> wrote:
Quoted 13 lines · Xiaoyao Li
Date: Wed, 28 May 2025 10:09:56 +0800
From: Xiaoyao Li <[email protected]>
Subject: Re: [PATCH v4 04/19] target/i386/cpu: RemoveX86CPU::check_cpuid
fieldQuoted 4 lines · Philippe Mathieu-Daudé
The
X86CPU::check_cpuidboolean was only set in the pc_compat_2_4[] array, via the 'check=off' property. We removed all machines using that array, lets remove that CPU property and simplifyx86_cpu_realizefn().No.
Quoted 7 lines
We cannot do this. Because it changes the behavior of QEMU.
'
check_cpuid' is true by default while 'enforce_cpuid' is false. So that QEMU emits warnings inx86_cpu_filter_features()by default when user requests unsupported CPU features. If remove "check" property and the internal 'check_cpuid', QEMU will not do it unless user setsenforce_cpuidexplicitly.Quoted 8 lines
One option would be to have
x86_cpu_filter_features()unconditionally turn on verbose and print warnings, but some people might want to turn off these warning prints, I don't know if anyone would, but it would be possible.The other option is still to keep the “check” property.
IMO, the latter option is the better way to reduce Philippe's burden.
we essentially loose warnings by default when some features aren't available, qemu still continues to run though.
Quoted 6 lines
Given that Daniel acked it from libvirt side, libvirt doesn't care about warnings (it does its has its own cpu model calculation). Likely other mgmt do not care about it either, and if they do they probably doing something wrong and should use QMP to get that data.
That leaves us with human users, for that case I'd say one should useenforce_cpuidif feature availability matters.
But with "check", it allows the VM to continue running with the unsupported bits cleared and warnings to inform users. This is really friendly.
Quoted 8 lines
so +1 to removal
Quoted 5 lines
Regards,
Zhao
Re: [PATCH v4 04/19] target/i386/cpu: Remove X86CPU::check_cpuid field
Quoted 45 lines · On Tue, 3 Jun 2025 23:54:19 +0800 Xiaoyao Li <[email protected]> wrote:
Quoted 37 lines · Igor Mammedov
Quoted 22 lines · On Wed, 28 May 2025 13:23:49 +0800 Zhao Liu <[email protected]> wrote:
Quoted 13 lines · Xiaoyao Li
Date: Wed, 28 May 2025 10:09:56 +0800
From: Xiaoyao Li <[email protected]>
Subject: Re: [PATCH v4 04/19] target/i386/cpu: RemoveX86CPU::check_cpuid
fieldQuoted 4 lines · Philippe Mathieu-Daudé
The
X86CPU::check_cpuidboolean was only set in the pc_compat_2_4[] array, via the 'check=off' property. We removed all machines using that array, lets remove that CPU property and simplifyx86_cpu_realizefn().No.
Quoted 7 lines
We cannot do this. Because it changes the behavior of QEMU.
'
check_cpuid' is true by default while 'enforce_cpuid' is false. So that QEMU emits warnings inx86_cpu_filter_features()by default when user requests unsupported CPU features. If remove "check" property and the internal 'check_cpuid', QEMU will not do it unless user setsenforce_cpuidexplicitly.Quoted 8 lines
One option would be to have
x86_cpu_filter_features()unconditionally turn on verbose and print warnings, but some people might want to turn off these warning prints, I don't know if anyone would, but it would be possible.The other option is still to keep the “check” property.
IMO, the latter option is the better way to reduce Philippe's burden.
we essentially loose warnings by default when some features aren't available, qemu still continues to run though.
Quoted 6 lines
Given that Daniel acked it from libvirt side, libvirt doesn't care about warnings (it does its has its own cpu model calculation). Likely other mgmt do not care about it either, and if they do they probably doing something wrong and should use QMP to get that data.
That leaves us with human users, for that case I'd say one should useenforce_cpuidif feature availability matters.
Quoted 3 lines
But with "check", it allows the VM to continue running with the unsupported bits cleared and warnings to inform users. This is really friendly.
it's friendly for human users (mostly developers), but for upper layers is doesn't make a difference, since it's noise in logs nobody reads until qemu somehow works.
the 1st category can and should use enforce flag instead to get what they ask for if they care about it.
If missing feature bits do not matter then it warnings shouldn't matter either. What I'm advocating for is being more strict/deterministic on QEMU side,
1. you get what you asked for with enforce or fix you CLI explicitly to be clear on what you are missing out.
2. you don't care if some features are missing, but then you don't really care about warnings either. (I'd still get missing features filtered out though, just silently). If one cares about missing features, one can use #1
I admit, It's a tiny bit of code but removing it, cleans up code a little bit and helps readability/in reviews. Doing such small cleanups here and there have a cumulative effect on the codebase.
Anyways it's not something I'd fight for, so if you insist on keeping it it's ok as well.
Quoted 9 lines
Quoted 8 lines
so +1 to removal
Quoted 5 lines
Regards,
Zhao
Re: [PATCH v4 04/19] target/i386/cpu: Remove X86CPU::check_cpuid field
Quoted 48 lines · Igor Mammedov
Quoted 45 lines · On Tue, 3 Jun 2025 23:54:19 +0800 Xiaoyao Li <[email protected]> wrote:
Quoted 43 lines · Igor Mammedov
Quoted 22 lines · On Wed, 28 May 2025 13:23:49 +0800 Zhao Liu <[email protected]> wrote:
Quoted 13 lines · Xiaoyao Li
Date: Wed, 28 May 2025 10:09:56 +0800
From: Xiaoyao Li <[email protected]>
Subject: Re: [PATCH v4 04/19] target/i386/cpu: RemoveX86CPU::check_cpuid
fieldQuoted 4 lines · Philippe Mathieu-Daudé
The
X86CPU::check_cpuidboolean was only set in the pc_compat_2_4[] array, via the 'check=off' property. We removed all machines using that array, lets remove that CPU property and simplifyx86_cpu_realizefn().No.
Quoted 7 lines
We cannot do this. Because it changes the behavior of QEMU.
'
check_cpuid' is true by default while 'enforce_cpuid' is false. So that QEMU emits warnings inx86_cpu_filter_features()by default when user requests unsupported CPU features. If remove "check" property and the internal 'check_cpuid', QEMU will not do it unless user setsenforce_cpuidexplicitly.Quoted 8 lines
One option would be to have
x86_cpu_filter_features()unconditionally turn on verbose and print warnings, but some people might want to turn off these warning prints, I don't know if anyone would, but it would be possible.The other option is still to keep the “check” property.
IMO, the latter option is the better way to reduce Philippe's burden.
we essentially loose warnings by default when some features aren't available, qemu still continues to run though.
Given that Daniel acked it from libvirt side, libvirt doesn't care about warnings (it does its has its own cpu model calculation). Likely other mgmt do not care about it either, and if they do they probably doing something wrong and should use QMP to get that data.
That leaves us with human users, for that case I'd say one should useenforce_cpuidif feature availability matters.
Quoted 7 lines
Quoted 3 lines
But with "check", it allows the VM to continue running with the unsupported bits cleared and warnings to inform users. This is really friendly.
it's friendly for human users (mostly developers), but for upper layers is doesn't make a difference, since it's noise in logs nobody reads until qemu somehow works.
The latter statement is accurate. In the libvirt case, this info ends up in the per-VM logfile, which is one of the key files humans look at first when triaging any QEMU related bug reports.
Quoted 4 lines
the 1st category can and should use enforce flag instead to get what they ask for if they care about it.
If missing feature bits do not matter then it warnings shouldn't matter either.
I don't think that's correct. The mgmt app can decide the missing features don't matter, but at the same time someone looking at the QEMU command line doesn't know the mgmt apps' intent and it is easy to overlook that QEMU will silently drop features. So the precense of the warnings in the log alerts the person debugging that the actual QEMU guest visible CPU is subtly different from what's on the command line, and that's useful IMHO for troubleshooting.
Quoted 7 lines
What I'm advocating for is being more strict/deterministic on QEMU side,
1. you get what you asked for with enforce or fix you CLI explicitly to be clear on what you are missing out.
2. you don't care if some features are missing, but then you don't really care about warnings either. (I'd still get missing features filtered out though, just silently). If one cares about missing features, one can use #1
Quoted 6 lines
I admit, It's a tiny bit of code but removing it, cleans up code a little bit and helps readability/in reviews. Doing such small cleanups here and there have a cumulative effect on the codebase.
Anyways it's not something I'd fight for, so if you insist on keeping it it's ok as well.
With regards,
Daniel
Signature · 3 lines
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
Re: [PATCH v4 04/19] target/i386/cpu: Remove X86CPU::check_cpuid field
In reply to Igor MammedovQuoted 34 lines · Igor Mammedov
Quoted 22 lines · On Wed, 28 May 2025 13:23:49 +0800 Zhao Liu <[email protected]> wrote:
Quoted 13 lines · Xiaoyao Li
Date: Wed, 28 May 2025 10:09:56 +0800
From: Xiaoyao Li <[email protected]>
Subject: Re: [PATCH v4 04/19] target/i386/cpu: RemoveX86CPU::check_cpuid
fieldQuoted 4 lines · Philippe Mathieu-Daudé
The
X86CPU::check_cpuidboolean was only set in the pc_compat_2_4[] array, via the 'check=off' property. We removed all machines using that array, lets remove that CPU property and simplifyx86_cpu_realizefn().No.
Quoted 7 lines
We cannot do this. Because it changes the behavior of QEMU.
'
check_cpuid' is true by default while 'enforce_cpuid' is false. So that QEMU emits warnings inx86_cpu_filter_features()by default when user requests unsupported CPU features. If remove "check" property and the internal 'check_cpuid', QEMU will not do it unless user setsenforce_cpuidexplicitly.Quoted 8 lines
One option would be to have
x86_cpu_filter_features()unconditionally turn on verbose and print warnings, but some people might want to turn off these warning prints, I don't know if anyone would, but it would be possible.The other option is still to keep the “check” property.
IMO, the latter option is the better way to reduce Philippe's burden.
Quoted 7 lines
we essentially loose warnings by default when some features aren't available, qemu still continues to run though.
Given that Daniel acked it from libvirt side, libvirt doesn't care about warnings (it does its has its own cpu model calculation). Likely other mgmt do not care about it either, and if they do they probably doing something wrong and should use QMP to get that data.
Acking it was a mistake on my part - I mis-interpreted the patch and so didn't notice we were loosing the verbose printing of missing features by default.
I'm actually curious why we made the 'check' feature tied to machine types at all. If it doesn't affect guest ABI, just causes verbose info on stderr, it feels like something we could have just had on all machine types new & old. Git history brings us back to
commit 3e68482224129c3ddc061af7c9d438b882ecfdd1
Author: Eduardo Habkost <[email protected]>
Date: Tue Nov 3 17:18:50 2015 -0200
target-i386: Set "check=off" by default on pc-*-2.4 and older
The default CPU model (qemu64) have some issues today: it enables some features (ABM and SSE4a) that are not present in many host CPUs. That means many hosts (but not all of them) had those features silently disabled in the default configuration in QEMU 2.4 and older.
With the new "check=on" default, this causes warnings to be printed in the default configuration, because of the lack of SSE4A on all Intel hosts, and the lack of ABM on Sandy Bridge and older hosts:
$ qemu-system-x86_64 -machine pc,accel=kvm
warning: host doesn't support requested feature: CPUID.80000001H:ECX.abm [bit 5]
warning: host doesn't support requested feature: CPUID.80000001H:ECX.sse4a [bit 6]
Those issues will be fixed in pc-*-2.5 and newer. But as we can't change the guest ABI in pc-*-2.4, disable "check" mode by default in pc-*-2.4 and older so we don't print spurious warnings.
IOW, we wanted to have 'check' unconditionally on by default, but had to do a temp hack to avoid spamming all configurations with the broken 'qemu64' CPU model design.
Quoted 2 lines
That leaves us with human users, for that case I'd say one should use
enforce_cpuidif feature availability matters.
IMHO even with mgmt apps, it is worth having 'check=on' by default as the log message has value in debugging scenarios. It could have the potential to highlight situations where an mgmt app has unwittingly done something wrong with CPU config. At the very least though its a warning to humans debugging that they should not trust the QEMU command line as a expressing the full CPU featureset.
With regards,
Daniel
Signature · 3 lines
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
Re: [PATCH v4 04/19] target/i386/cpu: Remove X86CPU::check_cpuid field
Quoted 82 lines · On Thu, 5 Jun 2025 10:24:34 +0100 Daniel P. Berrangé <[email protected]>…
Quoted 34 lines · Igor Mammedov
Quoted 22 lines · On Wed, 28 May 2025 13:23:49 +0800 Zhao Liu <[email protected]> wrote:
Quoted 13 lines · Xiaoyao Li
Date: Wed, 28 May 2025 10:09:56 +0800
From: Xiaoyao Li <[email protected]>
Subject: Re: [PATCH v4 04/19] target/i386/cpu: RemoveX86CPU::check_cpuid
fieldQuoted 4 lines · Philippe Mathieu-Daudé
The
X86CPU::check_cpuidboolean was only set in the pc_compat_2_4[] array, via the 'check=off' property. We removed all machines using that array, lets remove that CPU property and simplifyx86_cpu_realizefn().No.
Quoted 7 lines
We cannot do this. Because it changes the behavior of QEMU.
'
check_cpuid' is true by default while 'enforce_cpuid' is false. So that QEMU emits warnings inx86_cpu_filter_features()by default when user requests unsupported CPU features. If remove "check" property and the internal 'check_cpuid', QEMU will not do it unless user setsenforce_cpuidexplicitly.Quoted 8 lines
One option would be to have
x86_cpu_filter_features()unconditionally turn on verbose and print warnings, but some people might want to turn off these warning prints, I don't know if anyone would, but it would be possible.The other option is still to keep the “check” property.
IMO, the latter option is the better way to reduce Philippe's burden.
Quoted 7 lines
we essentially loose warnings by default when some features aren't available, qemu still continues to run though.
Given that Daniel acked it from libvirt side, libvirt doesn't care about warnings (it does its has its own cpu model calculation). Likely other mgmt do not care about it either, and if they do they probably doing something wrong and should use QMP to get that data.
Acking it was a mistake on my part - I mis-interpreted the patch and so didn't notice we were loosing the verbose printing of missing features by default.
