| Message ID | 20230721062421.12017-1-akihiko.odaki@daynix.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [v2] accel/kvm: Specify default IPA size for arm64 | expand |
Hi Akihiko, On 21/7/23 08:24, Akihiko Odaki wrote: > libvirt uses "none" machine type to test KVM availability. Before this > change, QEMU used to pass 0 as machine type when calling KVM_CREATE_VM. > > The kernel documentation says: >> On arm64, the physical address size for a VM (IPA Size limit) is >> limited to 40bits by default. The limit can be configured if the host >> supports the extension KVM_CAP_ARM_VM_IPA_SIZE. When supported, use >> KVM_VM_TYPE_ARM_IPA_SIZE(IPA_Bits) to set the size in the machine type >> identifier, where IPA_Bits is the maximum width of any physical >> address used by the VM. The IPA_Bits is encoded in bits[7-0] of the >> machine type identifier. >> >> e.g, to configure a guest to use 48bit physical address size:: >> >> vm_fd = ioctl(dev_fd, KVM_CREATE_VM, KVM_VM_TYPE_ARM_IPA_SIZE(48)); >> >> The requested size (IPA_Bits) must be: >> >> == ========================================================= >> 0 Implies default size, 40bits (for backward compatibility) >> N Implies N bits, where N is a positive integer such that, >> 32 <= N <= Host_IPA_Limit >> == ========================================================= > >> Host_IPA_Limit is the maximum possible value for IPA_Bits on the host >> and is dependent on the CPU capability and the kernel configuration. >> The limit can be retrieved using KVM_CAP_ARM_VM_IPA_SIZE of the >> KVM_CHECK_EXTENSION ioctl() at run-time. >> >> Creation of the VM will fail if the requested IPA size (whether it is >> implicit or explicit) is unsupported on the host. > https://docs.kernel.org/virt/kvm/api.html#kvm-create-vm > > So if Host_IPA_Limit < 40, such KVM_CREATE_VM will fail, and libvirt > incorrectly thinks KVM is not available. This actually happened on M2 > MacBook Air. > > Fix this by specifying 32 for IPA_Bits as any arm64 system should > support the value according to the documentation. > > Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com> > --- > V1 -> V2: Introduced an arch hook > > include/sysemu/kvm.h | 1 + > accel/kvm/kvm-all.c | 2 +- > target/arm/kvm.c | 2 ++ > target/i386/kvm/kvm.c | 2 ++ > target/mips/kvm.c | 2 ++ > target/ppc/kvm.c | 2 ++ > target/riscv/kvm.c | 2 ++ > target/s390x/kvm/kvm.c | 2 ++ > 8 files changed, 14 insertions(+), 1 deletion(-) My understanding of Peter's suggestion would be smth like: -- >8 -- diff --git a/include/sysemu/kvm.h b/include/sysemu/kvm.h index 115f0cca79..c0af15eb6c 100644 --- a/include/sysemu/kvm.h +++ b/include/sysemu/kvm.h @@ -201,10 +201,15 @@ typedef struct KVMCapabilityInfo { struct KVMState; +struct KVMClass { + AccelClass parent_class; + + int default_vm_type; +}; + #define TYPE_KVM_ACCEL ACCEL_CLASS_NAME("kvm") typedef struct KVMState KVMState; -DECLARE_INSTANCE_CHECKER(KVMState, KVM_STATE, - TYPE_KVM_ACCEL) +OBJECT_DECLARE_TYPE(KVMState, KVMClass, KVM_ACCEL) extern KVMState *kvm_state; typedef struct Notifier Notifier; diff --git a/accel/kvm/kvm-all.c b/accel/kvm/kvm-all.c index 373d876c05..fdd424e1a5 100644 --- a/accel/kvm/kvm-all.c +++ b/accel/kvm/kvm-all.c @@ -2458,12 +2458,13 @@ static int kvm_init(MachineState *ms) KVMState *s; const KVMCapabilityInfo *missing_cap; int ret; - int type = 0; + int type; uint64_t dirty_log_manual_caps; qemu_mutex_init(&kml_slots_lock); s = KVM_STATE(ms->accelerator); + type = KVM_GET_CLASS(s)->default_vm_type; /* * On systems where the kernel can support different base page diff --git a/target/arm/kvm.c b/target/arm/kvm.c index b4c7654f49..5c13594fdf 100644 --- a/target/arm/kvm.c +++ b/target/arm/kvm.c @@ -1064,4 +1064,8 @@ bool kvm_arch_cpu_check_are_resettable(void) void kvm_arch_accel_class_init(ObjectClass *oc) { + KVMClass *kc = KVM_CLASS(oc); + + /* Host_IPA_Limit ... */ + kc->default_vm_type = 32; } ---