I'm actually curious why we made the 'check' feature tied to machine types at all. If it doesn't affect guest ABI, just causes verbose info on stderr, it feels like something we could have just had on all machine types new & old. Git history brings us back to
commit
3e68482224129c3ddc061af7c9d438b882ecfdd1
Author: Eduardo Habkost <[email protected]>
Date: Tue Nov 3 17:18:50 2015 -0200target-i386: Set "check=off" by default on pc-*-2.4 and older
The default CPU model (qemu64) have some issues today: it enables some features (ABM and SSE4a) that are not present in many host CPUs. That means many hosts (but not all of them) had those features silently disabled in the default configuration in QEMU 2.4 and older.
With the new "check=on" default, this causes warnings to be printed in the default configuration, because of the lack of SSE4A on all Intel hosts, and the lack of ABM on Sandy Bridge and older hosts:
$ qemu-system-x86_64 -machine pc,accel=kvm
warning: host doesn't support requested feature: CPUID.80000001H:ECX.abm [bit 5]
warning: host doesn't support requested feature: CPUID.80000001H:ECX.sse4a [bit 6]Those issues will be fixed in pc-*-2.5 and newer. But as we can't change the guest ABI in pc-*-2.4, disable "check" mode by default in pc-*-2.4 and older so we don't print spurious warnings.
IOW, we wanted to have 'check' unconditionally on by default, but had to do a temp hack to avoid spamming all configurations with the broken 'qemu64' CPU model design.
Quoted 2 lines
That leaves us with human users, for that case I'd say one should use
enforce_cpuidif feature availability matters.
Quoted 6 lines
IMHO even with mgmt apps, it is worth having 'check=on' by default as the log message has value in debugging scenarios. It could have the potential to highlight situations where an mgmt app has unwittingly done something wrong with CPU config. At the very least though its a warning to humans debugging that they should not trust the QEMU command line as a expressing the full CPU featureset.
in that line of thought, maybe hardcode 'check=on' and drop conditional/property?
Quoted 3 lines
With regards,
Daniel
Re: [PATCH v4 04/19] target/i386/cpu: Remove X86CPU::check_cpuid field
In reply to Philippe Mathieu-DaudéQuoted 8 lines · On Mon, 12 May 2025 10:39:33 +0200 Philippe Mathieu-Daudé <[email protected]>…
The
X86CPU::check_cpuidboolean was only set in the pc_compat_2_4[] array, via the 'check=off' property. We removed all machines using that array, lets remove that CPU property and simplifyx86_cpu_realizefn().Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Zhao Liu <[email protected]>
Quoted 38 lines
--- target/i386/cpu.h | 1 - target/i386/cpu.c | 3 +-- 2 files changed, 1 insertion(+), 3 deletions(-)target/i386/cpu.h +0
-1@@ -2164,7 +2164,6 @@ struct ArchCPU {uint8_t hyperv_ver_id_sb;uint32_t hyperv_ver_id_sn; -bool check_cpuid;bool enforce_cpuid;/* * Force features to be enabled even if the host doesn't support them. target/i386/cpu.c +1
-2@@ -8174,7 +8174,7 @@ static void x86_cpu_realizefn(DeviceState *dev, Error **errp) } } -if (x86_cpu_filter_features (cpu, cpu->check_cpuid || cpu->enforce_cpuid)) { +if (x86_cpu_filter_features (cpu, cpu->enforce_cpuid)) {if (cpu->enforce_cpuid) { error_setg(&local_err, accel_uses_host_cpuid() ?@@ -8813,7 +8813,6 @@ static const Property x86_cpu_properties[] = {DEFINE_PROP_UINT8 ("hv-version-id-sbranch" ,X86CPU , hyperv_ver_id_sb,0 ),DEFINE_PROP_UINT32 ("hv-version-id-snumber" ,X86CPU , hyperv_ver_id_sn,0 ), -DEFINE_PROP_BOOL ("check" ,X86CPU , check_cpuid, true),DEFINE_PROP_BOOL ("enforce" ,X86CPU , enforce_cpuid, false),DEFINE_PROP_BOOL ("x-force-features" ,X86CPU , force_features, false), DEFINE_PROP_BOOL("kvm", X86CPU, expose_kvm, true),
PATCH v4 5/19 hw/core/machine: Remove hw_compat_2_4[] array
In reply to Philippe Mathieu-DaudéThe hw_compat_2_4[] array was only used by the pc-q35-2.4 and pc-i440fx-2.4 machines, which got removed. Remove it.
--- include/hw/boards.h | 3 --- hw/core/machine.c | 9 --------- 2 files changed, 12 deletions(-)
include/hw/boards.h +0 -3
@@ -866,7 +866,4 @@ extern const size_t hw_compat_2_6_len;extern GlobalProperty hw_compat_2_5[];extern const size_t hw_compat_2_5_len; -extern GlobalProperty hw_compat_2_4[]; -extern const size_t hw_compat_2_4_len; -#endif
hw/core/machine.c +0 -9
@@ -292,15 +292,6 @@ GlobalProperty hw_compat_2_5[] = { };const size_t hw_compat_2_5_len =G_N_ELEMENTS (hw_compat_2_5); -GlobalProperty hw_compat_2_4[] = { - {"e1000" ,"extra_mac_registers" ,"off" }, - {"virtio-pci" ,"x-disable-pcie" ,"on" }, - {"virtio-pci" ,"migrate-extra" ,"off" }, - {"fw_cfg_mem" ,"dma_enabled" ,"off" }, - {"fw_cfg_io" ,"dma_enabled" ,"off" } -}; -const size_t hw_compat_2_4_len =G_N_ELEMENTS (hw_compat_2_4); -MachineState *current_machine;static char *machine_get_kernel (Object *obj,Error **errp)
Signature · 1 line
2.47.1
Re: [PATCH v4 05/19] hw/core/machine: Remove hw_compat_2_4[] array
Quoted 6 lines · Philippe Mathieu-Daudé
The hw_compat_2_4[] array was only used by the pc-q35-2.4 and pc-i440fx-2.4 machines, which got removed. Remove it.
Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Zhao Liu <[email protected]>
Quoted 37 lines
--- include/hw/boards.h | 3 --- hw/core/machine.c | 9 --------- 2 files changed, 12 deletions(-)include/hw/boards.h +0
-3@@ -866,7 +866,4 @@ extern const size_t hw_compat_2_6_len;extern GlobalProperty hw_compat_2_5[];extern const size_t hw_compat_2_5_len; -extern GlobalProperty hw_compat_2_4[]; -extern const size_t hw_compat_2_4_len; -#endif hw/core/machine.c +0
-9@@ -292,15 +292,6 @@ GlobalProperty hw_compat_2_5[] = { };const size_t hw_compat_2_5_len =G_N_ELEMENTS (hw_compat_2_5); -GlobalProperty hw_compat_2_4[] = { - {"e1000" ,"extra_mac_registers" ,"off" }, - {"virtio-pci" ,"x-disable-pcie" ,"on" }, - {"virtio-pci" ,"migrate-extra" ,"off" }, - {"fw_cfg_mem" ,"dma_enabled" ,"off" }, - {"fw_cfg_io" ,"dma_enabled" ,"off" } -}; -const size_t hw_compat_2_4_len =G_N_ELEMENTS (hw_compat_2_4); -MachineState *current_machine;static char *machine_get_kernel (Object *obj,Error **errp)
PATCH v4 6/19 hw/net/e1000: Remove unused E1000_FLAG_MAC flag
In reply to Philippe Mathieu-DaudéE1000_FLAG_MAC was only used by the hw_compat_2_4[] array, via the 'extra_mac_registers=off' property. We removed all machines using that array, lets remove all the code around E1000_FLAG_MAC, including the MAC_ACCESS_FLAG_NEEDED enum, similarly to commit fa4ec9ffda7 ("e1000: remove old compatibility code").
--- hw/net/e1000.c | 93 ++++++++++++++++++++++---------------------------- 1 file changed, 40 insertions(+), 53 deletions(-)
hw/net/e1000.c +40 -53
@@ -127,10 +127,8 @@ struct E1000State_st {QEMUTimer *flush_queue_timer;/* Compatibility flags for migration to/from qemu 1.3.0 and older */ -#define E1000_FLAG_MAC_BIT 2#define E1000_FLAG_TSO_BIT 3#define E1000_FLAG_VET_BIT 4 -#define E1000_FLAG_MAC (1 << E1000_FLAG_MAC_BIT)#define E1000_FLAG_TSO (1 << E1000_FLAG_TSO_BIT)#define E1000_FLAG_VET (1 << E1000_FLAG_VET_BIT)
@@ -1212,52 +1210,51 @@ enum { NWRITEOPS = ARRAY_SIZE(macreg_writeops) };enum {MAC_ACCESS_PARTIAL =1 ,MAC_ACCESS_FLAG_NEEDED =2 }; -#define markflag (x) ((E1000_FLAG_##x << 2) | MAC_ACCESS_FLAG_NEEDED)/* In the array below the meaning of the bits is: [f|f|f|f|f|f|n|p] * f - flag bits (up to 6 possible flags) * n - flag needed -* p - partially implenented */ +* p - partially implemented */ static const uint8_t mac_reg_access[0x8000 ] = { - [IPAV ] =markflag (MAC ), [WUC ] =markflag (MAC ), - [IP6AT ] =markflag (MAC ), [IP4AT ] =markflag (MAC ), - [FFVT ] =markflag (MAC ), [WUPM ] =markflag (MAC ), - [ECOL ] =markflag (MAC ), [MCC ] =markflag (MAC ), - [DC ] =markflag (MAC ), [TNCRS ] =markflag (MAC ), - [RLEC ] =markflag (MAC ), [XONRXC ] =markflag (MAC ), - [XOFFTXC ] =markflag (MAC ), [RFC ] =markflag (MAC ), - [TSCTFC ] =markflag (MAC ), [MGTPRC ] =markflag (MAC ), - [WUS ] =markflag (MAC ), [AIT ] =markflag (MAC ), - [FFLT ] =markflag (MAC ), [FFMT ] =markflag (MAC ), - [SCC ] =markflag (MAC ), [FCRUC ] =markflag (MAC ), - [LATECOL ] =markflag (MAC ), [COLC ] =markflag (MAC ), - [SEQEC ] =markflag (MAC ), [CEXTERR ] =markflag (MAC ), - [XONTXC ] =markflag (MAC ), [XOFFRXC ] =markflag (MAC ), - [RJC ] =markflag (MAC ), [RNBC ] =markflag (MAC ), - [MGTPDC ] =markflag (MAC ), [MGTPTC ] =markflag (MAC ), - [RUC ] =markflag (MAC ), [ROC ] =markflag (MAC ), - [GORCL ] =markflag (MAC ), [GORCH ] =markflag (MAC ), - [GOTCL ] =markflag (MAC ), [GOTCH ] =markflag (MAC ), - [BPRC ] =markflag (MAC ), [MPRC ] =markflag (MAC ), - [TSCTC ] =markflag (MAC ), [PRC64 ] =markflag (MAC ), - [PRC127 ] =markflag (MAC ), [PRC255 ] =markflag (MAC ), - [PRC511 ] =markflag (MAC ), [PRC1023 ] =markflag (MAC ), - [PRC1522 ] =markflag (MAC ), [PTC64 ] =markflag (MAC ), - [PTC127 ] =markflag (MAC ), [PTC255 ] =markflag (MAC ), - [PTC511 ] =markflag (MAC ), [PTC1023 ] =markflag (MAC ), - [PTC1522 ] =markflag (MAC ), [MPTC ] =markflag (MAC ), - [BPTC ] =markflag (MAC ), + [IPAV ] =MAC_ACCESS_FLAG_NEEDED , [WUC ] =MAC_ACCESS_FLAG_NEEDED , + [IP6AT ] =MAC_ACCESS_FLAG_NEEDED , [IP4AT ] =MAC_ACCESS_FLAG_NEEDED , + [FFVT ] =MAC_ACCESS_FLAG_NEEDED , [WUPM ] =MAC_ACCESS_FLAG_NEEDED , + [ECOL ] =MAC_ACCESS_FLAG_NEEDED , [MCC ] =MAC_ACCESS_FLAG_NEEDED , + [DC ] =MAC_ACCESS_FLAG_NEEDED , [TNCRS ] =MAC_ACCESS_FLAG_NEEDED , + [RLEC ] =MAC_ACCESS_FLAG_NEEDED , [XONRXC ] =MAC_ACCESS_FLAG_NEEDED , + [XOFFTXC ] =MAC_ACCESS_FLAG_NEEDED , [RFC ] =MAC_ACCESS_FLAG_NEEDED , + [TSCTFC ] =MAC_ACCESS_FLAG_NEEDED , [MGTPRC ] =MAC_ACCESS_FLAG_NEEDED , + [WUS ] =MAC_ACCESS_FLAG_NEEDED , [AIT ] =MAC_ACCESS_FLAG_NEEDED , + [FFLT ] =MAC_ACCESS_FLAG_NEEDED , [FFMT ] =MAC_ACCESS_FLAG_NEEDED , + [SCC ] =MAC_ACCESS_FLAG_NEEDED , [FCRUC ] =MAC_ACCESS_FLAG_NEEDED , + [LATECOL ] =MAC_ACCESS_FLAG_NEEDED , [COLC ] =MAC_ACCESS_FLAG_NEEDED , + [SEQEC ] =MAC_ACCESS_FLAG_NEEDED , [CEXTERR ] =MAC_ACCESS_FLAG_NEEDED , + [XONTXC ] =MAC_ACCESS_FLAG_NEEDED , [XOFFRXC ] =MAC_ACCESS_FLAG_NEEDED , + [RJC ] =MAC_ACCESS_FLAG_NEEDED , [RNBC ] =MAC_ACCESS_FLAG_NEEDED , + [MGTPDC ] =MAC_ACCESS_FLAG_NEEDED , [MGTPTC ] =MAC_ACCESS_FLAG_NEEDED , + [RUC ] =MAC_ACCESS_FLAG_NEEDED , [ROC ] =MAC_ACCESS_FLAG_NEEDED , + [GORCL ] =MAC_ACCESS_FLAG_NEEDED , [GORCH ] =MAC_ACCESS_FLAG_NEEDED , + [GOTCL ] =MAC_ACCESS_FLAG_NEEDED , [GOTCH ] =MAC_ACCESS_FLAG_NEEDED , + [BPRC ] =MAC_ACCESS_FLAG_NEEDED , [MPRC ] =MAC_ACCESS_FLAG_NEEDED , + [TSCTC ] =MAC_ACCESS_FLAG_NEEDED , [PRC64 ] =MAC_ACCESS_FLAG_NEEDED , + [PRC127 ] =MAC_ACCESS_FLAG_NEEDED , [PRC255 ] =MAC_ACCESS_FLAG_NEEDED , + [PRC511 ] =MAC_ACCESS_FLAG_NEEDED , [PRC1023 ] =MAC_ACCESS_FLAG_NEEDED , + [PRC1522 ] =MAC_ACCESS_FLAG_NEEDED , [PTC64 ] =MAC_ACCESS_FLAG_NEEDED , + [PTC127 ] =MAC_ACCESS_FLAG_NEEDED , [PTC255 ] =MAC_ACCESS_FLAG_NEEDED , + [PTC511 ] =MAC_ACCESS_FLAG_NEEDED , [PTC1023 ] =MAC_ACCESS_FLAG_NEEDED , + [PTC1522 ] =MAC_ACCESS_FLAG_NEEDED , [MPTC ] =MAC_ACCESS_FLAG_NEEDED , + [BPTC ] =MAC_ACCESS_FLAG_NEEDED , - [TDFH ] =markflag (MAC ) |MAC_ACCESS_PARTIAL , - [TDFT ] =markflag (MAC ) |MAC_ACCESS_PARTIAL , - [TDFHS ] =markflag (MAC ) |MAC_ACCESS_PARTIAL , - [TDFTS ] =markflag (MAC ) |MAC_ACCESS_PARTIAL , - [TDFPC ] =markflag (MAC ) |MAC_ACCESS_PARTIAL , - [RDFH ] =markflag (MAC ) |MAC_ACCESS_PARTIAL , - [RDFT ] =markflag (MAC ) |MAC_ACCESS_PARTIAL , - [RDFHS ] =markflag (MAC ) |MAC_ACCESS_PARTIAL , - [RDFTS ] =markflag (MAC ) |MAC_ACCESS_PARTIAL , - [RDFPC ] =markflag (MAC ) |MAC_ACCESS_PARTIAL , - [PBM ] =markflag (MAC ) |MAC_ACCESS_PARTIAL , + [TDFH ] =MAC_ACCESS_FLAG_NEEDED |MAC_ACCESS_PARTIAL , + [TDFT ] =MAC_ACCESS_FLAG_NEEDED |MAC_ACCESS_PARTIAL , + [TDFHS ] =MAC_ACCESS_FLAG_NEEDED |MAC_ACCESS_PARTIAL , + [TDFTS ] =MAC_ACCESS_FLAG_NEEDED |MAC_ACCESS_PARTIAL , + [TDFPC ] =MAC_ACCESS_FLAG_NEEDED |MAC_ACCESS_PARTIAL , + [RDFH ] =MAC_ACCESS_FLAG_NEEDED |MAC_ACCESS_PARTIAL , + [RDFT ] =MAC_ACCESS_FLAG_NEEDED |MAC_ACCESS_PARTIAL , + [RDFHS ] =MAC_ACCESS_FLAG_NEEDED |MAC_ACCESS_PARTIAL , + [RDFTS ] =MAC_ACCESS_FLAG_NEEDED |MAC_ACCESS_PARTIAL , + [RDFPC ] =MAC_ACCESS_FLAG_NEEDED |MAC_ACCESS_PARTIAL , + [PBM ] =MAC_ACCESS_FLAG_NEEDED |MAC_ACCESS_PARTIAL , }; static void
@@ -1419,13 +1416,6 @@ static int e1000_tx_tso_post_load(void *opaque, int version_id)return 0 ; } -static bool e1000_full_mac_needed (void *opaque) -{ -E1000State *s = opaque; - -return chkflag (MAC ); -} -static bool e1000_tso_state_needed (void *opaque) {E1000State *s = opaque;
@@ -1451,7 +1441,6 @@ static const VMStateDescription vmstate_e1000_full_mac_state = { .name = "e1000/full_mac_state", .version_id = 1, .minimum_version_id = 1, - .needed = e1000_full_mac_needed, .fields = (const VMStateField []) {VMSTATE_UINT32_ARRAY (mac_reg, E1000State,0x8000 ),VMSTATE_END_OF_LIST ()
@@ -1679,8 +1668,6 @@ static void pci_e1000_realize(PCIDevice *pci_dev, Error **errp)static const Property e1000_properties[] = {DEFINE_NIC_PROPERTIES (E1000State, conf), -DEFINE_PROP_BIT ("extra_mac_registers" , E1000State, - compat_flags,E1000_FLAG_MAC_BIT , true),DEFINE_PROP_BIT ("migrate_tso_props" , E1000State, compat_flags,E1000_FLAG_TSO_BIT , true), DEFINE_PROP_BIT("init-vet", E1000State,
Signature · 1 line
2.47.1
PATCH v4 7/19 hw/virtio/virtio-pci: Remove VIRTIO_PCI_FLAG_MIGRATE_EXTRA definition
In reply to Philippe Mathieu-DaudéVIRTIO_PCI_FLAG_MIGRATE_EXTRA was only used by the hw_compat_2_4[] array, via the 'migrate-extra=true' property. We removed all machines using that array, lets remove all the code around VIRTIO_PCI_FLAG_MIGRATE_EXTRA.
--- include/hw/virtio/virtio-pci.h | 4 ---- hw/virtio/virtio-pci.c | 6 +----- 2 files changed, 1 insertion(+), 9 deletions(-)
include/hw/virtio/virtio-pci.h +0 -4
@@ -32,7 +32,6 @@ DECLARE_OBJ_CHECKERS(VirtioPCIBusState, VirtioPCIBusClass,enum {VIRTIO_PCI_FLAG_BUS_MASTER_BUG_MIGRATION_BIT ,VIRTIO_PCI_FLAG_USE_IOEVENTFD_BIT , -VIRTIO_PCI_FLAG_MIGRATE_EXTRA_BIT ,VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY_BIT ,VIRTIO_PCI_FLAG_DISABLE_PCIE_BIT ,VIRTIO_PCI_FLAG_PAGE_PER_VQ_BIT ,
@@ -57,9 +56,6 @@ enum {/* virtio version flags */ #define VIRTIO_PCI_FLAG_DISABLE_PCIE (1 << VIRTIO_PCI_FLAG_DISABLE_PCIE_BIT) -/* migrate extra state */ -#define VIRTIO_PCI_FLAG_MIGRATE_EXTRA (1 << VIRTIO_PCI_FLAG_MIGRATE_EXTRA_BIT) -/* have pio notification for modern device ? */ #define VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY \ (1 << VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY_BIT)
hw/virtio/virtio-pci.c +1 -5
@@ -146,9 +146,7 @@ static const VMStateDescription vmstate_virtio_pci = {static bool virtio_pci_has_extra_state (DeviceState *d) { -VirtIOPCIProxy *proxy =to_virtio_pci_proxy (d); - -return proxy->flags &VIRTIO_PCI_FLAG_MIGRATE_EXTRA ; +return true; }static void virtio_pci_save_extra_state (DeviceState *d,QEMUFile *f)
@@ -2349,8 +2347,6 @@ static void virtio_pci_bus_reset_hold(Object *obj, ResetType type)static const Property virtio_pci_properties[] = {DEFINE_PROP_BIT ("virtio-pci-bus-master-bug-migration" , VirtIOPCIProxy, flags,VIRTIO_PCI_FLAG_BUS_MASTER_BUG_MIGRATION_BIT , false), -DEFINE_PROP_BIT ("migrate-extra" , VirtIOPCIProxy, flags, -VIRTIO_PCI_FLAG_MIGRATE_EXTRA_BIT , true),DEFINE_PROP_BIT ("modern-pio-notify" , VirtIOPCIProxy, flags,VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY_BIT , false), DEFINE_PROP_BIT("x-disable-pcie", VirtIOPCIProxy, flags,
Signature · 1 line
2.47.1
PATCH v4 8/19 hw/virtio/virtio-pci: Remove VIRTIO_PCI_FLAG_DISABLE_PCIE definition
In reply to Philippe Mathieu-DaudéVIRTIO_PCI_FLAG_DISABLE_PCIE was only used by the hw_compat_2_4[] array, via the 'x-disable-pcie=false' property. We removed all machines using that array, lets remove all the code around VIRTIO_PCI_FLAG_DISABLE_PCIE (see commit 9a4c0e220d8 for similar VIRTIO_PCI_FLAG_* enum removal).
--- include/hw/virtio/virtio-pci.h | 4 ---- hw/virtio/virtio-pci.c | 5 +---- 2 files changed, 1 insertion(+), 8 deletions(-)
include/hw/virtio/virtio-pci.h +0 -4
@@ -33,7 +33,6 @@ enum {VIRTIO_PCI_FLAG_BUS_MASTER_BUG_MIGRATION_BIT ,VIRTIO_PCI_FLAG_USE_IOEVENTFD_BIT ,VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY_BIT , -VIRTIO_PCI_FLAG_DISABLE_PCIE_BIT ,VIRTIO_PCI_FLAG_PAGE_PER_VQ_BIT ,VIRTIO_PCI_FLAG_ATS_BIT , VIRTIO_PCI_FLAG_INIT_DEVERR_BIT,
@@ -53,9 +52,6 @@ enum {* vcpu thread using ioeventfd for some devices. */ #define VIRTIO_PCI_FLAG_USE_IOEVENTFD (1 << VIRTIO_PCI_FLAG_USE_IOEVENTFD_BIT) -/* virtio version flags */ -#define VIRTIO_PCI_FLAG_DISABLE_PCIE (1 << VIRTIO_PCI_FLAG_DISABLE_PCIE_BIT) -/* have pio notification for modern device ? */ #define VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY \ (1 << VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY_BIT)
hw/virtio/virtio-pci.c +1 -4
@@ -2349,8 +2349,6 @@ static const Property virtio_pci_properties[] = { VIRTIO_PCI_FLAG_BUS_MASTER_BUG_MIGRATION_BIT, false),DEFINE_PROP_BIT ("modern-pio-notify" , VirtIOPCIProxy, flags,VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY_BIT , false), -DEFINE_PROP_BIT ("x-disable-pcie" , VirtIOPCIProxy, flags, -VIRTIO_PCI_FLAG_DISABLE_PCIE_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,
@@ -2379,8 +2377,7 @@ static void virtio_pci_dc_realize(DeviceState *qdev, Error **errp)VirtIOPCIProxy *proxy =VIRTIO_PCI (qdev);PCIDevice *pci_dev = &proxy->pci_dev; -if (!(proxy->flags &VIRTIO_PCI_FLAG_DISABLE_PCIE ) && -virtio_pci_modern (proxy)) { +if (virtio_pci_modern (proxy)) { pci_dev->cap_present |=QEMU_PCI_CAP_EXPRESS ; }
Signature · 1 line
2.47.1
Re: [PATCH v4 08/19] hw/virtio/virtio-pci: Remove VIRTIO_PCI_FLAG_DISABLE_PCIE definition
Quoted 10 lines · Philippe Mathieu-Daudé
VIRTIO_PCI_FLAG_DISABLE_PCIEwas only used by the hw_compat_2_4[] array, via the 'x-disable-pcie=false' property. We removed all machines using that array, lets remove all the code aroundVIRTIO_PCI_FLAG_DISABLE_PCIE(see commit9a4c0e220d8for similar VIRTIO_PCI_FLAG_* enum removal).Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Thomas Huth <[email protected]>Reviewed-by Zhao Liu <[email protected]>
Quoted 50 lines
--- include/hw/virtio/virtio-pci.h | 4 ---- hw/virtio/virtio-pci.c | 5 +---- 2 files changed, 1 insertion(+), 8 deletions(-)include/hw/virtio/virtio-pci.h +0
-4@@ -33,7 +33,6 @@ enum {VIRTIO_PCI_FLAG_BUS_MASTER_BUG_MIGRATION_BIT ,VIRTIO_PCI_FLAG_USE_IOEVENTFD_BIT ,VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY_BIT , -VIRTIO_PCI_FLAG_DISABLE_PCIE_BIT ,VIRTIO_PCI_FLAG_PAGE_PER_VQ_BIT ,VIRTIO_PCI_FLAG_ATS_BIT , VIRTIO_PCI_FLAG_INIT_DEVERR_BIT,@@ -53,9 +52,6 @@ enum {* vcpu thread using ioeventfd for some devices. */ #define VIRTIO_PCI_FLAG_USE_IOEVENTFD (1 << VIRTIO_PCI_FLAG_USE_IOEVENTFD_BIT) -/* virtio version flags */ -#define VIRTIO_PCI_FLAG_DISABLE_PCIE (1 << VIRTIO_PCI_FLAG_DISABLE_PCIE_BIT) -/* have pio notification for modern device ? */ #define VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY \ (1 << VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY_BIT)hw/virtio/virtio-pci.c +1
-4@@ -2349,8 +2349,6 @@ static const Property virtio_pci_properties[] = { VIRTIO_PCI_FLAG_BUS_MASTER_BUG_MIGRATION_BIT, false),DEFINE_PROP_BIT ("modern-pio-notify" , VirtIOPCIProxy, flags,VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY_BIT , false), -DEFINE_PROP_BIT ("x-disable-pcie" , VirtIOPCIProxy, flags, -VIRTIO_PCI_FLAG_DISABLE_PCIE_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,@@ -2379,8 +2377,7 @@ static void virtio_pci_dc_realize(DeviceState *qdev, Error **errp)VirtIOPCIProxy *proxy =VIRTIO_PCI (qdev);PCIDevice *pci_dev = &proxy->pci_dev; -if (!(proxy->flags &VIRTIO_PCI_FLAG_DISABLE_PCIE ) && -virtio_pci_modern (proxy)) { +if (virtio_pci_modern (proxy)) { pci_dev->cap_present |=QEMU_PCI_CAP_EXPRESS ; }
PATCH v4 9/19 hw/i386/pc: Remove deprecated pc-q35-2.5 and pc-i440fx-2.5 machines
In reply to Philippe Mathieu-DaudéThese machines has been supported for a period of more than 6 years. According to our versioned machine support policy (see commit ce80c4fa6ff "docs: document special exception for machine type deprecation & removal") they can now be removed.