On Fri, 21 Jul 2023 at 08:30, Philippe Mathieu-Daudé <philmd@linaro.org> wrote: > > Hi Akihiko, > > On 21/7/23 08:24, Akihiko Odaki wrote: > > libvirt uses "none" machine type to test KVM availability. Before this > > change, QEMU used to pass 0 as machine type when calling KVM_CREATE_VM. > > > > The kernel documentation says: > >> On arm64, the physical address size for a VM (IPA Size limit) is > >> limited to 40bits by default. The limit can be configured if the host > >> supports the extension KVM_CAP_ARM_VM_IPA_SIZE. When supported, use > >> KVM_VM_TYPE_ARM_IPA_SIZE(IPA_Bits) to set the size in the machine type > >> identifier, where IPA_Bits is the maximum width of any physical > >> address used by the VM. The IPA_Bits is encoded in bits[7-0] of the > >> machine type identifier. > >> > >> e.g, to configure a guest to use 48bit physical address size:: > >> > >> vm_fd = ioctl(dev_fd, KVM_CREATE_VM, KVM_VM_TYPE_ARM_IPA_SIZE(48)); > >> > >> The requested size (IPA_Bits) must be: > >> > >> == ========================================================= > >> 0 Implies default size, 40bits (for backward compatibility) > >> N Implies N bits, where N is a positive integer such that, > >> 32 <= N <= Host_IPA_Limit > >> == ========================================================= > > > >> Host_IPA_Limit is the maximum possible value for IPA_Bits on the host > >> and is dependent on the CPU capability and the kernel configuration. > >> The limit can be retrieved using KVM_CAP_ARM_VM_IPA_SIZE of the > >> KVM_CHECK_EXTENSION ioctl() at run-time. > >> > >> Creation of the VM will fail if the requested IPA size (whether it is > >> implicit or explicit) is unsupported on the host. > > https://docs.kernel.org/virt/kvm/api.html#kvm-create-vm > > > > So if Host_IPA_Limit < 40, such KVM_CREATE_VM will fail, and libvirt > > incorrectly thinks KVM is not available. This actually happened on M2 > > MacBook Air. > > > > Fix this by specifying 32 for IPA_Bits as any arm64 system should > > support the value according to the documentation. > > > > Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com> > > --- > > V1 -> V2: Introduced an arch hook > > > > include/sysemu/kvm.h | 1 + > > accel/kvm/kvm-all.c | 2 +- > > target/arm/kvm.c | 2 ++ > > target/i386/kvm/kvm.c | 2 ++ > > target/mips/kvm.c | 2 ++ > > target/ppc/kvm.c | 2 ++ > > target/riscv/kvm.c | 2 ++ > > target/s390x/kvm/kvm.c | 2 ++ > > 8 files changed, 14 insertions(+), 1 deletion(-) > > My understanding of Peter's suggestion would be smth like: > > -- >8 -- > diff --git a/include/sysemu/kvm.h b/include/sysemu/kvm.h > index 115f0cca79..c0af15eb6c 100644 > --- a/include/sysemu/kvm.h > +++ b/include/sysemu/kvm.h > @@ -201,10 +201,15 @@ typedef struct KVMCapabilityInfo { > > struct KVMState; > > +struct KVMClass { > + AccelClass parent_class; > + > + int default_vm_type; The kernel docs say you need to check for the KVM_CAP_ARM_VM_IPA_SIZE before you can pass something other than zero to the KVM_CREATE_VM ioctl, so this needs to be a method, not just a value. (kvm_arm_get_max_vm_ipa_size() will do this bit for you.) If the machine doesn't provide a kvm_type method, we should default to "largest the host supports", I think. I was wondering if we could have one per-arch method for "actually create the VM" that both was a place for arm to set the default vm type and