Remove the now unused empty pc_compat_2_5[] array.
--- include/hw/i386/pc.h | 3 --- hw/i386/pc.c | 3 --- hw/i386/pc_piix.c | 13 ------------- hw/i386/pc_q35.c | 13 ------------- 4 files changed, 32 deletions(-)
include/hw/i386/pc.h +0 -3
@@ -298,9 +298,6 @@ extern const size_t pc_compat_2_7_len;extern GlobalProperty pc_compat_2_6[];extern const size_t pc_compat_2_6_len; -extern GlobalProperty pc_compat_2_5[]; -extern const size_t pc_compat_2_5_len; -#define DEFINE_PC_MACHINE (suffix, namestr, initfn, optsfn) \ static void pc_machine_##suffix##_class_init(ObjectClass *oc, \ const void *data) \
hw/i386/pc.c +0 -3
@@ -259,9 +259,6 @@ GlobalProperty pc_compat_2_6[] = { };const size_t pc_compat_2_6_len =G_N_ELEMENTS (pc_compat_2_6); -GlobalProperty pc_compat_2_5[] = {}; -const size_t pc_compat_2_5_len =G_N_ELEMENTS (pc_compat_2_5); -/* * @PC_FW_DATA: * Size of the chunk of memory at the top of RAM for the BIOS ACPI tables
hw/i386/pc_piix.c +0 -13
@@ -778,19 +778,6 @@ static void pc_i440fx_machine_2_6_options(MachineClass *m)DEFINE_I440FX_MACHINE (2 ,6 ); -static void pc_i440fx_machine_2_5_options (MachineClass *m) -{ -X86MachineClass *x86mc =X86_MACHINE_CLASS (m); - -pc_i440fx_machine_2_6_options (m); - x86mc->save_tsc_khz = false; - m->legacy_fw_cfg_order =1 ; -compat_props_add (m->compat_props, hw_compat_2_5, hw_compat_2_5_len); -compat_props_add (m->compat_props, pc_compat_2_5, pc_compat_2_5_len); -} - -DEFINE_I440FX_MACHINE (2 ,5 ); -#ifdef CONFIG_ISAPC static void isapc_machine_options (MachineClass *m) {
hw/i386/pc_q35.c +0 -13
@@ -672,16 +672,3 @@ static void pc_q35_machine_2_6_options(MachineClass *m) }DEFINE_Q35_MACHINE (2 ,6 ); - -static void pc_q35_machine_2_5_options (MachineClass *m) -{ -X86MachineClass *x86mc =X86_MACHINE_CLASS (m); - -pc_q35_machine_2_6_options (m); - x86mc->save_tsc_khz = false; - m->legacy_fw_cfg_order =1 ; -compat_props_add (m->compat_props, hw_compat_2_5, hw_compat_2_5_len); -compat_props_add (m->compat_props, pc_compat_2_5, pc_compat_2_5_len); -} - -DEFINE_Q35_MACHINE (2 ,5 );
Signature · 1 line
2.47.1
Re: [PATCH v4 09/19] hw/i386/pc: Remove deprecated pc-q35-2.5 and pc-i440fx-2.5 machines
Quoted 10 lines · Philippe Mathieu-Daudé
These machines has been supported for a period of more than 6 years. According to our versioned machine support policy (see commit
ce80c4fa6ff"docs: document special exception for machine type deprecation & removal") they can now be removed.Remove the now unused empty pc_compat_2_5[] array.
Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Zhao Liu <[email protected]>
PATCH v4 10/19 hw/i386/x86: Remove X86MachineClass::save_tsc_khz field
In reply to Philippe Mathieu-DaudéThe X86MachineClass::save_tsc_khz boolean was only used by the pc-q35-2.5 and pc-i440fx-2.5 machines, which got removed. Remove it and simplify tsc_khz_needed().
--- include/hw/i386/x86.h | 5 ----- hw/i386/x86.c | 1 - target/i386/machine.c | 5 ++--- 3 files changed, 2 insertions(+), 9 deletions(-)
include/hw/i386/x86.h +0 -5
@@ -27,13 +27,8 @@#include "qom/object.h" struct X86MachineClass { -/*< private >*/ MachineClass parent; -/*< public >*/ - -/* TSC rate migration: */ -bool save_tsc_khz;/* use DMA capable linuxboot option rom */ bool fwcfg_dma_enabled;/* CPU and apic information: */
hw/i386/x86.c +0 -1
@@ -382,7 +382,6 @@ static void x86_machine_class_init(ObjectClass *oc, const void *data) 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->save_tsc_khz = true; x86mc->fwcfg_dma_enabled = true; nc->nmi_monitor_handler = x86_nmi;
target/i386/machine.c +2 -3
@@ -1060,9 +1060,8 @@ static bool tsc_khz_needed(void *opaque) {X86CPU *cpu = opaque;CPUX86State *env = &cpu->env; -MachineClass *mc =MACHINE_GET_CLASS (qdev_get_machine ()); -X86MachineClass *x86mc =X86_MACHINE_CLASS (mc); -return env->tsc_khz && x86mc->save_tsc_khz; + +return env->tsc_khz; }static const VMStateDescription vmstate_tsc_khz = {
Signature · 1 line
2.47.1
Re: [PATCH v4 10/19] hw/i386/x86: Remove X86MachineClass::save_tsc_khz field
Quoted 8 lines · Philippe Mathieu-Daudé
The
X86MachineClass::save_tsc_khzboolean was only used by the pc-q35-2.5 and pc-i440fx-2.5 machines, which got removed. Remove it and simplifytsc_khz_needed().Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Thomas Huth <[email protected]>Reviewed-by Zhao Liu <[email protected]>
PATCH v4 11/19 hw/nvram/fw_cfg: Remove legacy FW_CFG_ORDER_OVERRIDE
In reply to Philippe Mathieu-DaudéThe MachineClass::legacy_fw_cfg_order boolean was only used by the pc-q35-2.5 and pc-i440fx-2.5 machines, which got removed. Remove it along with:
- FW_CFG_ORDER_OVERRIDE_* definitions
- fw_cfg_set_order_override()
- fw_cfg_reset_order_override()
- fw_cfg_order[]
- rom_set_order_override()
- rom_reset_order_override()
Simplify CLI and pc_vga_init() / pc_nic_init().
--- include/hw/boards.h | 3 +- include/hw/loader.h | 2 - include/hw/nvram/fw_cfg.h | 10 ---- hw/core/loader.c | 14 ----- hw/i386/pc.c | 7 +-- hw/nvram/fw_cfg.c | 109 +++----------------------------------- system/vl.c | 5 -- 7 files changed, 9 insertions(+), 141 deletions(-)
include/hw/boards.h +1 -2
@@ -286,8 +286,7 @@ struct MachineClass { no_parallel:1, no_floppy:1, no_cdrom:1, - pci_allow_0_address:1, - legacy_fw_cfg_order:1; + pci_allow_0_address:1;bool auto_create_sdcard;bool is_default;const char *default_machine_opts;
include/hw/loader.h +0 -2
@@ -270,8 +270,6 @@ int rom_add_elf_program(const char *name, GMappedFile *mapped_file, void *data, AddressSpace *as);int rom_check_and_register_reset (void );void rom_set_fw (FWCfgState *f); -void rom_set_order_override (int order); -void rom_reset_order_override (void );/** * rom_transaction_begin:
include/hw/nvram/fw_cfg.h +0 -10
@@ -42,14 +42,6 @@ struct FWCfgDataGeneratorClass {typedef struct fw_cfg_file FWCfgFile ; -#define FW_CFG_ORDER_OVERRIDE_VGA 70 -#define FW_CFG_ORDER_OVERRIDE_NIC 80 -#define FW_CFG_ORDER_OVERRIDE_USER 100 -#define FW_CFG_ORDER_OVERRIDE_DEVICE 110 - -void fw_cfg_set_order_override (FWCfgState *fw_cfg,int order); -void fw_cfg_reset_order_override (FWCfgState *fw_cfg); - typedef struct FWCfgFiles { uint32_t count; FWCfgFile f[];
@@ -75,8 +67,6 @@ struct FWCfgState {uint32_t cur_offset;Notifier machine_ready; -int fw_cfg_order_override; -bool dma_enabled;dma_addr_t dma_addr;AddressSpace *dma_as;
hw/core/loader.c +0 -14
@@ -1333,20 +1333,6 @@ void rom_set_fw(FWCfgState *f) fw_cfg = f; } -void rom_set_order_override (int order) -{ -if (!fw_cfg) -return ; -fw_cfg_set_order_override (fw_cfg, order); -} - -void rom_reset_order_override (void ) -{ -if (!fw_cfg) -return ; -fw_cfg_reset_order_override (fw_cfg); -} -void rom_transaction_begin (void ) {Rom *rom;
hw/i386/pc.c +1 -6
@@ -1033,7 +1033,6 @@ DeviceState *pc_vga_init(ISABus *isa_bus, PCIBus *pci_bus) {DeviceState *dev =NULL ; -rom_set_order_override (FW_CFG_ORDER_OVERRIDE_VGA );if (pci_bus) {PCIDevice *pcidev =pci_vga_init (pci_bus); dev = pcidev ? &pcidev->qdev :NULL ;
@@ -1041,7 +1040,7 @@ DeviceState *pc_vga_init(ISABus *isa_bus, PCIBus *pci_bus)ISADevice *isadev =isa_vga_init (isa_bus); dev = isadev ?DEVICE (isadev) :NULL ; } -rom_reset_order_override (); +return dev; }
@@ -1231,8 +1230,6 @@ void pc_nic_init(PCMachineClass *pcmc, ISABus *isa_bus, PCIBus *pci_bus)bool default_is_ne2k =g_str_equal (mc->default_nic,TYPE_ISA_NE2000 );NICInfo *nd; -rom_set_order_override (FW_CFG_ORDER_OVERRIDE_NIC ); -while ((nd =qemu_find_nic_info (TYPE_ISA_NE2000 , default_is_ne2k,NULL ))) {pc_init_ne2k_isa (isa_bus, nd, &error_fatal); }
@@ -1241,8 +1238,6 @@ void pc_nic_init(PCMachineClass *pcmc, ISABus *isa_bus, PCIBus *pci_bus)if (pci_bus) {pci_init_nic_devices (pci_bus, mc->default_nic); } - -rom_reset_order_override (); }void pc_i8259_create (ISABus *isa_bus,qemu_irq *i8259_irqs)
hw/nvram/fw_cfg.c +7 -102
145 lines not on this page · raw message
system/vl.c +0 -5
@@ -1192,10 +1192,7 @@ static int parse_fw_cfg(void *opaque, QemuOpts *opts, Error **errp)return -1 ; } } -/* For legacy, keep user files in a specific global order. */ -fw_cfg_set_order_override (fw_cfg,FW_CFG_ORDER_OVERRIDE_USER );fw_cfg_add_file (fw_cfg, name, buf, size); -fw_cfg_reset_order_override (fw_cfg);return 0 ; }
@@ -2745,7 +2742,6 @@ static void qemu_create_cli_devices(void) }/* init generic devices */ -rom_set_order_override (FW_CFG_ORDER_OVERRIDE_DEVICE );qemu_opts_foreach (qemu_find_opts ("device" ), device_init_func,NULL , &error_fatal); QTAILQ_FOREACH(opt, &device_opts, next) {
@@ -2756,7 +2752,6 @@ static void qemu_create_cli_devices(void)assert (ret_data ==NULL );/* error_fatal aborts */ loc_pop (&opt->loc); } -rom_reset_order_override (); }static bool qemu_machine_creation_done (Error **errp)
Signature · 1 line
2.47.1
Re: [PATCH v4 11/19] hw/nvram/fw_cfg: Remove legacy FW_CFG_ORDER_OVERRIDE
Quoted 11 lines · Philippe Mathieu-Daudé
The
MachineClass::legacy_fw_cfg_orderboolean was only used by the pc-q35-2.5 and pc-i440fx-2.5 machines, which got removed. Remove it along with:- FW_CFG_ORDER_OVERRIDE_* definitions
-fw_cfg_set_order_override()
-fw_cfg_reset_order_override()
- fw_cfg_order[]
-rom_set_order_override()
-rom_reset_order_override()
Quoted 5 lines
Simplify CLI and
pc_vga_init()/pc_nic_init().Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Zhao Liu <[email protected]>
PATCH v4 12/19 hw/core/machine: Remove hw_compat_2_5[] array
In reply to Philippe Mathieu-DaudéThe hw_compat_2_5[] array was only used by the pc-q35-2.5 and pc-i440fx-2.5 machines, which got removed. Remove it.
--- include/hw/boards.h | 3 --- hw/core/machine.c | 9 --------- 2 files changed, 12 deletions(-)
include/hw/boards.h +0 -3
@@ -862,7 +862,4 @@ extern const size_t hw_compat_2_7_len;extern GlobalProperty hw_compat_2_6[];extern const size_t hw_compat_2_6_len; -extern GlobalProperty hw_compat_2_5[]; -extern const size_t hw_compat_2_5_len; -#endif
hw/core/machine.c +0 -9
@@ -283,15 +283,6 @@ GlobalProperty hw_compat_2_6[] = { };const size_t hw_compat_2_6_len =G_N_ELEMENTS (hw_compat_2_6); -GlobalProperty hw_compat_2_5[] = { - {"isa-fdc" ,"fallback" ,"144" }, - {"pvscsi" ,"x-old-pci-configuration" ,"on" }, - {"pvscsi" ,"x-disable-pcie" ,"on" }, - {"vmxnet3" ,"x-old-msi-offsets" ,"on" }, - {"vmxnet3" ,"x-disable-pcie" ,"on" }, -}; -const size_t hw_compat_2_5_len =G_N_ELEMENTS (hw_compat_2_5); -MachineState *current_machine;static char *machine_get_kernel (Object *obj,Error **errp)
Signature · 1 line
2.47.1
Re: [PATCH v4 12/19] hw/core/machine: Remove hw_compat_2_5[] array
Quoted 6 lines · Philippe Mathieu-Daudé
The hw_compat_2_5[] array was only used by the pc-q35-2.5 and pc-i440fx-2.5 machines, which got removed. Remove it.
Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Zhao Liu <[email protected]>
PATCH v4 13/19 hw/block/fdc-isa: Remove 'fallback' property
In reply to Philippe Mathieu-DaudéThe "fallback" property was only used by the hw_compat_2_5[] array, as 'fallback=144'. We removed all machines using that array, lets remove ISA floppy drive 'fallback' property, manually setting the default value in isabus_fdc_realize().
--- hw/block/fdc-isa.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-)
hw/block/fdc-isa.c +1 -3
@@ -97,6 +97,7 @@ static void isabus_fdc_realize(DeviceState *dev, Error **errp) isa->iobase, fdc_portio_list, fdctrl, "fdc"); + fdctrl->fallback =FLOPPY_DRIVE_TYPE_288 ; fdctrl->irq =isa_bus_get_irq (bus, isa->irq); fdctrl->dma_chann = isa->dma;if (fdctrl->dma_chann !=-1 ) {
@@ -293,9 +294,6 @@ static const Property isa_fdc_properties[] = {DEFINE_PROP_SIGNED ("fdtypeB" , FDCtrlISABus, state.qdev_for_drives[1 ].type,FLOPPY_DRIVE_TYPE_AUTO , qdev_prop_fdc_drive_type, FloppyDriveType), -DEFINE_PROP_SIGNED ("fallback" , FDCtrlISABus, state.fallback, -FLOPPY_DRIVE_TYPE_288 , qdev_prop_fdc_drive_type, - FloppyDriveType), };static void isabus_fdc_class_init (ObjectClass *klass,const void *data)
Signature · 1 line
2.47.1
Re: [PATCH v4 13/19] hw/block/fdc-isa: Remove 'fallback' property
Quoted 11 lines · Philippe Mathieu-Daudé
The "fallback" property was only used by the hw_compat_2_5[] array, as 'fallback=144'. We removed all machines using that array, lets remove ISA floppy drive 'fallback' property, manually setting the default value in
isabus_fdc_realize().Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Zhao Liu <[email protected]>--- hw/block/fdc-isa.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-)
FWIW, this needs a fixup for iotest 172:
https://gitlab.com/thuth/qemu/-/jobs/10166450223#L466
Thomas
Re: [PATCH v4 13/19] hw/block/fdc-isa: Remove 'fallback' property
Quoted 13 lines · Thomas Huth
Quoted 11 lines · Philippe Mathieu-Daudé
The "fallback" property was only used by the hw_compat_2_5[] array, as 'fallback=144'. We removed all machines using that array, lets remove ISA floppy drive 'fallback' property, manually setting the default value in
isabus_fdc_realize().Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Zhao Liu <[email protected]>---
hw/block/fdc-isa.c| 4 +---1 file changed, 1 insertion(+), 3 deletions(-)
Quoted 3 lines
FWIW, this needs a fixup for iotest 172:
FYI, since I was testing this series anyway, I'll fix up this patch with:
tests/qemu-iotests/172.out +0 -3
@@ -14,7 +14,6 @@ Testing: dma = 2 (0x2) fdtypeA = "auto" fdtypeB = "auto" - fallback = "288" bus: floppy-bus.0 type floppy-bus dev: floppy, id ""
@@ -43,7 +42,6 @@ Testing: -fda TEST_DIR/t.qcow2 dma = 2 (0x2) fdtypeA = "auto" fdtypeB = "auto" - fallback = "288" bus: floppy-bus.0 type floppy-bus dev: floppy, id ""
@@ -79,7 +77,6 @@ Testing: -fdb TEST_DIR/t.qcow2 dma = 2 (0x2) fdtypeA = "auto" fdtypeB = "auto" - fallback = "288" bus: floppy-bus.0 type floppy-bus dev: floppy, id ""
etc.
and I will queue this series (without the "Remove X86CPU::check_cpuid field" patch as mentioned by Xiaoyao Li), unless Paolo or another x86 maintainer wants to do this instead (please let me know!).
Thomas
Re: [PATCH v4 13/19] hw/block/fdc-isa: Remove 'fallback' property
Quoted 19 lines · Thomas Huth
Quoted 13 lines · Thomas Huth
Quoted 11 lines · Philippe Mathieu-Daudé
The "fallback" property was only used by the hw_compat_2_5[] array, as 'fallback=144'. We removed all machines using that array, lets remove ISA floppy drive 'fallback' property, manually setting the default value in
isabus_fdc_realize().Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Zhao Liu <[email protected]>---
hw/block/fdc-isa.c| 4 +---1 file changed, 1 insertion(+), 3 deletions(-)Quoted 3 lines
FWIW, this needs a fixup for iotest 172:
FYI, since I was testing this series anyway, I'll fix up this patch
...
Quoted 1 line
and I will queue this series (without the "Remove
X86CPU::check_cpuidfield" patch as mentioned by Xiaoyao Li), unless Paolo or another x86 maintainer wants to do this instead (please let me know!).
Hmm, looking at this twice, hw/block/fdc-sysbus.c has a "fallback" property, too, and there are people out there who set this property manually, e.g.:
https://github.com/mariuz/linux-0.01/issues/4#issuecomment-974768898
So it seems like this property is meant as a config knob for the user, indeed, not just an internal one. I'll drop this patch for now.
Thomas
Re: [PATCH v4 13/19] hw/block/fdc-isa: Remove 'fallback' property
In reply to Thomas HuthQuoted 48 lines · Thomas Huth
Quoted 13 lines · Thomas Huth
Quoted 11 lines · Philippe Mathieu-Daudé
The "fallback" property was only used by the hw_compat_2_5[] array, as 'fallback=144'. We removed all machines using that array, lets remove ISA floppy drive 'fallback' property, manually setting the default value in
isabus_fdc_realize().Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Zhao Liu <[email protected]>---
hw/block/fdc-isa.c| 4 +---1 file changed, 1 insertion(+), 3 deletions(-)Quoted 3 lines
FWIW, this needs a fixup for iotest 172:
FYI, since I was testing this series anyway, I'll fix up this patch with:
tests/qemu-iotests/172.out +0
-3@@ -14,7 +14,6 @@ Testing: dma = 2 (0x2) fdtypeA = "auto" fdtypeB = "auto" - fallback = "288" bus: floppy-bus.0 type floppy-bus dev: floppy, id ""@@ -43,7 +42,6 @@ Testing: -fda TEST_DIR/t.qcow2 dma = 2 (0x2) fdtypeA = "auto" fdtypeB = "auto" - fallback = "288" bus: floppy-bus.0 type floppy-bus dev: floppy, id ""@@ -79,7 +77,6 @@ Testing: -fdb TEST_DIR/t.qcow2 dma = 2 (0x2) fdtypeA = "auto" fdtypeB = "auto" - fallback = "288" bus: floppy-bus.0 type floppy-bus dev: floppy, id ""
Quoted 7 lines
etc.
and I will queue this series (without the "Remove
X86CPU::check_cpuidfield" patch as mentioned by Xiaoyao Li), unless Paolo or another x86 maintainer wants to do this instead (please let me know!).Thomas
No, pls go ahead.
Re: [PATCH v4 13/19] hw/block/fdc-isa: Remove 'fallback' property
Quoted 20 lines · Michael S. Tsirkin
Quoted 17 lines · Thomas Huth
Quoted 16 lines · Thomas Huth
Quoted 11 lines · Philippe Mathieu-Daudé
The "fallback" property was only used by the hw_compat_2_5[] array, as 'fallback=144'. We removed all machines using that array, lets remove ISA floppy drive 'fallback' property, manually setting the default value in
isabus_fdc_realize().Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Zhao Liu <[email protected]>--- hw/block/fdc-isa.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-)FWIW, this needs a fixup for iotest 172:
Quoted 7 lines
Quoted 5 lines
and I will queue this series (without the "Remove
X86CPU::check_cpuidfield" patch as mentioned by Xiaoyao Li), unless Paolo or another x86 maintainer wants to do this instead (please let me know!).Thomas
No, pls go ahead.
Thanks both!
PATCH v4 14/19 hw/scsi/vmw_pvscsi: Remove PVSCSI_COMPAT_OLD_PCI_CONFIGURATION definition
In reply to Philippe Mathieu-DaudéPVSCSI_COMPAT_OLD_PCI_CONFIGURATION was only used by the hw_compat_2_5[] array, via the 'x-old-pci-configuration=on' property. We removed all machines using that array, lets remove all the code around PVSCSI_COMPAT_OLD_PCI_CONFIGURATION.
--- hw/scsi/vmw_pvscsi.c | 26 +++++++------------------- 1 file changed, 7 insertions(+), 19 deletions(-)
hw/scsi/vmw_pvscsi.c +7 -19
@@ -69,17 +69,11 @@ OBJECT_DECLARE_TYPE(PVSCSIState, PVSCSIClass, PVSCSI)/* Compatibility flags for migration */ -#define PVSCSI_COMPAT_OLD_PCI_CONFIGURATION_BIT 0 -#define PVSCSI_COMPAT_OLD_PCI_CONFIGURATION \ - (1 << PVSCSI_COMPAT_OLD_PCI_CONFIGURATION_BIT)#define PVSCSI_COMPAT_DISABLE_PCIE_BIT 1#define PVSCSI_COMPAT_DISABLE_PCIE \ (1 << PVSCSI_COMPAT_DISABLE_PCIE_BIT) -#define PVSCSI_USE_OLD_PCI_CONFIGURATION (s) \ - ((s)->compat_flags & PVSCSI_COMPAT_OLD_PCI_CONFIGURATION) -#define PVSCSI_MSI_OFFSET (s) \ - (PVSCSI_USE_OLD_PCI_CONFIGURATION(s) ? 0x50 : 0x7c) +#define PVSCSI_MSI_OFFSET (0x7c)#define PVSCSI_EXP_EP_OFFSET (0x40) typedef struct PVSCSIRingInfo {
@@ -1110,7 +1104,7 @@ pvscsi_init_msi(PVSCSIState *s)int res;PCIDevice *d =PCI_DEVICE (s); - res =msi_init (d,PVSCSI_MSI_OFFSET (s),PVSCSI_MSIX_NUM_VECTORS , + res =msi_init (d,PVSCSI_MSI_OFFSET ,PVSCSI_MSIX_NUM_VECTORS ,PVSCSI_USE_64BIT ,PVSCSI_PER_VECTOR_MASK ,NULL );if (res <0 ) {trace_pvscsi_init_msi_fail (res);
@@ -1158,15 +1152,11 @@ pvscsi_realizefn(PCIDevice *pci_dev, Error **errp)trace_pvscsi_state ("init" );/* PCI subsystem ID, subsystem vendor ID, revision */ -if (PVSCSI_USE_OLD_PCI_CONFIGURATION (s)) { -pci_set_word (pci_dev->config +PCI_SUBSYSTEM_ID ,0x1000 ); - }else { -pci_set_word (pci_dev->config +PCI_SUBSYSTEM_VENDOR_ID , -PCI_VENDOR_ID_VMWARE ); -pci_set_word (pci_dev->config +PCI_SUBSYSTEM_ID , -PCI_DEVICE_ID_VMWARE_PVSCSI ); -pci_config_set_revision (pci_dev->config,0x2 ); - } +pci_set_word (pci_dev->config +PCI_SUBSYSTEM_VENDOR_ID , +PCI_VENDOR_ID_VMWARE ); +pci_set_word (pci_dev->config +PCI_SUBSYSTEM_ID , +PCI_DEVICE_ID_VMWARE_PVSCSI ); +pci_config_set_revision (pci_dev->config,0x2 );/* PCI latency timer = 255 */ pci_dev->config[PCI_LATENCY_TIMER ] =0xff ;
@@ -1298,8 +1288,6 @@ static const VMStateDescription vmstate_pvscsi = {static const Property pvscsi_properties[] = {DEFINE_PROP_UINT8 ("use_msg" , PVSCSIState, use_msg,1 ), -DEFINE_PROP_BIT ("x-old-pci-configuration" , PVSCSIState, compat_flags, -PVSCSI_COMPAT_OLD_PCI_CONFIGURATION_BIT , false),DEFINE_PROP_BIT ("x-disable-pcie" , PVSCSIState, compat_flags,PVSCSI_COMPAT_DISABLE_PCIE_BIT , false), };
Signature · 1 line
2.47.1
Re: [PATCH v4 14/19] hw/scsi/vmw_pvscsi: Remove PVSCSI_COMPAT_OLD_PCI_CONFIGURATION definition
Quoted 9 lines · Philippe Mathieu-Daudé
PVSCSI_COMPAT_OLD_PCI_CONFIGURATIONwas only used by the hw_compat_2_5[] array, via the 'x-old-pci-configuration=on' property. We removed all machines using that array, lets remove all the code aroundPVSCSI_COMPAT_OLD_PCI_CONFIGURATION.Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Thomas Huth <[email protected]>Reviewed-by Zhao Liu <[email protected]>
Quoted 66 lines
--- hw/scsi/vmw_pvscsi.c | 26 +++++++------------------- 1 file changed, 7 insertions(+), 19 deletions(-)hw/scsi/vmw_pvscsi.c +7
-19@@ -69,17 +69,11 @@ OBJECT_DECLARE_TYPE(PVSCSIState, PVSCSIClass, PVSCSI)/* Compatibility flags for migration */ -#define PVSCSI_COMPAT_OLD_PCI_CONFIGURATION_BIT 0 -#define PVSCSI_COMPAT_OLD_PCI_CONFIGURATION \ - (1 << PVSCSI_COMPAT_OLD_PCI_CONFIGURATION_BIT)#define PVSCSI_COMPAT_DISABLE_PCIE_BIT 1#define PVSCSI_COMPAT_DISABLE_PCIE \ (1 << PVSCSI_COMPAT_DISABLE_PCIE_BIT) -#define PVSCSI_USE_OLD_PCI_CONFIGURATION (s) \ - ((s)->compat_flags & PVSCSI_COMPAT_OLD_PCI_CONFIGURATION) -#define PVSCSI_MSI_OFFSET (s) \ - (PVSCSI_USE_OLD_PCI_CONFIGURATION(s) ? 0x50 : 0x7c) +#define PVSCSI_MSI_OFFSET (0x7c)#define PVSCSI_EXP_EP_OFFSET (0x40) typedef struct PVSCSIRingInfo {@@ -1110,7 +1104,7 @@ pvscsi_init_msi(PVSCSIState *s)int res;PCIDevice *d =PCI_DEVICE (s); - res =msi_init (d,PVSCSI_MSI_OFFSET (s),PVSCSI_MSIX_NUM_VECTORS , + res =msi_init (d,PVSCSI_MSI_OFFSET ,PVSCSI_MSIX_NUM_VECTORS ,PVSCSI_USE_64BIT ,PVSCSI_PER_VECTOR_MASK ,NULL );if (res <0 ) {trace_pvscsi_init_msi_fail (res);@@ -1158,15 +1152,11 @@ pvscsi_realizefn(PCIDevice *pci_dev, Error **errp)trace_pvscsi_state ("init" );/* PCI subsystem ID, subsystem vendor ID, revision */ -if (PVSCSI_USE_OLD_PCI_CONFIGURATION (s)) { -pci_set_word (pci_dev->config +PCI_SUBSYSTEM_ID ,0x1000 ); - }else { -pci_set_word (pci_dev->config +PCI_SUBSYSTEM_VENDOR_ID , -PCI_VENDOR_ID_VMWARE ); -pci_set_word (pci_dev->config +PCI_SUBSYSTEM_ID , -PCI_DEVICE_ID_VMWARE_PVSCSI ); -pci_config_set_revision (pci_dev->config,0x2 ); - } +pci_set_word (pci_dev->config +PCI_SUBSYSTEM_VENDOR_ID , +PCI_VENDOR_ID_VMWARE ); +pci_set_word (pci_dev->config +PCI_SUBSYSTEM_ID , +PCI_DEVICE_ID_VMWARE_PVSCSI ); +pci_config_set_revision (pci_dev->config,0x2 );/* PCI latency timer = 255 */ pci_dev->config[PCI_LATENCY_TIMER ] =0xff ;@@ -1298,8 +1288,6 @@ static const VMStateDescription vmstate_pvscsi = {static const Property pvscsi_properties[] = {DEFINE_PROP_UINT8 ("use_msg" , PVSCSIState, use_msg,1 ), -DEFINE_PROP_BIT ("x-old-pci-configuration" , PVSCSIState, compat_flags, -PVSCSI_COMPAT_OLD_PCI_CONFIGURATION_BIT , false),DEFINE_PROP_BIT ("x-disable-pcie" , PVSCSIState, compat_flags,PVSCSI_COMPAT_DISABLE_PCIE_BIT , false), };
PATCH v4 15/19 hw/scsi/vmw_pvscsi: Remove PVSCSI_COMPAT_DISABLE_PCIE_BIT definition
In reply to Philippe Mathieu-DaudéPVSCSI_COMPAT_DISABLE_PCIE_BIT was only used by the hw_compat_2_5[] array, via the 'x-disable-pcie=on' property. We removed all machines using that array, lets remove all the code around PVSCSI_COMPAT_DISABLE_PCIE_BIT, including the now unused PVSCSIState::compat_flags field.
--- hw/scsi/vmw_pvscsi.c | 30 +----------------------------- 1 file changed, 1 insertion(+), 29 deletions(-)
hw/scsi/vmw_pvscsi.c +1 -29
@@ -68,11 +68,6 @@ struct PVSCSIClass { OBJECT_DECLARE_TYPE(PVSCSIState, PVSCSIClass, PVSCSI) -/* Compatibility flags for migration */ -#define PVSCSI_COMPAT_DISABLE_PCIE_BIT 1 -#define PVSCSI_COMPAT_DISABLE_PCIE \ - (1 << PVSCSI_COMPAT_DISABLE_PCIE_BIT) -#define PVSCSI_MSI_OFFSET (0x7c)#define PVSCSI_EXP_EP_OFFSET (0x40)
@@ -123,8 +118,6 @@ struct PVSCSIState {uint8_t msi_used;/* For migration compatibility */ PVSCSIRingInfo rings;/* Data transfer rings manager */ uint32_t resetting;/* Reset in progress */ - -uint32_t compat_flags; }; typedef struct PVSCSIRequest {
@@ -1224,21 +1217,8 @@ pvscsi_post_load(void *opaque, int version_id)return 0 ; } -static bool pvscsi_vmstate_need_pcie_device (void *opaque) -{ -PVSCSIState *s =PVSCSI (opaque); - -return !(s->compat_flags &PVSCSI_COMPAT_DISABLE_PCIE ); -} - -static bool pvscsi_vmstate_test_pci_device (void *opaque,int version_id) -{ -return !pvscsi_vmstate_need_pcie_device (opaque); -} -static const VMStateDescription vmstate_pvscsi_pcie_device = { .name ="pvscsi/pcie" , - .needed = pvscsi_vmstate_need_pcie_device, .fields = (const VMStateField []) {VMSTATE_PCI_DEVICE (parent_obj, PVSCSIState),VMSTATE_END_OF_LIST ()
@@ -1252,9 +1232,6 @@ static const VMStateDescription vmstate_pvscsi = { .pre_save = pvscsi_pre_save, .post_load = pvscsi_post_load, .fields = (const VMStateField []) { -VMSTATE_STRUCT_TEST (parent_obj, PVSCSIState, - pvscsi_vmstate_test_pci_device,0 , - vmstate_pci_device, PCIDevice),VMSTATE_UINT8 (msi_used, PVSCSIState),VMSTATE_UINT32 (resetting, PVSCSIState),VMSTATE_UINT64 (reg_interrupt_status, PVSCSIState),
@@ -1288,19 +1265,14 @@ static const VMStateDescription vmstate_pvscsi = {static const Property pvscsi_properties[] = {DEFINE_PROP_UINT8 ("use_msg" , PVSCSIState, use_msg,1 ), -DEFINE_PROP_BIT ("x-disable-pcie" , PVSCSIState, compat_flags, -PVSCSI_COMPAT_DISABLE_PCIE_BIT , false), };static void pvscsi_realize (DeviceState *qdev,Error **errp) {PVSCSIClass *pvs_c =PVSCSI_GET_CLASS (qdev);PCIDevice *pci_dev =PCI_DEVICE (qdev); -PVSCSIState *s =PVSCSI (qdev); -if (!(s->compat_flags &PVSCSI_COMPAT_DISABLE_PCIE )) { - pci_dev->cap_present |=QEMU_PCI_CAP_EXPRESS ; - } + pci_dev->cap_present |=QEMU_PCI_CAP_EXPRESS ; pvs_c->parent_dc_realize (qdev, errp); }
Signature · 1 line
2.47.1
Re: [PATCH v4 15/19] hw/scsi/vmw_pvscsi: Remove PVSCSI_COMPAT_DISABLE_PCIE_BIT definition
Quoted 10 lines · Philippe Mathieu-Daudé
PVSCSI_COMPAT_DISABLE_PCIE_BITwas only used by the hw_compat_2_5[] array, via the 'x-disable-pcie=on' property. We removed all machines using that array, lets remove all the code aroundPVSCSI_COMPAT_DISABLE_PCIE_BIT, including the now unusedPVSCSIState::compat_flagsfield.Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Thomas Huth <[email protected]>Reviewed-by Zhao Liu <[email protected]>
Quoted 82 lines
--- hw/scsi/vmw_pvscsi.c | 30 +----------------------------- 1 file changed, 1 insertion(+), 29 deletions(-)hw/scsi/vmw_pvscsi.c +1
-29@@ -68,11 +68,6 @@ struct PVSCSIClass { OBJECT_DECLARE_TYPE(PVSCSIState, PVSCSIClass, PVSCSI) -/* Compatibility flags for migration */ -#define PVSCSI_COMPAT_DISABLE_PCIE_BIT 1 -#define PVSCSI_COMPAT_DISABLE_PCIE \ - (1 << PVSCSI_COMPAT_DISABLE_PCIE_BIT) -#define PVSCSI_MSI_OFFSET (0x7c)#define PVSCSI_EXP_EP_OFFSET (0x40)@@ -123,8 +118,6 @@ struct PVSCSIState {uint8_t msi_used;/* For migration compatibility */ PVSCSIRingInfo rings;/* Data transfer rings manager */ uint32_t resetting;/* Reset in progress */ - -uint32_t compat_flags; }; typedef struct PVSCSIRequest {@@ -1224,21 +1217,8 @@ pvscsi_post_load(void *opaque, int version_id)return 0 ; } -static bool pvscsi_vmstate_need_pcie_device (void *opaque) -{ -PVSCSIState *s =PVSCSI (opaque); - -return !(s->compat_flags &PVSCSI_COMPAT_DISABLE_PCIE ); -} - -static bool pvscsi_vmstate_test_pci_device (void *opaque,int version_id) -{ -return !pvscsi_vmstate_need_pcie_device (opaque); -} -static const VMStateDescription vmstate_pvscsi_pcie_device = { .name ="pvscsi/pcie" , - .needed = pvscsi_vmstate_need_pcie_device, .fields = (const VMStateField []) {VMSTATE_PCI_DEVICE (parent_obj, PVSCSIState),VMSTATE_END_OF_LIST ()@@ -1252,9 +1232,6 @@ static const VMStateDescription vmstate_pvscsi = { .pre_save = pvscsi_pre_save, .post_load = pvscsi_post_load, .fields = (const VMStateField []) { -VMSTATE_STRUCT_TEST (parent_obj, PVSCSIState, - pvscsi_vmstate_test_pci_device,0 , - vmstate_pci_device, PCIDevice),VMSTATE_UINT8 (msi_used, PVSCSIState),VMSTATE_UINT32 (resetting, PVSCSIState),VMSTATE_UINT64 (reg_interrupt_status, PVSCSIState),@@ -1288,19 +1265,14 @@ static const VMStateDescription vmstate_pvscsi = {static const Property pvscsi_properties[] = {DEFINE_PROP_UINT8 ("use_msg" , PVSCSIState, use_msg,1 ), -DEFINE_PROP_BIT ("x-disable-pcie" , PVSCSIState, compat_flags, -PVSCSI_COMPAT_DISABLE_PCIE_BIT , false), };static void pvscsi_realize (DeviceState *qdev,Error **errp) {PVSCSIClass *pvs_c =PVSCSI_GET_CLASS (qdev);PCIDevice *pci_dev =PCI_DEVICE (qdev); -PVSCSIState *s =PVSCSI (qdev); -if (!(s->compat_flags &PVSCSI_COMPAT_DISABLE_PCIE )) { - pci_dev->cap_present |=QEMU_PCI_CAP_EXPRESS ; - } + pci_dev->cap_present |=QEMU_PCI_CAP_EXPRESS ; pvs_c->parent_dc_realize (qdev, errp); }
PATCH v4 16/19 hw/scsi/vmw_pvscsi: Convert DeviceRealize -> InstanceInit
In reply to Philippe Mathieu-DaudéSimplify replacing pvscsi_realize() by pvscsi_instance_init(), removing the need for device_class_set_parent_realize().
--- hw/scsi/vmw_pvscsi.c | 13 +++---------- 1 file changed, 3 insertions(+), 10 deletions(-)
hw/scsi/vmw_pvscsi.c +3 -10
@@ -1267,21 +1267,15 @@ static const Property pvscsi_properties[] = { DEFINE_PROP_UINT8("use_msg", PVSCSIState, use_msg, 1), }; -static void pvscsi_realize (DeviceState *qdev,Error **errp) +static void pvscsi_instance_init (Object *obj) { -PVSCSIClass *pvs_c =PVSCSI_GET_CLASS (qdev); -PCIDevice *pci_dev =PCI_DEVICE (qdev); - - pci_dev->cap_present |=QEMU_PCI_CAP_EXPRESS ; - - pvs_c->parent_dc_realize (qdev, errp); +PCI_DEVICE (obj)->cap_present |=QEMU_PCI_CAP_EXPRESS ; }static void pvscsi_class_init (ObjectClass *klass,const void *data) {DeviceClass *dc =DEVICE_CLASS (klass);PCIDeviceClass *k =PCI_DEVICE_CLASS (klass); -PVSCSIClass *pvs_k =PVSCSI_CLASS (klass);HotplugHandlerClass *hc =HOTPLUG_HANDLER_CLASS (klass); k->realize = pvscsi_realizefn;
@@ -1290,8 +1284,6 @@ static void pvscsi_class_init(ObjectClass *klass, const void *data) k->device_id =PCI_DEVICE_ID_VMWARE_PVSCSI ; k->class_id =PCI_CLASS_STORAGE_SCSI ; k->subsystem_id =0x1000 ; -device_class_set_parent_realize (dc, pvscsi_realize, - &pvs_k->parent_dc_realize);device_class_set_legacy_reset (dc, pvscsi_reset); dc->vmsd = &vmstate_pvscsi;device_class_set_props (dc, pvscsi_properties);
@@ -1306,6 +1298,7 @@ static const TypeInfo pvscsi_info = { .class_size = sizeof(PVSCSIClass), .instance_size = sizeof(PVSCSIState), .class_init = pvscsi_class_init, + .instance_init = pvscsi_instance_init, .interfaces = (const InterfaceInfo []) { {TYPE_HOTPLUG_HANDLER }, {INTERFACE_PCIE_DEVICE },
Signature · 1 line
2.47.1
Re: [PATCH v4 16/19] hw/scsi/vmw_pvscsi: Convert DeviceRealize -> InstanceInit
Quoted 6 lines · Philippe Mathieu-Daudé
Simplify replacing
pvscsi_realize()bypvscsi_instance_init(), removing the need fordevice_class_set_parent_realize().Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Zhao Liu <[email protected]>
Quoted 49 lines
--- hw/scsi/vmw_pvscsi.c | 13 +++---------- 1 file changed, 3 insertions(+), 10 deletions(-)hw/scsi/vmw_pvscsi.c +3
-10@@ -1267,21 +1267,15 @@ static const Property pvscsi_properties[] = { DEFINE_PROP_UINT8("use_msg", PVSCSIState, use_msg, 1), }; -static void pvscsi_realize (DeviceState *qdev,Error **errp) +static void pvscsi_instance_init (Object *obj) { -PVSCSIClass *pvs_c =PVSCSI_GET_CLASS (qdev); -PCIDevice *pci_dev =PCI_DEVICE (qdev); - - pci_dev->cap_present |=QEMU_PCI_CAP_EXPRESS ; - - pvs_c->parent_dc_realize (qdev, errp); +PCI_DEVICE (obj)->cap_present |=QEMU_PCI_CAP_EXPRESS ; }static void pvscsi_class_init (ObjectClass *klass,const void *data) {DeviceClass *dc =DEVICE_CLASS (klass);PCIDeviceClass *k =PCI_DEVICE_CLASS (klass); -PVSCSIClass *pvs_k =PVSCSI_CLASS (klass);HotplugHandlerClass *hc =HOTPLUG_HANDLER_CLASS (klass); k->realize = pvscsi_realizefn;@@ -1290,8 +1284,6 @@ static void pvscsi_class_init(ObjectClass *klass, const void *data) k->device_id =PCI_DEVICE_ID_VMWARE_PVSCSI ; k->class_id =PCI_CLASS_STORAGE_SCSI ; k->subsystem_id =0x1000 ; -device_class_set_parent_realize (dc, pvscsi_realize, - &pvs_k->parent_dc_realize);device_class_set_legacy_reset (dc, pvscsi_reset); dc->vmsd = &vmstate_pvscsi;device_class_set_props (dc, pvscsi_properties);@@ -1306,6 +1298,7 @@ static const TypeInfo pvscsi_info = { .class_size = sizeof(PVSCSIClass), .instance_size = sizeof(PVSCSIState), .class_init = pvscsi_class_init, + .instance_init = pvscsi_instance_init, .interfaces = (const InterfaceInfo []) { {TYPE_HOTPLUG_HANDLER }, {INTERFACE_PCIE_DEVICE },
PATCH v4 17/19 hw/net/vmxnet3: Remove VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS definition
In reply to Philippe Mathieu-DaudéVMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS was only used by the hw_compat_2_5[] array, via the 'x-old-msi-offsets=on' property. We removed all machines using that array, lets remove all the code around VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS.