also let us get the TARGET_S390X and TARGET_PPC ifdefs out of this bit of kvm-all.c, but maybe that would look just a bit too awkward: if (kc->create_vm(s, board_sets_kvm_type, board_kvm_type) < 0) { goto err; } where board_sets_kvm_type is a bool, true if board_kvm_type is valid, and board_kvm_type is whatever the board's mc->kvm_type method told us. (Default impl of the method: call KVM_CREATE_VM ioctl with retry-on-eintr, printing the simple error message; PPC and s390 versions similar but with their arch specific extra messages; arm version has a different default type if board_sets_kvm_type is false.) Not trying to do both of those things with one method would result in a simpler type = kc->get_default_kvm_type(s); API. thanks -- PMM
diff --git a/include/sysemu/kvm.h b/include/sysemu/kvm.h index 115f0cca79..7cc2eb1a8c 100644 --- a/include/sysemu/kvm.h +++ b/include/sysemu/kvm.h @@ -347,6 +347,7 @@ bool kvm_device_supported(int vmfd, uint64_t type); /* Arch specific hooks */ +extern const int kvm_arch_default_type; extern const KVMCapabilityInfo kvm_arch_required_capabilities[]; void kvm_arch_accel_class_init(ObjectClass *oc); diff --git a/accel/kvm/kvm-all.c b/accel/kvm/kvm-all.c index 373d876c05..914ade3ec3 100644 --- a/accel/kvm/kvm-all.c +++ b/accel/kvm/kvm-all.c @@ -2458,7 +2458,7 @@ static int kvm_init(MachineState *ms) KVMState *s; const KVMCapabilityInfo *missing_cap; int ret; - int type = 0; + int type = kvm_arch_default_type; uint64_t dirty_log_manual_caps; qemu_mutex_init(&kml_slots_lock); diff --git a/target/arm/kvm.c b/target/arm/kvm.c index b4c7654f49..f8203b9915 100644 --- a/target/arm/kvm.c +++ b/target/arm/kvm.c @@ -32,6 +32,8 @@ #include "hw/irq.h" #include "qemu/log.h" +const int kvm_arch_default_type = 32; + const KVMCapabilityInfo kvm_arch_required_capabilities[] = { KVM_CAP_LAST_INFO }; diff --git a/target/i386/kvm/kvm.c b/target/i386/kvm/kvm.c index ebfaf3d24c..00bccffdfc 100644 --- a/target/i386/kvm/kvm.c +++ b/target/i386/kvm/kvm.c @@ -87,6 +87,8 @@ static void kvm_init_msrs(X86CPU *cpu); +const int kvm_arch_default_type; + const KVMCapabilityInfo kvm_arch_required_capabilities[] = { KVM_CAP_INFO(SET_TSS_ADDR), KVM_CAP_INFO(EXT_CPUID), diff --git a/target/mips/kvm.c b/target/mips/kvm.c index c14e8f550f..754366874e 100644 --- a/target/mips/kvm.c +++ b/target/mips/kvm.c @@ -33,6 +33,8 @@ static int kvm_mips_fpu_cap; static int kvm_mips_msa_cap; +const int kvm_arch_default_type; + const KVMCapabilityInfo kvm_arch_required_capabilities[] = { KVM_CAP_LAST_INFO }; diff --git a/target/ppc/kvm.c b/target/ppc/kvm.c index a8a935e267..86f8b645a5 100644 --- a/target/ppc/kvm.c +++ b/target/ppc/kvm.c @@ -54,6 +54,8 @@ #define DEBUG_RETURN_GUEST 0 #define DEBUG_RETURN_GDB 1 +const int kvm_arch_default_type; + const KVMCapabilityInfo kvm_arch_required_capabilities[] = { KVM_CAP_LAST_INFO }; diff --git a/target/riscv/kvm.c b/target/riscv/kvm.c index 9d8a8982f9..fbe56ec3bb 100644 --- a/target/riscv/kvm.c +++ b/target/riscv/kvm.c @@ -764,6 +764,8 @@ void kvm_riscv_init_user_properties(Object *cpu_obj) kvm_riscv_destroy_scratch_vcpu(&kvmcpu); } +const int kvm_arch_default_type; + const KVMCapabilityInfo kvm_arch_required_capabilities[] = { KVM_CAP_LAST_INFO }; diff --git a/target/s390x/kvm/kvm.c b/target/s390x/kvm/kvm.c index a9e5880349..4b9347ffc8 100644 --- a/target/s390x/kvm/kvm.c +++ b/target/s390x/kvm/kvm.c @@ -144,6 +144,8 @@ static CPUWatchpoint hw_watchpoint; static struct kvm_hw_breakpoint *hw_breakpoints; static int nb_hw_breakpoints; +const int kvm_arch_default_type; + const KVMCapabilityInfo kvm_arch_required_capabilities[] = { KVM_CAP_LAST_INFO };