--- hw/net/vmxnet3.c | 20 ++++++-------------- 1 file changed, 6 insertions(+), 14 deletions(-)
hw/net/vmxnet3.c +6 -14
@@ -42,18 +42,13 @@#define VMXNET3_MSIX_BAR_SIZE 0x2000/* Compatibility flags for migration */ -#define VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS_BIT 0 -#define VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS \ - (1 << VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS_BIT)#define VMXNET3_COMPAT_FLAG_DISABLE_PCIE_BIT 1#define VMXNET3_COMPAT_FLAG_DISABLE_PCIE \ (1 << VMXNET3_COMPAT_FLAG_DISABLE_PCIE_BIT)#define VMXNET3_EXP_EP_OFFSET (0x48) -#define VMXNET3_MSI_OFFSET (s) \ - ((s)->compat_flags & VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS ? 0x50 : 0x84) -#define VMXNET3_MSIX_OFFSET (s) \ - ((s)->compat_flags & VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS ? 0 : 0x9c) +#define VMXNET3_MSI_OFFSET (0x84) +#define VMXNET3_MSIX_OFFSET (0x9c)#define VMXNET3_DSN_OFFSET (0x100)#define VMXNET3_BAR0_IDX (0)
@@ -61,8 +56,7 @@#define VMXNET3_MSIX_BAR_IDX (2)#define VMXNET3_OFF_MSIX_TABLE (0x000) -#define VMXNET3_OFF_MSIX_PBA (s) \ - ((s)->compat_flags & VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS ? 0x800 : 0x1000) +#define VMXNET3_OFF_MSIX_PBA (0x1000)/* Link speed in Mbps should be shifted by 16 */ #define VMXNET3_LINK_SPEED (1000 << 16)
@@ -2122,8 +2116,8 @@ vmxnet3_init_msix(VMXNET3State *s) &s->msix_bar,VMXNET3_MSIX_BAR_IDX ,VMXNET3_OFF_MSIX_TABLE , &s->msix_bar, -VMXNET3_MSIX_BAR_IDX ,VMXNET3_OFF_MSIX_PBA (s), - VMXNET3_MSIX_OFFSET(s), NULL); +VMXNET3_MSIX_BAR_IDX ,VMXNET3_OFF_MSIX_PBA , + VMXNET3_MSIX_OFFSET, NULL);if (0 > res) {VMW_WRPRN ("Failed to initialize MSI-X, error %d" , res);
@@ -2221,7 +2215,7 @@ static void vmxnet3_pci_realize(PCIDevice *pci_dev, Error **errp)/* Interrupt pin A */ pci_dev->config[PCI_INTERRUPT_PIN ] =0x01 ; - ret =msi_init (pci_dev,VMXNET3_MSI_OFFSET (s),VMXNET3_MAX_NMSIX_INTRS , + ret =msi_init (pci_dev,VMXNET3_MSI_OFFSET ,VMXNET3_MAX_NMSIX_INTRS ,VMXNET3_USE_64BIT ,VMXNET3_PER_VECTOR_MASK ,NULL );/* Any error other than -ENOTSUP(board's MSI support is broken) * is a programming error. Fall back to INTx silently on -ENOTSUP */
@@ -2472,8 +2466,6 @@ static const VMStateDescription vmstate_vmxnet3 = {static const Property vmxnet3_properties[] = {DEFINE_NIC_PROPERTIES (VMXNET3State, conf), -DEFINE_PROP_BIT ("x-old-msi-offsets" , VMXNET3State, compat_flags, -VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS_BIT , false),DEFINE_PROP_BIT ("x-disable-pcie" , VMXNET3State, compat_flags,VMXNET3_COMPAT_FLAG_DISABLE_PCIE_BIT , false), };
Signature · 1 line
2.47.1
Re: [PATCH v4 17/19] hw/net/vmxnet3: Remove VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS definition
Quoted 8 lines · Philippe Mathieu-Daudé
VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETSwas only used by the hw_compat_2_5[] array, via the 'x-old-msi-offsets=on' property. We removed all machines using that array, lets remove all the code aroundVMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS.Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Zhao Liu <[email protected]>
Quoted 68 lines
--- hw/net/vmxnet3.c | 20 ++++++-------------- 1 file changed, 6 insertions(+), 14 deletions(-)hw/net/vmxnet3.c +6
-14@@ -42,18 +42,13 @@#define VMXNET3_MSIX_BAR_SIZE 0x2000/* Compatibility flags for migration */ -#define VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS_BIT 0 -#define VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS \ - (1 << VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS_BIT)#define VMXNET3_COMPAT_FLAG_DISABLE_PCIE_BIT 1#define VMXNET3_COMPAT_FLAG_DISABLE_PCIE \ (1 << VMXNET3_COMPAT_FLAG_DISABLE_PCIE_BIT)#define VMXNET3_EXP_EP_OFFSET (0x48) -#define VMXNET3_MSI_OFFSET (s) \ - ((s)->compat_flags & VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS ? 0x50 : 0x84) -#define VMXNET3_MSIX_OFFSET (s) \ - ((s)->compat_flags & VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS ? 0 : 0x9c) +#define VMXNET3_MSI_OFFSET (0x84) +#define VMXNET3_MSIX_OFFSET (0x9c)#define VMXNET3_DSN_OFFSET (0x100)#define VMXNET3_BAR0_IDX (0)@@ -61,8 +56,7 @@#define VMXNET3_MSIX_BAR_IDX (2)#define VMXNET3_OFF_MSIX_TABLE (0x000) -#define VMXNET3_OFF_MSIX_PBA (s) \ - ((s)->compat_flags & VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS ? 0x800 : 0x1000) +#define VMXNET3_OFF_MSIX_PBA (0x1000)/* Link speed in Mbps should be shifted by 16 */ #define VMXNET3_LINK_SPEED (1000 << 16)@@ -2122,8 +2116,8 @@ vmxnet3_init_msix(VMXNET3State *s) &s->msix_bar,VMXNET3_MSIX_BAR_IDX ,VMXNET3_OFF_MSIX_TABLE , &s->msix_bar, -VMXNET3_MSIX_BAR_IDX ,VMXNET3_OFF_MSIX_PBA (s), - VMXNET3_MSIX_OFFSET(s), NULL); +VMXNET3_MSIX_BAR_IDX ,VMXNET3_OFF_MSIX_PBA , + VMXNET3_MSIX_OFFSET, NULL);if (0 > res) {VMW_WRPRN ("Failed to initialize MSI-X, error %d" , res);@@ -2221,7 +2215,7 @@ static void vmxnet3_pci_realize(PCIDevice *pci_dev, Error **errp)/* Interrupt pin A */ pci_dev->config[PCI_INTERRUPT_PIN ] =0x01 ; - ret =msi_init (pci_dev,VMXNET3_MSI_OFFSET (s),VMXNET3_MAX_NMSIX_INTRS , + ret =msi_init (pci_dev,VMXNET3_MSI_OFFSET ,VMXNET3_MAX_NMSIX_INTRS ,VMXNET3_USE_64BIT ,VMXNET3_PER_VECTOR_MASK ,NULL );/* Any error other than -ENOTSUP(board's MSI support is broken) * is a programming error. Fall back to INTx silently on -ENOTSUP */ @@ -2472,8 +2466,6 @@ static const VMStateDescription vmstate_vmxnet3 = {static const Property vmxnet3_properties[] = {DEFINE_NIC_PROPERTIES (VMXNET3State, conf), -DEFINE_PROP_BIT ("x-old-msi-offsets" , VMXNET3State, compat_flags, -VMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETS_BIT , false),DEFINE_PROP_BIT ("x-disable-pcie" , VMXNET3State, compat_flags,VMXNET3_COMPAT_FLAG_DISABLE_PCIE_BIT , false), };
PATCH v4 18/19 hw/net/vmxnet3: Remove VMXNET3_COMPAT_FLAG_DISABLE_PCIE definition
In reply to Philippe Mathieu-DaudéVMXNET3_COMPAT_FLAG_DISABLE_PCIE was only used by the hw_compat_2_5[] array, via the 'x-disable-pcie=on' property. We removed all machines using that array, lets remove all the code around VMXNET3_COMPAT_FLAG_DISABLE_PCIE.
--- hw/net/vmxnet3.c | 11 +---------- 1 file changed, 1 insertion(+), 10 deletions(-)
hw/net/vmxnet3.c +1 -10
@@ -41,11 +41,6 @@#define PCI_DEVICE_ID_VMWARE_VMXNET3_REVISION 0x1#define VMXNET3_MSIX_BAR_SIZE 0x2000 -/* Compatibility flags for migration */ -#define VMXNET3_COMPAT_FLAG_DISABLE_PCIE_BIT 1 -#define VMXNET3_COMPAT_FLAG_DISABLE_PCIE \ - (1 << VMXNET3_COMPAT_FLAG_DISABLE_PCIE_BIT) -#define VMXNET3_EXP_EP_OFFSET (0x48)#define VMXNET3_MSI_OFFSET (0x84)#define VMXNET3_MSIX_OFFSET (0x9c)
@@ -2466,8 +2461,6 @@ static const VMStateDescription vmstate_vmxnet3 = {static const Property vmxnet3_properties[] = {DEFINE_NIC_PROPERTIES (VMXNET3State, conf), -DEFINE_PROP_BIT ("x-disable-pcie" , VMXNET3State, compat_flags, -VMXNET3_COMPAT_FLAG_DISABLE_PCIE_BIT , false), };static void vmxnet3_realize (DeviceState *qdev,Error **errp)
@@ -2476,9 +2469,7 @@ static void vmxnet3_realize(DeviceState *qdev, Error **errp)PCIDevice *pci_dev =PCI_DEVICE (qdev);VMXNET3State *s =VMXNET3 (qdev); -if (!(s->compat_flags &VMXNET3_COMPAT_FLAG_DISABLE_PCIE )) { - pci_dev->cap_present |=QEMU_PCI_CAP_EXPRESS ; - } + pci_dev->cap_present |=QEMU_PCI_CAP_EXPRESS ; vc->parent_dc_realize (qdev, errp); }
Signature · 1 line
2.47.1
Re: [PATCH v4 18/19] hw/net/vmxnet3: Remove VMXNET3_COMPAT_FLAG_DISABLE_PCIE definition
Quoted 8 lines · Philippe Mathieu-Daudé
VMXNET3_COMPAT_FLAG_DISABLE_PCIEwas only used by the hw_compat_2_5[] array, via the 'x-disable-pcie=on' property. We removed all machines using that array, lets remove all the code aroundVMXNET3_COMPAT_FLAG_DISABLE_PCIE.Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Zhao Liu <[email protected]>
Quoted 40 lines
--- hw/net/vmxnet3.c | 11 +---------- 1 file changed, 1 insertion(+), 10 deletions(-)hw/net/vmxnet3.c +1
-10@@ -41,11 +41,6 @@#define PCI_DEVICE_ID_VMWARE_VMXNET3_REVISION 0x1#define VMXNET3_MSIX_BAR_SIZE 0x2000 -/* Compatibility flags for migration */ -#define VMXNET3_COMPAT_FLAG_DISABLE_PCIE_BIT 1 -#define VMXNET3_COMPAT_FLAG_DISABLE_PCIE \ - (1 << VMXNET3_COMPAT_FLAG_DISABLE_PCIE_BIT) -#define VMXNET3_EXP_EP_OFFSET (0x48)#define VMXNET3_MSI_OFFSET (0x84)#define VMXNET3_MSIX_OFFSET (0x9c)@@ -2466,8 +2461,6 @@ static const VMStateDescription vmstate_vmxnet3 = {static const Property vmxnet3_properties[] = {DEFINE_NIC_PROPERTIES (VMXNET3State, conf), -DEFINE_PROP_BIT ("x-disable-pcie" , VMXNET3State, compat_flags, -VMXNET3_COMPAT_FLAG_DISABLE_PCIE_BIT , false), };static void vmxnet3_realize (DeviceState *qdev,Error **errp)@@ -2476,9 +2469,7 @@ static void vmxnet3_realize(DeviceState *qdev, Error **errp)PCIDevice *pci_dev =PCI_DEVICE (qdev);VMXNET3State *s =VMXNET3 (qdev); -if (!(s->compat_flags &VMXNET3_COMPAT_FLAG_DISABLE_PCIE )) { - pci_dev->cap_present |=QEMU_PCI_CAP_EXPRESS ; - } + pci_dev->cap_present |=QEMU_PCI_CAP_EXPRESS ; vc->parent_dc_realize (qdev, errp); }
PATCH v4 19/19 hw/net/vmxnet3: Merge DeviceRealize in InstanceInit
In reply to Philippe Mathieu-DaudéSimplify merging vmxnet3_realize() within vmxnet3_instance_init(), removing the need for device_class_set_parent_realize().
--- hw/net/vmxnet3.c | 15 +-------------- 1 file changed, 1 insertion(+), 14 deletions(-)
hw/net/vmxnet3.c +1 -14
@@ -2238,6 +2238,7 @@ static void vmxnet3_instance_init(Object *obj)device_add_bootindex_property (obj, &s->conf.bootindex,"bootindex" ,"/ethernet-phy@0" ,DEVICE (obj)); +PCI_DEVICE (obj)->cap_present |=QEMU_PCI_CAP_EXPRESS ; }static void vmxnet3_pci_uninit (PCIDevice *pci_dev)
@@ -2463,22 +2464,10 @@ static const Property vmxnet3_properties[] = { DEFINE_NIC_PROPERTIES(VMXNET3State, conf), }; -static void vmxnet3_realize (DeviceState *qdev,Error **errp) -{ -VMXNET3Class *vc =VMXNET3_DEVICE_GET_CLASS (qdev); -PCIDevice *pci_dev =PCI_DEVICE (qdev); -VMXNET3State *s =VMXNET3 (qdev); - - pci_dev->cap_present |=QEMU_PCI_CAP_EXPRESS ; - - vc->parent_dc_realize (qdev, errp); -} -static void vmxnet3_class_init (ObjectClass *class,const void *data) {DeviceClass *dc =DEVICE_CLASS (class);PCIDeviceClass *c =PCI_DEVICE_CLASS (class); -VMXNET3Class *vc =VMXNET3_DEVICE_CLASS (class); c->realize = vmxnet3_pci_realize; c->exit = vmxnet3_pci_uninit;
@@ -2489,8 +2478,6 @@ static void vmxnet3_class_init(ObjectClass *class, const void *data) c->class_id =PCI_CLASS_NETWORK_ETHERNET ; c->subsystem_vendor_id =PCI_VENDOR_ID_VMWARE ; c->subsystem_id =PCI_DEVICE_ID_VMWARE_VMXNET3 ; -device_class_set_parent_realize (dc, vmxnet3_realize, - &vc->parent_dc_realize); dc->desc ="VMWare Paravirtualized Ethernet v3" ;device_class_set_legacy_reset (dc, vmxnet3_qdev_reset); dc->vmsd = &vmstate_vmxnet3;
Signature · 1 line
2.47.1
Re: [PATCH v4 19/19] hw/net/vmxnet3: Merge DeviceRealize in InstanceInit
Quoted 6 lines · Philippe Mathieu-Daudé
Simplify merging
vmxnet3_realize()withinvmxnet3_instance_init(), removing the need fordevice_class_set_parent_realize().Signed-off-by Philippe Mathieu-Daudé <[email protected]>Reviewed-by Daniel P. Berrangé <[email protected]>Reviewed-by Zhao Liu <[email protected]>
Quoted 48 lines
--- hw/net/vmxnet3.c | 15 +-------------- 1 file changed, 1 insertion(+), 14 deletions(-)hw/net/vmxnet3.c +1
-14@@ -2238,6 +2238,7 @@ static void vmxnet3_instance_init(Object *obj)device_add_bootindex_property (obj, &s->conf.bootindex,"bootindex" ,"/ethernet-phy@0" ,DEVICE (obj)); +PCI_DEVICE (obj)->cap_present |=QEMU_PCI_CAP_EXPRESS ; }static void vmxnet3_pci_uninit (PCIDevice *pci_dev)@@ -2463,22 +2464,10 @@ static const Property vmxnet3_properties[] = { DEFINE_NIC_PROPERTIES(VMXNET3State, conf), }; -static void vmxnet3_realize (DeviceState *qdev,Error **errp) -{ -VMXNET3Class *vc =VMXNET3_DEVICE_GET_CLASS (qdev); -PCIDevice *pci_dev =PCI_DEVICE (qdev); -VMXNET3State *s =VMXNET3 (qdev); - - pci_dev->cap_present |=QEMU_PCI_CAP_EXPRESS ; - - vc->parent_dc_realize (qdev, errp); -} -static void vmxnet3_class_init (ObjectClass *class,const void *data) {DeviceClass *dc =DEVICE_CLASS (class);PCIDeviceClass *c =PCI_DEVICE_CLASS (class); -VMXNET3Class *vc =VMXNET3_DEVICE_CLASS (class); c->realize = vmxnet3_pci_realize; c->exit = vmxnet3_pci_uninit;@@ -2489,8 +2478,6 @@ static void vmxnet3_class_init(ObjectClass *class, const void *data) c->class_id =PCI_CLASS_NETWORK_ETHERNET ; c->subsystem_vendor_id =PCI_VENDOR_ID_VMWARE ; c->subsystem_id =PCI_DEVICE_ID_VMWARE_VMXNET3 ; -device_class_set_parent_realize (dc, vmxnet3_realize, - &vc->parent_dc_realize); dc->desc ="VMWare Paravirtualized Ethernet v3" ;device_class_set_legacy_reset (dc, vmxnet3_qdev_reset); dc->vmsd = &vmstate_vmxnet3;
Quoted 11 lines · Philippe Mathieu-Daudé
(series reviewed)
Since v3:
- Fixed 2 issues noticed by Thomas (floppy fallback, e1000)Since v2:
- Removed qtest intest-x86-cpuid-compat.cSince v1:
- Fixed issues noticed by Thomas
Quoted 7 lines
The versioned 'pc' and 'q35' machines up to 2.12 been marked as deprecated two releases ago, and are older than 6 years, so according to our support policy we can remove them.
This series only includes the 2.4 and 2.5 machines removal, as it is a big enough number of LoC removed. Rest will follow. Highlight is the legacy
fw_cfgAPI removal :)
Overall it looks good to me, except the below 4 patches
Quoted 1 line
target/i386/cpu: Remove
X86CPU::check_cpuidfield
This one gets hard NAK because it changes the default behavior of QEMU.
Quoted 3 lines
hw/net/e1000: Remove unused
E1000_FLAG_MACflag hw/virtio/virtio-pci: RemoveVIRTIO_PCI_FLAG_MIGRATE_EXTRAdefinition hw/block/fdc-isa: Remove 'fallback' property
I'm not sure about the three. Because Unlike other properties removed by this series, the property name removed by above three patches don't have a 'x-' prefix.
It should be OK to remove properties with 'x-' prefix, but I'm not sure about the ones without it. There might be user using them explicitly. If so, remove them needs to go through standard deprecation process.
Quoted 25 lines · Xiaoyao Li
Quoted 11 lines · Philippe Mathieu-Daudé
(series reviewed)
Since v3:
- Fixed 2 issues noticed by Thomas (floppy fallback, e1000)Since v2:
- Removed qtest intest-x86-cpuid-compat.cSince v1:
- Fixed issues noticed by ThomasQuoted 7 lines
The versioned 'pc' and 'q35' machines up to 2.12 been marked as deprecated two releases ago, and are older than 6 years, so according to our support policy we can remove them.
This series only includes the 2.4 and 2.5 machines removal, as it is a big enough number of LoC removed. Rest will follow. Highlight is the legacy
fw_cfgAPI removal :)Overall it looks good to me, except the below 4 patches
Quoted 1 line
target/i386/cpu: Remove
X86CPU::check_cpuidfieldThis one gets hard NAK because it changes the default behavior of QEMU.
Quoted 7 lines
Quoted 3 lines
hw/net/e1000: Remove unused
E1000_FLAG_MACflag hw/virtio/virtio-pci: RemoveVIRTIO_PCI_FLAG_MIGRATE_EXTRAdefinition hw/block/fdc-isa: Remove 'fallback' propertyI'm not sure about the three. Because Unlike other properties removed by this series, the property name removed by above three patches don't have a 'x-' prefix.
It should be OK to remove properties with 'x-' prefix, but I'm not sure about the ones without it. There might be user using them explicitly. If so, remove them needs to go through standard deprecation process.
As discussed in another thread, we have a lot of properties that are meant for internal use, though they are exposed to the user (and "x-" was originally meant for experimental properties, not for internal ones). I assume these properties here are such internal ones, so I think it should be ok to remove them now without explicit deprecation. We did this in the past already for the older pc machine types, too, and so far nobody ever complained AFAIK. So I suggest to remove them now, and if someone comlains, we can still revert the corresponding patch.
Thomas
Quoted 30 lines · Thomas Huth
Quoted 29 lines · Xiaoyao Li
Quoted 11 lines · Philippe Mathieu-Daudé
(series reviewed)
Since v3:
- Fixed 2 issues noticed by Thomas (floppy fallback, e1000)Since v2:
- Removed qtest intest-x86-cpuid-compat.cSince v1:
- Fixed issues noticed by ThomasQuoted 7 lines
The versioned 'pc' and 'q35' machines up to 2.12 been marked as deprecated two releases ago, and are older than 6 years, so according to our support policy we can remove them.
This series only includes the 2.4 and 2.5 machines removal, as it is a big enough number of LoC removed. Rest will follow. Highlight is the legacy
fw_cfgAPI removal :)Overall it looks good to me, except the below 4 patches
Quoted 1 line
target/i386/cpu: Remove
X86CPU::check_cpuidfieldThis one gets hard NAK because it changes the default behavior of QEMU.
Quoted 3 lines
hw/net/e1000: Remove unused
E1000_FLAG_MACflag hw/virtio/virtio-pci: RemoveVIRTIO_PCI_FLAG_MIGRATE_EXTRAdefinition hw/block/fdc-isa: Remove 'fallback' property
Quoted 5 lines
Quoted 3 lines
I'm not sure about the three. Because Unlike other properties removed by this series, the property name removed by above three patches don't have a 'x-' prefix.
It should be OK to remove properties with 'x-' prefix, but I'm not sure about the ones without it. There might be user using them explicitly. If so, remove them needs to go through standard deprecation process.
As discussed in another thread, we have a lot of properties that are meant for internal use, though they are exposed to the user (and "x-" was originally meant for experimental properties, not for internal ones). I assume these properties here are such internal ones, so I think it should be ok to remove them now without explicit deprecation. We did this in the past already for the older pc machine types, too, and so far nobody ever complained AFAIK. So I suggest to remove them now, and if someone comlains, we can still revert the corresponding patch.
I'm fine with it since nobody explicitly rejects the removal as I did for "target/i386/cpu: Remove X86CPU::check_cpuid field"
Quoted 2 lines
Thomas
Quoted 1 line · Philippe Mathieu-Daudé
(series reviewed)
Thomas, since you wanted to do it, pls feel free to merge.
Quoted 66 lines
Since v3:
- Fixed 2 issues noticed by Thomas (floppy fallback, e1000)Since v2:
- Removed qtest intest-x86-cpuid-compat.cSince v1:
- Fixed issues noticed by ThomasThe versioned 'pc' and 'q35' machines up to 2.12 been marked as deprecated two releases ago, and are older than 6 years, so according to our support policy we can remove them.
This series only includes the 2.4 and 2.5 machines removal, as it is a big enough number of LoC removed. Rest will follow. Highlight is the legacy
fw_cfgAPI removal :)Philippe Mathieu-Daudé (19):
hw/i386/pc: Remove deprecated pc-q35-2.4 and pc-i440fx-2.4 machines hw/i386/pc: RemovePCMachineClass::broken_reserved_endfield hw/i386/pc: Remove pc_compat_2_4[] array
target/i386/cpu: RemoveX86CPU::check_cpuidfield hw/core/machine: Remove hw_compat_2_4[] array hw/net/e1000: Remove unusedE1000_FLAG_MACflag hw/virtio/virtio-pci: RemoveVIRTIO_PCI_FLAG_MIGRATE_EXTRAdefinition hw/virtio/virtio-pci: RemoveVIRTIO_PCI_FLAG_DISABLE_PCIEdefinition hw/i386/pc: Remove deprecated pc-q35-2.5 and pc-i440fx-2.5 machines hw/i386/x86: RemoveX86MachineClass::save_tsc_khzfield hw/nvram/fw_cfg: Remove legacyFW_CFG_ORDER_OVERRIDEhw/core/machine: Remove hw_compat_2_5[] array hw/block/fdc-isa: Remove 'fallback' property hw/scsi/vmw_pvscsi: RemovePVSCSI_COMPAT_OLD_PCI_CONFIGURATION
definition
hw/scsi/vmw_pvscsi: RemovePVSCSI_COMPAT_DISABLE_PCIE_BITdefinition hw/scsi/vmw_pvscsi: Convert DeviceRealize -> InstanceInit hw/net/vmxnet3: RemoveVMXNET3_COMPAT_FLAG_OLD_MSI_OFFSETSdefinition hw/net/vmxnet3: RemoveVMXNET3_COMPAT_FLAG_DISABLE_PCIEdefinition hw/net/vmxnet3: Merge DeviceRealize in InstanceInitinclude/hw/boards.h | 9 +-- include/hw/i386/pc.h | 7 -- include/hw/i386/x86.h | 5 -- include/hw/loader.h | 2 - include/hw/nvram/fw_cfg.h | 10 --- include/hw/virtio/virtio-pci.h | 8 -- target/i386/cpu.h | 1 - hw/block/fdc-isa.c | 4 +- hw/core/loader.c | 14 ---- hw/core/machine.c | 18 ----- hw/i386/pc.c | 42 ++--------- hw/i386/pc_piix.c | 26 ------- hw/i386/pc_q35.c | 26 ------- hw/i386/x86.c | 1 - hw/net/e1000.c | 93 ++++++++++-------------- hw/net/vmxnet3.c | 44 ++--------- hw/nvram/fw_cfg.c | 109 ++-------------------------- hw/scsi/vmw_pvscsi.c | 67 +++-------------- hw/virtio/virtio-pci.c | 11 +-- system/vl.c | 5 -- target/i386/cpu.c | 3 +- target/i386/machine.c | 5 +- tests/qtest/test-x86-cpuid-compat.c | 14 ---- 23 files changed, 77 insertions(+), 447 deletions(-)Signature · 1 line
2.47.1